Laminin-alpha6beta1 integrin interaction enhances survival and proliferation and modulates steroidogenesis of ovine granulosa cells.

Abstract

This study aimed to determine the physiological role of laminin (LN) and its receptor, alpha(6)beta(1) integrin, in controlling the functions of granulosa cells (GC) during follicular development in sheep ovary. Immunohistochemistry experiments showed the presence of increasing levels of LN (P<0.0001), and high levels of mature alpha(6)beta(1) integrin in GC layers of healthy antral follicles during the follicular and the preovulatory phases of the estrous cycle. In vitro, the addition of a function-blocking antibody raised against alpha(6) subunit (anti-alpha(6) IgG) to the medium of ovine GC cultured on LN impaired cell spreading (P<0.0001), decreased the proliferation rate (P<0.05) and increased the apoptosis rate (P<0.05). Furthermore, addition of anti-alpha(6) IgG enhanced estradiol (E2) secretion by GC in the presence or absence of follicle-stimulating hormone (FSH), luteinizing hormone or insulin-like growth factor-I in culture medium (P<0.0001), and inhibited progesterone (P4) secretion in basal conditions or in the presence of low (0.5 ng/ml) FSH concentrations only (P<0.0001). The anti-alpha(6) IgG effect was specific to an interaction of LN with alpha(6)beta(1) integrin since it was ineffective on GC cultured on heat-denatured LN, RGD (arginine-glycine-aspartic acid) peptides and non-coated substratum. Hence, this study established that alpha(6)beta(1) integrin 1) was expressed in GC of antral follicles, 2) mediated the actions of LN on survival, proliferation and steroidogenesis of GC, and 3) was able to dramatically modulate P4 and E2 secretion by GC in vitro. It is suggested that during the follicular and the preovulatory phases of the estrous cycle, the increasing levels of LN in GC of large antral follicles might support their final development to ovulation.

