Modulation of Hen Granulosa Cell Steroidogenesis and Plasminogen Activator Activity by Transforming Growth Factor Alpha Growth Factors Downloaded from informahealthcare.com by Flinders University of South Australia on 01/18/15 For personal use only.
J L TILLY* and A L JOHNSON
Studies were conducted with chicken granulosa cells to evaluate the relative efficacy of human recombinant transforming growth factor alpha (TGFa)versus murine epidermal growth factor (EGF)to affect cyclic adenosine monophosphate (CAMP) accumulation and progesterone production stimulated by luteinizing hormone (LH) or steroid output induced by a CAMP analogue, and to modulate plasminogen activator (PA) activity. Increasing concentrations of EGF (33-328 nM) and TGFa (0.04-18 nM) were found t o inhibit cAMP formation stimulated by LH in a dose-dependent manner, with calculated halfmaximal inhibitory doses (ID?,,’s) of 97.1 and 0.27 nM, respectively. Similarly, a 470-fold difference in the ability of TGFa (ID5,,=0.13nM) versus EGF (IDJ,,=h1.3nM) t o halfmaximally suppress LH-induced progesterone production was observed in the same cells. Progesterone production stimulated by a cAMP analogue (8-bromo-cAMP, 1 mM) was also attenuated by EGF (ID?,,= 75.9 nM) and TGFa (ID?,,= 0.08 nM), suggesting a post-CAMP site of inhibition by these growth factors on steroidogenesis. Finally, a 260-fold to 330-fold difference in the efficacy of TGFa versus EGF to half-maximally stimulate cell-associated and secreted PA activity was observed. From these data, we propose that TGFa may serve an important role in regulating follicular growth and maturation in the domestic hen via its ability to affect granulosa cell steroidogenesis and plasminogen activator activity.
KEYWORDS:
TGFa,TGFD,CAMP, progesterone, plasminogen activator, ovary
enclosed mouse oocytes, suggesting a role for ECF in the final steps of oocyte maturation. In granulosa cells of the domestic hen, EGF has Previous studies have indicated that growth factors may play a significant role in the regulation of been shown to inhibit cyclic adenosine monoovarian cell differentiation and development in both phosphate (CAMP) and progesterone production mammalian and avian species. For instance, a con- stimulated by luteinizing hormone (LH), and prosiderable amount of evidence has been generated gesterone output induced by a cAMP analogue, concerning the ability of several growth factors to collectively indicating an inhibitory effect of EGF on influence gonadotropin-induced differentiation in steroidogenesis both prior and distal to cAMP forrat, bovine and porcine ovarian cells (Hsueh et al., mation (Pulley and Marrone, 1986). More recently, 1984; Adashi et al., 1985; Hill, 1989; Carson et al., Yoshimura and Tamura (1988) demonstrated that 1989). Additionally, Downs, Daniel and Eppig (1988) EGF stimulates the iii vitro proliferation of chicken demonstrated that epidermal growth factor (EGF) granulosa cells, whereas data from our laboratory induces germinal vesicle breakdown in cumulus cell- has indicated that EGF and several other growth factors increase plasminogen activator (PA) activity in dispersed granulosa cells (Tilly and Johnson, 1987; 1990). *Corresponding author
INTRODUCTION
247
Growth Factors Downloaded from informahealthcare.com by Flinders University of South Australia on 01/18/15 For personal use only.
248
TlLLY AND IOHNSON
The primary structure of the chicken EGF receptor and its binding properties have recently been deduced using a complementary deoxyribonucleic acid (DNA) clone containing the complete coding sequence for the receptor (Lax et a]., 1988). Results from these studies indicated that human recombinant transforming growth factor alpha (TGFa) bound to the chicken EGF receptor with an approximate 100-fold greater affinity than murine EGF. Furthermore, human TGFa was found to be 100-fold more effective in the induction of DNA synthesis as compared to murine EGF. In the light of these findings, we evaluated the relative efficacy of human TGFa versus murine EGF to alter steroidogenesis and plasminogen activator activity in dispersed granulosa cells collected from the largest preovulatory follicle in the ovary of the domestic hen. Additionally, we examined the ability of transforming growth factor beta (TGFB) to modulate the actions of TGFa on granulosa cell function.
