| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLES |
Vincent Center for Reproductive Biology, Department of Obstetrics and Gynecology, (Y.M., X.-J.T., S.M., J.L.T.), and Pediatric Surgical Research Laboratories, Department of Pediatric Surgery (T.F.M., P.K.D.), Massachusetts General Hospital/Harvard Medical School, Boston, Massachusetts 02114
Address all correspondence and requests for reprints to: Jonathan L. Tilly Ph.D., Massachusetts General Hospital, VBK137E-GYN, 55 Fruit Street, Boston, Massachusetts 02114. E-mail: tilly.jonathan{at}mgh.harvard.edu
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Using monolayer cultures of murine PGC maintained in vitro
without and with somatic feeder cells, several molecules capable of
regulating PGC survival have been identified over the past several
years (reviewed in Refs. 1, 7, 8). Based on these investigations and
others demonstrating expression of several specific cytokines and their
cognate receptors in PGC as well as in fetal gonads, it has been
proposed that apoptosis in germ cells during fetal development is
initiated, at least in part, due to insufficient levels of survival
factors, such as stem cell factor (SCF) or leukemia inhibitory factor
(LIF), derived from fetal ovarian somatic cells (reviewed in Refs. 1, 8). For example, early genetic mutation analyses revealed gonadal
dysgenesis and sterility in male and female mice lacking functional
expression of either SCF or the SCF receptor, c-kit (9). These studies
support more recent evidence from studies of cultured murine PGC
indicating that SCF can independently promote survival of these cells
in vitro (reviewed in Refs. 1, 8). By comparison, LIF
gene knock-out mice remain fertile, albeit defects related to embryo
implantation failure were noted in female mice (10). Although LIF has
also been reported to promote PGC survival in vitro
(reviewed in Refs. 1, 8), it may be that loss of the antiapoptotic
actions of LIF in the germline can be compensated for by another
cytokine in LIF-null mice. Indeed, work with SCF and LIF in the context
of germ cell survival has been paralleled by a number of other studies
implicating such factors as interleukin-4 (11), basic fibroblast growth
factor (12), and tumor necrosis factor-
(13) as being antiapoptotic
and/or mitogenic in PGC maintained in vitro. Thus, fetal
germ cell apoptosis is probably initiated, in vivo, due to
competition among germ cells for limited supplies of these and other as
yet unidentified survival factors in the developing gonad.
A major complicating issue when interpreting the available literature in the field of germ cell death during fetal development, however, is the relative paucity of information regarding the regulation of germ cell apoptosis once PGC have differentiated into female and male gonocytes. This can, in all likelihood, be attributed to the lack of appropriate and validated model systems to investigate these types of questions in fetal ovaries and testes, starkly contrasting the well characterized and widely used system of monolayer cultures of isolated PGC. In one study of testicular development, work of van Dissel-Emilani et al. reported on the use of gonocytes and Sertoli cells, isolated from fetal male rats at day 20 postcoitum (in addition to newborn and postpartum day 3 pups), in a coculture system to elucidate the effects of fibroblast growth factor-2 (FGF-2) on cellular survival (14). Although FGF-2 was identified as a potent Sertoli cell survival factor, as well as a survival and mitogenic factor for postnatally derived gonocytes, unfortunately no indication was made as to the actions of this growth factor on postcoitum day 20 gonocytes. In a subsequent study of fetal rat testes in organ culture, Olaso et al. (15) presented a comprehensive analysis of the pro-apoptotic effects of transforming growth factor-ß (TGF-ß) on male gonocytes in testes collected on days 13.5 and 17.5 postcoitum and maintained in vitro for up to 48 h. These data are in agreement with a previous report that TGF-ß inhibits proliferation of murine PGC collected on e8.5 and cultured on fibroblast feeder layers (16).
