Differentiation-Dependent Expression of Connective Tissue Growth ...

0013-7227/01/$03.00/0 Endocrinology Copyright © 2001 by The Endocrine Society

Vol. 142, No. 3 Printed in U.S.A.

Differentiation-Dependent Expression of Connective Tissue Growth Factor and Lysyl Oxidase Messenger Ribonucleic Acids in Rat Granulosa Cells* ROGER B. SLEE, STEPHEN G. HILLIER, PAWLINA LARGUE, CHRISTOPHER R. HARLOW, GINO MIELE, AND MICHAEL CLINTON Department of Reproductive and Developmental Sciences, University of Edinburgh (R.B.S., S.G.H., P.L., C.R.H.), Edinburgh, United Kingdom EH3 9ET; and Department of Gene Expression and Development, Roslin Institute (G.M., M.C.), Roslin, United Kingdom EH25 9PS ABSTRACT Searching for novel genes involved in tissue remodeling during ovarian folliculogenesis, we carried out differential display RT-PCR (DDRT-PCR) on RNA from gonadotropin-stimulated rat granulosa cells (GC). GC from preantral and early antral follicles in immature rat ovaries were cultured in serum-free medium containing no hormone (control), recombinant human FSH (10 ng/ml), 5␣-dihydrotestosterone (DHT; 10⫺6 M), or FSH plus DHT. Total cellular RNA was extracted from cells at 6, 12, 24, and 48 h of treatment for DDRT-PCR analysis, corresponding to an estimated 60% saturation of the messenger RNA (mRNA) population. Six distinct complementary DNA clones were obtained that reproduced the DDRT-PCR profile on a Northern blot of the corresponding RNA samples. Two of these clones detected transcripts that were strongly down-regulated by FSH. One corresponded to connective tissue growth factor (CTGF), a cysteinerich secreted protein related to platelet-derived growth factor that is implicated in mitogenesis and angiogenesis, and a second was iden-

O

VARIAN FOLLICULAR growth requires a tightly coordinated interplay among gonadotropin-stimulated follicular cells, extracellular matrix (ECM), and factors in serum (1). The lamina basalis separating thecal and granulosa cells and connective tissue encapsulating the theca must undergo constant development and remodeling to accommodate follicular expansion (2). Previous research has shown that the regulation of granulosa cell (GC) gene expression by FSH is modulated by androgens and polypeptide growth factors produced by LH-stimulated thecal cells, suggesting a mechanism for paracrine regulation of GC responsiveness to FSH throughout the follicular maturation process (3). Conversely, FSH-stimulated granulosa cells presumably secrete factors that promote thecal development and ECM turnover (4), including proteases, antiproteases (5), and angiogenic growth factors (6). Searching for candidate genes involved in this intimate cell dialogue, we examined differences in gene expression caused by exposure of rat granulosa cell cultures to FSH and

Received July 21, 2000. Address all correspondence and requests for reprints to: Stephen G. Hillier, Ph.D., Department of Reproductive and Developmental Sciences, University of Edinburgh, 37 Chalmers Street, Edinburgh, United Kingdom EH3 9ET. E-mail: [email protected]. * Presented in part at the 82nd Annual Meeting of The Endocrine Society, Toronto, Canada, 2000. This work was supported by Medical Research Council Program Grant 8929853 (to S.G.H.).

tical to lysyl oxidase (LO), a key participant in extracellular matrix deposition. In detailed expression studies, Northern analysis revealed a single, approximately 2.5-kb CTGF transcript maximally suppressed within 3 h of exposure to FSH with or without DHT and two LO transcripts (⬃3.8 and ⬃5.2 kb) maximally suppressed at 6 h. DHT alone did not affect CTGF mRNA, but strongly enhanced LO mRNA relative to the control value. In vivo, CTGF and LO transcripts were significantly suppressed in GC 48 h after equine CG injection (10 IU, ip) compared with untreated controls and were further reduced 12 h after administration of additional 10 IU hCG to induce luteinization. In situ hybridization confirmed GC in preantral/early antral follicles as principal sites of CTGF and LO mRNA expression. We conclude that expression of CTGF and LO mRNAs is inversely related to GC differentiation. The encoded proteins probably have roles in the regulation of tissue remodeling and extracellular matrix formation during early follicular development. (Endocrinology 142: 1082–1089, 2001)

androgen using differential display RT-PCR (DDRT-PCR) (7). An immature GC culture system was selected to reveal differentiation-related changes under experimentally defined conditions in vitro. Here, we report evidence that connective tissue growth factor (CTGF) and lysyl oxidase (LO) are members of the FSH/androgen-regulated gene repertoire expressed in mammalian granulosa cells. As CTGF is implicated in the regulation of connective tissue synthesis (8, 9) and LO in the formation of ECM (10, 11), both genes are likely to play physiologically significant roles in the regulation of ovarian follicular development. Materials and Methods Hormones and tissue culture reagents Recombinant human FSH (3860 IU FSH/mg) was donated by Dr. C. Howles (Serono Laboratories, Inc., Welwyn Garden City, UK). Equine CG (eCG), hCG, steroids, diethylstilbestrol (DES), l-glutamine, penicillin, streptomycin, and BSA were obtained from Sigma (Poole, Dorset, UK). The culture medium was medium 199 supplemented with 25 mm HEPES, 2 mm l-glutamine, penicillin (50 IU/ml), streptomycin (50 mg/ ml), and 0.1% (wt/vol) BSA. Donor calf serum, medium 199, and Dulbecco’s PBS were supplied by Life Technologies, Inc. (Paisley, UK).