MATERIALS AND METHODS
Animals Single comb white Leghorn hens, 24-40 weeks of age and laying regular sequences of at least five eggs, were used in all studies. Birds were housed individually in laying batteries, had free access to feed (Agway Layer Mash, Agway, Inc., Bordentown, NJ) and water, and were exposed to an artificial photoperiod of 15 h of light:9 h of dark with lights on at 2300 h. Individual laying cycles were monitored daily by the timing of oviposition.
Source and Preparation of Reagents Stock solutions of all treatments utilized were diluted with medium 199/HEPES (supplemented as described; Tilly and Johnson, 1987) to the desired concentration for each experiment. Ovine luteinizing hormone (NIAMDD-oLH-23; National Pituitary Agency) was diluted from a stock solution prepared in 0.01 M phosphate-buffered saline/2% bovine serum albumin (v:w). Phorbol 12-myristate 13acetate (Sigma Chemical Co., St Louis, MO) was diluted from a stock solution in 95% ethanol; 8-bromo-CAMP (Sigma) was dissolved directly into incubation medium prior to use. Epidermal growth factor (murine, receptor grade; Collaborative Research Inc., Bedford, MA) and transforming
growth factor alpha (human, recombinant; Bachem Bioscience Inc., Torrance, CA) were diluted from stock solutions in distilled, deionized water. Transforming growth factor-beta (p) (human; Collaborative Research Inc.) was prepared from a 5mM hydrogen chloride stock solution stored in a dichlorodimethylsilane-treated glass vial. All PA assay reagents (human glu-plasminogen, Desafib-fibrinogen and Spectrozyme-PL) were obtained from American Diagnostica, Inc (Greenwich, CT).
Collection and Preparation of Granulosa Cells On the morning of an experiment, hens were sacrificed by cervical dislocation 17-16 h prior to ovulation of the second or third egg of a sequence. The largest preovulatory follicle was removed and immediately placed in ice-cold 1%saline. Granulosa cells were collected and dispersed as previously described (Tilly and Johnson, 1987). Cell number and viability (routinely > 90Y0) were determined with a haemacytometer using the trypan blue exclusion technique. The cell suspension was then diluted with an appropriate volume of sterile M199/HEPES incubation medium to a final concentration of 5 ~ 1 0 ~ viable cells/ml medium.
Cell Incubations/Preparation Aliquots (10' cells) of the granulosa cell suspension were added to 12 x75 mm polypropylene (CAMPand steroid assays) or borosilicate (PA assays) tubes, combined with treatments and/or incubation medium to give a final volume of 0.5 ml, and incubated at 39°C for 4 h (CAMP and steroid determinations) or 18 h (PA assays). In all studies conducted with TGFp, incubation tubes and pipette tips were first treated with dichlorodimethylsilane (Eastman Kodak Co., Rochester, NY), washed thoroughly and rinsed well in order to prevent nonspecific adsorption of TGFB. For cAMP and steroid analyses (experiment l), cells were pelleted by centrifugation (200 xg, 20 min, 4°C) following incubation. A 0.2ml aliquot of the supernatant was then removed and analysed for progesterone content, whereas the cell pellet and remaining supernatant were immediately frozen at -70°C to be assayed for cAMP at a later date. For steroid measurements only (experiments 2 and 3), the cells and medium were frozen at -70°C immediately following incubation until assayed for progesterone levels. For PA assays (experiments 4 and 5), cells were' pelleted by centri-
TRANSFORMING GROWTH FACTORS IN THE H F h OVARY
fugation (200 x ~ 20 , min, 4°C) following incubation. The supernatant was carefully removed and analysed for secreted PA activity whereas the cell pellet was assayed for cell-associated enzyme activity.
Growth Factors Downloaded from informahealthcare.com by Flinders University of South Australia on 01/18/15 For personal use only.