Very little is also currently known of the intracellular effectors used by the ligand-activated cytokine receptors that provide antiapoptotic signals in germ cells. In many somatic cell lineages, growth factor-promoted activation of phosphatidylinositol-3'-kinase (PI3K) appears to be critical for cellular survival because cotreatment of cells with either of two PI3K inhibitors, wortmannin (17) or LY294002 (18), ameliorates the growth factor response leading to rapid apoptosis (for examples, see Refs. 19, 20, 21). One likely downstream candidate for PI3K-mediated phosphorylation is the serine/threonine kinase, c-Akt (also referred to as protein kinase B or rac-PK), an enzyme thought to couple cytokine signaling to an evolutionarily-conserved central checkpoint in the cell death pathway controlled by the Bcl-2 family of proteins (reviewed in Refs. 22, 23). A second possible mediator of cytokine-initiated cellular responses is p70 S6 kinase (p70S6K), a rapamycin-sensitive signaling enzyme that may also be a downstream target for activated PI3K (24, 25, 26). Since PI3K and p70S6K appear essential for many SCF-induced responses in various somatic cell lineages (for examples, see Refs. 24, 27, 28, 29), it is possible that these intracellular kinases serve a similar role in cytokine-supported germ cell survival.
In the present report, we developed and validated an in vitro organ culture model to determine if the reported antiapoptotic actions of SCF and LIF, alone or in combination, in PGC could be extended to oogonia and oocytes in the fetal mouse ovary, and to elucidate the role played by PI3K and/or p70S6K in cytokine-promoted female germ cell survival. Furthermore, based on the well-characterized function of insulin-like growth factor-I (IGF-I) as a survival factor for a variety of cell types (30; reviewed in Ref. 31), including somatic granulosa cells of ovarian follicles (reviewed in Ref. 32), as well on the reported expression of the IGF-I receptor in rodent oocytes (33), experiments were also designed to evaluate the role of IGF-I in modulating oogonium and oocyte apoptosis in the fetal mouse ovary. Lastly, we determined if the pro-apoptotic function of TGF-ß recently ascribed to fetal male mouse gonocytes (15) was conserved in female germ cells as well.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Isolation of genital ridges
Dissection of genital ridges was performed under sterile
conditions on e13.5, essentially as described for studies of rodent
Müllerian duct regression and fetal lung development (34, 35).
Briefly, the gravid uterus was removed from the anesthetized mother and
placed in BGJb medium (Gibco BRL Life Technologies, Grand Island, NY) supplemented with 100 U/ml
penicillin, 100 µg/ml streptomycin and 1.3 µg/ml amphotericin-B
(referred to hereafter as culture medium). Fetuses were removed from
the uterus suspended by the umbilical cord, fixed supine on a sterile
translucent Tygon square, and hemisected at the level of the diaphragm.
The bowel and liver were retracted upward to expose the
retroperitoneum, and the remainder of the dissection was completed
under 36x magnification. The mouse fetus at e13.5, which measures
about 12–15 mm in crown-rump length, has clearly visible genital
ridges with Müllerian and Wolffian ducts lateral to each gonad.
The testis can be identified by a characteristic circuitous capsular
vessel that becomes the testicular artery, and its transverse sex cords
that become the seminiferous tubules. The ovary can be distinguished
from the testis because the female gonad is longer and thinner and
lacks characteristic vessels or cords. From female fetuses, the gonad,
mesonephros, and indifferent ducts from each side were dissected
en bloc, floated off in a meniscus of culture medium created
between the curved blades of a pair of fine jeweler’s forceps, and
transferred to the organ culture dish (see below).
Organ cultures
Each genital ridge, which measures on average 0.2 x 1.2
mm, was placed on a sterile preformed 2% agarose square (approximately
1 cm2) positioned on a triangular stainless steel grid
suspended within the inner well of a Falcon 3037 organ culture dish.
