Heparan Sulfate Proteoglycans Regulate Responses to Oocyte ...

REPRODUCTION-DEVELOPMENT

Heparan Sulfate Proteoglycans Regulate Responses to Oocyte Paracrine Signals in Ovarian Follicle Morphogenesis Laura N. Watson, David G. Mottershead, Kylie R. Dunning, Rebecca L. Robker, Robert B. Gilchrist, and Darryl L. Russell Robinson Institute, School of Paediatrics and Reproductive Health, The University of Adelaide, Adelaide 5005, Australia

In the ovarian follicle, oocyte-secreted factors induce cumulus-specific genes and repress mural granulosa cell specific genes to establish these functionally distinct cell lineages. The mechanism establishing this precise morphogenic pattern of oocyte signaling within the follicle is unknown. The present study investigated a role for heparan sulphate proteoglycans (HSPG) as coreceptors mediating oocyte secreted factor signaling. In vitro maturation of cumulus oocyte complexes in the presence of exogenous heparin, which antagonizes HSPG signaling, prevented cumulus expansion and blocked the induction of cumulus-specific matrix genes, Has2 and Tnfaip6, whereas conversely, the mural granulosa-specific genes, Lhcgr and Cyp11a1, were strongly up-regulated. Heparin also blocked phosphorylation of SMAD2. Exogenous growth differentiation factor (GDF)-9 reversed these heparin effects; furthermore, GDF9 strongly bound to heparin sepharose. These observations indicate that heparin binds endogenous GDF9 and disrupts interaction with heparan sulphate proteoglycan coreceptor(s), important for GDF9 signaling. The expression of candidate HSPG coreceptors, Syndecan 1– 4, Glypican 1– 6, and Betaglycan, was examined. An ovulatory dose of human chorionic gonadotropin down-regulated Betaglycan in cumulus cells, and this regulation required GDF9 activity; conversely, Betaglycan was significantly increased in luteinizing mural granulosa cells. Human chorionic gonadotropin caused very strong induction of Syndecan 1 and Syndecan 4 in mural granulosa as well as cumulus cells. Glypican 1 was selectively induced in cumulus cells, and this expression appeared dependent on GDF9 action. These data suggest that HSPG play an essential role in GDF9 signaling and are involved in the patterning of oocyte signaling and cumulus cell function in the periovulatory follicle. (Endocrinology 153: 4544 – 4555, 2012)

O

varian follicles coordinately respond to maternal endocrine and oocyte paracrine signals to mediate the maturation of oocytes and cyclic production of female reproductive hormones. The follicle comprises layers of somatic cells that proliferate and differentiate forming two phenotypically and functionally different cell populations: cumulus cells, which envelope the oocyte forming the cumulus oocyte complex (COC), and mural granulosa cells (GC), which line the follicle wall. These two cell types play distinct roles in the function of the follicle, together providing a specialized microenvironment in which the

oocyte matures. Throughout folliculogenesis cumulus cells regulate the oocyte’s microenvironment and energy supply via expression of genes required for carbohydrate metabolism and cholesterol biosynthesis. Mural granulosa cells participate in endocrine communication through expression of genes encoding steroidogenic enzymes and the LH receptor. Mural granulosa cells thus sense the LH surge and respond by activating ovulation and oocyte maturation cascades. Epidermal growth factor-like ligands released by granulosa cells trigger cumulus cell induction of specific cumulus matrix genes and oocyte meiotic resump-

ISSN Print 0013-7227 ISSN Online 1945-7170 Printed in U.S.A. Copyright © 2012 by The Endocrine Society doi: 10.1210/en.2012-1181 Received February 15, 2012. Accepted June 12, 2012. First Published Online July 3, 2012

Abbreviations: Alk, Activin receptor-like kinase; BMP, bone morphogenetic protein; CEI, cumulus expansion index; COC, cumulus oocyte complex; eCG, equine chorionic gonadotropin; ECM, extracellular matrix; EGF, epidermal growth factor; FGF, fibroblast growth factor; GC, granulosa cell; GDF, growth differentiation factor; hCG, human chorionic gonadotropin; HSPG, heparan sulfate proteoglycan; IVM, in vitro maturation; OSF, oocytesecreted factor; STAT, signal transducer and activator of transcription.

