0013-7227/01/$03.00/0
Printed in U.S.A.
Endocrinology 142(10):4203– 4211
Copyright © 2001 by The Endocrine Society
Basic Fibroblast Growth Factor Maintains Calcium
Homeostasis and Granulosa Cell Viability by Stimulating
Calcium Efflux via a PKC␦-Dependent Pathway
J. J. PELUSO, A. PAPPALARDO,
AND
G. FERNANDEZ
Departments of Physiology (J.J.P., A.P., G.F.) and Obstetrics and Gynecology (J.J.P.), University of
Connecticut Health Center, Farmington, Connecticut 06030
Previous studies have demonstrated that basic fibroblast
growth factor prevents granulosa cell apoptosis. The following
six observations provide insight into the mechanism by which
basic fibroblast growth factor mediates its antiapoptotic action.
First, loading granulosa cells with 1,2 bis(2-aminophenoxy)ethane-N,N,Nⴕ,Nⴕ-tetraacetic acid, an intracellular calcium chelator, prevented apoptosis when granulosa cells were deprived
of basic fibroblast growth factor. Second, treatment with thapsigargin, an agent known to increase intracellular free calcium,
induced granulosa cell apoptosis even in the presence of basic
fibroblast growth factor. Third, an activator of PKC mimicked,
whereas PKC inhibitors blocked, basic fibroblast growth factor’s antiapoptotic action. Fourth, continuous basic fibroblast
growth factor exposure maintained relatively constant levels of
intracellular free calcium, and a PKC inhibitor induced a sustained 2- to 3-fold increase in intracellular free calcium. Fifth,
granulosa cells, as well as spontaneously immortalized granulosa cells, were shown to express PKC␦, -␭, and -␨. Finally, the
PKC␦-specific inhibitor, rottlerin, blocked basic fibroblast
growth factor’s antiapoptotic action in granulosa cells and spontaneously immortalized granulosa cells. These studies suggest
that basic fibroblast growth factor regulates intracellular free
calcium through a PKC␦-dependent mechanism and that a sustained increase in intracellular free calcium is sufficient to induce and is required for granulosa cell apoptosis.
A
LTHOUGH BASIC FIBROBLAST growth factor (bFGF)
and its receptors are expressed by granulosa cells
(GCs) of healthy follicles at all stages of development (1), and
bFGF prevents GC apoptosis (2– 4), the mechanism through
which bFGF mediates GC survival is unknown. We do know
that ligand activation of the FGF receptor induces receptor
dimerization, tyrosine kinase activity, and autophosphorylation (5). Further, studies have shown that genistein, a tyrosine kinase inhibitor, blocks bFGF’s ability to prevent GC
apoptosis (3). Although this is consistent with bFGF’s known
mechanism of action, studies have not been conducted to
elucidate the antiapoptotic signaling events downstream of
tyrosine phosphorylation of the FGF receptor.
It is likely that the downstream signaling events involve
both acute and genomic actions (2, 6 – 8). Although the
Abbreviations: BAPTA, 1,2 bis(2-aminophenoxy)ethane-N,N,N⬘,N⬘tetraacetic acid; bFGF, Basic fibroblast growth factor; [Ca2⫹]i, intracellular free calcium; DAG, diacylglycerol; GC, granulosa cell; NCX,
sodium/calcium exchanger; NKA, sodium/potassium-adenosine
triphosphatase; PMCA, plasma membrane calcium-adenosine triphosphatase; SIGC, spontaneously immortalized granulosa cells; TPA,
12-O-tetradecanoylphorbol-13-acetate.
Additional studies demonstrated that in spontaneously immortalized granulosa cells, basic fibroblast growth factor increased PKC␦ activity by 60% within 2.5 min compared with
serum-free control levels. Rottlerin attenuated basic fibroblast growth factor’s ability to stimulate PKC␦ activity and to
maintain intracellular free calcium. Further, intracellular
free calcium levels in spontaneously immortalized granulosa
cells transfected with a PKC␦ antibody in the presence of
basic fibroblast growth factor were 2-fold higher than those
spontaneously immortalized granulosa cells transfected with
IgG. Similarly, transfecting spontaneously immortalized
granulosa cells with a specific PKC␦-substrate increased intracellular free calcium compared with spontaneously immortalized granulosa cells transfected with a specific substrate for PKC⑀. Moreover, basic fibroblast growth factor
increased and rottlerin attenuated 45Ca efflux by 50% compared with that in basic fibroblast growth factor-treated
cells. Finally, an inhibitor of the plasma membrane calciumadenosine triphosphatase pump suppressed 45Ca efflux, elevated intracellular free calcium, and induced apoptosis. Collectively, these studies demonstrate that basic fibroblast
growth factor activates PKC␦, which, in turn, stimulates calcium efflux, accounting in part for basic fibroblast growth factor’s ability to maintain calcium homeostasis and, ultimately,
granulosa cell viability. (Endocrinology 142: 4203– 4211, 2001)
genomic actions are important, the acute actions are critical,
as bFGF deprivation for just 30 min commits spontaneously
immortalized granulosa cells (SIGCs) to undergo apoptosis
(8). These acute actions of bFGF appear to be related to
changes in intracellular free calcium ([Ca2⫹]i) (8). This is
based on the observations that an increase in [Ca2⫹]i occurs
before SIGCs undergo apoptosis (8), and bFGF prevents this
increase in [Ca2⫹]i (8). In SIGCs the antiapoptotic effect of
bFGF is mimicked by 12-O-tetradecanoylphorbol-13-acetate
(TPA) and attenuated by general PKC inhibitors (8). Further,
PKC inhibitors abrogate the ability of bFGF to maintain normal basal [Ca2⫹]i (8). As PKC␦ is the only PKC isotype that
is expressed by SIGCs (8) that can be activated by TPA (9, 10),
PKC␦ has been implicated as the mediator of bFGF’s action.
