Download Differential Expression of Four Connexin Genes, Cx-26, Cx

0013-7227/96/503.00/O
Endocrinology
Copyright 0 1996 by The Endocrine
Vol. 137, No. 11
Printed in U.S.A.
Society
Differential
Expression of Four Connexin Genes, Cx-26,
Cx-30.3, Cx-32, and Cx-43, in the Porcine Ovarian Follicle
KOJI
ITAHANA”,
YOKO
Graduate School of Biological
630-01, Japan
MORIKAZU,
AND
TATSUO
Sciences, Nara Institute
TAKIZYA
of Science and Technology,
ABSTRACT
Connexin genes expressed in porcine ovaries were isolated by reverse transcription-PCR,
and the expression of four connexin genes
(Cx-43, -32, -30.3, and -26) was detected by in situ hybridization
in
internal and surrounding
compartments
of large antral follicles.
Cx-43 and Cx-30.3 were expressed in the theta interna as well as in
the granulosa cell compartment. However, two zones were observed
in the theta interna by probing with Cx-43; about a quarter of the
region close to the lamina basalis appeared devoid of Cx-43 messenger
RNA (mRNA) and protein as well. Cx-30.3, in contrast, seemed to be
expressed ubiquitously in these two compartments. Cx-26 and Cx-32
G
Al’ JUNCTIONS are specialized plasma membrane associations containing intercellular channels that coordinate tissue function by allowing cell to cell passage of
ions and small molecules (l-3). Gap junction channels (GJC)
are formed by the association of two hemichannels (connexons), one provided by each of the communicating cells. Each
connexon is composed of 6 identical subunits, or connexins,
that are essential to form a central pore and are products of
members of a growing multigene family; at present, 12 different genes in mammals have been identified that code for
different members of the connexin family (1, 2). Some connexins have been cloned from multiple species, and their
sequences have turned out to be fairly conserved throughout
the species; e.g. connexin-43 (Cx-43) (4) appears to be the
most ubiquitous connexin, sharing 97% amino acid identity
among mammals (2). In addition to the conserved primary
structures in the transmembrane domains, the molecular
diversity of different connexins in the cytoplasmic domains
could be an indicator of the potential structural and regulatory complexities of GJC. Indeed, connexins are expressed
in a variety of cells, and each cell type has its own characteristic pattern of expression of the connexin gene (2, 3).
In ovarian follicles, intercellular communications between
oocytes and somatic cell components through GJC are involved in folliculogenesis and oogenesis (5). First, gap junctions permit the transfer of metabolites such as nucleotides,
amino acids, and sugars from the granulosa cells to the
oocyte as needed for growth and development. Second, the
Received April 29, 1996.
Address all correspondence and requests for reprints to: Dr. Tatsuo
Takeya, Graduate School of Biological Sciences, Nara Institute of
Science and Technology,
Ikoma, Nara 630-01, Japan. E-mail:
[email protected].
*On leave of absence from the Institute for Chemical Research, Kyoto
University (Kyoto, Japan).
5036
Ikoma, Nara
were expressed only in the thecal cells. Both Cx-43 and -30.3 also gave
positive signals in the cumulus cells to the same extent as obtained
in the granulosa cells with the respective probes, whereas the expression of Cx-32 and -26 was undetectable. Cx-43 mRNA drastically
decreased in the functioning stage of the corpus luteum, whereas
Cx-30.3 mRNA increased significantly
in the early stage of luteinization in the estrus cycles. These results suggest that Cx-43 is not the
only gap junction protein to be expressed in the ovarian follicle; other
proteins, such as Cx-32, -30.3, and -26 seem to be expressed, and their
are regulated
differently.
(Endocrinology
137:
expressions
5036~5044,1996)
communication through GJC may be very important in the
regulation of meiosis. Mammalian oocytes are known to
become arrested in prophase I of meiosis during the fetal and
neonatal periods until the LH surge resumes the meiotic
arrest and eventually causes germinal vesicle breakdown at
puberty. During this process, the disassembly of gap junctions correlates temporally with the initiation of germinal
vesicle breakdown (6-B), and maturation was initiated when
the cumulus-oocyte complex was experimentally separated
from the follicle well (5,9,10). Although the exact role of LH
during the process is not yet clear, one well established role
of LH is its involvement, in concert with another gonadotropin, FSH, in follicular steroidogenesis. According to a
proposed mechanism described in the two cell-two gonadotropin hypothesis (ll), an intercellular communication between two types of cells, granulosa and thecal cells, may be
essential. Taking into consideration the finding that several
types of cells are involved in both folliculogenesis and oogenesis, it is reasonable to suggest that multiple connexin
genes participate in these processes, and it is important to
extend the analysis to the surrounding thecal cell compartment as well. To date, Cx-43 has been the only connexin gene
reported to be expressed in granulosa cells (12, 13); no information is available on thecal cells.