Experiments In experiment 1, cells were incubated for 4 h with 10 ng/tube LH and 0.1 mM IBMX in the absence and presence of EGF (33-328 nM) or TGFa (0.04-18 nM), after which levels of cAMP and progesterone were determined. Experiment 2 was conducted to evaluate the effects of EGF and TGFa on steroidogenesis stimulated by the cAMP analogue, 8-bromo-CAMP. Progesterone production was determined following exposure of granulosa cells to 8-bromo-CAMP (1mM) without and with increasing concentrations of EGF (33-328nM) or TGFa (0.04-18nM). In experiments 1 and 2, cells were also treated with LH (long) and IBMX (0.1mM) or 8-bromo-CAMP (1mM) in the absence and presence of the tumourpromoting phorbol ester, phorbol 12-myristate 13acetate (PMA: a known activator of protein kinase C; Castagna et al., 1982). Based on our previous findings (Tilly and Johnson, 1988a; 1989), a maximally effective dose of the phorbol ester (162 nM) was chosen in order to compare the ability of growth factors and PMA to affect the cAMP and steroid response of granulosa cells to LH and the cAMP analogue. The effect of TGFP on steroidogenesis and its interaction with TGFa was evaluated in the next series of studies (experiment 3). Cells were treated with increasing concentrations of TGFP (0.08-800pM) in the absence and presence of LH (10ng/tube) or LH plus TGFa (18nM), after which levels of progesterone were determined. In experiment 4, the effects of EGF (0.33-164 nM) and TGFa (1.8-1800 pM) on granulosa cell plasminogen activator activity were evaluated following an 18 h incubation. Additionally, the actions of TGFP on granulosa cell PA activity and its interactions with T G F a were also examined (experiment 5). Cells were treated with increasing concentrations of TGFP (0.08-800 pM) in the absence and presence of TGFa (18 pM), after which levels of cellassociated and secreted PA activity were determined.
Radioimmunoassays I’rogesterone and cAMP levels were determined by
2-19
specific radioimmunoassays as previously described (Tilly and Johnson, 1989; Johnson and Tilly, 1988).
PA Activity Assay Cell-associated and secreted PA activity were determined from the cell pellet and supernatant, respectively, using the S-2251 colorimetric assay system as previously described (Tilly and Johnson, 1987, 1988b) with one minor modification. Upon termination of the PA assay reaction, 0.2 ml aliquots were transferred to 96-well Corning ELISA plates (Fisher Scientific, Springfield, NJ) and read in a Bio-Tek automatic plate reader at an absorbance of 405 nm. Statistics
All experiments, containing three replicate tubes per treatment, were repeated at least three times. Data, presented as the mean+SEM of results from the replicate experiments, were analysed by a one-way analysis of variance (ANOVA) with significant interactions ( p < 0.05) partitioned by the Newman-Keuls multiple range test. Estimated median effective doses (ED,,) and inhibitory doses (IDsu) were calculated from the combined data of the replicate experiments using the ALLFIT Program (DeLean, Munson and Rodbard, 1978).
RESULTS In the first experiment, LH in the presence of lBMX stimulated an approximate 13-fold increase in cAMP accumulation as compared to basal levels ( p < 0.001, Fig. 1). Coincubation of cells with 33-328 nM EGF did not affect basal cAMP levels (p>0.10, data not shown) but caused a significant suppression of LHinduced cAMP formation with an apparent IDi,, of 97.1 k 14.6 nM (see Fig. 1, upper panel). The highest concentration of EGF (328 nM) inhibited LHstimulated cAMP production by approximately 28% (pO.10 for all concentrations of
TGFP plus LH and TGFa us LH and TGFa alone, )I+).