Approximately 700 µl of culture medium were then added to the inner
well to reach the base of the agarose square, thus allowing the tissue
to absorb medium without or with treatments through the agarose square
without the need for submersion culture (34, 35). Afterwards, 1.5 ml of
prewarmed (37 C) sterile water were added to the outer well. Once all
cultures were prepared, 2–3 genital ridges were immediately fixed (see
below), and the remaining genital ridges were cultured for 24, 48, or
72 h at 37 C in a humidified chamber gassed with 5%
CO2-95% air. Treatment groups consisted of culture medium
without (controls) or with 10% FBS (Gibco BRL), 100 ng/ml
murine recombinant SCF (R&D, Minneapolis, MN), 100 ng/ml murine
recombinant LIF (R&D), 50 ng/ml human recombinant IGF-I (Promega Corp., Madison, WI), IGF-I plus SCF, IGF-I plus LIF, or IGF-I
plus SCF plus LIF. In the second set of experiments, ovaries were
cultured for 72 h in the absence or presence of 100 ng/ml SCF plus
100 ng/ml LIF or with 50 ng/ml IGF-I without or with 25 ng/ml human
recombinant TGF-ß1 (R&D). To test the role of PI3K as a downstream
mediator of cytokine-stimulated oogonium and oocyte survival,
additional cultures were conducted in medium supplemented without or
with 100 ng/ml SCF plus 100 ng/ml LIF or with 50 ng/ml IGF-I in the
absence or presence of one of two inhibitors of PI3K prepared as
concentrated stock solutions in dimethylsulfoxide (DMSO), LY294002 (5
µM; Sigma Chemical Co.) or wortmannin (50
nM; Sigma Chemical Co.). To test the role of
p70S6K in cytokine-mediated germ cell survival, cultures
were also carried out in medium supplemented without and with 100 ng/ml
SCF plus 100 ng/ml LIF or with 50 ng/ml IGF-I in the absence or
presence of the p70S6K inhibitor, rapamycin (25 ng/ml;
Sigma Chemical Co.). The doses of inhibitors chosen for
the present studies were based on those used in previous studies to
block PI3K or p70S6K (19, 20, 21, 24, 25, 26), as well as on
results from preliminary dose-response studies in our laboratory with
fetal mouse ovaries in organ culture (data not shown). Furthermore, in
all experiments with inhibitors, controls consisted of culture medium
with vehicle (
0.1% DMSO final concentration). A 6–0 silk suture was
placed perpendicular to the proximal end of the ovaries to facilitate
orientation during histologic preparation. A small piece of abdominal
wall muscle was also placed along the medial aspect of the gonad to
stabilize the position of the gonad and the silk suture.
Histology
Freshly isolated (0 h, on grid, before culture) genital ridges
or cultured genital ridges at the conclusion of the experimental
manipulation were covered with 2% low-melting temperature agarose
maintained at 44 C, and the agarose was allowed to harden. The
agarose-coated tissue was then fixed in neutral-buffered 4%
formaldehyde with 5% Bouin’s fluid, dehydrated in ethanol, cleared in
xylene, embedded in paraffin, and serially sectioned at 6 µm
thickness. In some cases, every serial section of the ovary was aligned
in order on glass microscope slides for hematoxylin and eosin (H/E)
staining. These sections were used for general histologic analysis of
cellular morphology, as well as for determination of germ cell counts
(see below). In H/E-stained sections, cells possessing lightly stained
round nuclei along with a maintenance of cytoplasmic volume and easily
discernible spherical plasma membranes were considered nonapoptotic,
whereas cells showing nuclear condensation (basophilia), cytoplasmic
shrinkage and convoluted plasma membranes were considered apoptotic
(36).
Germ cell counts
As indicated above, nonapoptotic germ cells were identified by
the presence of large, spherical, lightly stained nuclei containing
fine chromatin granules, and by clearly visible and contiguous plasma
membranes. The total number of nonapoptotic germ cells in sections
taken at sites approximately one-third, one-half, and two-thirds
through the fetal ovary, along the long axis, were counted. Each ovary
was given a numerical code so that all germ cell counts were conducted
without knowledge of treatment group. After all counts were completed,
the mean number of germ cells per section was determined for each ovary
by taking the mean of the values from the three sections, each ovary
was decoded, and the values were then assigned to the corresponding
treatment group.
