Relaxin-Induced Deoxyribonucleic Acid Synthesis in Porcine ...

2 downloads 0 Views 768KB Size Report
relaxin was shown to enhance the trophic effects of insulin and IGF-I on porcine ... were donated by the National Hormone and Pituitary Pro- gram, NIDDK ...
BIOLOGY OF REPRODUCTION 53, 1286-1292 (1995)

Relaxin-Induced Deoxyribonucleic Acid Synthesis in Porcine Granulosa Cells Is Mediated by Insulin-Like Growth Factor- 1 ' Kathleen M. Ohleth and Carol A. Bagnell 2

Departmentof Animal Sciences, Rutgers University, New Brunswick, New Jersey 08903 ABSTRACT Relaxin stimulates in vitro DNA synthesis and cell proliferation of porcine granulosa cells (GC) and theca cells. The objective of the study reported here was to determine whether components of the ovarian insulin-like growth factor (IGF) system mediate relaxin's growthpromoting effects on porcine GC in vitro. In small follicle GC, relaxin (1-100 ng/ml) significantly (p < 0.05) increased IGF-I secretion to 25-34% above control. Hormonal responsiveness of GC was shown by incubation with FSH (200 ng/ml), which resulted in 125% stimulation of IGF-I secretion relative to that in cells incubated alone. When IGF-I activity in the GC cultures was neutralized with a specific IGF-I antibody, relaxin (10 and 100 ng/ml)-induced [3H]thymidine incorporation was inhibited (p < 0.05). Coincubation with IGF-I antibody also suppressed basal and IGF-I (10 ng/ml)-induced [3H]thymidine incorporation into GC DNA, but had no effect on insulin (1 ig/ml)-induced DNA synthesis, demonstrating the specificity and lack of toxicity of the IGF-I antibody. Ligand blot analysis showed no change in secretion of GC IGF binding protein (IGFBP) in response to relaxin (1, 10, and 100 ng/ml). In contrast, IGF-I (10 ng/ml) increased secretion of IGFBP3 and -5,whereas FSH (200 ng/ml) decreased IGFBP-3 secretion and increased IGFBP-4 secretion (p < 0.05). In IGF-I receptor competition studies, IGF-I, but not relaxin, displaced [125l]IGF-I from the GC IGF-I receptor. These studies provide direct evidence for an interaction of relaxin and the ovarian IGF system. They are the first to show 1)a stimulatory effect of relaxin on IGF-I secretion; 2)the necessity of IGFI activity for relaxin-induced GC DNA synthesis; and 3)the absence of an effect of relaxin on GC IGFBPs or IGF-I receptor. These findings support a paracrine role for relaxin in the porcine follicle and show that relaxin acts indirectly to promote follicle growth by stimulating GC IGF-I secretion. INTRODUCTION Relaxin, a member of the insulin-like family of hormones, promotes growth in a number of reproductive tissues. In rodents, relaxin stimulates growth of the uterus [1, 2]; of the cervix [3, 4]; and of mammary ducts [5], alveoli [5], and epithelial cells [6]. In pigs, uterine [7, 8], cervical [9, 10], and mammary [11] growth increases in response to relaxin. In the porcine follicle, relaxin induces growth of granulosa [12] and theca [13] cells by increasing DNA synthesis and cell proliferation in vitro. Since relaxin is a product of porcine theca interna cells [14, 15], this growth-promoting action of relaxin in the porcine follicle supports both a paracrine and autocrine role for thecal relaxin. The mechanism(s) by which relaxin stimulates growth in porcine follicle cells is unknown. Relaxin may act directly through its own receptor, initiating a cascade of second messengers leading to increases in DNA synthesis and cell proliferation. Alternatively, relaxin may act indirectly to promote growth by influencing growth factor production and/ or availability. An interaction of relaxin with the ovarian insulin-like growth factor (IGF) system was suggested when relaxin was shown to enhance the trophic effects of insulin and IGF-I on porcine granulosa cells (GC) in vitro [12].