4544

endo.endojournals.org

Endocrinology, September 2012, 153(9):4544 – 4555


tion (1). Appropriate development of these functionally distinct somatic cell compartments is essential to healthy oocyte development and subsequent maturation and ovulation (2). The divergence of cumulus cell and mural granulosa cell lineages, which occurs during early antral follicle development, and the subsequent expansion or mucification of the cumulus cells, which occurs after the LH-surge, both require oocyte signaling via oocyte-secreted paracrine factors (3, 4). Growth factors of the TGF␤ [growth differentiation factor (GDF)-9 and bone morphogenetic protein (BMP)15] and the fibroblast growth factor (FGF)-8B superfamily produced by oocytes promote the cumulus cell lineage only in somatic cells most closely surrounding the oocyte (5–7). Spatial patterning of oocyte secreted factor signals is observed through restricted cumulus cell expression of genes induced by oocytes (8 –10), whereas mural cell-specific genes such as those for LH receptor and steroidogenic enzymes as well as several others are repressed by oocytes, thereby restricting their expression to the mural granulosa layers (11, 12). How the partitioning of responses in cumulus and mural cells is achieved is unknown. The proximity to the oocyte ensures the most intense concentration of oocyte derived ligands occur within the adjacent cumulus cell layers, but because mural granulosa cells do readily respond when exposed to oocyte-secreted factors (OSF) (13, 14), a more active mechanism may be required to establish the distinct pattern of responsiveness to oocyte signals between the different ovarian somatic cell compartments. One such mechanism may involve localized sequestering of oocyte secreted ligands via specific molecular interactions. Ligands of the TGF␤ and FGF superfamily commonly require interaction with heparan sulfate proteoglycan (HSPG) coreceptors. HSPG are known to maintain morphogenic gradients in developmental processes [reviewed by Hacker et al. (15)]. In Drosophila the TGF␤ homolog decapentaplegic requires the HSPG, dally and dly, to give organized spatial signaling patterns (16). Mammalian TGF␤ isoforms 1 and 2 (17) and activin (18) bind heparan sulfate, and BMP4 is sequestered to cell surfaces by interaction with HSPG (19). HSPG consist of a protein core and one or more covalently attached glycosaminoglycan chains. The major families of HSPG are the syndecans and the glypicans. These are cell surface proteoglycans that are involved in numerous biological processes, including growth factor regulation, adhesion, cell proliferation, and differentiation (20). Distinct lengths and sulfation patterns of oligosaccharide side chains determine the specificity of interaction of the HSPG (21–23). Cell surface HSPG can act to sequester a ligand at its site of action, thereby increasing its local con-


4545

centration, or by presenting a ligand to its signaling receptor, acting as a coreceptor. Alternatively, HSPG can prevent ligand-receptor interaction through competitive binding or by mediating protein internalization by endocytosis [reviewed by Bernfield et al. (20)]. The presence of HSPG therefore has a profound influence on cell differentiation and patterning of responses to these signals. Heparin is a glycosaminoglycan with qualitatively similar structure to heparan sulfates, the general distinction being in the degree and pattern of sulfation of the polysaccharide chains. Soluble heparin can either mimic endogenous HSPG action promoting growth factor dimerization and receptor interaction (24) or act as a competitive antagonist of endogenous HSPG action (17, 22, 25). Whether GDF-9 and BMP15 interact directly with HSPG has not been determined. However, heparin is known to block cumulus expansion in cultured COC, without blocking oocyte meiotic resumption (26, 27), and that recombinant GDF9 and BMP15 protein are released from human embryonic kidney-293 cell surfaces by the addition of heparin (28, 29), suggesting these ligands are anchored to cell surfaces by heparan sulfate interactions. We hypothesized that HSPG binding may participate in patterning the response to oocyte derived signals in the follicle through sequestering ligands in specific spatial and temporal patterns in the ovarian follicle. Expansion of the COC in response to the LH surge is dependent on the action of oocyte-derived signals, particularly GDF9, and this physiological process provides a sensitive system for investigating oocyte-cumulus cell signaling interactions. Here we demonstrate that GDF9 signaling in cumulus cells is disrupted by heparin treatment and restored when exogenous mature GDF9 is added concomitantly with heparin. This indicates that interaction with a HSPG is required for normal endogenous GDF9 action within the maturing COC. We identified appreciable expression and distinct regulation patterns of all members of the syndecan and glypican families of HSPG. Glypican 1 mRNA and protein were selectively induced in COC by both in vivo and in vitro oocyte maturation stimuli. During in vitro maturation (IVM), oocyte signaling is deficient (30, 31), and we found that Syndecan 1, Syndecan 4, and Glypican 1 are dysregulated in IVM. Together our data suggest that GDF9 signaling mediating cumulus expansion is dependent on the interaction with HSPG coreceptors that may sequester ligand and/or facilitate receptor-ligand interaction within the COC.

Materials and Methods Materials Equine chorionic gonadotropin (eCG) was purchased from Organon, Australia (Sydney, New South Wales, Australia). Hu-

4546

Watson et al.