In situ 3'-end-labeling (ISEL) of DNA
The occurrence of apoptosis in germ cells was also assessed by
monitoring the presence of DNA fragmentation in situ, as
described previously (37). Briefly, paraffin sections prepared as
described above (see "Histology") were deparaffinized and
rehydrated, and then incubated in 2% H2O2
(vol:vol in 95% methanol) for 1 min at 20 C to quench endogenous
peroxidase activity. Samples were next deproteinated by incubation with
10 µg/ml proteinase-K for 30 min at 37 C. After washing in PBS, free
3'-ends of the DNA were labeled with 50 µM
biotin-16-deoxy-UTP (dUTP; Gibco BRL) using 125 U terminal
deoxynucleotidyl transferase (Boehringer Mannheim,
Indianapolis, IN) for 15 min at 37 C. Sections were then washed and
blocked for 30 min with 3% BSA (wt/vol), and subsequently incubated
with avidin-biotin horseradish peroxidase complex (ABC kit;
Vector Laboratories, Inc., Burlingame, CA) at 20 C for 15
min. Sites of biotinylated-dUTP incorporation were then detected by
incubating slides with 0.5 mg/ml 3,3'-diaminobenzidine and 0.03%
hydrogen peroxide for 5 min at 20 C, and colorimetric reactions
(generation of a brown reaction product) were terminated by placing the
slides in a buffer consisting of 10 mM Tris-HCl and 1
mM EDTA (pH 8.0). Negative controls, conducted by omitting
terminal deoxynucleotidyl transferase, yielded no reaction product
(data not shown). Slides were analyzed by conventional light microscopy
after light counterstaining with hematoxylin, and cells exhibiting
brown-staining were considered positive for apoptosis-associated DNA
fragmentation (37).
Analysis of germ cell proliferation
To ascertain if cytokines altered germ cell dynamics via effects
on mitosis, fetal ovaries were cultured without and with 100 ng/ml SCF
plus 100 ng/ml LIF or with 50 ng/ml IGF-I for 24 h, after which
5-bromo-2'-deoxyuridine (BrdU; Sigma Chemical Co.) was
added to each well at a final concentration of 30 µM. As
a positive control for proliferation, parallel cultures were conducted
with 0.1 µM all-trans retinoic acid (RA;
Sigma Chemical Co.) included in the culture medium because
RA is known to be a potent stimulator of germ cell proliferation in the
mouse (38). All cultures were continued for an additional 2 h at
37 C (pulse-labeling), after which tissues were fixed, embedded in
paraffin, sectioned, and analyzed by immunohistochemistry for the
occurrence of BrdU incorporation as a marker of new DNA synthesis
associated with cellular proliferation (39). To detect BrdU, sections
were rehydrated and subjected to high temperature antigen unmasking
(40), as detailed in our previous studies (41, 42), before
immunoanalysis using 6 µg/ml of a mouse monoclonal antibody against
BrdU (clone BMC9318; Boehringer Mannheim). Chromogenic detection of the
sites of BrdU-primary antibody complexes was performed by incubating
sections for 1 h with a 1:200 dilution of a biotinylated horse
antimouse IgG antibody (Vector Laboratories, Inc.),
followed by addition of avidin-biotin horseradish peroxidase complex
components (ABC kit; Vector Laboratories, Inc.) at 20 C
for 45 min. Sections were then washed and incubated with 0.5 mg/ml
3,3'-diaminobenzidine and 0.03% hydrogen peroxide for 1 min at 20 C,
and colorimetric reactions (generation of a brown reaction product)
were terminated by placing the slides in a buffer consisting of 10
mM Tris-HCl and 1 mM EDTA (pH 8.0). Negative
controls, conducted by omitting the primary antibody, yielded no
reaction product (data not shown). Slides were analyzed by conventional
light microscopy after light counterstaining with hematoxylin.