Animals Immature (21-day-old) female Wistar rats from Charles River Laboratories, Inc. UK (Margate, UK), were kept in a temperature-controlled room on a 12-h light, 12-h dark cycle and fed rat chow and water ad libitum. To provide ovaries containing proliferating but essentially non-

1082

CTGF AND LO GENE EXPRESSION IN GRANULOSA CELLS differentiated granulosa cells for tissue culture (see below), animals were given daily sc injections of DES (2 mg/day) for 2 days to stimulate preantral/early antral follicular development. For expression studies in vivo, gonadotropin-stimulated ovaries were obtained 48 h after ip injection of eCG (10 IU) or 12 h after ip injection of hCG, given 48 h after injection of 10 IU eCG. Controls received no hormone treatment.

GC isolation and culture Ovaries were removed for GC isolation after killing the animals by asphyxiation with CO2. All handling and treatment of animals were performed according to guidance issued by the British Home Office. Follicles were punctured with a 25-gauge hypodermic needle, and GC were gently expelled into culture medium. The cells were harvested by centrifugation, resuspended in fresh medium, and counted in a hemocytometer. Cell viability assessed by trypan blue staining was 30% or more. For tissue culture studies (cells from DES-treated animals), 24-well tissue culture grade plastic dishes (Corning, Inc., Corning, NY) were precoated with donor calf serum and washed twice with Dulbecco’s PBS before inoculation with 0.25 ml culture medium containing approximately 2 ⫻ 105 viable cells (12). After overnight preincubation at 37 C in a humidified incubator gassed with 5% CO2 in air, prewarmed culture medium (0.25 ml) containing hormone(s) was added to start the following treatments: FSH (1 or 10 ng/ml), DHT (1 ␮m), or FSH plus DHT. Control cultures contained no hormone. Thirty-six wells were used for each treatment at each time point. Incubation was performed for 0 – 48 h at 37 C. The index response was aromatase activity, measured by incubating washed cell monolayers for an additional 6 h at 37 C in 0.5 ml medium containing 1 ␮m testosterone as an aromatase substrate. This medium was collected and assayed for estradiol content by RIA (12).

RNA isolation and RNA gel blot analysis Total RNA was isolated using RNAzol B (Tel-Test, Friendswood, TX) following the manufacturer’s recommendations. To obtain RNA from cultured granulosa cell monolayers, RNAzol B (150 ␮l) was added to the culture wells after removing spent culture medium; for freshly isolated granulosa cells, RNAzol was added after collecting the cells by centrifugation in 1.5-ml Eppendorf (Eppendorf AG, Hamburg, Germany) tubes. RNA was fractionated by electrophoresis on 1.0% agarose-formaldehyde gels. Northern hybridization was carried out according to standard methods using nylon filters and hybridizing in 7% SDS, 0.5 m sodium phosphate (pH 7) at 65 C. Posthybridization washes were two washes for 5 min each time in 2 ⫻ SSC (standard saline citrate) and two washes for 15 min each time in 0.1 ⫻ SSC in presence of 0.1% SDS at 65 C. The probes were 32P-labeled amplicons yielded by DDRT-PCR analysis of GC RNA (see below), rat cytochrome P450 aromatase (P450arom) complementary DNA (cDNA; provided by Dr. J. S. Richards, Baylor College of Medicine, Houston, TX), and rat 18S ribosomal RNA (rRNA) cDNA (Ambion, Inc., Austin, TX). The filters were exposed to autoradiographic film (XAR-5, Eastman Kodak Co., Rochester, NY) for 1–2 days at ⫺70 C using an intensifying screen.

1083

performed using the Promega Corp. cycle sequencing kit (Promega Corp. Europe, Southampton, UK) according to the manufacturer’s instructions, using M13 forward and reverse primers (14, 15). To determine the identity of isolated DDRT-PCR cDNAs, all sequence data were checked against nonredundant GENBANK/EMBL protein and nucleotide databases, using the National Center for Biotechnology Information worldwide web implementations of the BLAST algorithm (16).

In situ hybridization Ovaries were fixed in 4% paraformaldehyde in PBS and embedded in paraffin blocks. After digestion with proteinase K and acetylation, sections (10 ␮m) were hybridized (overnight incubation at 55 C) with complementary RNA (cRNA) probes generated from cDNA templates labeled with [35S]UTP (Amersham International, Aylesbury, UK) using an RNA transcription kit (Promega Corp.). Slides were washed in buffers of decreasing salt concentrations, dehydrated through ethanol gradients, and processed for liquid emulsion autoradiography (Kodak NTB2). After exposure for 3 weeks at 4 C, slides were developed, counterstained with hematoxylin, and taken for photomicrography. Sense cRNA probes were used as a control for nonspecific binding.

Data analysis Northern blots were quantified by electronic autoradiography (Instant Imager, Packard, Downers Grove, IL) with normalization of mRNA abundance to 18S rRNA. Data were analyzed using one-way ANOVA with Student’s t test.

Results Validation of granulosa cell culture system

Enzymatic assay of aromatase activity (Fig. 1) and Northern analysis of P450arom mRNA (data not shown) confirmed hormone-responsive gene expression with synergy between FSH and DHT in the GC culture system from which input RNA was obtained for DDRT-PCR analysis. Differential display RT-PCR reveals CTGF and LO mRNA expression by cultured GC

Differential display RT-PCR analysis identified 12 bands that reproducibly exhibited hormonally responsive expres-