Epidermal growth factor stimulated granulosa cell PA activity in a dose-dependent manner, with an apparent ED,,, of 4.7 f 0.3 and 23.8 k 1.7 nM for cellassociated and secreted PA activity, respectively (Fig. 4, upper panel). The highest concentration of EGF (164 nM) increased cell-associated and secreted PA activity 4.3-fold and 7.9-fold, respectively, over basal levels ( p
J L TILLY* and A L JOHNSON
Studies were conducted with chicken granulosa cells to evaluate the relative efficacy of human recombinant transforming growth factor alpha (TGFa)versus murine epidermal growth factor (EGF)to affect cyclic adenosine monophosphate (CAMP) accumulation and progesterone production stimulated by luteinizing hormone (LH) or steroid output induced by a CAMP analogue, and to modulate plasminogen activator (PA) activity. Increasing concentrations of EGF (33-328 nM) and TGFa (0.04-18 nM) were found t o inhibit cAMP formation stimulated by LH in a dose-dependent manner, with calculated halfmaximal inhibitory doses (ID?,,’s) of 97.1 and 0.27 nM, respectively. Similarly, a 470-fold difference in the ability of TGFa (ID5,,=0.13nM) versus EGF (IDJ,,=h1.3nM) t o halfmaximally suppress LH-induced progesterone production was observed in the same cells. Progesterone production stimulated by a cAMP analogue (8-bromo-cAMP, 1 mM) was also attenuated by EGF (ID?,,= 75.9 nM) and TGFa (ID?,,= 0.08 nM), suggesting a post-CAMP site of inhibition by these growth factors on steroidogenesis. Finally, a 260-fold to 330-fold difference in the efficacy of TGFa versus EGF to half-maximally stimulate cell-associated and secreted PA activity was observed. From these data, we propose that TGFa may serve an important role in regulating follicular growth and maturation in the domestic hen via its ability to affect granulosa cell steroidogenesis and plasminogen activator activity.
KEYWORDS:
TGFa,TGFD,CAMP, progesterone, plasminogen activator, ovary
enclosed mouse oocytes, suggesting a role for ECF in the final steps of oocyte maturation. In granulosa cells of the domestic hen, EGF has Previous studies have indicated that growth factors may play a significant role in the regulation of been shown to inhibit cyclic adenosine monoovarian cell differentiation and development in both phosphate (CAMP) and progesterone production mammalian and avian species. For instance, a con- stimulated by luteinizing hormone (LH), and prosiderable amount of evidence has been generated gesterone output induced by a cAMP analogue, concerning the ability of several growth factors to collectively indicating an inhibitory effect of EGF on influence gonadotropin-induced differentiation in steroidogenesis both prior and distal to cAMP forrat, bovine and porcine ovarian cells (Hsueh et al., mation (Pulley and Marrone, 1986). More recently, 1984; Adashi et al., 1985; Hill, 1989; Carson et al., Yoshimura and Tamura (1988) demonstrated that 1989). Additionally, Downs, Daniel and Eppig (1988) EGF stimulates the iii vitro proliferation of chicken demonstrated that epidermal growth factor (EGF) granulosa cells, whereas data from our laboratory induces germinal vesicle breakdown in cumulus cell- has indicated that EGF and several other growth factors increase plasminogen activator (PA) activity in dispersed granulosa cells (Tilly and Johnson, 1987; 1990). *Corresponding author
INTRODUCTION
247
Growth Factors Downloaded from informahealthcare.com by Flinders University of South Australia on 01/18/15 For personal use only.
248
TlLLY AND IOHNSON
The primary structure of the chicken EGF receptor and its binding properties have recently been deduced using a complementary deoxyribonucleic acid (DNA) clone containing the complete coding sequence for the receptor (Lax et a]., 1988). Results from these studies indicated that human recombinant transforming growth factor alpha (TGFa) bound to the chicken EGF receptor with an approximate 100-fold greater affinity than murine EGF. Furthermore, human TGFa was found to be 100-fold more effective in the induction of DNA synthesis as compared to murine EGF. In the light of these findings, we evaluated the relative efficacy of human TGFa versus murine EGF to alter steroidogenesis and plasminogen activator activity in dispersed granulosa cells collected from the largest preovulatory follicle in the ovary of the domestic hen. Additionally, we examined the ability of transforming growth factor beta (TGFB) to modulate the actions of TGFa on granulosa cell function.