Data presentation and statistical analysis
In each experiment, two to three genital ridges were used for
each treatment group, and all experiments were independently replicated
at least three times. Therefore, all quantitative data represent the
mean ± SEM of combined results obtained from analysis
of a minimum of six ovaries in each treatment group. One-way ANOVA was
used to compare mean values of the various treatment groups, followed
by Scheffé’s F test to determine significant
differences at P < 0.05. Where appropriate,
photomicrographs are presented to depict representative tissue
histology (H/E-staining), extent of DNA fragmentation (ISEL), or BrdU
incorporation (immunohistochemistry) observed in ovaries from the
replicate experiments.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
) and ISEL of DNA (Fig. 1B). In
vitro culture of ovaries in the absence of serum or cytokines
resulted in a time-dependent increase in the number of germ cells
displaying chromatin condensation, cytoplasmic shrinkage and DNA
fragmentation, such that most remaining germ cells present at 72 h
were apoptotic (Fig. 1, C and D). Identical results were obtained when
we used BGJb medium containing 10% FBS (data not shown;
see Germ cell counts below). However, apoptosis in oogonia
and oocytes was markedly inhibited by treatment of fetal ovaries with
SCF (100 ng/ml) together with LIF (100 ng/ml), even after 72 h of
culture (Fig. 1, E and F) (data not shown for the 24 and 48 h time
points). Furthermore, the antiapoptotic actions of SCF/LIF were
mimicked by treatment of fetal ovaries with 50 ng/ml IGF-I (Fig. 1, G
and H). We also noted that the magnitude of apoptosis was more
reflected by analysis of cellular and nuclear morphology
(H/E-staining), as compared with analysis of DNA cleavage (ISEL),
consistent with recent reports that major DNA cleavage detected by
in situ end-labeling procedures is a late event in apoptosis
that occurs well after morphological signs of cellular demise (43).
|
). In fetal ovaries cultured without
trophic support, the mean number of healthy gonocytes significantly
decreased to 78 ± 4%, 38 ± 5% and 10 ± 2% of the
preculture mean number after 24, 48, and 72 h of culture,
respectively (Fig. 2). Surprisingly, provision of 10% FBS had no
effect on germ cell maintenance (Fig. 3).
|
|
, E and F, and Fig. 3). The number of germ cells in
SCF/LIF-treated ovaries was 5.5 ± 0.7-fold greater than that of
control ovaries after 72 h in culture, although this remained
approximately one-half of the number of germ cells present in freshly
isolated e13.5 ovaries (Fig. 3). Interestingly, relative to the
combined effects of SCF plus LIF, only a marginal antiapoptotic effect
was observed in germ cells of fetal ovaries when LIF was added
separately (Fig. 3), and SCF alone was completely ineffective in
preventing oogonium and oocyte loss in cultured fetal ovaries (Fig. 3).
Inclusion of IGF-I in the culture medium also significantly attenuated
germ cell apoptosis over the 72-h culture period (Fig. 1, G and H, and
Fig. 3). Morphometric analysis indicated that the number of healthy
oogonia and oocytes remaining in IGF-I-treated ovaries after 72 h
of culture was 5.3 ± 0.6-fold greater than that in untreated
control ovaries cultured in parallel (Fig. 3). However, like SCF plus
LIF, germ cell numbers in ovaries cultured in the presence of IGF-I
remained significantly less than the preculture mean number (Fig. 3),
suggesting that a combination of survival factors is necessary to fully
suppress oogonium and oocyte apoptosis in this model system. In support
of this, we noted that, although there were no improvements in germ
cell viability when ovaries were cotreated with IGF-I and either SCF or
LIF vs. the effects obtained with IGF-I alone (Fig. 3),
provision of all three factors (IGF-I plus SCF plus LIF) to cultured
ovaries maintained germ cell numbers to levels not significantly
different than the preculture mean number (Fig. 3).