In this study, we extended the search for connexin genes
by examining their expression in porcine ovarian follicles
and the surrounding thecal region. We detected transcripts
of Cx-32, -30.3, and -26 in addition to Cx-43 by reverse transcription-PCR (RT-PCR) and confirmed their expressions by
in situ hybridization. Cx-43 and the other three genes were
differentially expressed in the respective compartments. To
further analyze the expression in estrous cycles, messenger
RNA (mRNA) levels of these connexin genes in follicular and
luteal phases of the porcine ovaries were analyzed by Northern blot analysis and RT-PCR.
CONNEXIN
Animals,
Materials
tissues, and cells
mRNA IN THE OVARIAN
and Methods
Pig ovaries were obtained from 4- to 12-month-old pigs at a local
slaughterhouse (14). Follicular contents were sucked out by syringes
from small or large follicles (-1-7 mm in diameter) and used directly as
a source of follicular cells containing granulosa cells and oocytes. Cumulus-oocyte complexes could be isolated under the microscope. Corpora lutea were classified into three stages in addition to corpus albicans:
early, functional, and regressing stages based on the morphological
characteristics of the surface (14).
RNA extraction
oligonucleotides
and RT-PCR cloning
with degenerate
Total RNA was extracted by acid guanidinium-phenol
chloroform
procedure from follicles directly or from corpora lutea (15), and possible
contaminated genomic DNA was removed by digestion with deoxyribonuclease free of ribonuclease (RQl DNase, Promega, Madison, WI).
Ten micrograms of total RNA were reverse transcribed into complementary DNA (cDNA) by Superscript II (BRL) with oligo-(deoxythymidine),,_,, primer for RT-I’CR according to the supplier’s instructions.
Two pairs of degenerate oligonucleotides (primers Fl and F2 and
primers Rl and R2; F stands for forward primers and R for reverse
primers) corresponding to the conserved regions among the connexin
family were synthesized (Fig. 1, A and B) (3). Primer Fl corresponds to
the amino-terminal region, and primer F2 includes a conserved amino
acid sequence, VCYD, within the first extracellular domain. Primer R2
includes a conserved amino acid sequence, VDCF, in the second extracellular domain, and Rl corresponds to the carboxy-end portion within
the same domain. To minimize probe degeneracy, not all of the possible
sequence variations revealed in the alignment were represented in each
probe. RT-PCR reactions (30 ~1 in volume) contained 0.2 pg RT products,
0.2 FM of primer F or R, and 200 PM deoxy-NTP with Tuq Extender Buffer
(Stratagene, La Jolla, CA). The addition of AmpliTaq (2 U; I’erkin-Elmer,
A
Cytoplasmic
Membrane
Extracellular
3
t
t
3
F2
R2
t Rl
FOLLICLE
Norwalk, CT) and Tuq Extender PCR Additive (2 U; Stratagene) was
performed after the reaction reached 94 C to minimize undesired priming during the initial cycle. Thirty cycles at 94 C were performed for 30
set, 60 C for 2 min, and 72 C for 3 min. This was followed by a final
extension for 15 min at 72 C. Reaction products were separated and
visualized on agarose gel electrophoresis. The products were finally
subcloned in pT7Blue-T cloning vector (Novagen) and subjected to DNA
sequence analysis by AB13700 (Perkin-Elmer).
Tissue sections and hematoxylin
primerF1
:
ATGGGTGACTGGAG(C/T)(G/T)(C/T)C(C/T)T(G/A)G
primer F2 :
GGCTG(C/T)(AG)A(G/C)AA(C/T)GTCTGCTA(T/C)G
primer Rl :
ACCACCA(G/A)CAT(G/A)AAGA(T/C)(G/A)ATGAAG
primer RZ :
GCC(T/G)GGA(G/A)A(C/T)(G/A)AA(G/A)CAGTCCAC
FIG. 1. Primers and reactions of RT-PCR. A, Domains of the connexins are defined according to the method of Bennett et al. (3) and are
indicated: amino-terminus
(N), cytoplasmic loop (Cy), carboxy-terminus (C), membrane spanning(Ml-M4),
and extracellular loop (El and
E2). The locations of four primers constructed for RT-PCR are shown
side by side (Fl and F2; Rl and R2). B, Nucleotide sequences of each
primer.
leosin staining
Ovarian sections were prepared for in situ hybridization essentially
as described previously (15). In brief, the removed ovaries were covered
in OCT embedding compound (Miles, Elkhart, IN) and snap-frozen in
liquid nitrogen. Cryosections (8 pm thick) were prepared using a Cryostat HM500-OM (Microm, Germany), collected on silane-coated slides,
and stored at -80 C until hybridization. For fixation, slides were dipped
into freshly prepared 4% paraformaldehyde in PBS for 60 min at room
temperature and washed in PBS (10 min), followed by digestion with
protenase K (1 pg/ml) for 10 min at 37 C. The slides were washed in 0.1
M triethanolamine
for 1 min, rinsed in distilled water, and dipped into
0.1 M triethanolamine
containing 0.25% acetic anhydride for 1 min.
Sections were washed finally in PBS and sequentially dehydrated for 1
min each at 70%, 80%, 90%, and 100% ethanol (followed by air-drying).