MATERIALS AND METHODS
Animals Single comb white Leghorn hens, 24-40 weeks of age and laying regular sequences of at least five eggs, were used in all studies. Birds were housed individually in laying batteries, had free access to feed (Agway Layer Mash, Agway, Inc., Bordentown, NJ) and water, and were exposed to an artificial photoperiod of 15 h of light:9 h of dark with lights on at 2300 h. Individual laying cycles were monitored daily by the timing of oviposition.
Source and Preparation of Reagents Stock solutions of all treatments utilized were diluted with medium 199/HEPES (supplemented as described; Tilly and Johnson, 1987) to the desired concentration for each experiment. Ovine luteinizing hormone (NIAMDD-oLH-23; National Pituitary Agency) was diluted from a stock solution prepared in 0.01 M phosphate-buffered saline/2% bovine serum albumin (v:w). Phorbol 12-myristate 13acetate (Sigma Chemical Co., St Louis, MO) was diluted from a stock solution in 95% ethanol; 8-bromo-CAMP (Sigma) was dissolved directly into incubation medium prior to use. Epidermal growth factor (murine, receptor grade; Collaborative Research Inc., Bedford, MA) and transforming
growth factor alpha (human, recombinant; Bachem Bioscience Inc., Torrance, CA) were diluted from stock solutions in distilled, deionized water. Transforming growth factor-beta (p) (human; Collaborative Research Inc.) was prepared from a 5mM hydrogen chloride stock solution stored in a dichlorodimethylsilane-treated glass vial. All PA assay reagents (human glu-plasminogen, Desafib-fibrinogen and Spectrozyme-PL) were obtained from American Diagnostica, Inc (Greenwich, CT).
Collection and Preparation of Granulosa Cells On the morning of an experiment, hens were sacrificed by cervical dislocation 17-16 h prior to ovulation of the second or third egg of a sequence. The largest preovulatory follicle was removed and immediately placed in ice-cold 1%saline. Granulosa cells were collected and dispersed as previously described (Tilly and Johnson, 1987). Cell number and viability (routinely > 90Y0) were determined with a haemacytometer using the trypan blue exclusion technique. The cell suspension was then diluted with an appropriate volume of sterile M199/HEPES incubation medium to a final concentration of 5 ~ 1 0 ~ viable cells/ml medium.
Cell Incubations/Preparation Aliquots (10' cells) of the granulosa cell suspension were added to 12 x75 mm polypropylene (CAMPand steroid assays) or borosilicate (PA assays) tubes, combined with treatments and/or incubation medium to give a final volume of 0.5 ml, and incubated at 39°C for 4 h (CAMP and steroid determinations) or 18 h (PA assays). In all studies conducted with TGFp, incubation tubes and pipette tips were first treated with dichlorodimethylsilane (Eastman Kodak Co., Rochester, NY), washed thoroughly and rinsed well in order to prevent nonspecific adsorption of TGFB. For cAMP and steroid analyses (experiment l), cells were pelleted by centrifugation (200 xg, 20 min, 4°C) following incubation. A 0.2ml aliquot of the supernatant was then removed and analysed for progesterone content, whereas the cell pellet and remaining supernatant were immediately frozen at -70°C to be assayed for cAMP at a later date. For steroid measurements only (experiments 2 and 3), the cells and medium were frozen at -70°C immediately following incubation until assayed for progesterone levels. For PA assays (experiments 4 and 5), cells were' pelleted by centri-
TRANSFORMING GROWTH FACTORS IN THE H F h OVARY
fugation (200 x ~ 20 , min, 4°C) following incubation. The supernatant was carefully removed and analysed for secreted PA activity whereas the cell pellet was assayed for cell-associated enzyme activity.
Growth Factors Downloaded from informahealthcare.com by Flinders University of South Australia on 01/18/15 For personal use only.