IGF-I is detected in porcine follicular fluid and produced by porcine GC in vitro [16], and IGF-I mRNA is expressed in pig follicles at all stages of development [17]. In vivo studies indicate an increase in follicular fluid IGF-I in response to gonadotropin priming of prepubertal pigs [18]. FSH also augments secretion and de novo production of IGF-I from porcine GC in vitro [19]. Whether relaxin influences IGF-I secretion from porcine GC in vitro has not been investigated. IGF binding proteins (IGFBPs) have been reported to inhibit or potentiate IGF-I activity [20]. FSH and IGF-I regulate IGFBP secretion by porcine GC [21-23]. Relaxin stimulates basal and progesterone-induced IGFBP-1 secretion from cultured human endometrial stromal cells [24], as well as human decidual IGFBP-1 secretion [251. Whether relaxin alters porcine GC IGFBP secretion or interacts with follicle IGFBPs is unknown. The objective of the present study was to determine whether relaxin's growth-promoting effects on porcine GC are mediated indirectly by components of the ovarian IGF system. Cell culture and RIA were used to examine whether relaxin alters IGF-I secretion from porcine GC in vitro. The role of IGF-I in mediating the trophic effects of relaxin on porcine GC was investigated by neutralization of relaxininduced DNA synthesis with a specific IGF-I antibody. Ligand blotting was used to determine whether relaxin influences porcine GC secretion of IGFBPs and/or interacts with any IGFBPs. Although relaxin is not reported to compete for insulin [26] or IGF-I receptors [27, 28] in other tissues. whether relaxin binds to the porcine GC IGF-I receptor has not been reported and was included in this study.

Accepted August 14, 1995. Received May 31, 1995. ISupport: USDA Award #93-37203-8979 and NJ Agricultural Experimental Station Grant #D-06125-1-95. 2 Correspondence: Carol A. Bagnell, Dept. of Animal Sciences, P.O. Box 231, Cook College, Rutgers University, New Brunswick, NJ 08903. FAX: (908) 932-6996: e-mail: [email protected].

1286

RELAXIN'S TROPHIC EFFECTS ARE MEDIATED BY IGF-I

MATERIALS AND METHODS Materials Porcine relaxin was a gift from Dr. G.D. Bryant-Greenwood (University of Hawaii, Honolulu, HI). A monoclonal anti-human IGF-I antibody for immunoneutralization studies was purchased from Upstate Biotechnology Inc. (Lake Placid, NY). This monoclonal IGF-I antibody was previously designated Sml.2 and originated from Van Wyk et al. [29]. A control mouse IgG was purchased from Zymed Laboratories, Inc. (S. San Francisco, CA). Porcine FSH for cell culture and a polyclonal IGF-I antibody (UB3-189) for RIA were donated by the National Hormone and Pituitary Program, NIDDK (Baltimore, MD). Human recombinant IGF-I for cell culture and RIA standards were purchased from BACHEM (Torrance, CA) and R&D Systems (Minneapolis, MN), respectively. Porcine insulin was donated by Lilly Research Laboratories (Indianapolis, IN). Anti-rabbit gamma globulin and normal rabbit serum were obtained from Antibodies, Inc. (Davis, CA) and Vector Labs, Inc. (Burlingame, CA), respectively. [125 I]IGF-I and [125I]IGF-II were purchased from Amersham Corporation (Arlington Heights, IL). [3 H]Thymidine and [125I]Na iodide were obtained from NEN Research Products (Boston, MA). Monotyrosylated relaxin was a gift from Dr. R.V. Anthony (Colorado State University, Fort Collins, CO). Anti-mouse IgG horse-radish peroxidase conjugate was obtained from Transduction Laboratories (Lexington, KY). BSA was purchased from Sigma Chemical Company (St. Louis, MO). Fetal calf serum (FCS), Medium 199 (M199) and all other cell culture reagents were obtained from Life Technologies (Grand Island, NY). Cell Culture Ovaries from immature pigs were collected from local slaughterhouses. Granulosa cells from small follicles (1-3 mm) were needle-aspirated and washed twice with M199 containing 25 mM HEPES, 100 U/ml penicillin, 100 pg/ml streptomycin, 10 gpg/ml gentamicin, and 0.5 pg/ml fungizone. Viable cells (1 x 106 cell/well) were inoculated into 16-mm wells and incubated in M199 containing 10% FCS (serumM199) at 37°C in a humidified atmosphere of 5% CO2:95% air. After 48 h of attachment, cells were treated with hormones in serum-free M199 containing 5 pg/ml transferrin and 0.1% BSA (serum-free M199) for 48 h, and IGF-I secretion into media was measured. To monitor IGFBP secretion, GC (1.5 X 106 cells/well) were plated in 16-mm wells and, after attachment, incubated in serum-free M199 with hormones and aprotinin (50 gg/ml) in the absence of 0.1% BSA. After 72 h, IGFBP secretion into media was measured [21]. IGF-IRIA IGF-I content of media samples was measured by RIA according to the method of Lee et al. [30]. Briefly, IGF-I was