Full text

Page 1

Laminin–?6?1integrin interaction enhances survival and proliferation
and modulates steroidogenesis of ovine granulosa cells
F Le Bellego, C Pisselet, C Huet, P Monget and D Monniaux
Physiologie de la Reproduction et des Comportements, UMR 6073 INRA-CNRS-Université François Rabelais de Tours, INRA 37380 Nouzilly, France
(Requests for offprints should be addressed to D Monniaux; Email: monniaux@tours.inra.fr)
(C Huet is now at The Burnham Institute, La Jolla, California 92037, USA)
Abstract
This study aimed to determine the physiological role of
laminin (LN) and its receptor, ?6?1integrin, in controlling
the functions of granulosa cells (GC) during follicular
development in sheep ovary. Immunohistochemistry
experiments showed the presence of increasing levels of
LN (P<0·0001), and high levels of mature ?6?1integrin in
GC layers of healthy antral follicles during the follicular
and the preovulatory phases of the estrous cycle. In vitro,
the addition of a function-blocking antibody raised against
?6subunit (anti-?6IgG) to the medium of ovine GC
cultured on LN impaired cell spreading (P<0·0001),
decreased the proliferation rate (P<0·05) and increased the
apoptosis rate (P<0·05). Furthermore, addition of anti-?6
IgG enhanced estradiol (E2) secretion by GC in the
presence or absence of follicle-stimulating hormone (FSH),
luteinizing hormone or insulin-like growth factor-I in
culture medium (P<0·0001), and inhibited progesterone
(P4) secretion in basal conditions or in the presence of low
(0·5 ng/ml) FSH concentrations only (P<0·0001). The
anti-?6IgG effect was specific to an interaction of LN
with ?6?1integrin since it was ineffective on GC cultured
on heat-denatured LN, RGD (arginine-glycine-aspartic
acid) peptides and non-coated substratum. Hence, this
study established that ?6?1integrin 1) was expressed in
GC of antral follicles, 2) mediated the actions of LN on
survival, proliferation and steroidogenesis of GC, and 3)
was able to dramatically modulate P4 and E2 secretion by
GC in vitro. It is suggested that during the follicular and the
preovulatory phases of the estrous cycle, the increasing
levels of LN in GC of large antral follicles might support
their final development to ovulation.
Journal of Endocrinology (2002) 172, 45–59
Introduction
In the ovary, terminal development of antral follicles to
ovulation is characterized by important transitions in the
functions of granulosa cells (GC). First, GC from antral
follicles gradually lose their proliferative activity and dif-
ferentiate into estradiol (E2)-secreting cells. Thereafter, in
preovulatory follicles, GC luteinize into progesterone
(P4)-secreting cells in response to the preovulatory surge
of the gonadotropins follicle-stimulating hormone (FSH)
and luteinizing hormone (LH). It has previously been
hypothesized that extracellular matrix (ECM) components
present in ovarian follicles might modulate the actions of
gonadotropins on these transitions (Gospodarowicz et al.
1980, Savion et al. 1981, Amsterdam et al. 1989, Aten
et al. 1995, Aharoni et al. 1996, Huet et al. 1997, Rodgers
et al. 1999). Within the ovarian follicles, the outer layer of
GC rests on a basement membrane separating the intra-
follicular compartment from the theca (Luck 1994). This
basement membrane is believed to play a role in influenc-
ing proliferation and differentiation of GC (Amsterdam
et al. 1989, Luck 1994) and in protecting them from
entering apoptosis in early atretic follicles (Monniaux
et al. 1999). Laminin (LN) has previously been shown to
be one of the most abundant ECM components within
the basement membrane (Wordinger et al. 1983,
Leardkamolkarn & Abrahamson 1992, Zhao & Luck
1995, Lee et al. 1996, Huet et al. 1997, van Wezel et al.
1998, Rodgers et al. 1999, Alexopoulos et al. 2000). In
vitro, purified LN used as a culture substratum improves
GC survival in rat (Aharoni et al. 1996) and sheep (Huet
et al. 2001) and stimulates their proliferation in sheep
(Huet et al. 2001). However, the effects of LN on GC
steroidogenesis are still the object of controversy since LN
enhances P4 secretion in rat (Aten et al. 1995, Aharoni
et al. 1996) and pig (Sites et al. 1996), but inhibits it in
human (Fujiwara et al. 1997). Moreover, the functional
importance of LN in GC differentiation to an E2-secreting
cell type during terminal follicular development, and then
to a P4-secreting cell type during the luteinization process
has not been established.
LN exerts its biological effects acting on target cells via
specific cell surface ?? heterodimeric receptors named
integrins (Vuori 1998, Giancotti & Ruoslahti 1999). At
45
Journal of Endocrinology (2002) 172, 45–59
0022–0795/02/0172–045
? 2002 Society for Endocrinology
Printed in Great Britain
Online version via http://www.endocrinology.org