However, the studies that implicate PKC␦ as a mediator of
bFGF’s antiapoptotic action have all been conducted on
SIGCs and not primary GCs. Therefore, the present studies
were undertaken to determine whether PKC␦ mediates
bFGF’s antiapoptotic action in primary GCs. This is likely,
because in GCs bFGF increases the level of diacylglycerol
(DAG), an endogenous PKC activator, by 2- to 3-fold within
2 min of exposure (11).
4203
4204
Endocrinology, October 2001, 142(10):4203– 4211
Peluso et al. • bFGF Regulates Calcium Efflux through a PKC␦-Dependent Pathway
Subsequent studies were conducted to elucidate the mechanism through which bFGF-activated PKC␦ regulates
[Ca2⫹]i. These studies will demonstrate that bFGF-activated
PKC␦ ultimately stimulates calcium efflux. In this manner,
bFGF maintains [Ca2⫹]i within a physiological range and
thereby promotes GC viability.
centration of 10 ␮m (8). The cells were incubated for 10 min at 37 C and
then observed at a magnification of ⫻200 under fluorescent optics using
the fluorescein isothiocyanate filter set. The number of fluorescent cells
(i.e. apoptotic cells) in a field was counted. The total number of cells in
that field was also counted under phase optics. The process was continued until at least 100 cells/well were counted. The percentage of
apoptotic cells was then calculated.
Materials and Methods
Reagents and materials
[Ca2⫹]i measurements
DMEM/F-12 was used for the SIGC cultures, and RPMI 1640 was
used for the GC cultures (Sigma, St. Louis, MO). These media were
supplemented with the following reagents depending on the experimental design. bFGF (R&D Systems, Inc., Minneapolis, MN) and TPA
(Sigma) were used at final concentrations of 0.6 and 10 nm, respectively.
The PKC inhibitors, cherlerythrine chloride, rottlerin, bisindoylmaleimide II, and its inactive analog, bisindoylmaleimide V, were purchased
from Calbiochem (La Jolla, CA). Cherlerythrine chloride was added at
a final concentration of 1 ␮m, and bisindoylamleimide II and V were
added at a final concentration of 0.05 ␮m. Rottlerin was used at a final
concentration of 5 ␮m. Thapsigargin (25 ␮m) was purchased from Sigma.
Lanthanum chloride (10 ␮m) and eosin (50 ␮m) were also purchased
from Sigma In experiments involving 1,2 bis(2-aminophenoxy)ethaneN,N,N⬘,N⬘-tetraacetic acid (BAPTA), GCs were cultured in serumsupplemented medium for 75 min and then exposed to BAPTA-AM
(5 ␮m; Molecular Probes, Inc., Eugene, OR) in serumsupplemented
medium at 37 C for 45 min. The serum was then removed, and the cells
were treated according to the experimental design.
Granulosa cell isolation and culture
Immature female Wistar rats (22 d of age) were obtained from Charles
River Laboratories, Inc. (Wilmington, MA), and housed under controlled conditions of temperature, humidity, and photoperiod (12 h of
light, 12 h of darkness; lights on at 0700 h). On the day of the experiment,
immature animals were 23 or 24 d of age. The rats were cervically
dislocated between 0930 and 1000 h, the ovaries were removed, and GCs
were isolated. This protocol was approved by the animal care committee
of the University of Connecticut Health Center.
GCs were isolated according to the procedure of Lederer et al. (12).
Once collected, small and large GC populations were separated by
Percoll gradient centrifugation as previously described (12). The large
GC population was composed of approximately 80% large GCs and 20%
small GCs. Only the large GC population was used in these experiments,
because they undergo apoptosis when deprived of growth factors (13).
Unless otherwise stated, this cell population was washed, resuspended
in RPMI 1640, and used in various protocols as outlined in subsequent
sections. For the apoptosis studies the freshly isolated GCs were plated
with the various treatments in 0.5 ml serum-free medium in Lab-Tek
slides at a density of 1.25 ⫻ 105 cells/ml, and cultures were maintained
in 5% CO2 at 37 C for 5 h.
SIGC culture
SIGCs were cultured in DMEM/F-12 supplemented with 5% FBS as
previously described (8). The SIGCs were routinely maintained in Falcon
T-flasks (Becton Dickinson and Co., Lincoln Park, NJ). SIGCs were
plated in 100-mm glass (Kimax, Fisher Scientific, Pittsburgh, PA) culture
dishes at a density of 4 ⫻ 105 cells/ml in 12 ml medium. SIGCs were
plated in 0.5 ml medium at 1.25 ⫻ 105 cells/ml in 8-chamber glass
Lab-Tek slides (Nunc Inc., Naperville, IL). Unless otherwise stated, the
cells were initially cultured in DMEM/F-12 supplemented with 5% FBS
for 24 h. The serum-supplemented medium was removed, and the cells
were cultured in serum-free DMEM/F-12 with various reagents for up
to 5 additional h.
Identification of apoptotic cells
Apoptosis was assessed by in situ staining using the nuclear dye,
YOPRO-1 (Molecular Probes, Inc.) (8, 14). To stain apoptotic cells,
YOPRO-1 was added directly into each culture chamber at a final con-
Before loading, GCs were plated on cover glass in 35-mm dishes for
2 h in serum-supplemented medium. SIGCs were plated on cover glass
in 35-mm dishes in serum-supplemented for 24 h, then for 24 h with
serum-free bFGF-supplemented medium. After the culture period, both
GCs and SIGCs were then loaded at room temperature with fluo-4/AM,
a calcium dye indicator, in the presence of bFGF (8). After loading, the
cover glass was washed and placed in a coverslip clamp culture chamber
(ALA Scientific Instruments, Inc., Westbury, NY). The cells were incubated at room temperature in 0.5 ml Krebs-HEPES buffer supplemented
with bFGF. To allow the cells to establish a baseline level, the images
from the first 3.5 min were discarded. Fluorescent images were collected
at 30-sec intervals from cells as indicated for each experimental design.