Lastly, culture of fetal ovaries for 72 h with TGF-ß alone did
not alter germ cell numbers vs. untreated controls (Fig. 4). However, the antiapoptotic effects of
SCF/LIF or IGF-I in cultured fetal ovaries were significantly
suppressed, but not abolished, by cotreatment with TGF-ß (Fig. 4).
|
). The antiapoptotic actions of
IGF-I in fetal ovarian germ cells were similarly abrogated by
cotreatment with PI3K inhibitors (Fig. 5; data not shown for LY294002).
In the absence of cytokines or growth factors, germ cell counts in
ovaries treated with either of the two PI3K inhibitors were comparable
to those obtained in control cultures without inhibitor (Fig. 5),
confirming specificity of the inhibitor response to only those cell
populations receiving an external stimulus coupled to PI3K activation
and cellular survival (i.e. SCF/LIF or IGF-I). The
p70S6K inhibitor, rapamycin, did not alter basal rates of
germ cell apoptosis and, in contrast to the effects noted for LY294002
and wortmannin, rapamycin did not antagonize the ability of SCF/LIF or
IGF-I to prevent apoptosis (Fig. 5).
|
) or with SCF/LIF (Fig. 6B), we could
not detect evidence of BrdU incorporation in cells of any section
analyzed. Provision of IGF-I resulted in an extremely low but
detectable incidence of germ cell mitogenesis, with a single
BrdU-labeled cell observed periodically in some sections (Fig. 6C).
However, ovaries cultured in the presence of RA possessed many
BrdU-positive germ cells (Fig. 6D), confirming fidelity of the assay
for detection of proliferating germ cells in fetal ovaries exposed to a
known mitogenic stimulus.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Second, and in agreement with a large volume of data implicating IGF-I as survival factor for many different cell types (30, 31, 46, 47), we also showed that IGF-I alone is a potent inhibitor of germ cell apoptosis in cultured fetal ovaries. Interestingly, however, our findings contrast the reported inability of IGF-I to affect survival of murine PGC cultured in vitro (48). This observation provides a second example of clear differences in the endocrine control of apoptosis in PGC vs. oogonia and oocytes (see above). Consequently, the changes that occur in hormonal responsiveness of PGC during development into gender-specific gonocytes indicate that data obtained from analysis of PGC cultured in vitro are not directly applicable to differentiating gonocytes, and underscore the need for such in-depth analyses of the regulation of apoptosis specifically in oogonia and oocytes.
As proposed by Cooke et al. (11), it is important to point out that germ cell numbers during gonadal development are probably regulated more by "a complex combination of positive and negative regulatory factors" as opposed to the absence or presence of any one factor. In agreement with this hypothesis, our third observation is that TGF-ß, although without effect on basal rates of apoptosis, antagonized the survival actions of SCF/LIF or IGF-I in oogonia and oocytes of cultured fetal ovaries. These data support and extend recent work of Olaso et al. that identified a pro-apoptotic function for TGF-ß in male gonocytes during fetal testicular development in rats (15), and concur with a negative influence of TGF-ß on murine PGC numbers in vitro via antiproliferative actions (16). Moreover, expression of TGF-ß is detectable in fetal rat gonads throughout fetal life from e14.5, and is known to be produced by the presumptive granulosa cells at the time of primordial follicle formation (49). These findings, coupled with reports that murine oocytes also produce TGF-ß (50), suggest that TGF-ß plays an important paracrine or autocrine role in modulating germ cell apoptosis during fetal ovarian development.