DDRT-PCR analysis The source material for DDRT-PCR analysis was total RNA isolated from cultured GC from DES-treated immature rats incubated for 6, 12, 24, and 48 h with and without FSH, DHT, or FSH plus DHT. DDRT-PCR was performed essentially as previously described (7) with minor modifications (13–15). The DDRT-PCR screen was performed using 60 separate primer combinations (3 anchor primers and 20 random primers) corresponding to an estimated 60% saturation of the messenger RNA (mRNA) population. Radiolabeled DDRT-PCR products were loaded onto a 6% nondenaturing HR-1000 GenomyxLR polyacrylamide gel (Genomyx, Foster City, CA). Samples were run for 2 h and 15 min at 2700 V (50 C) on a GenomyxLR DNA analyzer (Beckman Coulter, Fullerton, CA). A representative subset of bands exhibiting a range of hormone response profiles was selected for further processing, as previously described (13). After purification (14), cDNAs were PCR reamplified and cloned into the EcoRI site of pBluescript SK⫹ (Stratagene Europe, Amsterdam, The Netherlands). The authenticity of isolated DDRT-PCR candidate cDNAs was confirmed (15) before expression studies by probing duplicate Southern-blotted DDRT-PCR products. Sequencing was

FIG. 1. Hormone-dependent GC differentiation in vitro, illustrated by expression of P450arom enzyme (aromatase) activity. GC from immature female rat ovaries were cultured for 48 h at 37 C without hormones (Control) or with FSH (10 ng/ml), DHT (10⫺6 M), or FSH plus DHT. Aromatase enzyme activity, based on formation of estradiol during a subsequent 6-h incubation of cell monolayers with 10⫺6 M testosterone as an aromatase substrate. Bars indicate the mean (⫾SEM) levels in culture medium determined by RIA in four separate culture experiments. a, significantly higher (P ⬍ 0.01) than control and DHT; b, significantly higher (P ⬍ 0.01) than all other treatments.

1084

CTGF AND LO GENE EXPRESSION IN GRANULOSA CELLS

Endo • 2001 Vol. 142 • No. 3

sion profiles (data not shown). Further processing of 6 bands generated 6 distinct cDNA clones that reproduced the DDRT-PCR profile on a Northern blot of the corresponding RNA samples. Two of these clones are the subject of the present report (Fig. 2). Both clones detected transcripts that were strongly down-regulated by FSH (Fig. 3). One corresponded to nucleotides 934-1340 of full-length rat CTGF cDNA (17), which encodes a secreted protein implicated in tissue remodeling and angiogenesis (18). The second was identical to nucleotides 1868 –2428 of rat lysyl oxidase cDNA (19), which encodes a key participant in ECM deposition (20).

Gonadotropin treatment in vivo down-regulates GC CTGF and LO mRNA expression

Granulosa cell CTGF mRNA expression is suppressed by FSH in vitro

In situ hybridization analysis of CTGF and LO mRNA expression in relation to follicle development

Northern analysis revealed time- and hormone-dependent effects of FSH on CTGF mRNA levels. Freshly isolated granulosa cells abundantly expressed an approximately 2.5-kb CTGF mRNA, which was also expressed fairly constantly throughout culture in the absence of hormones and in the presence of DHT alone (Fig. 3A). Within 3 h of exposure to FSH, the CTGF transcript was strongly suppressed relative to the control value. The response to FSH was not measurably affected by the additional presence of DHT. DHT alone had no effect relative to the control.

To localize the cellular compartments expressing CTGF and LO mRNA in vivo, in situ hybridization was performed on ovarian sections from immature animals stimulated by eCG with or without hCG. GC in preantral/early antral follicles of ovaries from control animals showed specific CTGF (Fig. 6A) and LO (Fig. 6E) mRNA signals. The CTGF signal was more intense than that of LO and was expressed in many primary and secondary follicles from which LO signaling was absent. Consistent with the overall lack of follicular maturity in control animals, P450arom expression was uniformly negative (Fig. 6I). GC in large antral follicles induced by treatment with eCG abundantly expressed P450arom mRNA (Fig. 6J), but showed reduced expression of both CTGF (Fig. 6B) and LO (Fig. 6F) mRNA. Additional 12-h treatment with hCG to induce GC luteinization further suppressed CTGF (Fig. 6C) and LO (Fig. 6G). The CTGF mRNA signal persisted in the most antrally located granulosa layer and cumulus GC of luteinizing antral follicles (Fig. 7).

GC LO mRNA expression is suppressed by FSH, but enhanced by androgen, in vitro

Freshly isolated GC also expressed an approximately 5.2-kb LO mRNA transcript and a less abundant approximately 3.8-kb transcript, both of which declined and then gradually recovered during culture in the absence of hormones (Fig. 3B). Within 6 h of treatment with FSH in the presence and absence of DHT, both transcripts were fully suppressed. Conversely, treatment with DHT alone markedly increased the expression of both transcripts relative to the control.

FIG. 2. Sections of differential display autoradiograms revealing expression of CTGF and LO mRNA by rat GC. Cells were cultured for the periods of time indicated in the absence (C) and presence of FSH (F), DHT (D), or FSH plus DHT (F⫹D), as described in Materials and Methods. Total cellular RNA was isolated and reverse transcribed using 60 separate primer combinations (3 anchor primers and 20 random primers). In both examples shown, the anchor primer was 5⬘-d(TTTTTTTTTTTTMC)3⬘, where M ⫽ A, G, or C. Amplicons (arrows) corresponding to CTGF (upper panel) and LO (lower panel) were generated using 5⬘-d(GAGCAGACCT)-3⬘ and 5⬘-d(GCGACCCATG)-3⬘, respectively, as arbitrary upstream primers. Note that both products are clearly suppressed at all time points by the treatments with F and F⫹D.

To determine whether the differentiation-associated changes observed in vitro reflected physiological events in vivo, immature female rats were treated with eCG with or without hCG 48 h later to promote preovulatory follicular development. Treatment with eCG alone caused approximately 50% (P ⬍ 0.01) down-regulation of granulosa cell CTGF and LO mRNA expression relative to controls. Additional treatment with hCG to induce luteinization caused further significant suppression of both mRNAs (Figs. 4 and 5).