Unless otherwise stated, the intensity of the fluo-4 fluorescence was
expressed as a fold change compared with the 3.5-min value.
Western blot analysis of PKC isotypes
PKC expression in primary GCs was assessed immediately after
isolation. SIGCs were harvested from serum-supplemented medium.
After the cells were collected, they were processed for Western blot
analysis as previously described (8). Briefly, 10 ␮g lysate was loaded
onto each lane, and the sample was electrophoresed on a 10% polyacrylamide gel at 100 V, then transferred to nitrocellulose. The nitrocellulose blot was incubated for 2 h with agitation at room temperature
with monoclonal antibodies built against various PKC isotypes (Transduction Laboratories, Inc., Lexington, KY) at the dilution recommended
by the manufacturer. The blot was then process as previously described
(8). The specific protein was detected by chemiluminescence using the
LumiGLO detection system (Kirkegaard & Perry Laboratories, Gaithersburg, MD). Specific staining was assessed by omitting the primary
antibody from the Western blot protocol.
bFGF and PKC␦ activity
PKC␦ activity was assessed using a phosphospecific PKC antibody
(Cell Signaling, Beverly, MA). When PKC␦ is activated, it undergoes
autophosphorylation on Ser662. Therefore, the amount of phosphorylated PKC␦ is a direct assessment of PKC␦ activity (15). In this study
SIGCs were cultured for 24 h in serum, then for 24 h in serum-free
medium with bFGF, then either bFGF was removed or rottlerin was
added to the cultures. After 2.5 min at room temperature, the cells were
harvested. A crude membrane preparation (16) was made and used for
Western blot analysis. The blots were first probed with a phosphospecific PKC antibody. This antibody detects PKC␣, -␤, -␥, and -␦ that are
phosphorylated at Ser662. As SIGCs express PKC␦ and not PKC␣ or
PKC␤, the amount of phospho-PKC is a direct assessment of PKC␦
activity in SIGCs. The blots were then stripped and reprobed with a
PKC␦-specific antibody.
A quantitative estimate of PKC␦ activity was obtained by scanning
the films and determining the density of the phospho-PKC bands using
IPGel software (Scanalytics, Inc., Vienna, VA). To ensure valid quantitation of the phospho-PKC bands, several exposures of each Western
blot were prepared. The exposures selected for analysis possessed gray
scale values from 101–194 on a linear gray scale of 0 –255. All treatments
from each experiment were run on the same gel and therefore subjected
to the same Western blot procedure (i.e. incubation times, etc.). For
analysis, the background from each film was subtracted from each band
to yield a specific density. The specific density of each band was then
divided by the specific density of the control treatment, resulting in a
fold increase from the control. This was done to correct for the relatively
small variation between experiments.
Peluso et al. • bFGF Regulates Calcium Efflux through a PKC␦-Dependent Pathway
Effect of PKC␦ antibody or PKC␦-specific substrate peptide
on [Ca2⫹]i
To establish a cause and effect relationship between PKC␦ and the
maintenance of [Ca2⫹]i, either a PKC␦ antibody or peptide substrate
was delivered into the cells using protein transfection. To demonstrate the feasibility and effectiveness of this approach, SIGCs were
exposed to tetramethylrhodamine B isothiocyanate-labeled IgG in the
presence or absence of the protein transfection agent, Chariot, according to manufacturer’s instructions (Active Motif, Carlsbad, CA).
After transfection, the cells were loaded with the calcium indicator
dye, fluo-4. The percentage of cells that incorporated tetramethylrhodamine B isothiocyanate-IgG and their ability to load with Fluo-4
were then assessed.
Once the manufacturer’s protocol was shown to be effective for
SIGCs, it was used in the following study. First, either the PKC␦ antibody
or IgG was mixed with transfection agent at a ratio of 1 part antibody
to 30 parts transfecting reagent. The SIGCs were then incubated with the
antibody mixture for a 30 min at 22 C and then for 3 h at 37 C. The
antibody mixture was then removed, and the cells were loaded with
fluo-4 as previously described (8). Similar studies were conducted with
the substrate peptides for either PKC␦ or PKC⑀ (Calbiochem, San Diego,
CA). Each peptide was used at a concentration of 500 ng/transfection
reaction.
After fluo-4 loading, cells were observed for a 3.5-min base line
period. At 3.5 min an image of the fluo-4 fluorescent intensity was
captured to determine the fluo-4 fluorescence (F). Then calcium ionophore, A23187 (50 ␮m; Sigma), was added to determine maximum
fluorescent intensity (Fmax). EGTA (7 mm; Sigma) was then added, and
the minimum fluorescent intensity was determined (Fmin). [Ca2⫹]i was
estimated by the following equation: intracellular free calcium (nm) ⫽
345 nm (F ⫺ Fmin)/(Fmax ⫺ F) (17). This approach to estimate [Ca2⫹]i
levels was used to allow comparisons between treatment groups.
Calcium efflux measurements
SIGCs were loaded in the presence of bFGF and 45Ca (10 ␮Ci/ml) for
18 h and washed with 45Ca-free medium as described by Husain and
associates (18). The bFGF-treated cultures were placed at room temperature, and a 100-␮l aliquot was taken. Then either vehicle or test agent
was added. Samples were taken at 15- or 30-sec intervals over the next
3 min. The samples were added to liquid scintillation cocktail and
counted for 1 min in an LKB 6500 liquid scintillation counter (Rockville,
MD). The values were corrected for the removal of medium, and the
cumulative amount of 45Ca extruded was determined. Cumulative 45Ca
efflux was expressed as the fold increase from the value obtained 15 sec
after the addition of vehicle or test agent.