1287

separated from its binding proteins by acidification with an equal volume of 1.73% aqueous trifluoroacetic acid (TFA) for 10 min. The acidified samples were passed through preconditioned C18 Sep-Pak cartridges (Waters, Milford, MA), and the IGFBPs were removed with 0.1% TFA. IGF-I was eluted with 0.1% TFA in acetonitrile, dried, and reconstituted in 0.5 ml assay buffer. This extraction procedure removes 80% of sample IGFBPs [30]. Standard amounts of IGF-I added to media were measured by RIA, before and after extraction, to determine an IGF-I recovery of 75%. Sample aliquots were diluted in assay buffer (300 p1 total) and incubated overnight at 4°C with polyclonal IGF-I antibody (UB3-189; 100 pl; final concentration 1:20 000) and [125I]IGF-I (100 l; 15 000 cpm/tube). Subsequently, normal rabbit serum (100 pl; final concentration 1:700) and antirabbit gamma globulin (100 pl; final concentration 1:70) were added and incubated for 4 h at 4C. Samples were centrifuged at 3000 rpm for 30 min, and the pellets were counted. Intra- and interassay variation was 4.2% and 7.2%, respectively. Relaxin alone in media did not cross-react with the IGF-I antibody in the RIA (data not shown). IGF-I secretion was calculated on a per-well basis, since preliminary experiments showed no significant change in cell number after 2 days of hormone treatment. [3 HlThymidine Incorporationand IGF-I Immunoneutralization [3H]Thymidine incorporation studies were performed as previously described [13] with modifications. Granulosa cells (8 X 104 cells/well) were plated in 6-mm wells. After 48-h attachment in serum-M199, GC were incubated in serum-free M199 with hormones in the presence or absence of IGF-I monoclonal antibody or mouse IgG for 48 h. Hormone treatments were preincubated with the IGF-I antibody or mouse IgG for 2 h at 4C before addition to the GC culture. [3H]Thymidine (4 ipCi/ml) was added for the last 8 h of the treatment period. Cells were lifted from the wells with 0.25% trypsin/5 mM EDTA and collected onto glass fiber filters by means of a multiwell cell harvester (Cambridge Technology, Inc., Watertown, MA), and [3 H]thymidine incorporation into the DNA was assessed. Previous studies from this laboratory show that [3H]thymidine incorporation into porcine GC measured by this method represents DNA synthesis and not DNA repair [12]. The IGF-I antibody used for the immunoneutralization studies has 5% cross-reactivity with human IGF-II and 1% cross-reactivity with rat IGF-II and does not bind insulin at 10 - 6 M [29, 31]. To rule out the possibility of a relaxin-IGFI antibody interaction in the immunoneutralization studies, relaxin binding to the monoclonal IGF-I antibody was investigated. Increasing concentrations of IGF-I and relaxin (100-1000 ng) were blotted onto a polyvinylidene fluoride (PVDF; Millipore Corp., Bedford, MA) membrane, which

1288

OHLETH AND BAGNELL

was blocked in 5% nonfat dry milk (NFDM; Bio-Rad Labs., Richmond, CA) in PBS and rinsed 5 min in wash buffer (PBS/0.1% Tween-20). The membrane was incubated with the IGF-I monoclonal antibody (2 glg/ml in PBS/1% NFDM) for 1 h at room temperature, washed, and incubated with an anti-mouse IgG horse-radish peroxidase conjugate (1:2500) for 1 h at room temperature. Diaminobenzidine (0.67 mg/ml 0.05 M Tris HCl, pH 7.6) was used to visualize IGF-I antibody binding. Ligand Blotting Ligand blotting was performed as previously reported by Mondschein et al. [21] with modifications. Media were collected (4 wells/sample), concentrated [21] 8-10 X with Centricon-10 (Amicon, Beverly, MA), assayed for protein (BioRad Labs., Melville, NY), and stored at -20°C until electrophoresis. Aliquots of concentrated media (10 tg protein) in loading buffer were run on a 12% SDS-PAGE gel under nonreducing conditions. Proteins were transferred onto a PVDF membrane at 400-500 mA for 1 h on ice. The membrane was dried at 37 0C for 5 min and washed with buffers as reported by Hossenlopp et al. [32]. The membrane was subsequently incubated overnight at 4C with [125I]IGF-II (500 000 cpm/50 cm 2 membrane), washed, allowed to dry, and exposed to Hyperfilm-MP (Amersham) for 2-7 days at - 80°C. Densitometry was used to quantify changes in IGFBP secretion. To determine whether relaxin interacts with any IGFBPs, [12 5I]relaxin ligand blotting was performed on samples from the IGF ligand blotting experiments. Monotyrosylated relaxin was iodinated according to Hall et al. [8] with Iodogen (Sigma) to a specific activity of 55 ICi/pg. The samples were treated as described above except that the membranes were incubated with [125 I]relaxin (750 000 cpm/cm 2 membrane) instead of [125I]IGF-II overnight at 4C. Autoradiographs were monitored for [' 25 I]relaxin binding to IGFBPs. IGF-IReceptor Competition Receptor binding studies were performed as previously described [33] with modifications. Briefly, GC (5 X 106 cells/ tube) from medium follicles (4-6 mm) were incubated with ['25 I]IGF-I (50 000 cpm; specific activity 263 mCi/mg) and increasing concentrations of IGF-I (0.5-100 ng/ml) or relaxin (1-10 000 ng/ml) for 2 h at room temperature. The incubation was terminated by the addition of 1 ml ice-cold PBS, the cell suspensions were centrifuged at 1500 rpm for 30 min, and the cell pellet was counted for [125I]IGF-I binding. Statistics The results presented are the mean + SEM of at least three experiments. Data are presented as percentages relative to control values, with the control equal to 100%. The