Hematoxylin-eosin staining was performed essentially according to the
procedure described in Ref. 15.
In situ hybridization
and probes
DNA regions containing unique sequences of the respective connexin
genes (Fig. 2) were subcloned into the RNA transcription vector pBluescript II SK+ or pT7Blue T-vector. Complementary RNA probes on the
antisense were prepared by in vitro transcription of the linearized constructs using SP6 or T7 RNA polymerase in the presence of digoxigeninUTP. After RNA polymerase reactions, DNA templates were hydrolyzed with RQl deoxyribonuclease.
Procedures for the in situ hybridization
were essentially based on
those of Ausubel et al. (15). Each probe was diluted in hybridization
buffer containing 50% formamide, 0.6 M NaCl, 10 mM Tris (pH 7.4), 1 mM
EDTA (pH 8.0), 0.25% SDS, 10% dextran sulfate, 1 X Denhardt’s solution,
and 200 ~g/ml Esckevickia coli transfer RNA. Probe solution was heated
at 85 C for 3 min, chilled on ice, and added to each slide in a volume of
30 ~1. The slides were incubated overnight at 50 C in a covered container
humidified with 50% formamide and drained; they were then immersed
in 5 X SSC (standard saline citrate), followed by incubation in 2 X SSC
containing 50% formamide for 30 min at 50°C. The slides were washed
with TEN [0.5 M NaCl, 10 mM Tris (pH 8.0), and 1 mM EDTA] for 10 min
at 37 C, then treated with 1 pg/ml ribonuclease A in TEN for 30 min at
37 C and finally with 2 X SSC for 20 min at 50 C. The slides were then
washed with 0.2 x SSC at 50 C twice. Sections were stained by the use
of a digoxigenin detection kit (Boehringer Mannheim, Mannheim,
Germany) according to the supplier’s instructions,
mRNA levels: Northern
B
5037
blotting
and RT-PCR
Total RNA samples (20 pg) from follicular cells or the respective
stages of corpora lutea (14) were denatured with formaldehyde, electrophoresed in 1.2% agarose gels, and transferred onto nylon membranes (Immobilon, Millipore Corp., Bedford, MA) as described previously (10, 14). Hybridization
was performed under conditions of high
stringency; filters were washed three times with 2 X SSC-0.1% SDS for
15 min each time, followed by washing in 0.2 X SSC-0.1% SDS at 37 C.
A probe used for filter hybridizations was derived from the Cx-43 cDNA
clone used for in situ hybridization described above, but the probe was
labeled with [32P]deoxy-CTP instead of digoxigenin.
The components of RT-PCR reactions were identical to those of reactions with the degenerate oligonucleotides described above unless
otherwise indicated, whereas reaction conditions were slightly modified; 35 cycles at 94 C were performed for 30 set, 60 C for 1 min, and 72
C for 2 min, followed by a final extension for 5 min at 72 C. Reaction
products were separated and visualized on acrylamide gel electrophoresis. For quantitative analysis, the products were stained with CYBR
Green I (FMC Bioproducts), and the intensities were measured by a
CONNEXIN mRNA IN THE OVARIAN FOLLICLE
5038
POT cx43
Hum Cx43
Rat cx43
II
I,-- ----------
KL LDKVQAYSTA GGKVWLSVLF IFRILLLGTA
Primer Fl -- ----------------------------
1
--- ------- ----------
Endo.
Vol137.No
VESAWGDEQS AFRCNTQQPG CENVCYDKSF PISHVRFWVL QIIFVSVPTL
---------------------------__________
-----_----
----------
__________
--_-------
11
LYLAHVFYVM
_---__--__
----------
----------
----__----
FKSVFEVAFL
---I------
LIQWYIYGFS
_---------
LSAVYTCKRD PCPHQVDCFL
----------------_-_
---v------
----_-----
----------
I
RKEEKLNKKE EELKVAQTDG VNVEMHLKQI EIKKFKYGIE
---D-----------------_-----------------_------
---_------
---E------
----------
EHGKVKMRGG LLRTYIISIL
__________
----___-__
----------
----_-----
----------
SRPTEKT
---_-----_---
POT Cx32
Hum Cx32
Rat Cx32
0
PrimerF2
HF FPISHVRLWS LQLILVSTPA
__ __---------_---____
Q----_----__________
LLVAMHVAHQ QHIEKKMLRL
__________
----_-----_--______
----------
EGHADPLHLE EVKRHKVHIS GTLWWTYVIS WFRLLFEAA
---G-----__________
______------------v
---G-v---_-_-____-_
------------------V
FMYVFYLLYP
_------------------
GYAMVRLVKC EAYPCPNT
_-__-------F-------------DV------
EF FPLSHVRLWA LQLILVTCPS
__ --.J--------____----
Par cx30.3
Mus Cx30.3
Por Cx26
Hum Cx26
Rat Cx26
LLWMHVAYR QERERKHRLK HGPDARSLYD NPNKKRGGLW WTYLLSLIFK
---------E-----------N-PA--~
-LS------_---------
rl-- ---------primerF2
HY FPISHIRLWA
_-
-_--------
LQLIFVSTPA
----M-----------s--
LLVAMHVA~R RHEKKRKFIK
-----__----------M__________
_______---
GEMKSEFKDI EEIKTQKVRI
--I-N-------_-_____
--I--------_-_-__--
AAVDAGFLYI
v---S-----
EGALWWTYTG SIFFRILFEA
--s------T
-----VI----s------s
-----VI---
FHRLYRNYDM
--(-I-Kj-,---
AFMYVFYVMY
V------I------_----
SGFSMQRLVK CNAWPCPNT
N--F ______ --------PrimerR
p---------------0
FIG. 2. Predicted amino acid sequences of cDNA clones amplified by RT-PCR on porcine ovarian follicle mRNA. Amplified products with the
primers shown in Fig. 1B were subcloned into pT’IBlue-T vector and subjected to DNA sequence analysis. A search for amino acid sequences
thus determined was made in the SWISS-PROT database. Each amplified region of porcine counterpart is aligned with the corresponding
region(s) of the mammalian gene(s): human Cx-43 (17), rat Cx-43 (4), human Cx-32 (18), rat Cx-32 (19), mouse Cx-30.3 (20), human Cx-26 (21),
and rat Cx-26 (22). Amino acid sequences identical to those of the porcine genes are indicated by dashes. Open. boxes indicate the primer sites.