Statistical analysis
All experiments were repeated several times, as indicated in the
figures. The experiments in which apoptosis was assessed by YOPRO-1
staining were performed in quadruplicate, with each experiment replicated two or three times. These data were pooled and analyzed by
one-way ANOVA, followed by Student-Newman-Keuls test when ap-
FIG. 1. The effect of [Ca2⫹]i on primary
GC apoptosis. A, [Ca2⫹]i were maintained by treating GCs with BAPTA,
whereas [Ca2⫹]i levels were elevated by
thapsigargrin (TG) in the presence of
bFGF. Apoptosis was assessed after 5 h
of treatment using the nuclear dye
YOPRO-1, as outlined in Materials and
Methods. These experiments were conducted in quadruplicate and conducted
on 2 different d. Data were pooled
and analyzed by ANOVA, followed by
Student-Newman-Keuls post-hoc test.
*, Significantly different from control
(P ⬍ 0.05).
Endocrinology, October 2001, 142(10):4203– 4211 4205
propriate. Comparisons between two groups were made by t test. Regardless of the test, P ⬍ 0.05 was considered significant.
Results
In primary GCs a prolonged increase in [Ca2⫹]i occurs
before apoptosis (13). To determine whether this increase in
[Ca2⫹]i induces apoptosis, GCs were loaded with an intracellular calcium chelator, BAPTA-AM. Under these conditions serum depletion resulted in an increase in apoptosis,
but this increase was attenuated by pretreatment with
BAPTA (Fig. 1A). In addition, treatment with thapsigargin,
an agent known to increase [Ca2⫹]i, induced GC apoptosis
even in the presence of bFGF (Fig. 1B).
As shown in Fig. 2A, both bFGF and TPA reduced the
percentage of apoptotic GCs to approximately 35%. Given
that the percentage of apoptotic before culture is about 30 –
35%, this indicates that both bFGF and TPA are extremely
effective in suppressing GC apoptosis. Further, these observations confirm previous work with SIGCs (8) that suggest
a major role for PKC in regulating GC viability. This concept
was further supported by the observation that in GCs the
ability of bFGF to prevent apoptosis was attenuated by the
general PKC inhibitor, bisindolylmaleimide II, but not by its
inactive analog, bisindolylamleimide V (Fig. 2B). In addition,
bFGF’s ability to maintain [Ca2⫹]i levels was blocked by
chelerythrine chloride, another general PKC inhibitor (Fig. 3).
Western blot analysis revealed that PKC␦, -␭, and -␨
were expressed by GCs (Fig. 4). As TPA mimics bFGF’s
action, and PKC␦ is responsive to TPA (9, 10), it is likely
that PKC␦ mediates bFGF’s action. To test this hypothesis,
the effect of rottlerin, a PKC␦-specific inhibitor, on GC and
SIGC apoptosis was determined. In this study rottlerin
inhibited the ability of bFGF to preserve GC and SIGC
viability (Fig. 5, A and B).
PKC␦ activity was monitored using a phospho-PKC antibody. As shown in Fig. 6, PKC␦ activity was 1.4- to 1.7-fold
higher in bFGF-treated cells than in cells cultured for 2.5 min
in serum-free medium. Further, rottlerin, a PKC␦-specific
inhibitor, blocked the ability of bFGF to increase PKC␦ activity. Although phospho-PKC␦ levels were altered, the
overall amount of PKC␦ was similar in all treatment groups
(Fig. 6).
The role of PKC␦ in mediating bFGF’s ability to maintain
normal basal [Ca2⫹]i was also examined by exposing
SIGCs to rottlerin in the presence of bFGF. Under these
4206
Endocrinology, October 2001, 142(10):4203– 4211
Peluso et al. • bFGF Regulates Calcium Efflux through a PKC␦-Dependent Pathway
FIG. 2. The effect of a PKC activator
and inhibitor on GC apoptosis. A, The
effect of the PKC activator, TPA. B, The
effect of the general PKC inhibitor,
bisindoylmaleimide II (Bis II), and its
inactive analog, bisindoylmaleimide V
(Bis V), on GC apoptosis. These experiments were conducted in quadruplicate and conducted on 2 different d.
Data were pooled and analyzed by
ANOVA, followed by Student-NewmanKeuls post-hoc test. *, Significantly different from control (P ⬍ 0.05).
FIG. 3. The effect of the general PKC inhibitor, chelerythrine chloride, on [Ca2⫹]i levels in GCs. In these studies bFGF was continuously present, and chelerythrine chloride was added after the
3.5-min equilibration period. Values were expressed as a fold increase compared with those at the 3.5-min point. The data in this
graph represent the mean of 10 cells treated with either DMSO
(vehicle control) or chelerythrine chloride. The SE associated with
each time point was less than 0.3 and is not shown. The bFGF and
bFGF plus chelerythrine chloride treatments were replicated six
and four times, respectively.
conditions rottlerin induced an increase in [Ca2⫹]i (Fig. 7).
The second approach used to determine the relationship
between bFGF, PKC␦ and [Ca2⫹]i involved transfecting
SIGCs with either IgG or a PKC␦ antibody. To demonstrate
the feasibility of this approach, SIGCs were exposed to
TRITC-labeled IgG in the presence or absence of the protein transfection agent, Chariot. After transfection, the
cells were loaded with the calcium indicator dye, fluo-4.
TRITC-IgG was not incorporated into the SIGCs unless the
transfection agent was present (data not shown). In the
presence of the transfection agent the cells appeared
healthy, as judged by phase optics (Fig. 8A, upper panel),
and approximately 90% of these cells incorporated TRITCIgG (Fig. 8A, middle panel). These cells were also able to be
loaded with fluo-4 (Fig. 8A, lower panel).