data were analyzed by ANOVA, and the means were compared by Duncan's Multiple Range test by use of the SAS program for Windows [34]. P values < 0.05 were accepted as significant. RESULTS IGF-I Secretion in Response to Relaxin or FSH Relaxin at 1, 10, and 100 ng/ml significantly (p < 0.05) increased IGF-I secretion by small follicle GC (Fig. 1). This relaxin-induced increase was 25 + 6.7% to 34 + 7.8% above control. Since FSH is reported to stimulate porcine GC IGF-I secretion in vitro [19], the responsiveness of the cultured GC was verified by FSH (200 ng/ml) treatment, which significantly (p < 0.05) enhanced IGF-I secretion 125 + 17.1% above control. Effect of IGF-IAntibody on [3 HIThymidine Incorporation To examine whether IGF-I mediates relaxin-induced growth of porcine GC, a monoclonal IGF-I antibody was used in an attempt to inhibit relaxin-induced [3H]thymidine incorporation by GC. Since IGF-I (3 ng/ml), insulin (1 g/ ml), and relaxin (10 and 100 ng/ml) stimulate DNA synthesis by porcine GC [12], these concentrations of hormones were used to monitor GC [3 H]thymidine incorporation into DNA in the presence or absence of a specific monoclonal IGF-I antibody. Figure 2 illustrates that, in the presence of the IGFI antibody (1.5 jg/ml), the increase in GC [3 H]thymidine incorporation in response to IGF-I (3 ng/ml; 221 + 12.0% above control) was attenuated (48 11.7% above control; p < 0.05). However, the inhibitory effects of the IGF-I antibody were reversed when 10 ng/ml of IGF-I was added to the incubation and DNA synthesis was restored (Fig. 2; 242 + 21.4% above control). The specificity and lack of cytotoxicity of the IGF-I antibody was further demonstrated by the absence of an inhibitory effect of the antibody on insulin-dependent GC DNA synthesis (Fig. 2). Furthermore, incubation of GC with a control mouse IgG (1.5 g/ml) of the same subtype (IgG kappa), did not affect control, IGFI-induced, or insulin-induced DNA synthesis (Fig. 2). On the other hand, relaxin-induced increases in GC [3 H]thymidine incorporation were significantly (p < 0.05) suppressed below control levels in the presence of the IGF-I antibody (Fig. 3). The decreases in DNA synthesis in response to relaxin at 10 and 100 ng/ml were 82% and 73%, respectively, with IGF-I antibody coincubation, compared to incubation of GC with relaxin alone. In addition, under basal conditions, there was a 25% suppression in control GC [3H]thymidine incorporation when the IGF-I antibody was included in the incubation (Figs. 2 and 3; p < 0.05). Immunoblotting results showed no evidence of relaxin binding to the monoclonal IGF-I antibody (data not shown).