Fluor Imager (Molecular Dynamics, Sunnyvale, CA). A portion of nucleotide sequences of the glyceraldehyde-3-phosphate
dehydrogenase
(GAPDH) gene that were conserved among the mammals (16) was
chosen to construct primers for PCR; the reaction was expected to produce a 122-bp fragment, which was used for preparation of 32P-labeled
GAPDH probe as well.
Immunocytochemistry
Frozen sections were rehydrated into PBS to remove OCT compound
and incubated in PBS containing 10% skim milk. Incubation with antiCx-43 monoclonal antibody (Chemicon International) was performed
overnight at 4 C, followed by washing three times with PBS. The bound
antibodies were revealed with antimouse Ig-fluorescein (5-bromo-2’deoxyuridine labeling and detection kit I, Boehringer Mannheim) according to the manufacturer’s instructions.
R.esldts
Domain structure of the connexin family
of primers for PCR
and construction
A search was made for connexin genes expressed in the
porcine ovarian follicles by applying RT-PCR. To design
primers for that purpose, domain structures identified in
the connexin family were considered. The general archi-
tecture of the connexin genes can be outlined, as shown in
Fig. 1, based on the sequence similarities between previously characterized connexin genes and the evidence obtained by biochemical analyses (3); the N- and C-terminals
are cytoplasmic (designated N and C, respectively), and
there are four membrane-spanning regions (Ml-M4), with
a cytoplasmic loop (Cy) and two extracellular loops (El
and E2). Of these domains, Cy and C are relatively unique
regions for an individual gene, and regions other than
these two are well conserved in the family. However, there
exist some portions within the N, E2, and M4 regions
where identities are shared only by genes such as Cx-32,
-31, and -26, making it possible to divide the family into
two groups (1,3,11): group I (or /3) consists of genes such
as Cx-32, -31, and -26, whereas group II (or) consists of
genes such as Cx-43, -46, and -38.
Two degenerate oligonucleotide primers, Fl and F2, corresponding to the sense strands of N and El, respectively,
and two other degenerate primers, Rl and R2, corresponding
to the antisense strand sequences of the 5’- and 3’-portions
of E2, respectively, were prepared (Fig. 1B). Fl and Rl are
CONNEXIN mRNA IN THE OVARIAN FOLLICLE
specific to group II, whereas F2 and R2 are common to all
connexin genes.
cDNA cloning and DNA sequence analyses
RT-PCR amplification with combinations of the abovedescribed primers at a 60 C annealing temperature produced
distinct respective bands after agarose gel electrophoresis
(data not shown). The products were recovered from gel,
subcloned in pT7Blue-T vector, and subjected to DNA sequence analyses. Products with combinations of primers
other than with F2 and R2 were found to encode single
species of the connexin gene Cx-43, as far as sequenced; a
predicted amino acid sequence from a deduced DNA sequence was identical to that of rat Cx-43 except for two
residues (Fig. 2) (4). In contrast, products with F2 and R2
were found to contain a mixture of porcine counterparts of
mammalian Cx-32, Cx-30.3, and Cx-26 in addition to Cx-43
(Fig. 2) (17-22). All cloned cDNA contained the Cy region
that is unique to each connexin gene in addition to conserved
M2 and M3 regions (Fig. lA), as expected. The degree of
sequence identity in each gene was fairly high, i.e. at least
90%, among the mammals.
In situ hybridization
As Cx-43 is the only type of gap junction protein reported
to be expressed in the granulosa cell compartment (12, 13),
we wanted to confirm and examine further the expression of
connexin genes other than Cx-43 in the corresponding and
neighboring compartments. To do so, a digoxigenin-labeled
probe specific to an individual gene was prepared and hybridized with thin sections of porcine ovary (Fig. 3, lanes A
and B). To begin with, cross-reactivities between each probe
were examined, and their specificities were confirmed (data
not shown).