Having established appropriate transfection conditions, a
study was conducted in which either PKC␦ antibody or
PKC␦-specific substrate was transfected into SIGCs. Under
these conditions, the PKC␦ antibody increased [Ca2⫹]i nearly
2-fold compared with IgG. A smaller, but significant, increase in [Ca2⫹]i levels was observed after transfecting SIGCs
with a specific PKC␦ substrate peptide compared with a
specific PKC⑀ substrate peptide (Fig. 8B).
FIG. 4. The expression of PKC isotypes in both primary GCs and
SIGCs. Lysates were prepared from primary GCs immediately after
isolation. Lysates from SIGCs, which were used as a positive control,
were made from SIGCs that were cultured in serum-supplemented
medium.
There are several possible mechanisms through which
PKC␦ could regulate normal basal [Ca2⫹]i. However, it is
likely that the molecular targets of PKC␦ are within the
plasma membrane, because activated PKC␦ is often localized
to the plasma membrane (9, 10). Given this, one possibility
is that PKC␦ influences the activity of proteins that regulate
calcium efflux. As shown in Fig. 9, bFGF stimulated 45Ca
efflux compared with that in serum-free controls. Both rottlerin and lanthanum, an inhibitor of the plasma membrane
calcium-adenosine triphosphatase (PMCA) (19), suppressed
bFGF-induced 45Ca efflux to the level in serum-free controls
(Fig. 9, A and B).
Lanthanum also caused a gradual increase in [Ca2⫹]i in the
presence of bFGF (Fig. 10A). However, [Ca2⫹]i levels only
increased by 25–50% after a 10-min exposure, compared with
a nearly 2-fold increase associated with rottlerin treatment
(compare Figs. 7 and 10A). Finally, both lanthanum chloride
and eosin, another PMCA inhibitor (20), attenuated bFGF’s
ability to maintain viable SIGCs (Fig. 10B).
Discussion
Although previous studies have shown that bFGF inhibits
GCs and SIGCs from undergoing apoptosis (2– 4, 8, 21), the
signaling events downstream of ligand activation of the FGF
receptor have not been clearly identified. Our initial studies,
Peluso et al. • bFGF Regulates Calcium Efflux through a PKC␦-Dependent Pathway
Endocrinology, October 2001, 142(10):4203– 4211 4207
FIG. 5. The effect of the PKC␦-specific
inhibitor, rottlerin, on bFGF-regulated
GC (A) and SIGC (B) apoptosis. These
experiments were conducted in quadruplicate and conducted on 2 different
days. Data were pooled and analyzed by
ANOVA, followed by Student-NewmanKeuls post-hoc test. *, Significantly different from control (P ⬍ 0.05); **, significantly different from both control
and bFGF groups (P ⬍ 0.05).
which were conducted exclusively on SIGCs, suggest that
PKC␦ mediates bFGF’s antiapoptotic action by maintaining
[Ca2⫹]i within a physiological range (8). However, for these
findings to be physiologically relevant, a similar PKC␦dependent mechanism would also have to be involved in the
signal transduction cascade through which bFGF maintains
GCs. Specifically, it is essential to determine whether GCs
undergo calcium-dependent apoptosis and whether bFGF
regulates [Ca2⫹]i through a PKC␦-dependent mechanism.
Previous studies have shown that an increase in [Ca2⫹]i
occurs before GCs undergo apoptosis (13). That the dysregulation of [Ca2⫹]i is an essential component of the apoptotic
cascade is illustrated by the fact that BAPTA, an intracellular
calcium chelator, prevents GCs from undergoing apoptosis
in the absence of bFGF. In addition, thapsigargin, which
increases [Ca2⫹]i by releasing calcium from its intracellular
stores, induces apoptosis even in the presence of bFGF. These
data demonstrate that GCs undergo calcium-induced
apoptosis.
The present studies also show that bFGF maintains relatively constant levels of [Ca2⫹]i, and the PKC inhibitor, chelerythrine chloride, attenuates bFGF’s ability to regulate
[Ca2⫹]i in GCs. Further, PKC␦ is implicated in regulating
[Ca2⫹]i and GC viability by two sets of observations. First, the
PKC␦-specific inhibitor, rottlerin, abrogates bFGF’s antiapoptotic action. Second, an activator of PKC, the phorbol
ester TPA, prevents GC apoptosis. Like SIGCs, GCs express
PKC␦, -␭, and -␨. Of these three PKC isotypes, only PKC␦
possesses a phorbol ester-binding site and is capable of being
activated by TPA (9, 10). Moreover, the fact that GCs express
PKC␦ confirms the in situ hybridization studies of Hunnzicker-Dunn and associates (22). These in situ hybridization
studies reveal that PKC␦ is expressed predominately by GCs
of healthy follicles (22). The observation that PKC␦ is selectively expressed by GCs of healthy follicles is consistent with
our hypothesis that PKC␦ plays an essential role in regulating the viability of GCs. As these findings in GCs are identical
to the observations made in SIGCs, they indicate that SIGCs
accurately mimic the physiological responses of GCs to
bFGF. This validates the use of SIGCs for the more mechanistic biochemical studies that require large numbers of cells.
In our first biochemical study with SIGCs, PKC␦ activity
was monitored using a phosphospecific PKC antibody.
When PKC␦ is activated, it undergoes autophosphorylation
on Ser662 (15). Therefore, the amount of phosphorylated
PKC␦ is a direct assessment of PKC␦ activity (15). This study
supports our hypothesis by revealing that bFGF stimulated
and rottlerin inhibited the activation of PKC␦. Collectively,
these data suggest that bFGF regulates GC and SIGCs viability by maintaining [Ca2⫹]i through a PKC␦-dependent
mechanism.