Since the absolute number of germ cells present in the fetal ovary, like cell numbers in any tissue, is determined by the extent of both proliferation and apoptosis, we next tested whether or not SCF/LIF or IGF-I maintains germ cell numbers in cultured fetal ovaries by not only suppressing germ cell apoptosis but also by promoting oogonium mitosis. As anticipated, fetal ovaries cultured without trophic hormone support failed to show evidence of BrdU incorporation as a marker of new DNA synthesis associated with proliferation. A comparable lack of mitogenesis was observed in those cultures receiving SCF/LIF, indicating that the primary, if not exclusive action, of this cytokine combination in cultured fetal ovaries is to promote germ cell survival. By comparison, inclusion of IGF-I in the culture medium promoted a low level of germ cell mitogenesis, as evidenced by the presence of one or two BrdU-positive germ cells in some of the sections analyzed. However, many germ cells in fetal ovaries treated in parallel with a potent germ cell mitogen, retinoic acid (38), exhibited positive-staining for BrdU immunoreactivity. Thus, high levels of germ cell mitogenesis could be experimentally induced and detected using this model system, but the cytokines and growth factors tested were relatively ineffective in this regard. These findings agree with the hypothesis of Raff who suggested the central importance of active repression of apoptosis by external growth factors in most cells as the primary mechanism (as opposed to proliferation) to regulate cell number, and of the creation of a competition for survival factors in limited availability may be one means of selection of the "fittest" cells (51).
At present, essentially nothing is known regarding the pathways activated in germ cells following stimulation with either cytokines or growth factors. In somatic cells, binding of SCF with its receptor, c-kit, induces rapid receptor dimerization (27). Activation of the intrinsic tyrosine kinase domain of the SCF receptor then occurs, leading to receptor autophosphorylation as well as to phosphorylation of a number of intracellular substrates (24, 52, 53). Importantly, c-kit autophosphorylation recruits a class of cytoplasmic signal transduction proteins containing Src-homology-2 domains to the receptor, including PI3K which binds efficiently to the activated c-kit protein (54, 55). The exact role of PI3K in SCF receptor signaling remains to be elucidated. However, PI3K appears to be essential for the SCF-mediated mitogenic response in porcine aortic endothelial cells transfected with c-kit (27). Serve et al. also showed a dependency on PI3K activity for SCF receptor-mediated cell adhesion (28), and in human mast cells PI3K has been implicated in SCF-promoted histamine release (29). In rat mast cells, PI3K has been implicated in SCF-promoted proliferation, and this response appears to require activation of p70S6K as a downstream effector of PI3K action (24). The role of PI3K in SCF-induced cellular responses appears specific since activation of phosphatidylinositol-specific phospholipase C, a second classical signal transduction pathway, does not occur in rat mast cells (56) or NIH 3T3 cells (53) following SCF stimulation. In contrast to what is known of the involvement of PI3K in SCF-initiated signaling events, the role of PI3K in LIF signaling has not been studied in depth. Following receptor interaction in somatic cells, LIF is known to trigger tyrosine phosphorylation of various proteins (57), as well as to recruit mitogen-activated protein kinases and p70S6K for downstream signaling events (25, 58).
In the present study, inhibition of PI3K with LY294002 or wortmannin, two chemicals widely used for their ability to suppress specifically activity of this key signal transduction enzyme (17, 18, 19, 20, 21), almost completely reversed the survival effects of SCF plus LIF in germ cells of cultured fetal ovaries. Moreover, consistent with IGF-I-promoted activation of PI3K as a critical step for survival of many somatic cell lineages (46, 47), cotreatment of fetal ovaries with wortmannin (as well as with LY294002; data not shown) also effectively blocked the antiapoptotic actions of IGF-I in germ cells. By comparison, inhibition of p70S6K with rapamycin did not alter the efficacy of SCF/LIF or of IGF-I to suppress germ cell death in fetal ovaries. Although p70S6K is known to be a component of various cellular responses elicited by IGF-I and PI3K (59), our findings with female germ cells agree with previous work using PC12 cells (46) and Rat-1 fibroblasts (60) that activation of p70S6K appears to be either dispensable or not involved in events related to specifically the survival actions of SCF/LIF or IGF-I following ligand-receptor interaction.