Cx-43 was expressed in the granulosa cell compartment
(the follicle granulosa cells within the basement membrane) and in the region surrounding the follicles as well
(Fig. 3). These two regions were clearly separated by cell
layers where no signal was detected; based on the profile
of hematoxylin-eosin staining, both stained and unstained
compartments were assumed to be included in the theta
interna, indicating that the theta interna could be divided
into two zones and that about a quarter of internal thecal
cell layers adjacent to the basal membrane was devoid of
the Cx-43 mRNA. This expression profile was further confirmed at the protein level by immunoblotting
using an
anti-Cx-43 antibody as well as in a different follicle (Fig.
4, A-C). The expression profile of Cx-30.3 was different
from that of Cx-43 in two respects: 1) Cx-30.3 mRNA was
detected in both the granulosa cell and the surrounding
thecal compartments without interruption as observed in
the case of Cx-43; and 2) the signal with Cx-30.3 was more
intense in thecal cells than in granulosa cells. The difference in intensities between Cx-43 and Cx-30.3 does not
reflect the difference in their mRNA expression levels,
because the specific activities and/ or lengths of these two
probes were not identical. Cx-26 and Cx-32, in contrast,
were expressed exclusively in the thecal cell compartment
(Fig. 3). Therefore, Cx-43 seems not to be the only gap
5039
junction protein expressed in the ovarian follicle; other
genes, such as Cx-26, -30.3, and -32, appear to be expressed
as well.
Finally, ribonuclease treatment of the sections before hybridization failed to give any signals with any probe, indicating that hybridization reactions were actually between
probes and intrinsic unique transcripts. Furthermore, the
expression of all Cx genes examined were restricted to within
the follicle and its surrounding thecal regions and not in the
regions of surrounding stroma.
Expression
of connexin genes in cumulus-oocyte
complexes
Besides GJC that connect single species of cells, intercellular couplings between granulosa cells and cumulus
cells or between cumulus cells and oocytes are believed to
play important roles in the regulation of both oocyte
growth and meiosis (5). We, therefore, examined connexin
gene expression in the cumulus cells (Fig. 3, lane C). Due
to the fixation procedures or for other reasons, as porcine
sections used in the above experiments were derived from
either fully grown antral follicles or out-centered portions
of follicles, no cumulus-oocyte complexes (COC) were recovered or detected in those preparations. We thereby
recovered COC from the total follicle contents, spun them
down onto slides glasses, and probed them with respective
digoxigenin-labeled DNAs. Four or five COC were examined with each probe, and the typical profiles obtained are
shown in Fig. 3 (lane C).
Both the Cx-43 and Cx-30.3 probes gave intense signals in
the cumulus cells, whereas signals were undetectable with
the Cx-32 and Cx-26 probes. It is difficult to conclude that
there was a total absence of Cx-32 and Cx-26 transcripts in
COC, but it was clear that the intensities given by Cx-43 or
Cx-30.3 were at least comparable in the granulosa cell compartment and the cumulus cells, whereas the signal was
significantly faint in cumulus cells in the case of Cx-32 and
Cx-26 (Fig. 3). The same reasoning may be possible with the
oocytes themselves, where no significant signals were obtained with any probes; staining was seen in the central
region with the Cx-30.3 probe derived from the surrounding
cumulus cells, not from the oocyte.
Connexin
mRNA levels in estrus cycle
The expression profiles of connexin genes in luteinizing
ovaries have not yet been reported. As Cx-43 and Cx-30.3
were found in the present study to be expressed in granulosa
cells, we prepared total RNA from three stages of corpus
luteum (early stage, functional stage, and regressing stage)
and corpus albicans in addition to RNA from both small (cl
mm in diameter) and large (>5 mm in diameter) follicles and
examined their mRNA levels by Northern blot analysis and
RT-PCR (Fig. 5).
The expression of Cx-43 mRNA was clearly detected by
Northern blot analysis; the expression level was relatively
high in follicle cells and during the regression stage, whereas
the expression was significantly down-regulated in the functional stage of the corpus luteum (Fig. 5A). The same expression profile of Cx-43 mRNA was confirmed by RT-PCR
(data not shown). The expression of Cx-30.3 mRNA, on the
CONNEXIN mRNA IN THE OVARIAN FOLLICLE
Endo . 1996
Vol 137 . No 11
FIG. 3. Expression of Cx-43, -32, -30.3, and -26 mRNAs in the porcine large follicles. Sections (8 pm) adjacent to each other were hybridized
with the respective digoxigenin-labeled
probes. Lane A shows a representative follicle (magnification,
X20); lane B shows the same sample with
higher magnification (X 100); lane C shows a representative profile of cumulus-oocyte complex with the respective probes (bar indicates 25 km).