To definitively establish a cause and effect relationship
between the activation of PKC␦ and bFGF’s ability to maintain normal basal [Ca2⫹]i levels, three separate approaches
were used. First, the PKC␦-specific inhibitor, rottlerin, was
shown to suppress PKC␦ activity and promote an increase in
[Ca2⫹]i even in the presence of bFGF. The second approach
involved transfecting SIGCs with either IgG or a PKC␦ antibody. This study demonstrates that even in the presence of
bFGF the PKC␦ antibody increased [Ca2⫹]i levels nearly
2-fold compared with IgG. Finally, a specific PKC␦ substrate
was transfected into SIGCs. This substrate competes with
endogenous PKC␦ substrates, thereby blocking the physiological responses to PKC␦ activation. By using this approach,
a smaller, but significant, increase in [Ca2⫹]i levels was observed after transfecting SIGCs with a specific PKC␦ substrate peptide compared with a specific PKC⑀ substrate peptide. Taken together, these studies clearly demonstrate that
a cause and effect relationship exists between the bFGF activation of PKC␦ and the maintenance of normal basal
[Ca2⫹]i.
How, then, might PKC␦ regulate normal basal [Ca2⫹]i
levels? There are at least five possible sites of actions (23, 24).
First, PKC␦ could stimulate the sacroplasmic/endoplasmic
reticulum calcium-adenosine triphosphatase pump, which
would lower [Ca2⫹]i by pumping intracellular calcium into
its intracellular stores (23). Second, PKC␦ could inhibit IP3
receptors, thereby stopping IP3-induced calcium release
from intracellular stores. Type 1 and 3 IP3 receptors are
expressed by both GCs and SIGCs, but these receptors were
not localized to the plasma membrane (Peluso, J. J., unpublished observations). Because of their cellular localization, it
4208
Endocrinology, October 2001, 142(10):4203– 4211
Peluso et al. • bFGF Regulates Calcium Efflux through a PKC␦-Dependent Pathway
FIG. 7. The effect of the PKC␦-specific inhibitor, rottlerin, on [Ca2⫹]i
levels in SIGCs. In these studies bFGF was continuously present, and
rottlerin was added after the 3.5-min equilibration period. Values
were expressed as the fold increase compared with those at the 3.5min point. The data in this graph represent the mean of 18 and 16 cells
treated with either dimethylsulfoxide (vehicle) or rottlerin, respectively. The SE associated with each time point was less than 0.07 and
is not shown. This entire experiment was replicated three times.
FIG. 6. The effects of bFGF and the PKC␦-specific inhibitor, rottlerin,
on PKC␦ levels and activity in SIGCs. A, The total amount of PKC␦
and the amount of activated (phosphorylated) PKC␦ are shown. B,
The relative PKC␦ activity was assessed by desensitizing the band
associated with phospho-PKC. Relative PKC␦ activity was expressed
as a percentage of the control. These experiments were conducted in
duplicate on 3 different days. Values were pooled and expressed as the
mean ⫾ SE. *, Significantly different from control and bFGF plus
rottlerin treatment groups (P ⬍ 0.05).
is unlikely that DAG-activated PKC␦, which is localized to
the plasma membrane, could regulate either the sacroplasmic/endoplasmic reticulum calcium-adenosine triphosphatase pump or the IP3 receptors.
The third possible site is action could involve voltageoperated calcium channels, which are present in the plasma
membrane of GCs (25). These channels can be inhibited by
TPA (i.e. an activator of PKC) (24). Similarly, channels that
regulate capacitative calcium entry represent a fourth possibility (24). These channels are activated by the depletion of
calcium stores. These channels are independent of voltage,
stimulated by IP3, highly selective for calcium, and inhibited
by PKC (24). It is possible that DAG-activated PKC␦ could
inhibit either of these types of channels.
The fifth site of PKC␦’s action could involve the regulation
of proteins that promote calcium efflux. In vascular smooth
muscle cells, normal basal [Ca2⫹]i levels are maintained in
part by removing excess calcium from the cytoplasm (23).
The removal of excess cytoplasmic calcium is mediated by
two mechanisms. The first is the PMCA pump (26). The
PMCA pump has a high affinity for calcium. However, in
many cells it has a low capacity because it is expressed at a
low level. The second mechanism involves two proteins: the
sodium/potassium-adenosine triphosphatase (NKA) pump
(27) and the sodium/calcium exchanger (NCX) (28, 29). The
NKA moves sodium out of the cell. This provides the gradient that activates the NCX, which pumps calcium out and
sodium into the cell. The NCX has a high capacity to produce
calcium efflux. Both the PMCA pump and the NCX/NKA
are localized to the plasma membrane, and their activities are
influenced by PKC (26 –29).
Although PKC␦ could regulate that function of voltageoperated calcium channels and channels that regulate capacitative calcium entry, the present data suggest the calcium
efflux is regulated by bFGF in a PKC␦-dependent manner.
This is based on the observations that bFGF stimulates calcium efflux compared with that in control cells, and rottlerin
inhibits bFGF-induced calcium efflux. Further, lanthanum,
an inhibitor of PMCA, not only suppresses bFGF-induced
calcium efflux, but also results in a gradual increase in [Ca2⫹]i
and, ultimately, apoptosis. These findings demonstrate that
inhibiting calcium efflux results in a sustained increase in
[Ca2⫹]i that is sufficient to induce apoptosis after 5 h. It is
important to note that the lanthanum-induced increase in
[Ca2⫹]i is considerably less than the 2- to 3-fold increase in
[Ca2⫹]i observed after PKC inhibition with rottlerin. This
could indicate that another component, possibly NCX and
NKA, may be regulated by bFGF-activated PKC␦.