Page 2

least nine different integrins have been described as LN
receptors (Kühn & Eble 1994, Mercurio 1995). Interest-
ingly, GC from different animal species express very high
levels of ?6integrin subunit (human: Honda et al. 1995;
marmoset: Giebel et al. 1996, Giebel and Rune 1997; pig:
Fujiwara et al. 1995; mouse: Nakamura et al. 1997) which
has been reported to associate with ?1or ?4integrin
subunit, forming highly specific LN receptors in various
cell types (Sonnenberg et al. 1988). In vivo, GC of primates
strongly express ?1integrin subunit (Giebel et al. 1996,
Giebel & Rune 1997), suggesting that a functional ?6?1
LN receptor may be present in these cells. However, the
role of ?6?1integrin in mediating the actions of LN on GC
survival, proliferation and steroidogenesis during follicular
development and luteinization remains to be established.
The present study aimed to assess the role of both LN
and its receptor, ?6?1integrin, in controlling GC functions
during terminal follicular development in sheep ovary. For
this purpose, changes in LN and ?6?1integrin levels were
studied in sheep antral follicles larger than 1 mm diameter
during the follicular and preovulatory phases of the estrous
cycle in vivo. In further experiments in vitro, the effects of
the addition of a function-blocking antibody specifically
raised against ?6integrin, that has been reported to inhibit
interaction between ?6?1integrin and all LN isoforms
(Sonnenberg et al. 1990), were investigated on GC shape,
apoptosis, proliferation, and secretion of E2 and P4.
Results suggest that LN–integrin ?6?1interaction has an
important role to play in modulating survival, growth and
steroidogenesis of GC in sheep antral follicles and thereby
might participate in the regulation of terminal follicular
development to the preovulatory stage.
Materials and Methods
Reagents and chemicals
Fluorogestone acetate sponges used to synchronize estrous
cycles, and human chorionic gonadotropin (hCG) used for
injections to animals were obtained from Intervet (Angers,
France). Porcine FSH (pFSH) from pituitary extract
(pFSH activity=1·15 times activity of NIH pFSH-P1)
used for injections to animals, and purified ovine LH
(oLH) (CY 1085, oLH activity=1·38 times activity of
NIH S21) were obtained from Dr Y Combarnous
(Nouzilly, France). Purified ovine FSH-20 (oFSH) (lot no.
AFP-7028D, 4453 IU/mg, FSH activity=75 times
activity of oFSH-S1) used for culture treatment was a gift
from NIDDK, National Hormone Pituitary Program,
Bethesda, MD, USA. Recombinant human insulin-like
growth factor-I (IGF-I) was a gift from Dr P Swift
(Ciba-Geigy, Saint Aubin, Switzerland). B2 medium was
prepared according to Menezo (1976). Rat monoclonal
antibody GoH3 raised against human ?6integrin subunit
(anti-?6IgG) was purchased from Serotec (Oxford, Oxon,
UK), and rabbit polyclonal antibody anti-LN-1 (anti-LN)
from Engelbreth-Holm-Swarm (EHS) mouse sarcoma was