The sample on the bottom ro’owindicates a follicle stained with hematoxylin-eosin;
an area surrounded by an open &XC is the region shown in
lane B.
other hand,
was difficult
to detect quantitatively
by Northern
blot; hence, we examined its expression by RT-PCR on the
same RNA preparations as those studied for Cx-43 expres-
sion (Fig. 5B). The condition used for PCR produced amplification in a linear fashion; a 5-fold increase in the template
caused an equivalent increase in the product, and a serial
CONNEXIN
mRNA
IN THE
OVARIAN
FOLLICLE
5041
ne antral follicles. A, A section (8 pm) containing a follicle different from that used in Fig. 3 was stained
with anti-Cx-43 antibody as a primary antibody and visualized by fluorescein isothiocyanate-labeled
antimouse IgG as a second antibody. B,
An adjacent section was visualized with the same second antibody, but without the treatment with anti-Cx-43 antibody. C, A section adjacent
to that used in A was hybridized with the digoxigenin-labeled
Cx-43 probe. D, A hematoxylin-eosin-stained
section. Magnification,
x200.
decrease in the template resulted in significant reductions of
the products to the level where any quantification was difficult (Fig. 58, lanes 3 and 8-10). The peak of Cx-30.3 mRNA
expression was seen in large follicles and the early stage of
the corpus luteum, and the authenticities of the bands were
confirmed by Southern blot with the 32P-labeled Cx-30.3
probe (data not shown).
Discussion
In this study, we identified the expression of four connexin
genes (Cx-43, Cx-32, Cx-30.3, and Cx-26) in the porcine ovary
by applying RT-PCR and characterized their expression profiles by in situ hybridization, Northern blot analysis, and
RT-PCR. The results revealed that these four types of con-
CONNEXIN mRNA IN THE OVARIAN FOLLICLE
5042
Endo. 1996
Vol 137. No 11
123456
Ml2345678910
cx-30.3
Ml2345678910
GAPDH
GAPDH \
100
7
80
z!
$
2
60
40
2
20
0
123456
123456
FIG. 5. Expression profiles of Cx-43 and Cx-30.3 mRNAs in the estrus cycle. A, Total RNA (20 pg) isolated from each stage in the estrus cycle
of porcine ovary were run on agarose gel, transferred onto nylon membrane (Immobilon N, Millipore), and probed with the 32P-labeled Cx-43
cDNA. Data are representative of two experiments. The positions of the 28s and 18s ribosomal RNAs are indicated by arrows (top panel). The
same filter was probed with the 32P-labeled GAPDH DNA (middle panel). The expression level at each stage was normalized with the densities
given by GAPDH mRNA, and their relative estimates against the lane 1 are shown as a linear chart (bottom panel). Each point represents the
mean level for two experiments. Lane 1, Small follicles; lane 2, large follicles; lane 3, corpus luteum (early stage); lane 4, corpus luteum
(functioning stage); lane 5, corpus luteum (regressing stage); lane 6, corpus albicans. B, The expression of Cx-30.3 was examined by RT-PCR
using the same RNA preparations as those described above, and the product profiles (131 bp) obtained by acrylamide gel electrophoresis are
shown (top panel). The RT products were examined by PCR with GAPDH primers producing 122-bp fragments (middle panel). Data are
representative of two experiments. The expression levels were normalized as described in A, and the level of lane 3 is indicated as 100%; lane
numbers correspond to those of the top and middle panels (bottom panel). Each point represents the mean level for two experiments. Lane M,
HaeIII-digested pUC18 DNA, lanes 1-6, the same as in A, lane 7, no template; lane 8, same RNA but 5 times as much as in lane 3; lane 9,
l/&h the RNA of lane 3; lane 10, 1/25th RNA of lane 3.
nexin genes were actually expressed within the follicles and
the surrounding thecal cells, and their expression patterns
were differentially regulated both temporally and spatially,
as schematically summarized in Fig. 6. The signals thus obtained with the respective probes were not simply due to
cross-hybridization with other genes, because the specificity
of each probe was examined against those of other probes,
and no cross-reactivities were detected, as described above.
However, in general, the possibility of the existence of a
gene(s) that is still unidentified but possesses structural commonalities with the connexin family cannot be excluded.
The connexin family genes could be further classified into
two groups based on their primary structures: group I (p),
including Cx32, Cx-30.3, and Cx-26, and group II (a), Cx-43.
Although it is not yet clear whether this classification by
structural features reflects any physiological significance, it
is clear that both types of connexin genes are expressed in
ovarian follicles. Cx-32 and Cx-26 seem to be expressed predominantly in the thecal cell compartment, distinguishing
these two genes from Cx-30.3 within group II.