Discussion

This study identifies CTGF and LO as FSH-suppressible genes in GC. CTGF is a chemotactic and mitogenic factor


1085

FIG. 3. Differentiation-related expression of CTGF and LO mRNAs by rat GC in vitro. GC isolated from immature female rat ovaries were established as monolayer cell cultures in serum-free medium, as described in Materials and Methods. Sixteen hours later, expression studies commenced based on incubation for up to 48 h in the absence (C) and presence of FSH (F), DHT (D), or FSH plus DHT (F⫹D). Total RNA was isolated for Northern analysis from freshly isolated (⫺16 h) and cultured cells at the times indicated. Total RNA (5 ␮g) was loaded in each lane, electrophoresed, blotted, and probed for CTGF mRNA and LO mRNA (upper panels) using amplicons isolated by DD-RT-PCR (see Fig. 2) as probes. Blots were stripped and reprobed for 18S rRNA to control for differences in loading and/or blotting. A, Autoradiogram showing the approximately 2.5-kb CTGF transcript that is abundantly expressed in freshly isolated cells and, after an initial decline, in untreated cell cultures. Exposure to F with or without D suppresses the CTGF mRNA transcript within 3 h. B, Autoradiogram showing strongly expressed approximately 5.2-kb LO mRNA transcript and the weaker approximately 3.8-kb transcript that both decline and then recover during culture without hormones. Treatment with F with or without D suppresses LO mRNA expression within 6 h, whereas D alone is stimulatory. Representative results from three experiments are shown.

initially isolated from endothelial cells (8), with proposed roles in wound repair and/or fibrosis (9) and angiogenesis (18). LO catalyzes the final known enzymatic step in crosslinking and insolubilization of collagen and elastin precursors in the formation of a mature and functional ECM (10, 11, 21). Thus, the encoded products of both genes are likely to play biologically important roles in the ovary. Expression of CTGF mRNA in rat ovarian cells has not previously been described, although a recent report (22) demonstrated CTGF mRNA and protein in follicles and corpora lutea of porcine ovary. CTGF protein has also been detected in human follicular fluid (23). The DDRT-PCR product used here to detect ovarian CTGF mRNA expression was a 716-nucleotide sequence corresponding to the region of the rat CTGF cDNA open reading frame (16) that is predicted to encode the whole of domain 4 and the C-terminal half of domain 3. As in other tissues that express CTGF mRNA, only a single approximately 2.5-kb transcript was observable by Northern analysis, consistent with the full-length (2345-bp) CTGF transcript. CTGF is a member of the connective tissue growth factor/

cysteine-rich 61/nephroblastoma-overexpressed (CCN) family of growth factors predicted to have arisen from a common ancestral gene more than 40 million yr ago (see Ref. 9 for review). Other members of the family include Fisp 12 (murine ortholog of human CTGF), human and murine Cyr61, the chicken ortholog of Cyr61, Cef10, and human and Xenopus Nov. With the exception of Nov, these are immediate early genes induced by serum, growth factors, or certain oncogenes. The encoded products are cysteine-rich secretory proteins organized into four distinct motifs. The first contains an insulin-like growth factor-binding domain that is common to the seven known insulin-like growth factor-binding proteins, the second contains a von Willebrand factor type C module thought to be involved in oligomerization, the third is a thrombospondin type I repeat thought to mediate cell attachment and binding to ECM, and the fourth is a C-terminal domain implicated in heparin binding and dimerization (9). Recently, a novel CTGF-like (CTGF-L) gene, structurally related to CTGF but lacking the C-terminal domain, was identified and cloned from human osteoblasts (24). CTGF-L mRNA was also expressed in fibroblasts, heart,

1086


FIG. 4. Northern analysis reveals differentiation-dependent expression of CTGF mRNA by rat GC in response to gonadotropin treatment in vivo. GC were obtained from the ovaries of immature female rats 48 h after ip injection of eCG (10 IU) or 12 h after ip injection of hCG, given 48 h after injection of 10 IU eCG (eCG⫹hCG). Controls (CON) received no hormone treatment. Northern analysis was performed on total RNA from freshly isolated cells, which was electrophoresed (5 ␮g/lane), blotted, and probed using the CTGF cDNA isolated by DDRT-PCR (see Fig. 2). Blots were stripped and reprobed for 18S rRNA (18S) to control for differences in loading and/or blotting. A, Autoradiogram from a typical experiment showing that the abundantly expressed approximately 2.5-kb CTGF transcript (CTGF) in freshly isolated cells from untreated animals is down-regulated after treatment with eCG and further down-regulated after additional treatment with hCG. B, Composite data from four separate experiments. 䡺, CTGF mRNA intensity relative to 18S rRNA (mean ⫾ SEM) measured by electronic autoradiography. a, Significantly lower (P ⬍ 0.01) than CON; b, significantly lower (P ⬍ 0.01) than both the CON and eCG treatments.