These latter findings add to the increasing volume of literature that many growth factors appear to inhibit apoptosis by a PI3K-dependent cascade of events. For example, PI3K is required for the prevention of neuronal apoptosis by nerve growth factor (21), and those growth factors capable of activating a primary downstream effector of PI3K, the serine/threonine kinase c-Akt, are effective antiapoptotic molecules (19, 20, 61). The connection between PI3K and c-Akt is even more intriguing in light of recent evidence that couples c-Akt to a central checkpoint in the cell death pathway controlled by the Bcl-2 family of pro- and antiapoptotic proteins (62; reviewed in Refs. 22, 23). Furthermore, it was reported that the PI3K/c-Akt signaling pathway transduces a survival signal that ultimately prevents activation of caspases (19), a family of proapoptotic proteases responsible for dismantling the cell during death (63). Although a connection between SCF/LIF- or IGF-I-initiated PI3K signaling and altered bioactivity of Bcl-2 family members or caspases in germ cells has not yet been made, previous studies of the genetic pathways that underlie apoptosis in the ovary (reviewed in Ref. 32), including in-depth analyses of ovaries of female mice genetically deficient in expression of Bcl-2 (5), Bax (64), and caspase-2 (65), have confirmed that these proteins indeed serve as central checkpoints in germ cell death pathways. Therefore, future studies that employ these and other genetically manipulated mice for use with the fetal ovarian organ culture system described herein will allow us to continue to dissect the molecular basis of how key endocrine factors, such as SCF, LIF, IGF-I and TGF-ß, mediate germ cell survival and death during fetal gonadal development.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 On leave from the Department of Obstetrics and Gynecology, Faculty
of Medicine, University of Tokyo, Tokyo 113, Japan, and supported by
the Japanese Society for the Promotion of Science.
Received May 27, 1998.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
(TNF-) stimulates proliferation of mouse
primordial germ cells in culture. Dev Biol 161:91–95[CrossRef][Medline]
This article has been cited by other articles:
 |
|
 |
|
  F. H Thomas, R. S Ismail, J.-Y. Jiang, and B. C Vanderhyden Kit Ligand 2 Promotes Murine Oocyte Growth In Vitro Biol Reprod, January 1, 2008; 78(1): 167 - 175. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A M Lobascio, F G Klinger, M L Scaldaferri, D Farini, and M De Felici Analysis of programmed cell death in mouse fetal oocytes Reproduction, August 1, 2007; 134(2): 241 - 252. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Kawamura, N. Kawamura, J. Kumagai, J. Fukuda, and T. Tanaka Tumor Necrosis Factor Regulation of Apoptosis in Mouse Preimplantation Embryos and Its Antagonism by Transforming Growth Factor Alpha/Phosphatidylionsitol 3-Kinase Signaling System Biol Reprod, April 1, 2007; 76(4): 611 - 618. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. J. Hutt, E. A. McLaughlin, and M. K. Holland KIT/KIT Ligand in Mammalian Oogenesis and Folliculogenesis: Roles in Rabbit and Murine Ovarian Follicle Activation and Oocyte Growth Biol Reprod, September 1, 2006; 75(3): 421 - 433. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
I. B Carlsson, M. P E Laitinen, J. E Scott, H. Louhio, L. Velentzis, T. Tuuri, J. Aaltonen, O. Ritvos, R. M L Winston, and O. Hovatta Kit ligand and c-Kit are expressed during early human ovarian follicular development and their interaction is required for the survival of follicles in long-term culture. Reproduction, April 1, 2006; 131(4): 641 - 649. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K.J. Hutt, E.A. McLaughlin, and M.K. Holland Kit ligand and c-Kit have diverse roles during mammalian oogenesis and folliculogenesis Mol. Hum. Reprod., February 1, 2006; 12(2): 61 - 69. [Abstract] [Full Text] [PDF] |