HSPG Regulated Oocyte Secreted Factor Signaling

man chorionic gonadotropin (hCG) was purchased from Calbiochem, Australia (Alexandrina, New South Wales, Australia). FSH was purchased from Dr. A. F. Parlow (National Hormone and Peptide Program, National Institutes of Diabetes and Digestive and Kidney Diseases, Torrance, CA), and epidermal growth factor (EGF) was purchased from Sigma Aldrich (Sydney, New South Wales, Australia). Culture media were purchased from Gibco, Invitrogen (Sydney, New South Wales, Australia). Unless otherwise stated, all reagents were purchased from Sigma Aldrich.

Isolation and culture of mouse COC and granulosa cells All mice (F1 C57BL/6 ⫻ CBA) were maintained on a 12-h day, 12-h night cycle with rodent chow and water provided ad libitum. All experiments were approved by the University of Adelaide’s Animal Ethics Committee and were conducted in accordance with the Australian Code of Practice for the Care and Use of Animals for Scientific Purposes. For COC and granulosa cell collections, mice were injected ip at 22–25 d of age with 5 IU eCG. For IVM, COC were isolated from preovulatory follicles 44 h after eCG by puncturing large antral follicles with a 30gauge needle and washed in complete ␣MEM (supplemented with 2% fetal calf serum, 100 U/ml penicillin G, 100 mg/ml streptomycin sulfate, and 1 ␮M cilostamide). COC were cultured in drops of complete ␣MEM supplemented with 50 mIU/ml FSH, 10 ng/ml EGF, and 0.23 mM pyruvate overlayed with sterile mineral oil for 6, 8, 12, or 16 h at 37 C in 5% CO2-95% air. In the indicated experiments, IVM cultures used either the above control conditions, or included heparin (300 ␮g/ml; Sigma catalog no. H3149) or heparin ⫹ GDF9 (R&D Systems, Minneapolis, MN; catalog no. 739-G9, 500 ng/ml). In vivo matured COC and granulosa cells were collected from mice injected ip with 5 IU eCG and then 44 h later with 5 IU hCG. COC and GC were isolated by follicle puncture at 0 or 8 h after hCG. After 16 h COC were collected from oviducts by blunt dissection. Corresponding GC were collected by repeated puncturing of large follicles to express cells that were washed in fresh serum-free media and pelleted and then snap frozen for subsequent RNA or protein extraction. For tissue localization of proteins, ovaries were collected at the indicated times after eCG or hCG stimulation. Detailed immunohistochemistry and Western blot methods are described in Supplemental Methods, published on The Endocrine Society’s Journals Online web site at http://endo.endojournals.org.

Real-time RT-PCR RNA was extracted from COC and granulosa cells using Trizol (Invitrogen) as per the manufacturer’s instructions with the inclusion of 7.5 ␮g of glycoblue (Ambion Inc., Austin, TX) during precipitation overnight at ⫺20 C. Total RNA was then treated with 1 U of deoxyribonuclease (Ambion) as per the manufacturer’s instructions to remove any contaminating DNA. First-strand cDNA was synthesized from total RNA using random hexamer primers (Geneworks, Hindmarsh, South Australia) and Superscript III reverse transcriptase (Invitrogen Australia Pty. Ltd., (Sydney, New South Wales, Australia)). Genespecific primers for Syndecan 1 and Glypican 1 real-time PCR were designed against published mRNA sequences (National Center for Biotechnology Information PubMed database) and


synthesized by Sigma Genosys (Sigma Aldrich) (Supplemental Table 1). Real-time PCR was performed in triplicate for each sample on a Corbett Rotor Gene 6000 (QIAGEN, Valencia, CA). Each reaction included 10 ng of total RNA, 0.25 ␮M each primer, 10 ␮l of SYBR green master mix, and H2O to a final volume of 20 ␮l. PCR cycling conditions were 50 C for 2 min and 95 C for 10 min, followed by 40 amplification cycles of 95 C for 15 sec and 60 C for 1 min. Taqman gene expression assays were purchased from Applied Biosystems (Foster City, CA) (Supplemental Table 1), and reactions were run in triplicate on an AB7900HT PCR system (Applied Biosystems) using the manufacturer’s specified amplification settings. Each reaction used 0.5 ␮l of Taqman assay, 5 ␮l of Taqman master mix, 5 ng cDNA, and H2O to a final volume of 10 ␮l. For all real-time PCR, controls included omission of the cDNA template or reverse transcriptase enzyme in otherwise complete reactions. Gene expression was normalized to the expression Rpl19 housekeeping gene control. All values were then expressed relative to calibrator samples using the 2-(⌬⌬CT) method. After real-time PCR, analysis of dissociation curves confirmed that a single product was amplified in all reactions.