The present data indicate that bFGF-activated PKC␦ stimulates calcium efflux. There is one study in swine granulosa
Peluso et al. • bFGF Regulates Calcium Efflux through a PKC␦-Dependent Pathway
Endocrinology, October 2001, 142(10):4203– 4211 4209
FIG. 8. Effect of transfecting specific antibodies and peptides on [Ca2⫹]i levels in SIGCs. A, SIGCs were transfected with TRITC-IgG using the
Chariot transfection protocol. After transfection, the SIGCs appeared healthy, as judged by phase microscopy (upper panel), with TRITC-IgG
detected within the cytoplasm of approximately 90% of these cells (middle panel). These cells were subsequently loaded with fluo-4, demonstrating that [Ca2⫹]i measurements can be made on these cells. B, The effect of anti-PKC antibody and a specific PKC␦ substrate of [Ca2⫹]i.
The values represent the means of between 20 –70 cells for each treatment group. The observations were taken from experiments that were
conducted on 3 different days. Data were pooled and analyzed by ANOVA, followed by Student-Newman-Keuls post-hoc test. *, Significantly
different from control (P ⬍ 0.05); **, significantly different from both control and either the IgG- or PKC⑀-specific substrate (P ⬍ 0.05).
FIG. 9. The effect of bFGF on 45Ca efflux in SIGCs. A, The rate of 45Ca efflux
is shown for representative cultures
treated with vehicle, bFGF, bFGF plus
rottlerin (RL), or bFGF plus lanthanum
chloride (La). Values shown in B are the
mean ⫾ SE of the fold increase in
cumulative 45Ca efflux after 3 min of
culture (n ⫽ 8 –12/treatment group).
Experiments were conducted on 5 different days. Data were pooled and
analyzed by ANOVA, followed by
Student-Newman-Keuls post-hoc test.
*, Significantly different from all other
groups (P ⬍ 0.05).
cells that describes the kinetics of calcium efflux. In this study
TPA was shown to inhibit calcium efflux (30). However, this
study monitored calcium efflux over the course of 4 h. Under
these experimental conditions, two different pools of intracellular calcium contribute to the overall amount of calcium
that is extruded. The first compartment involves the rapid
release of calcium and is not inhibited by TPA. The second
compartment influences the later or slow phase of calcium
efflux and is attenuated by TPA. It is difficult to interpret this
finding because exposure to TPA for just a few hours can
down-regulate PKC levels (8). If TPA depleted endogenous
PKC levels in swine GCs, then the previously published
report on calcium efflux would indicate that PKC is involved
in stimulating calcium efflux, which would be consistent
with the data of the present studies.
Based on these studies, we propose that ligand activation
of the FGF receptor increases intracellular levels of DAG (11).
DAG remains associated with the plasma membrane, where
it activates PKC␦. Once activated, PKC␦ functions to serine/
threonine phosphorylate various molecular targets that may
include calcium efflux regulators, such as PMCA. As has
been shown in other cell types, TPA-dependent phosphorylation of PMCA enhances the rate of calcium efflux, thereby
maintaining calcium homeostasis (26). Future studies will be
directed to test this hypothesis.
Finally, the present in vitro studies imply that by regulating
GC viability, bFGF promotes follicular development and inhibits follicular atresia in vivo. The findings that GCs of developing follicles synthesize bFGF and express high affinity
FGF receptors (1) are consistent with this hypothesis. Fur-
4210
Endocrinology, October 2001, 142(10):4203– 4211
Peluso et al. • bFGF Regulates Calcium Efflux through a PKC␦-Dependent Pathway
FIG. 10. The effect of inhibitors of PMCA on [Ca2⫹]i and apoptosis of SIGCs. A, The effect of lanthanum on bFGF- regulated [Ca2⫹]i levels. Values
represent a mean of 16 cells for bFGF and 16 cells for bFGF plus lanthanum. The SEs were less than 0.05 and are not shown. These experiments
were repeated four times for bFGF and seven times for bFGF plus lanthanum with similar results. B, The effects of the PMCA inhibitors,
lanthanum (Lan) and eosin, on SIGCs apoptosis are shown. Apoptosis was assessed by YOPRO-1 staining for the lanthanum-treated cells and
by 4,6-diamidino-2-phenylindole (DAPI) staining for the eosin-treated cells. This was necessary because eosin fluoresces at the same wavelength
as YOPRO-1. Data were pooled and analyzed by ANOVA, followed by Student-Newman-Keuls post-hoc test. *, Significantly different from the
control (P ⬍ 0.05).
ther, PKC␦ is selectively expressed by GCs of healthy follicles
of all sizes (22). Our in vitro studies establish a causal relationship among bFGF, PKC␦, and GC viability by demonstrating that bFGF’s antiapoptotic action is mediated by
PKC␦. These findings therefore provide insights into the
putative mechanism by which bFGF prevents GC apoptosis
and follicular atresia in vivo.
Acknowledgments
The authors thank Dr. Robert Burghardt of Texas A&M University for
providing the SIGC cells, and Dr. Bruce White of University of Connecticut Health Center for his helpful discussions.
Received May 8, 2001. Accepted July 6, 2001.
Address all correspondence and requests for reprints to: John J. Peluso, Ph.D., Department of Physiology, University of Connecticut Health
Center, Farmington, Connecticut 06030. E-mail: [email protected].
This work was supported by NIH Grant HD-33467.
8.
9.
10.
11.
12.
13.
14.
References
1. Ojeda SR, Dissen GA 1994 Developmental regulation of the ovary via growth
factor tyrosine kinase receptors. Trends Endocrinol Metab 5:317–323
2. Aharoni D, Dantes A, Oren M, Amsterdam A 1995 cAMP-mediated signals
as determinants for apoptosis in primary granulosa cells. Exp Cell Res 218:
271–282
3. Tilly JL, Billig H, Kowalski KI, Hsueh AJ 1992 Epidermal growth factor and
basic fibroblast growth factor suppress the spontaneous onset of apoptosis in
cultured rat ovarian granulosa cells and follicles by a tyrosine kinase-dependent mechanism. Mol Endocrinol 6:1942–1950
4. Trolice MP, Pappalardo A, Peluso JJ 1997 Basic fibroblast growth factor and
N-cadherin maintain rat granulosa cell and ovarian surface epithelial cell
viability by stimulating the tyrosine phosphorylation of the fibroblast growth
factor receptors. Endocrinology 138:107–113
5. Szebenyi G, Fallon JF 1999 Fibroblast growth factors as multifunctional signaling factors. Int Rev Cytol 185:45–106
6. Amsterdam A, Gold RS, Hosokawa K, et al. 1999 Crosstalk among multiple
signaling pathways controlling ovarian cell death. Trends Endocrinol Metab
10:255–262
7. Hosokawa K, Aharoni D, Dantes A, et al. 1998 Modulation of Mdm2 expres-
15.