from the Institut Pasteur (Lyon, France). Mouse mono-
clonal antibody raised against human integrin ?1subunit
(anti-?1IgG) was a gift from Dr C Boucheix (Paris,
France). Anti-rat and anti-rabbit IgG antibodies coupled
to horseradish peroxidase were purchased from Interchim
(Montluçon, France), and anti-mouse IgG antibody
coupled to horseradish peroxidase was obtained from
Diagnostic Pasteur (Marnes-la-Coquette, France). The
following reagents were purchased from Sigma (L’Isle
d’Abeau Chesnes, France): McCoy’s 5a medium with
bicarbonate,penicillin/streptomycin,
albumin (BSA tissue culture grade) used for culture
medium, transferrin, selenium, bovine insulin, andro-
stenedione, LN from EHS tumor (mainly LN-1; Mercurio
1995), anti-rat IgG (whole molecule)-agarose, Igepal,
phenylmethylsulfonyl fluoride
epoxysuccinyl-leucylamide-(4-guanidino)-butane (E64),
pepstatin A and aprotinin. Hepes, -glutamine, fungizone
and trypsin were purchased from GIBCO BRL (Cergy-
Pontoise, France). Pronectin F (RGD peptides: arginine-
glycine-asparticacid sequence)
Interchim, Tissue-Tek from Miles Laboratories (Elkhart,
IN, USA), 3-aminopropyltriethoxy-silane (silane) and
saponin from Fluka Biochemika (Mulhouse, France), BSA
used for immunochemistry from Bayer (Puteaux, France),
Vectastain Elite ABC kit from Vector Laboratories
(AbCys, Paris, France), 3,3?-diaminobenzidine tetra-
hydrochloride dihydrate (DAB) from Aldrich Chimie
(L’Isle d’Abeau Chesnes, France) and Schiff reagent for
Feulgen staining from Merck (Schuchardt, Germany).
ECL(enhancedchemiluminescence)
obtained from Amersham Pharmacia Biotech (Les Ulis,
France). Sterile 96-well plates (Nunclon Delta) were
obtained from Nunc (Naperville, IL, USA) and plastic
tissue culture chamber slides from Poly-Labo (Strasbourg,
France). [3H]Thymidine (specific activity 6·7 Ci/nmol)
was obtained from Dupont De Nemours (Les Ulis, France)
and NTB2 emulsion for autoradiography from Integra
Bioscience (Cergy-Pontoise, France). Nitrocellulose filter
(0·2 µm pore size) was purchased from Schleicher and
Schuell (Ecquevilly, France). The BioRad Protein Assay
kit was purchased from BioRad (Ivry s/Seine, France).
bovineserum
(PMSF),
-trans-
was obtainedfrom
reagents were
Animals
All procedures were approved by the Agricultural Agency
and the Scientific Research Agency (approval number A
37801) and conducted in accordance with the guidelines
for Care and Use of Agricultural Animals in Agricultural
Research and Teaching. Seventy-eight adult Romanov
ewes were treated with intravaginal sponges impregnated
with progestagen (fluorogestone acetate, 40 mg) for
15 days to mimic a luteal phase.
For immunohistochemistry experiments, 10 ewes were
slaughtered at different times of the follicular phase. A first
F LE BELLEGO and others·
Actions of laminin on ovine granulosa cells
46
www.endocrinology.org
Journal of Endocrinology (2002) 172, 45–59