RELAXIN'S TROPHIC EFFECTS ARE MEDIATED BY IGF-I Effect of Hormones on GC IGFBPs and IGF-I Receptor Binding To monitor changes in IGFBP secretion, GC were incubated with relaxin (1, 10, and 100 ng/ml), FSH (200 ng/ml), or IGF-I (10 ng/ml), and results of densitometric analysis of the ligand blot were compared. Relaxin did not significantly affect GC secretion of IGFBP-3, -4, or -5 at any concentration. In contrast, IGF-I increased GC secretion of IGFBP-3 and -5 to 122 + 10.3% and 1051 + 302.1% above control, respectively (p < 0.05). In addition, FSH decreased IGFBP3 by 80% (p < 0.05) and increased IGFBP-4 GC secretion to 389 + 217.7% above control (p < 0.05). [125I]Relaxin ligand blotting revealed that [12 5I]relaxin did not bind any membrane-bound IGFBPs (data not shown). IGF-I (0.5-100 ng/ml) displaced [125I]IGF-I from the GC IGF-I receptor in a dose-dependent manner (Fig. 4). Conversely, relaxin (110 000 ng/ml) did not compete for the IGF-I receptor, even at supraphysiological doses (Fig. 4). DISCUSSION Although the trophic effects of relaxin have been documented in a variety of reproductive tissues, including the ovarian follicle, a mechanism to explain precisely how relaxin promotes growth has not been forthcoming. The present study is the first to provide evidence that IGF-I is a key element in mediating the trophic effects of relaxin. Data presented here indicate that relaxin stimulates IGF-I secretion at concentrations similar to those that promote GC DNA synthesis and proliferation [12]. In addition, relaxin-induced [3 H]thymidine incorporation into GC DNA is suppressed by neutralization of IGF-I activity with a specific IGF-I antibody. Finally, the absence of an effect of relaxin on GC IGFBP secretion and the lack of relaxin binding to GC IGFBPs or IGF-I receptor point to the importance of IGF-I itself in mediating the trophic effects of relaxin on GC. These findings are consistent with a growing body of evidence that growth factors are important mediators of the trophic effects of hormones. In the pig ovary, it is suggested that FSH and estradiol induce in vivo GC growth not by direct stimulation of DNA synthesis, but by indirect mediation through locally produced growth factors, which act on GC to induce proliferation in an autocrine or paracrine manner [35]. In addition, Bley et al. [36] conclude that locally produced mitogens mediate estrogen-induced growth of rat GC. Furthermore, immunoneutralization of epidermal growth factor (EGF) blocks FSH-induced [3 H]thymidine incorporation in hamster preantral follicle cells, suggesting that EGF mediates FSH-induced hamster follicle growth [37]. Likewise, we suggest here that relaxin-induced follicle cell growth is mediated indirectly by relaxin stimulation of local IGF-I production in the porcine ovary. The physiological significance of a 25-34% increase in

1289

FIG. 1. Effect of relaxin on IGF-I secretion from small follicle GC. After 48 h of attachment inserum-M199, GC were incubated with relaxin (1, 10, and 100 ng/ml) in serum-free M199 for 48 h. Media were collected, IGF-I separated from IGFBPs, and IGF-I content measured by RIA. Bars with different letters are significantly different (p < 0.05). Control media contained 121 pg IGF-I/1 x 106 cells/48 h.

IGF-I secretion in response to relaxin, in terms of mediating relaxin-induced growth, was supported by IGF-I immunoneutralization studies. Relaxin-induced DNA synthesis was blocked when IGF-I activity in GC cultures was suppressed by coincubation with a specific monoclonal IGF-I antibody. Evidence that IGF-I mediates the growth effects of relaxin in the porcine follicle is consistent with reports that IGF-I mediates the trophic effects of growth-promoting peptides

FIG. 2. Influence of IGF-I antibody on control, IGF-I-, and insulin-induced thymidine incorporation by small follicle GC. After 48-h attachment, GC were incubated with IGF-I (3 ng/ml) or insulin (1 ig/ml) in presence or absence of monoclonal IGF-I antibody (1.5 gg/ml) or mouse IgG (1.5 Ig/ml) in serum-free M199 for 48 h.To reverse effects of IGF-I antibody, IGF-I (10 ng/ml) was added to the incubation. 3 [3 HIThymidine was added for last 8 h of treatment, and [ Hlthymidine incorporation into DNA was determined. Bars with different letters within each treatment are significantlydifferent (p < 0.05). Ab, antibody.

1290

OHLETH AND BAGNELL 200

o

150

.n

100

E.

cells

·

IGF Ab

b

t

alone

CZ~U1

c

b

C

c

I

50

I___ 0

10 Relaxin

100

FIG. 3. Effect of IGF-I antibody on relaxin-induced thymidine incorporation by small follicle GC:. GC were cultured as described in Figure 2 legend and incubated with relaxin (10, 100 ng/ml) in presence or absence of monoclonal IGF-I antibody, and thymidine ircorporation into DNA was assessed. Bars with different letters are significantly diff erent (p