The differential expressions of these genes between granulosa and thecal cells may reflect their differences in physiological roles; one possible role is in the biogenesis of steroid
hormones. Granulosa and thecal cells have been speculated
to have different roles in steroidogenic capacities. One factor
to be considered is differences in the cells’ spatial localizations; the granulosa cells of the follicles are avascular, and the
granulosa cells are not accessible to low density lipoprotein,
which is the principal source of cholesterol used for steroidogenesis. In contrast, thecal intema cells can become uptake
precursors for steroidogenesis and respond to LH. These
possibilities have been examined by using isolated granulosa
cells and thecal cells, leading to the two cell-two gonadotropin theory (ll), which proposed that in response to LH,
the theta interna cells convert C,, steroids to Cl9 steroids, and
that FSH stimulates granulosa cells to aromatize the preformed C,, steroids to estrogen. To support such a mechanism and make it possible, not only GJC, which might be
specific for each cell type to create intercellular synchronization but also another type of GJC that can bridge different
CONNEXIN
mRNA
IN THE
Cx32
cx43
TFu- 0
‘I
BM
GC
\
cx30.3
CUM(--
Cx26
-u-0
FIG. 6. Species and localizations
porcine antral follicles
expression of connexin
ing. TE, Theta externa,
granulosa cells; CUM,
of connexin genes expressed in the
are schematically shown. Regions where the
genes were identified are illustrated by shadTI, theca interna; BM, basal membrane; GC,
cumulus cells; 00, oocytes.
OVARIAN
5043
FOLLICLE
cruitment of dominant follicles to be matured or ovulated
(27).
Observations by Medui et al. (23) and in our study also
strongly suggest that Cx-43 and LHR are coexpressed in the
rest of the compartments. It has been established, however,
that an LH surge causes down-regulation of Cx-43 mRNA
expression (28,29) (Itahana, K., and T. Takeya, unpublished
observation). The coexpression of LHR and Cx-43 in antral
or preovulatory follicles thus suggests that the LH level during the follicle phase of the estrus cycle is not sufficient to
repress Cx-43 mRNA expression directly or induce such
inhibitory activity, whereas the increased LH level at the
surge could be sufficient to trigger the down-regulatory
mechanism.
In the present study, we newly identified the expression
of Cx-30.3 in cumulus granulosa cells. The expression of
Cx-43 in this region has been reported (13,30,31). Valdimarsson et al. (30) further demonstrated the expression of Cx-32
in COC; they identified the products at both the protein and
mRNA levels. In contrast, the expression of Cx-32 in cumulus
cells seemed much fainter than that in thecal cells and was
almost undetectable in this study. This does not necessarily
mean that Cx-32 is not expressed in cumulus cells, because
Valdimarsson et al. (30) used both RT-PCR and an anti-Cx-32
antibody to detect the product, and thereby the sensitivity of
their methods may differ from ours.
References
types of cells might exist. Our present findings on the identities and the expression profiles of connexin genes are relevant to this question.
Although Cx-43 and Cx-30.3 are coexpressed in both granulosa and theta interna cells, they showed distinct differences, in that Cx-43 was not expressed specifically in the
theta interna region close to the lamina basalis. The immunostaining of large antral follicles by using anti-LH receptor
(LHR) antibody previously revealed the presence of two
clear-cut zones in the theta interna (23); about one third of
the inner layers of theta interna cells adjacent to the basal
lamina were not stained, whereas the rest of the internal
thecal cells were significantly stained. The functional significance of the existence of the zone devoid of LHR is not yet
clear, but the existence of thecal cell layers where key enzymes for steroidogenesis, such as 3-hydroxysteroid dehydrogenase, P450-17 anhydrase, and aromatase could not be
detected in human follicles, has also been reported (24-26).
Based on the species and stage specificities of the sources of
the materials and on the present hematoxylin-eosin staining
profile, this LHR-devoid zone seems to correspond to the
portion of theta interna detected in this study as a region
devoid of Cx-43 mRNA.
The existence of the zone lacking Cx-43 between the theta
and granulosa cell compartments in the large antral follicles
implies the unique role of Cx-43 in folliculogenesis and/or
oogenesis. This zone may function as a blockade against the
follicle compartment by interrupting the transfer of signal
molecules produced in the theta cells by LH stimulation, and
we suspect that this mechanism may be crucial for the re-
1. Kumar
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
NM, Gilula NB 1996 The zap iunction communication channel. Cell
84381-388
Beyer EC 1993 Gap junctions. Int Rev Cyt 137C:l-37
Bennett MVL, Barrio LC, Bargiello TA, Spray DC, Hertzberg E, Saez JC 1991
Gap junctions: new tools, new answers, new questions. Ne&on 6:3Og320
Beyer EC, Paul DL, Goodenough DA 1987 Connexin 43: a protein from rat
heart homologous to a gap junction protein from liver. J Cell Biol105:2621-2629
Eppig JJ 1991 Intercommunication
between mammalian oocytes and companion somatic cells. BioEssays 13:569-574
Gilula NB, Epstein ML, Beers W 1978 Cell-to-cell communication and ovul&ion; a study of the cumulus-oocyte complex. J Cell Biol 78:58-75
Larsen WJ, Wert SE, Brunner GD 1986 A dramatic loss of cumulus cell gap
junction is correlated with germinal vesicle breakdown in rat oocytes. Dev Biol
113:517-521
Wert SE, Larsen WJ 1990 Preendocytotic alterations in cumulus cell gap
junctions precede meiotic resumption in the rat cumulus-oocyte complex.