Page 3

group of 3 ewes was slaughtered in the early follicular
phase (12 h after sponge removal). A second group of 3
ewes was slaughtered in the late follicular phase (36 h after
sponge removal). A third group of 4 ewes was slaughtered
in the preovulatory phase (20 h after an intramuscular
injection of 1000 IU hCG performed 36 h after sponge
removal to mimic a preovulatory LH surge). After
slaughter,ovarieswere dissected
embedded in Tissue-Tek and frozen in liquid nitrogen for
subsequent immunohistochemical analysis.
Alternatively, cells for culture and immunoblotting
were collected from animals in the luteal phase of the
estrous cycle (10 days after sponge removal, plus intra-
muscular injections of 6 IU and 5 IU pFSH 24 h and 12 h
prior to slaughter respectively). This treatment allowed
recovery of GC with a viability ranging between 60 and
80%, as estimated by trypan blue exclusion. Ovaries from
3 ewes were used for each culture experiment. A total
number of 12 and 18 cultures of GC from small and large
follicles, respectively, were performed.
andimmediately
Immunohistochemistry
Ovarian sections (10 µm) were prepared with a cryostat
and attached to silane-coated slides. The size and the
degree of atresia of each follicle were estimated on serial
sections fixed in methanol–formaldehyde–acetic acid
(80:15:5) and subsequently stained with Feulgen. Follicles
were classified according to their size: small follicles (1–
3 mm in diameter) and large follicles (3·5–7 mm in diam-
eter). The degree of atresia of each follicle was assessed by
microscopic examination of GC: healthy follicles with
frequent mitosis and no pyknosis, atretic follicles with
pyknosis and no mitosis, and late atretic follicles with rare
healthy GC and widespread pyknosis.
Sections were fixed for 10 min in ice-cold 4% para-
formaldehyde in phosphate-buffered saline (PBS) contain-
ing 0·1% saponin (PBS/saponin) and washed in PBS/
saponin. Immunohistochemistry was then carried out
using the Vectastain ABC kit according to the manu-
facturer’s instructions with few modifications. Briefly,
sections were treated with 0·3% H2O2 for 30 min,
saturated in rabbit or goat serum, and incubated overnight
with primary antibodies in PBS/saponin containing 0·1%
BSA (final dilutions were 10 ng/ml for anti-?6IgG,
1:100 000 for anti-LN and 1·5 ng/ml for anti-?1IgG).
Control sections were incubated with the same dilution of
purified rat IgG or rabbit non-immune serum diluted in
the same buffer. All sections were subsequently incubated
for 4 h at room temperature with biotinylated rabbit
anti-rat or goat anti-rabbit IgG (final dilution 1:800 in
PBS/saponin/BSA), washed in PBS/saponin, and finally
incubated for 30 min with the avidin–biotin–horseradish
peroxidase complex. The staining was obtained by incu-
bating sections in Tris–HCl (50 mM, pH 7·8) containing
0·2 mg/ml DAB and 0·0036% H2O2for 4 min at room
temperature.
For LN and integrin ?6subunit, quantification of
immunostaining was performed by densitometric analysis
using a microscope-linked personal computer-based image
analyzer (SAMBA TM 2005, Alcatel TITN Meylan,
France) as previously described (Huet et al. 1997). Each
section was analyzed with a ?100 objective. For each
follicle, the mean optical density of staining (expressed in
arbitrary units) was estimated from measurements of 20
microscopical square fields of constant area (45 µm2)
chosen at random in the wall of GC. Quantitative analysis
was performed on follicles larger than 1 mm in diameter
present in one ovary from each of the 10 ewes. In total,
150 follicles were analyzed: 73 healthy follicles, 41 atretic
follicles and 36 late atretic follicles. To assess the specificity
of staining, additional measurements were performed in
follicular antrum, in non-vascularized areas of interstitial
tissue close to the studied follicles, and in ovarian control
sections incubated with rat IgG or rabbit non-immune
serum.
Isolation of GC
Immediately after slaughter, ovaries from 3 ewes were
immersed for 15 min in isotonic solution containing
fungizone and antibiotics (penicillin and streptomycin).
Then ovaries were placed in B2 medium and follicles
larger than 1 mm diameter were dissected within 1 h of
slaughter. Follicular fluid from large follicles (>4 mm) was
aspirated with a 26-gauge needle. Each follicle was then
slit open in B2 medium and GC were removed by gently
scraping the interior surface of the follicle with a platinum
loop. GC suspensions were pooled according to the size
class (small: 1–3 mm, or large: >4 mm) of follicles. The
two resulting cell suspensions were centrifuged at 300?g
for 7 min and re-suspended in culture medium (McCoy’s
5a containing bicarbonate supplemented with 20 mmol/l
Hepes, 100 kUI/l penicillin, 0·1 g/l streptomycin,
3 mmol/l -glutamine, 0·1% BSA (w/v), 100 µg/l insulin,
0·1 µmol/l androstenedione, 5 mg/l transferrin, 20 µg/l
selenium). Cells were counted and the viability was
estimated by trypan blue exclusion. For immunoprecipi-
tation and western blotting experiments, suspensions were
centrifuged at 300?g for 7 min. The pellets were stored
at ?20 ?C until experiments.
Immunoprecipitation
GC were lysed for 30 min on ice in 1 ml RIPA buffer
containing 50 mM Tris–HCl (pH 8), 150 mM NaCl, 1%
Igepal, 0·5% sodium deoxycholate, 0·1% SDS and protease
inhibitors (2 mM PMSF, 10 µM aprotinin, 29·2 µM
pepstatin A). Insoluble material was removed by a quick
centrifugation at 20 000?g. Anti-?6IgG (25 µg/ml) and
anti-rat IgG agarose beads (1/10 v/v) were added to the
Actions of laminin on ovine granulosa cells
·
F LE BELLEGO and others47
www.endocrinology.org
Journal of Endocrinology (2002) 172, 45–59