testes, and, most abundantly, ovary, but cell localization in the ovary was not reported. In view of their functional properties in other tissues where they are expressed and their pattern of expression during GC differentiation, it seems highly likely that CTGF, CTGF-L, and possibly other members of the CCN growth factor family serve significant paracrine roles in the ovary. Coexpression of CTGF and LO mRNAs in GC is consistent with connective tissue biosynthesis and ECM deposition occurring hand in hand during follicular development (1, 2). A functional link between the two genes is further suggested by the observation that CTGF can stimulate LO enzyme activity and insoluble collagen formation in human gingival fibroblasts (25). ECM synthesis involves deamination of peptidyl lysine residues to allow covalent inter- or intrachain crosslinking of collagen and elastin fibrils catalyzed by LO (26). The degree of collagen and elastin cross-linking varies between tissues and is constantly altering in the lamina basalis of developing ovarian follicles (1, 27). Here we documented strong LO mRNA expression by GC in preantral and early antral follicles that subsided in response to gonadotropic stimulation in vitro and in vivo. The abundant, approximately 5.2-kb and the rarer, approximately 3.8-kb GC LO mRNA transcripts correspond to similarly sized LO transcripts in rat vascular smooth muscle cells (19, 28). However, the regulation, secretion, and extracellular activities of LO at any cellular site remain poorly understood. LO enzyme activity in rabbit ovarian follicles is reported to be increased after hCGinduced ovulation (29), and LO mRNA expression is up-

Endo • 2001 Vol. 142 • No. 3

FIG. 5. Northern analysis reveals differentiation-dependent expression of LO mRNA by rat GC in response to gonadotropin treatment in vivo. GC were obtained from the ovaries of immature female rats 48 h after ip injection of eCG (10 IU) or 12 h after ip injection of hCG, given 48 h after injection of 10 IU eCG (eCG⫹hCG). Controls (CON) received no hormone treatment. Total RNA from freshly isolated cells was electrophoresed (5 ␮g/lane), blotted, and probed using LO cDNA isolated by DD-RT-PCR (see Fig. 2). Blots were stripped and reprobed for 18S rRNA (18S) to control for differences in loading and/or blotting. A, Autoradiogram from typical experiment showing that the major (⬃5.2-kb) LO transcript (LO) expressed in freshly isolated cells from untreated animals is down-regulated after treatment with eCG and further down-regulated after additional treatment with hCG. B, Composite data from four separate experiments. 䡺, LO mRNA signal intensity relative to 18S rRNA (mean ⫾ SEM) measured by electronic autoradiography. a, Significantly lower (P ⬍ 0.01) than CON; b, significantly lower (P ⬍ 0.01) than both CON and eCG treatments.

regulated at the time of ovulation in perch ovary (30); otherwise, there is no scientific literature on ovarian LO. Therefore, it remains to be determined whether LO of GC origin has a particular role in ECM deposition during folliculogenesis, as predicted by the pattern of LO gene expression observed here. CTGF and LO mRNAs are sequentially repressed in GC under conditions of culture that promote cellular differentiation in vitro. The rapid (3-h) suppression of CTGF mRNA upon exposure to FSH is consistent with the designation of the CTGF gene as immediate-early in fibroblasts, endothelial cells, vascular smooth muscle cells, epithelial cells, chondrocytes, and glioblastoma cells (9, 31). However, CTGF mRNA expression at these sites is typically up-regulated by serum enrichment or exposure to tissue growth factors such as transforming growth factor-␤ (TGF␤), platelet-derived growth factor (PDGF), or basic fibroblast growth factor. It remains to be determined what effects these or any other growth/differentiation factors might exert on GC CTGF expression, but FSH is suppressive. LO mRNA expression in immature GC was also down-regulated by FSH, albeit less abruptly than CTGF. Again, this contrasts with LO expression profiles in other tissues, such as lung fibroblasts (32, 33) and rat vascular smooth muscle cells (34, 35), where PDGF, angiotensin II, or serum enrichment induces LO mRNA. An exception to this rule is interferon-␥, a proinflammatory cytokine present in aneurysm and arteriosclerotic plaque rupture that down-regulates LO gene expression in rat aortic smooth muscle cells (36). Down-regulation by FSH of CTGF and LO gene expression in GC is presumably mediated by cAMP-mediated postre-


1087

FIG. 6. In situ hybridization analysis of CTGF (A–D), LO (E–H), and P450arom (I–L) mRNAs in rat ovary. Ovaries were obtained from immature female rats 48 h after ip injection of eCG (10 IU; B, F, and J) or 12 h after ip injection of hCG, given 48 h after injection of 10 IU eCG (C, G, and K). Controls (A, E, and I) received no hormone treatment. After fixation in 4% paraformaldehyde and embedding in paraffin blocks, sections (10 ␮m) were hybridized with [35S]cRNA probes. Sense cRNA probes were used as a control for nonspecific binding (D, H, and L). Note the abundant CTGF (A) and LO (E), with no expression of P450arom (I) mRNA signals in the GC of preantral/early antral follicles (solid arrows). However, many primary and secondary follicles that express CTGF mRNA do not express LO mRNA (open arrowheads). GC in large follicles induced by treatment with eCG show much reduced CTGF (B) and LO (F) signals, whereas P450arom (J) is enhanced (asterisks). After additional treatment with hCG, the CTGF signal (C) is further reduced, but remains visible in the innermost luteinizing GC layer (arrows), whereas both LO (G) and P450arom disappear. 䡺, 700 ␮M.

ceptor signaling. Elevation of cAMP levels within target cells by a variety of methods blocks the induction of CTGF by TGF␤1 (37, 38). Moreover, TGF␤1-induced LO mRNA expression in human lung fibroblasts is inhibited by PGE2 via increased formation of cAMP (33, 39), and cAMP-mediated differentiation of preadipocytes into adipocytes is associated with down-regulation of LO expression (40). As the differentiation-inducing action of FSH on GC is also cAMP mediated (41), it seems safe to conclude that postreceptor sig-

naling via this route explains the associated down-regulation of CTGF and LO gene expression. However, this remains to be confirmed experimentally. The nonaromatizable androgen DHT directly stimulates GC LO gene expression. A precedent for androgenic regulation of LO expression is the stimulation by testosterone of LO enzymatic activity in bovine aorta smooth muscle cells reported by Bronson et al. (42). Androgen treatment in vivo also reduces elastase expression in rat hearts, probably by