Tissue Cell 22:827-851
Tsfriri A, Charming Cl’ 1975 An inhibitory influence of granulosa cells and
follicular fluid upon oocyte meiosis in vim. Endocrinology 96:922-927
Racowsky C, Baldwin KV 1989 In vitro and in vim studies reveal that hamster
oocyte m&tic arrest is maintained only transiently by follicular fluid, but
persistently
by membrana/cumulus
granulosa cell contact. Dev Biol
134297-306
Gore-Langton RE, Armstrong D 1994 Follicular steroidogenesis and its control. In: Knobil E, Neil1 JD (eds) The Physiology of Reproduction. ed 2. Raven
Press, New York, vol 1:571-627
Risek B, Guthrie S, Kumar N, Gilula NB 1990 Modulation of gap junction
transcript and protein expression during pregnancy in the rat. J Cell Biol
110:269-282
Wiesen JF, Midgley Jr R 1993 Changes in expression of connexin 43 gap
junction messenger ribonucleic acid and protein during ovarian follicular
growth. Endocrinology 133:741-746
Tanaka T, Andoh N, Takeya T, Sato E 1992 Differential screening of ovarian
cDNA libraries detected the expression of the porcine collagenase inhibitor
gene in junctional corpora lutea. Mol Cell Endocrinol83:65-71
Ausubel FM, Brent R, Kingston RE, Moore DD, Smith JA, Seidman JG,
Shuhl K 1988 Current Protocols in Molecular Biology, suppl 7. Wiley Interscience, New York
Tokunaga K, Nakamura Y, Sakata K, Fujimori K, Ohkubo M, Sawada K,
Sakiyama S 1987 Enhanced expression of a glyceraldehyde-3-phosphate
dehydrogenase gene in human lung cancers. Cancer Res 47:5616-5619
Fishman GI, Spray DC, Leinwand LA 1990 Molecular characterization and
VI
I
5044
18.
19.
20.
21.
22.
23.
24.
CONNEXIN
mRNA IN THE OVARIAN
functional expression of the human cardiac gap junction channel. J Cell Biol
111:589-598
Kumar NM, Gilula NB 1986 Cloning and characterization of human and rat
liver cDNAs coding for a gap junction protein. J Cell Biol 103:767-776
Paul DL 1986 Molecular cloning of cDNA for rat liver gap junction protein.
J Cell Biol 103:123-134
Hennemann
H, Dahl E, White JB, Schwarz H-J, Lalley PA, Chang S,
Nicholson BJ, Willecke K 1992 Two gap junction genes, connexin 31.1 and
30.3, are closely linked on mouse chromosome 4 and preferentially expressed
in skin. J Biol Chem 26717225-17233
Lee SW, Tomasetto C, Paul D, Keyomarsi K, Sager R 1992 Transcriptional
downregulation
of gap junction proteins blocks junctional communication in
human mammary tumor cell lines. J Cell Biol 188:1213-1221
Zhang J-T, Nicholson BJ 1989 Sequence and tissue distribution of a second
protein of hepatic gap junctions, Cx2.6, as deduced from its cDNA. J Cell Biol
109:3391-3401
Meduri G, Vuhai-Luuthi
MT, Jolivet A, Milgrom E 1992 New functional
zonation in the ovary as shown by immunohistochemistry
of luteinizing hormone receptor. Endocrinology 131:366-373
Inkster SE, Brodie A 1991 Expression of aromatase cytochrome P-450 in
premenopausal and postmenopausal human ovaries: an immunocytochemical
FOLLICLE
Endo. 1996
Vol 137 . No 11
study. J Clin Endocrionol Metab 73:717-726
25. Sasano H, Mori T, Sasano N, Negora H, Mason JI 1990 Immunolocalization
of 3-B-hydroxysteroid
dehydrogenase
in human ovary. J Reprod Fertil
89:743-751
26. Zlotkin T, Fargash Y, Orly I 1986 Cell-specific expression of immunoreactive
cholesterol side-chain cleavage cytochrome P-450 during follicular development in the rat ovary. Endocrinology
119:2809-2820
27. Fortune JE 1994 Ovarian follicular growth and development in mammals, Biol
Reprod 501225-232
28. Schreiber JR, Beckmann MW, Polacek D, Davies PF 1993 Changes in gap
junction connexin-43 messenger ribonucleic acid levels associated with rat
ovarian follicular development as demonstrated by in situ hybridiazation. Am J
Obstet Gynecol 168:1094-1104
29. Granot I, Dekel N 1994 Phosphorylation and expression of connexin-43 ovarian gap junction protein are regulated by luteinizing hormone. J Biol Chem
269:30502-30509
30. Valdimarsson G, De Sousa PA, Kidder GM 1993 Coexpression of gap junction
proteins in the cumulus-oocyte complex. Mol Reprod Dev 367-15
31. Sutovsky P, Flechon JE, Flechon B, Motlik J, Peynot N, Chesne P, Heyman
Y 1993 Dynamic changes of gap junctions and cytoskeleton during in vitro
culture of cattle oocyte cumulus complexes. Biol Reprod 49:1277-1287