Page 4

supernatant and samples were incubated for 5 h at 4 ?C
under constant agitation. Agarose beads were pelleted,
washed 5 times in RIPA buffer, and re-suspended in 20 µl
Laemmli buffer without ?-mercaptoethanol.
Western immunoblotting
Whole ovaries and GC were crushed at 4 ?C in 25 mM
Tris–HCl (pH 7·5) containing 250 mM sucrose and pro-
tease inhibitors (1 mM PMSF, 2 nM E64 and 10 µM
aprotinin). Cell membranes were precipitated by centri-
fugation at 30 000?g, re-suspended in the same buffer
without protease inhibitors and stored at ?20 ?C. Protein
concentration was determined with the BioRad Protein
Assay.
Ovary homogenates and GC membranes were respect-
ively submitted to electrophoresis on 8% and 10% (w:v)
SDS-polyacrylamide gel under reducing conditions. The
proteins were then electrotransferred onto nitrocellulose
filters overnight at 4 ?C. Filters were incubated for 2 h at
room temperature with 20 mM Tris-buffered saline (TBS)
(pH 7·6), containing 10% nonfat dry milk powder
(NFDMP) and 0⋅2% Tween-20 to saturate non-specific
sites. Thereafter, filters were incubated for 1 h at 37 ?C
with anti-?6IgG (final dilution 1 µg/ml), anti-?1IgG
(final dilution 1·5 µg/ml) or anti-LN (final dilution
1:1000) antibodies in TBS/5% NFDMP/0·05% Tween-
20. After washing in TBS/1% NFDMP/0·2% Tween-20,
nitrocellulose filters were incubated for 1 h at 37 ?C with
a horseradish peroxidase-conjugated anti-rat, anti-mouse
or anti-rabbit IgG (final dilution 1:10 000) in TBS/5%
NFDMP/0·05% Tween-20. After washing in TBS/1%
NFDMP/0·2% Tween-20, the signal was detected by
ECL.
Granulosa cell culture
GC culture was performed according to the method of
Campbell et al. (1996). Cells were cultured in 96-well
tissue-culture plates coated with LN, heat-denatured LN,
RGD peptides, or non-coated plates (plastic). The differ-
ent substrata were prepared 72 h before use and allowed to
dry at room temperature. LN was used at 5 µg/cm2in
distilled water. Heat-denatured LN was used at the same
concentration after a 1-h heating treatment at 80 ?C. This
treatment induced LN fragmentation, as assessed by
western immunoblotting (data not shown). RGD peptides
were used at 0·4 µg/cm2in PBS. GC suspensions from
small and large follicles were seeded at 105viable cells/well
and cultured at 37 ?C in a humidified atmosphere with 5%
CO2, in serum-free culture medium (see Isolation of GC).
The effects of anti-?6IgG or control rat IgG were tested at
different doses in the range 0·01 to 5 µg/ml, in combi-
nation with oFSH (0·5 or 5 ng/ml), IGF-I (100 ng/ml)
and oLH (5 ng/ml). Each condition was tested in triplicate
in each culture. Culture media were partially replaced
(175 µl over 250 µl) every 48 h and the spent medium was
stored at ?20 ?C until assay. At 48 h or 144 h of culture,
cells were detached with trypsin and counted.
The time required for GC isolation was such that, to
maintain GC viability, each culture was performed with
pooled GC recovered from only 3 ewes. As this did not
provide a sufficient number of cells to allow comparison of
all experimental conditions within a single culture, distinct
culture experiments were designed to investigate the
effect of the substrata (LN, heat-denatured LN, RGD
peptides, plastic) and the effect of the treatments (FSH,
IGF-I, LH). However, the effects of anti-?6IgG were
always compared with the control rat IgG within the same
culture experiment.
Determination of thymidine labeling index, pyknotic index
and cell nuclear diameter
GC proliferation and survival were assessed by measuring
the thymidine labeling index (percentage of labeled cells)
and the pyknotic index (percentage of pyknotic cells)
respectively, after 48 h of culture. In parallel, the spreading
of GC on LN was estimated by measuring the diameter of
the cellular nuclei after 72 h of culture.
To determine the thymidine labeling index, cells
were washed with B2 medium without thymine, then
incubated with [3H]thymidine (0·25 µCi/ml) at 37 ?C
for 2 h. After 2 washes with B2 medium (with
thymine), cells were detached with 1% trypsin, pelleted
and fixed in 3% glutaraldehyde for 1·5 h at room
temperature. Cells were then smeared onto histological
slides by cytocentrifugation. Smears were stained with
Feulgen, dipped in NTB2 emulsion, air-dried, and
exposed for autoradiography for 6 days at 4 ?C. The
thymidine labeling index was estimated by counting the
number of labeled and unlabeled cells in 20 different
microscopic fields (objective ?100).
To determine the pyknotic index, cells were detached
by trypsin, fixed in glutaraldehyde, submitted to cyto-
centrifugation and smears were stained with Feulgen as
described above. The pyknotic index was estimated by
counting the number of pyknotic cells, i.e. cells with
condensed or fragmented nuclear chromatin (Aharoni et al.
1995), and non-pyknotic cells in 20 different microscopic
fields (objective ?100). Previous results established that
the presence of pyknotic cells was associated with DNA
fragmentation characteristic of the apoptotic process in
cultured GC (Huet et al. 2001).
For both thymidine labeling index and pyknotic index,
calculations were made on 500–1000 cells per slide.
To evaluate cell spreading, GC were grown on plastic
culture chamber slides coated with LN in the presence or
absence of anti-?6 IgG. Cell nuclear diameter was
measured on 40 Feulgen-stained cells per well with a
reticule under a microscope (objective ?100).
F LE BELLEGO and others·
Actions of laminin on ovine granulosa cells
48
www.endocrinology.org
Journal of Endocrinology (2002) 172, 45–59