1088


FIG. 7. In situ hybridization analysis of CTGF mRNA in luteinizing GC. Ovaries were obtained from immature female rats 12 h after ip injection of hCG, given 48 h after injection of 10 IU eCG, After fixation in 4% paraformaldehyde and embedding in paraffin blocks, sections (10 ␮m) were hybridized with antisense (A) or sense (B) [35S]cRNA CTGF probes. Note residual CTGF signal in antrally located and cumulus GC (arrows). A strong CTGF signal persists in the GC of the immature follicles in the top right corner (asterisk). 䡺, 100 ␮M.

enhancement of LO activity (43). GC express androgen receptors through which androgen modulates the expression of FSH-inducible genes such as P450arom. It remains to be determined whether DHT induction of GC LO is androgen receptor mediated. The interaction between FSH and androgen also needs to be clarified. The DDRT-PCR screen predicted that DHT would interact with FSH to increase downregulation of CTGF and LO mRNA expression relative to FSH alone, which was not borne out by the in vitro expression studies. A comprehensive study of the relative contributions of FSH, LH, sex steroids, and other regulatory factors on GC CTGF and LO gene expression in vivo and in vitro is therefore currently under way. The pattern of expression of CTGF and LO mRNAs in GC suggests that the encoded proteins might play roles in establishing and/or maintaining early follicular cell phenotypes, similar to the postulated role of CTGF in osteoporosis (17). CTGF mRNA is expressed in many tissues, with kidney (44) and brain (45) previously cited as the sites of highest expression. However, direct comparison of kidney and im-

Endo • 2001 Vol. 142 • No. 3

mature rat ovary by in situ hybridization reveals far higher levels of CTGF mRNA in the latter (Hillier, S. G., et al. unpublished). In the rat ovary, the CTGF gene is switched on at the very earliest stages of follicular development when GC have just begun to proliferate. A similar pattern of expression is observed in pig ovary, where CTGF has been hypothesized to promote ovarian cell growth and blood vessel formation during follicular and luteal development (22). Based on its expression profile in the ovary and biological properties in other tissue systems, it is tempting to suggest that CTGF might contribute to the process of thecal cell recruitment, which is a crucial process in folliculogenesis (46). CTGF is structurally and functionally related to PDGF and vascular endothelial growth factor (47), growth factors that are mitogenic for mesenchymally derived cells in blood, muscle, bone/cartilage, and connective tissue. Thus, a role for CTGF as a thecal cell mitogen is also likely. Thecal cells express PDGF␤ receptor, through which PDGF activates phosphatidylinositol-3-kinase-Akt/protein kinase B and Ras extracellular signal-regulated kinase-1/2 signaling, leading to thecal cell proliferation and enhanced LH-responsive steroidogenesis (48). CTGF reacts with antiserum produced against PDGF and serves as a ligand for the PDGF receptor (8). However, it is unclear which receptors and/or signaling pathways might be activated by CTGF, because a specific CTGF receptor may also exist (49). GC also express PDGF receptors (50), which presumably transduce the mitogenic action of PDGF on these cells (51). Therefore, CTGF could also be an autocrine GC mitogen, particularly during the FSH-independent preantral phase of follicular development when the CTGF gene is most abundantly expressed. It remains to be established whether the proteins encoded by CTGF and LO mRNAs follow the same differentiationrelated pattern of expression in rat GC, although this could be predicted from the parallel expression of mRNA and protein observed during follicular development in the porcine ovary (22). In situ hybridization analyses reveal both genes to be transcriptionally suppressed during preovulatory follicular development induced by eCG, and more so after luteinization induced by hCG. We note that CTGF mRNA remains highly expressed in some antrally located GC for up to 12 h after injection of eCG. Cells in this layer are intimately involved in the formation of the corpus luteum, acting as foci of connective tissue synthesis and neovascularization after ovulation. If, as seems likely (22), biologically active CTGF is produced at this site, an important physiological role in corpus luteum formation is readily envisaged. In summary, CTGF and LO are gonadotropin-regulated genes in GC. mRNAs encoding these proteins are abundantly expressed in immature follicles at levels inversely related to GC differentiation. LO is up-regulated by androgen in vitro. Cytodifferentiation induced by FSH in vitro or by eCG in vivo is associated with rapid down-regulation of both CTGF and LO mRNAs. This pattern of gene expression predicts that both encoded products are likely to have biologically important roles in the regulation of ovarian follicular development and function. Acknowledgments We are grateful to Dr. David Armstrong, Roslin Institute (Edinburgh, UK), for help with the in situ hybridization analysis.