Experiments In experiment 1, cells were incubated for 4 h with 10 ng/tube LH and 0.1 mM IBMX in the absence and presence of EGF (33-328 nM) or TGFa (0.04-18 nM), after which levels of cAMP and progesterone were determined. Experiment 2 was conducted to evaluate the effects of EGF and TGFa on steroidogenesis stimulated by the cAMP analogue, 8-bromo-CAMP. Progesterone production was determined following exposure of granulosa cells to 8-bromo-CAMP (1mM) without and with increasing concentrations of EGF (33-328nM) or TGFa (0.04-18nM). In experiments 1 and 2, cells were also treated with LH (long) and IBMX (0.1mM) or 8-bromo-CAMP (1mM) in the absence and presence of the tumourpromoting phorbol ester, phorbol 12-myristate 13acetate (PMA: a known activator of protein kinase C; Castagna et al., 1982). Based on our previous findings (Tilly and Johnson, 1988a; 1989), a maximally effective dose of the phorbol ester (162 nM) was chosen in order to compare the ability of growth factors and PMA to affect the cAMP and steroid response of granulosa cells to LH and the cAMP analogue. The effect of TGFP on steroidogenesis and its interaction with TGFa was evaluated in the next series of studies (experiment 3). Cells were treated with increasing concentrations of TGFP (0.08-800pM) in the absence and presence of LH (10ng/tube) or LH plus TGFa (18nM), after which levels of progesterone were determined. In experiment 4, the effects of EGF (0.33-164 nM) and TGFa (1.8-1800 pM) on granulosa cell plasminogen activator activity were evaluated following an 18 h incubation. Additionally, the actions of TGFP on granulosa cell PA activity and its interactions with T G F a were also examined (experiment 5). Cells were treated with increasing concentrations of TGFP (0.08-800 pM) in the absence and presence of TGFa (18 pM), after which levels of cellassociated and secreted PA activity were determined.
Radioimmunoassays I’rogesterone and cAMP levels were determined by
2-19
specific radioimmunoassays as previously described (Tilly and Johnson, 1989; Johnson and Tilly, 1988).
PA Activity Assay Cell-associated and secreted PA activity were determined from the cell pellet and supernatant, respectively, using the S-2251 colorimetric assay system as previously described (Tilly and Johnson, 1987, 1988b) with one minor modification. Upon termination of the PA assay reaction, 0.2 ml aliquots were transferred to 96-well Corning ELISA plates (Fisher Scientific, Springfield, NJ) and read in a Bio-Tek automatic plate reader at an absorbance of 405 nm. Statistics
All experiments, containing three replicate tubes per treatment, were repeated at least three times. Data, presented as the mean+SEM of results from the replicate experiments, were analysed by a one-way analysis of variance (ANOVA) with significant interactions ( p < 0.05) partitioned by the Newman-Keuls multiple range test. Estimated median effective doses (ED,,) and inhibitory doses (IDsu) were calculated from the combined data of the replicate experiments using the ALLFIT Program (DeLean, Munson and Rodbard, 1978).
RESULTS In the first experiment, LH in the presence of lBMX stimulated an approximate 13-fold increase in cAMP accumulation as compared to basal levels ( p < 0.001, Fig. 1). Coincubation of cells with 33-328 nM EGF did not affect basal cAMP levels (p>0.10, data not shown) but caused a significant suppression of LHinduced cAMP formation with an apparent IDi,, of 97.1 k 14.6 nM (see Fig. 1, upper panel). The highest concentration of EGF (328 nM) inhibited LHstimulated cAMP production by approximately 28% (pO.10 for all concentrations of
TGFP plus LH and TGFa us LH and TGFa alone, )I+).
Epidermal growth factor stimulated granulosa cell PA activity in a dose-dependent manner, with an apparent ED,,, of 4.7 f 0.3 and 23.8 k 1.7 nM for cellassociated and secreted PA activity, respectively (Fig. 4, upper panel). The highest concentration of EGF (164 nM) increased cell-associated and secreted PA activity 4.3-fold and 7.9-fold, respectively, over basal levels ( p