16.
17.
18.
19.
20.
21.
sion and p53-induced apoptosis in immortalized human ovarian granulosa
cells. Endocrinology 139:4688 – 4700
Lynch K, Fernandez G, Pappalardo A, Peluso JJ 2000 Basic fibroblast growth
factor inhibits apoptosis of spontaneously immortalized granulosa cells by
regulating intracellular free calcium levels through a protein kinase Cdeltadependent pathway. Endocrinology 141:4209 – 4217
Kampfer S, Uberall F, Giselbrecht S, Hellbert K, Baier G, Grunicke HH 1998
Characterization of PKC isozyme specific functions in cellular signaling. Adv
Enzyme Regul 38:35– 48
Kanashiro CA, Khalil RA 1998 Signal transduction by protein kinase C in
mammalian cells. Clin Exp Pharmacol Physiol 25:974 –985
Oury F, Faucher C, Rives I, Bensaid M, Bouche G, Darbon JM 1992 Regulation
of cyclic adenosine 3⬘,5⬘-monophosphate-dependent protein kinase activity
and regulatory subunit RII␤ content by basic fibroblast growth factor (bFGF)
during granulosa cell differentiation: possible implication of protein kinase C
in bFGF action. Biol Reprod 47:202–212
Lederer KJ, Luciano AM, Pappalardo A, Peluso JJ 1995 Proliferative and
steroidogenic capabilities of rat granulosa cells of different sizes. J Reprod
Fertil 103:47–54
Luciano AM, Pappalardo A, Ray C, Peluso JJ 1994 Epidermal growth factor
inhibits large granulosa cell apoptosis by stimulating progesterone synthesis
and regulating the distribution of intracellular free calcium. Biol Reprod 51:
646 – 654
Idziorek T, Estaquier J, De Bels F, Ameisen JC 1995 YOPRO-1 permits
cytofluorometric analysis of programmed cell death (apoptosis) without interfering with cell viability. J Immunol Methods 185:249 –258
Peters CA, Maizels ET, Hunzicker-Dunn M 1999 Activation of PKC␦ in the
rat corpus luteum during pregnancy: potential role of prolactin signaling. J Biol
Chem 274:37499 –37505
Chakravarthy BR, Bussey A, Whitfield JF, Sikorska M, Williams RE, Durkin
JP 1991 The direct measurement of protein kinase C (PKC) activity in isolated
membranes using a selective peptide substrate. Anal Biochem 196:144 –150
Simpson AWM 1999 Fluorescent measurements of calcium: basic practical
considerations. Methods Mol Biol 114:3–30
Husain M, Jiang L, See V, et al. 1997 Regulation of vascular smooth muscle
cell proliferation by plasma membrane Ca(2⫹)-ATPase. Am J Physiol 272:
C1947–1959
Herscher CJ, Rega AF 1997 On the mechanism of inhibition of the PMCa(2⫹)ATPase by lanthanum. Ann NY Acad Sci 834:407– 409
Gatto C, Hale CC, Xu W, Milanick MA 1995 Eosin, a potent inhibitor of the
plasma membrane Ca pump, does not inhibit the cardiac Na-Ca exchanger.
Biochemistry 34:965–972
Amsterdam A, Keren-Tal I, Aharoni D 1996 Cross-talk between cAMP and
p53-generated signals in induction of differentiation and apoptosis in steroidogenic granulosa cells. Steroids 61:252–256
Peluso et al. • bFGF Regulates Calcium Efflux through a PKC␦-Dependent Pathway
22. Cutler Jr RE, Maizels ET, Hunzicker-Dunn M 1994 Delta protein kinase-C in the
rat ovary: estrogen regulation and localization. Endocrinology 135:1669 –1678
23. Marin J, Encabo A, Briones A, Garcia-Cohen EC, Alonso MJ 1999 Mechanisms involved in the cellular calcium homeostasis in vascular smooth muscle:
calcium pumps. Life Sci 64:279 –303
24. Putney Jr JW, Ribeiro CM 2000 Signaling pathways between the plasma
membrane and endoplasmic reticulum calcium stores. Cell Mol Life Sci 57:
1272–1286
25. Lieberherr M, Grosse B, Machelon V 1999 Phospholipase C-␤ and ovarian sex
steroids in pig granulosa cells. J Cell Biochem 74:50 – 60
26. Garcia ML, Strehler EE 1999 Plasma membrane calcium ATPases as critical
27.
28.
29.
30.
Endocrinology, October 2001, 142(10):4203– 4211 4211
regulators of calcium homeostasis during neuronal cell function. Front Biosci
4:D869 –D882
Therien AG, Blostein R 2000 Mechanisms of sodium pump regulation. Am J
Physiol Cell Physiol 279:C541–C566
Blaustein MP, Lederer WJ 1999 Sodium/calcium exchange: its physiological
implications. Physiol Rev 79:763– 854
Philipson KD, Nicoll DA 2000 Sodium-calcium exchanger: a molecular prespective. Annu Rev Physiol 62:111–133
Veldhuis JD, Demers LM 1986 An inhibitory role for the protein kinase C
pathway in ovarian steroidogenesis. Studies with cultured swine granulosa
cells. Biochem J 239:505–511