CTGF AND LO GENE EXPRESSION IN GRANULOSA CELLS References 1. Rodgers RJ, van Wezel IL, Irving-Rodgers HF, Lavranos TC, Irvine CM, Krupa M 1999 Roles of extracellular matrix in follicular development. J Reprod Fertil [Suppl] 54:343–352 2. Werb Z, Chin JR 1998 Extracellular matrix remodeling during morphogenesis. Ann NY Acad Sci 857:110 –118 3. Hillier SG 1999 Intragonadal regulation of male and female reproduction. Ann Endocrinol (Paris) 60:111–117 4. Yada H, Hosokawa K, Tajima K, Hasegawa Y, Kotsuji F 1999 Role of ovarian theca and granulosa cell interaction in hormone production and cell growth during the bovine follicular maturation process. Biol Reprod 61:1480 –1486 5. Smith MF, McIntush EW, Ricke WA, Kojima FN, Smith GWJ 1999 Regulation of ovarian extracellular matrix remodelling by metalloproteinases and their tissue inhibitors: effects on follicular development, ovulation and luteal function. Reprod Fertil [Suppl] 54:367–381 6. Jaffe RB 2000 Importance of angiogenesis in reproductive physiology. Semin Perinatol 24:79 – 81 7. Liang PL, Pardee AB 1992 Differential display of eukaryotic messenger RNA by means of the polymerase chain reaction. Science 257:967–971 8. Bradham DM, Igarashi A, Potter RL, Grotendorst GR 1991 Connective tissue growth factor: a cysteine-rich mitogen secreted by human vascular endothelial cells is related to the SRC-induced immediate early gene product CEF-10. J Cell Biol 114:1285–1294 9. Brigstock DR 1999 The connective tissue growth factor/cysteine-rich 61/ nephroblastoma overexpressed (CCN) family. Endocr Rev 20:189 –206 10. Krawetz SA 1994 The origin of lysyl oxidase. Comp Biochem Physiol Biochem Mol Biol 108:117–119 11. Smith-Mungo LI, Kagan HM 1998 Lysyl oxidase: properties, regulation and multiple functions in biology. Matrix Biology 16:387–398 during development and tumor formation. APMIS 107:11–18 12. Hillier SG, de Zwart FA 1982 Androgen/antiandrogen modulation of cyclic AMP-induced steroidogenesis during granulosa cell differentiation in tissue culture. Mol Cell Endocrinol 28:347–361 13. Miele G, MacRae L, McBride D, Manson J, Clinton M 1998 Elimination of false positives generated through PCR re-amplification of differential display cDNA. BioTechniques 25:138 –144 14. Miele G, MacRae L, McBride D, Manson J, Clinton M 1999 Elimination of false positives generated through PCR re-amplification of differential display cDNA. In: McClelland M, Pardee AN (eds) Expression Genetics: Differential Display. Eaton, Natick, pp 433– 444 15. Miele G, Slee R, Manson J, Clinton M 1999 A rapid protocol for the authentication of isolated differential display RT-PCR cDNAs. Prep Biochem Biotechnol 29:245–255 16. Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ 1990 Basic local alignment search tool. J Mol Biol 15:403– 410 17. Xu J, Smock SL, Safadi FF, Rosenzweig AB, Odgren PR, Marks Jr SC, Owen TA, Popoff SN Cloning the full-length cDNA for rat connective tissue growth factor: implications for skeletal development. J Cell Biochem 77:103–115 18. Babic AM, Chen CC, Lau LF 1999 Fisp12/mouse connective tissue growth factor mediates endothelial cell adhesion and migration through integrin ␣v␤3, promotes endothelial cell survival, and induces angiogenesis in vivo. Mol Cell Biol 19:2958 –2966 19. Trackman PC, Pratt AM, Wolanski A, Tang SS, Offner GD, Troxler RF, Kagan HM 1990 Cloning of rat aorta lysyl oxidase cDNA: complete codons and predicted amino acid sequence. Biochemistry 29:4863–70 [published erratum appears in Biochemistry 30:8282, 1991] 20. Ravid K, Smith-Mungo LI, Zhao Z, Thomas KM, Kagan HM 1999 Upregulation of lysyl oxidase in vascular smooth muscle cells by cAMP: role for adenosine receptor activation. J Cell Biochem 75:177–185 21. Kagan HM, Vaccaro CA, Bronson RE, Tang SS, Brody JS 1986 Ultrastructural immunolocalization of lysyl oxidase in vascular connective tissue. J Cell Biol 103:1121–1128 22. Wandji S-A, Gadsby JE, Barber JA, Hammond JM 2000 Messenger ribonucleic acids for MAC25 and connective tissue growth factor (CTGF) are inversely regulated during folliculogenesis and early luteogenesis. Endocrinology 141:2648 –2657 23. Yang DH, Kim HS, Wilson EM, Rosenfeld RG, Oh Y 1998 Identification of glycosylated 38-kDa connective tissue growth factor (IGFBP-related protein 2) and proteolytic fragments in human biological fluids, and up-regulation of IGFBP-rP2 expression by TGF-␤ in Hs578T human breast cancer cells. J Clin Endocrinol Metab 83:2593–2596 24. Kumar S, Hand AT, Connor JR, Dodds RA, Ryan PJ, Trill JJ, Fisher SM, Nuttall ME, Lipshutz DB, Zou C, Hwang SM, Votta BJ, James IE, Rieman DJ, Gowen M, Lee JC 1999 Identification and cloning of a connective tissue growth factor-like cDNA from human osteoblasts encoding a novel regulator of osteoblast functions. J Biol Chem 274:17123–17131 25. Hong HH, Uzel MI, Duan C, Sheff MC, Trackman PC 1999 Regulation of lysyl oxidase, collagen, and connective tissue growth factor by TGF-␤1 and detection in human. Lab Invest 79:1655–1667 26. Kagan HM, Trackman PC 1991 Properties and function of lysyl oxidase. Am J Respir Cell Mol Biol 5:206 –210 27. Oksjoki S, Sallinen S, Vuorio E, Anttila L 1999 Cyclic expression of mRNA

28.

29. 30. 31.

32. 33.

34.

35.

36. 37.

38.

39.

40. 41. 42. 43. 44. 45. 46. 47.

48.

49.

50. 51.