Page 5

Estradiol-17? and progesterone radioimmunoassays
The concentrations of E2 and P4 in the culture medium
of GC from large follicles were measured after 144 h
of culture by radioimmunoassay protocols previously
described (Saumande 1981, Saumande et al. 1985,
Saumande 1991) and adapted to directly measure steroids
in cell culture media. The limit of detection of E2 was
1·5 pg/tube (7·5 pg/well) and the intra- and interassay
coefficients of variation were less than 7% and 9% respect-
ively. The limit of detection of P4 was 12 pg/tube
(60 pg/well) and the intra- and interassay coefficients of
variation were less than 10% and 11% respectively. Results
are expressed as the amount of steroids secreted between
96 h and 144 h of culture per 50 000 cells recovered at the
end of the culture period.
Statistical analysis
All experimental data are presented as the mean?...
In immunohistochemistry experiments, statistical com-
parisons of means were performed using ANOVA
followed by Tukey-Kramer tests. In GC culture exper-
iments, the effects of increasing doses of anti-?6IgG were
analyzed using one-way ANOVA for repeated measures
followed by Tukey-Kramer tests. Data were fitted to
sigmoidal curves with GraphPrad PRISM software (San
Diego, CA, USA). For a given substratum and combi-
nation of treatments, the effects of anti-?6IgG on cell
numbers, thymidine labeling index and pyknotic index
were analyzed by paired t-test. The effects of anti-?6IgG
on cell spreading were analyzed by t-test. The effects of
anti-?6IgG on GC steroidogenesis were analyzed using
two-way ANOVA in order to appreciate the effect of
anti-?6IgG as well as the culture effect resulting from
variations between both animals and quality of the ovarian
follicles dissected for each culture.
Results
LN and its receptor, ?6?1integrin, are expressed in GC of
antral ovarian follicles in vivo
LN, ?6and ?1integrin subunits were detected by western
immunoblotting in the whole ovary and GC (Fig. 1). LN
was detected as a heterotrimer consisting of a heavy chain
(band at 400 kDa) and 2 light chains (2 bands at approxi-
mately 200 kDa) (Fig. 1A, lanes 1 and 2). The ?6integrin
subunit was detected as a single band at 120 kDa (Fig. 1A,
lanes 3 and 4) and the ?1integrin subunit as two bands
corresponding to the mature (125 kDa) and the precursor
(110 kDa) forms as previously described (Rubinstein et al.
1997) (Fig. 1A, lanes 5 and 6). In immunoprecipitation
experiments performed on GC, the ?6integrin sub-
unit co-immunoprecipitated with the mature ?1integrin
subunit but not with the precursor form (Fig. 1B).
By immunohistochemistry, LN was detected in the
follicular basement membrane, theca and GC layers, and
interstitial tissue (Fig. 2A,B,C) as previously described
(Huet et al. 1997). In all follicles, the basement membrane
was strongly stained. In contrast, the intensity of staining
was very low in follicular antrum (Fig. 3B) and similar to
values obtained in ovarian tissue of control sections incu-
bated with non-immune serum (data not shown). In GC
layers, the intensity of LN staining did not vary with the
size of healthy antral follicles (Fig. 3A), but increased with
atresia (Fig. 3B) (P<0·01, late atretic vs healthy small
follicles) as previously described (Huet et al. 1997). In
addition, regardless of the follicular size, the intensity of
LN staining in GC of healthy follicles increased
(P<0·0001) during the follicular phase and after hCG
administration in the preovulatory phase of the cycle
(Fig. 3A).
The ?6integrin subunit was clearly detected in GC and
endothelial cells of blood vessels present in theca and
interstitium (Fig. 2D,E,F). The intensity of the ?6integrin
subunit staining was high in GC, compared with staining
in non-vascularized areas of interstitial tissue (Fig. 3D). In
contrast, the intensity of staining was very low in follicular
antrum (Fig. 3D) and similar to values obtained on ovarian
tissue in control sections incubated with non-immune IgG
(data not shown). In GC layers, the intensity of staining
was lower in large, compared with small healthy follicles
Figure 1 (A) Western immunoblotting analysis of LN, ?6and ?1
integrin subunits in GC (lanes 2, 4, 6) and in whole ovary
homogenates (lanes 1, 3, 5). Lanes 1 and 2: detection of the
3 chains of LN; lanes 3 and 4: detection of ?6integrin subunit;
lanes 5 and 6: detection of ?1integrin subunit (upper band) and
its precursor (lower band). (B) Association between ?6and ?1
integrin subunits. GC lysates were submitted to immunoprecipitation
(IP) with anti-?6IgG, then western immunoblotting analysis was
performed with anti-?1IgG, as described in Materials and Methods.
Lane 1: without immunoprecipitation, the two bands corresponding
to the mature (upper band) and the precursor (lower band) forms of
?1integrin subunit were detected. Lane 2: immunoprecipitation with
anti-?6IgG, only the mature form of ?1integrin subunit was detected.
Lane 3: control immunoprecipitation with non-specific rat IgG, ?1
integrin subunit was undetectable.
Actions of laminin on ovine granulosa cells
·
F LE BELLEGO and others49
www.endocrinology.org
Journal of Endocrinology (2002) 172, 45–59