1089

transcripts for connective tissue components in the mouse ovary. Mol Hum Reprod 5:803– 808 Green RS, Lieb ME, Weintraub AS, Gacheru SN, Rosenfield CL, Shah S, Kagan HM, Taubman MB 1995 Identification of lysyl oxidase and other platelet-derived growth factor-inducible genes in vascular smooth muscle cells by differential screening. Lab Invest 73:476 – 482 Himeno N 1986 Effect of prostaglandins on collagen synthesis in rabbit ovarian follicles during the ovulatory process. Nippon Naibunpi Gakkai Zasshi 62:1181–193 Langenau DM, Goetz FW, Roberts SB 1999 The upregulation of messenger ribonucleic acids during 17␣,20␤-dihydroxy-4-pregnen-3-one-induced ovulation in the perch ovary. J Mol Endocrinol 23:137–152 Iyer VR, Eisen MB, Ross DT, Schuler G, Moore T, Lee JCF, Trent JM, Staudt LM, Hudson J Jr, Boguski MS, Lashkari D, Shalon D, Botstein D, Brown PO 1999 The transcriptional program in the response of human fibroblasts to serum. Science 283:83– 87 Boak AM, Roy R, Berk J, Taylor L, Polgar P, Goldstein RH, Kagan HM 1994 Regulation of lysyl oxidase expression in lung fibroblasts by transforming growth factor-␤1 and prostaglandin E2. Am J Respir Cell Mol Biol 11:751–755 Roy R, Polgar P, Wang Y, Goldstein RH, Taylor L, Kagan HM 1996 Regulation of lysyl oxidase and cyclooxygenase expression in human lung fibroblasts: interactions among TGF-␤, IL-1␤, and prostaglandin E. J Cell Biochem 62:411– 417 Gacheru SN, Thomas KM, Murray SA, Csiszar K, Smith-Mungo LI, Kagan HM 1997 Transcriptional and post-transcriptional control of lysyl oxidase expression in vascular smooth muscle cells: effects of TGF-␤1 and serum deprivation. J Cell Biochem 65:395– 407 Shanley CJ, Gharaee-Kermani M, Sarkar R, Welling TH, Kriegel A, Ford JW, Stanley JC, Phan SH 1997 Transforming growth factor-␤1 increases lysyl oxidase enzyme activity and mRNA in rat aortic smooth muscle cells. J Vasc Surg 25:446 – 452 Song YL, Ford JW, Gordon D, Shanley CJ 2000 Regulation of lysyl oxidase by interferon-␥ in rat aortic smooth muscle cells. Arterioscler Thromb Vasc Biol 20:982–988 Kothapalli D, Hayashi N, Grotendorst GR 1998 Inhibition of TGF-␤-stimulated CTGF gene expression and anchorage-independent growth by cAMP identifies a CTGF-dependent restriction point in the cell cycle. FASEB J 12:1151–1161 Duncan MR, Frazier KS, Abramson S, Williams S, Klapper H, Huang X, Grotendorst GR 1999 Connective tissue growth factor mediates transforming growth factor ␤-induced collagen synthesis: down-regulation by cAMP. FASEB J 13:1774 –1786 Choung J, Taylor L, Thomas K, Zhou X, Kagan H, Yang X, Polgar P 1998 Role of EP2 receptors and cAMP in prostaglandin E2 regulated expression of type I collagen alpha1, lysyl oxidase, and cyclooxygenase-1 genes in human embryo lung fibroblasts. J Cell Biochem 71:254 –263 Dimaculangan DD, Chawla A, Boak A, Kagan HM, Lazar MA 1994 Retinoic acid prevents downregulation of ras recision gene/lysyl oxidase early in adipocyte differentiation. Differentiation 58:47–52 Richards JS, Russell DL, Robker RL, Dajee M, Alliston TN 1998 Molecular mechanisms of ovulation and luteinization. Mol Cell Endocrinol 145:47–54 Bronson RE, Calaman SD, Traish AM, Kagan HM 1987 Stimulation of lysyl oxidase (EC 1.4.3.13) activity by testosterone and characterization of androgen receptors in cultured calf aorta smooth-muscle cells. Biochem J 244:317–323 LeGros T, McConnell D, Murry T, Edavettal M, Racey-Burns LA, Shepherd RE, Burns AH 2000 The effects of 17␣-methyltestosterone on myocardial function in vitro. Med Sci Sports Exerc 32:897–903 Ryseck RP, Macdonald-Bravo H, Mattei MG, Bravo R 1991 Structure, mapping, and expression of fisp-12, a growth factor-inducible gene encoding a secreted cysteine-rich protein. Cell Growth Differ 2:225–2233 Oemar BS, Luscher TF 1997 Connective tissue growth factor. Friend or foe? Arterioscler Thromb Vasc Biol 17:1483–9 Parrott JA, Skinner MK 2000 Kit ligand actions on ovarian stromal cells: effects on theca cell recruitment and steroid production. Mol Reprod Dev 55:55– 64 Tischer E, Gospodarowicz D, Mitchell R, Silva M, Schilling J, Lau K, Crisp T, Fiddes JC, Abraham JA 1989 Vascular endothelial growth factor: a new member of the platelet-derived growth factor gene family. Biochem Biophys Res Commun 165:1198 –1206 Taylor CC 2000 Platelet-derived growth factor activates porcine thecal cell phosphatidylinositol-3-kinase-Akt/PKB and ras-extracellular signal-regulated kinase-1/2 kinase signaling pathways via the platelet-derived growth factor-␤ receptor. Endocrinology 141:1545–1553 Nishida T, Nakanishi T, Shimo T, Asano M, Hattori T, Tamatani T, Tezuka K, Takigawa M 1998 Demonstration of receptors specific for connective tissue growth factor on a human chondrocytic cell line (HCS-2/8). Biochem Biophys Res Commun 247:905–909 McWilliam R, Leake RE, Coutts JR 1995 Growth factors in human ovarian follicle fluid and growth factor receptors in granulosa-luteal cells. Int J Biol Markers 10:216 –220 May JV, Frost JP, Bridge AJ 1990 Regulation of granulosa cell proliferation: facilitative roles of platelet-derived growth factor and low density lipoprotein. Endocrinology 126:2896 –2905