Comparison of global gene expression changes in heparin vs. SB431542-treated COC To further confirm that heparin broadly blocks GDF9 signaling consequences, we performed a microarray analysis of the global gene expression changes in COC caused by heparin treatment compared with those caused by blocking GDF9 signaling with the Alk4/5/7 inhibitor compound SB431542 (Sigma Aldrich), which blocks SMAD-2/3 phosphorylation. Isolated COC from 10 eCG 44 h-treated mice were distributed randomly into IVM cultures (80 –100 COC per treatment) either with standard conditions (control; ␣MEM ⫹ 50 mIU/ml of FSH, 10 ng/ml of EGF) or standard conditions plus 300 ␮g/ml heparin or with 2 ␮M SB431542. After 12 h culture, RNA was isolated from COC by Trizol (Invitrogen) extraction, and RNA integrity was verified and quantity determined on an Agilent bioanalyzer with RNA pico chips (Agilent Technologies, Forrest Hill, Victoria, Australia). Samples from four independent replicate experiments were labeled and hybridized to Affymetrix Gene 1.0 microarrays (Santa Clara, CA) and washed and scanned according to the manufacturer’s instructions in the University of Adelaide Microarray Centre. Statistical analysis of the differential gene expression was compared among the three groups using the Partek Genomics Suite, version 6.4 (Partek Inc., St. Louis, MO). An ANOVA was applied with post hoc step-up adjustment for false discovery due to multiple measures. Genes accepted as significantly regulated by the treatment were those with greater than 2-fold difference from control and P ⬍ 0.05 after ANOVA.

Heparin binding of recombinant GDF9 Expression cassettes encoding human and mouse GDF9 DNA sequence were synthesized (Genscript USA, Inc., Piscataway, NJ) incorporating the rat serum albumin signal sequence at the 5⬘ end followed by a His8 tag and a Strep II epitope tag at the N terminus of the GDF9 pro-region, which was transferred to the pEF-IRES expression vector. Stable transfected human embryonic kidney293T cell lines producing either mouse or human pro-GDF9 were established, and the serum-free conditioned media were subjected to immobilized Ni2⫹-based ion-affinity purification


targeting the His-tag as described previously (28, 29). The purified recombinant mouse or human GDF9 was incubated with 20 ␮l of heparin-sepharose resin (GE Healthcare, Rydalmere, New South Wales, Australia) and mixed by rotation for 1 h at 4 C. The flowthrough was collected and the resin washed with successive 100 ␮l washes of NaCl at concentrations of 0.25 M (⫻2), 0.5 M (⫻2), 1.0 M (⫻2) and 2.0 M, in Na phosphate buffer (10 mM Na phosphate; 150 mM NaCl, pH 7.4). Aliquots from each wash and elution were separated on reducing PAGE for Western blot using mAb-Gdf9 –53 (Serotec, Oxford, UK; catalog no. MCA5635GA) primary antibody at 3.4 ng/ml and detection with the Odyssey infrared imaging system (Licor Bioscience, Lincoln, NE). Detailed Western blot methods are described in the Supplemental Methods.

Statistical analyses All experiments were repeated at least three times using COC pooled from 10 –20 mice and mean results from three independent experimental replicates subjected to statistical analysis. Experiments were analyzed for statistical significance using GraphPad Prism software version 5.01 (GraphPad, San Diego, CA), using one- or two-way ANOVA with Tukey’s post hoc test as indicated in Results and figure legends. A P ⬍ 0.05 was considered statistically significant.

Results Inhibition of cumulus expansion by heparin is restored by GDF9 treatment A robust signal from the oocyte is required to maintain the cumulus lineage gene expression profile including induction of matrix genes required for COC expansion and repression of mural granulosa cell genes. To test whether the action of OSF during COC expansion is modulated by interaction between the ligands and HSPG, we conducted IVM of COC in the presence of exogenous heparin. After 16 h IVM in the presence of FSH (50 mIU/ml) and EGF (10 ng/ml), the cumulus expansion index (CEI) scored by a blinded observer was reduced in a dose-dependent fashion


4547

by treatment with heparin (Supplemental Fig. 1, A and B). In repeated experiments using COC treated with heparin at 300 ␮g/ml, the CEI was significantly (P ⬍ 0.001) reduced by heparin treatment (Fig. 1), whereas cotreatment with heparin and GDF9 at 500 ng/ml significantly reversed the effect of heparin alone and was able to restore cumulus expansion to control levels (Fig. 1, C and D). Heparin blocks cumulus-specific gene expression and signal transduction mediated by GDF9 To further determine whether the effect of heparin in blocking COC expansion occurs via a disruption of normal oocyte secreted factor signaling, we tested cumulus and mural granulosa cell-specific gene expression in heparin- or heparin ⫹ GDF9-treated COC. The levels of two cumulus matrix markers normally induced by GDF9, Has2 and Tnfaip6, were strongly reduced by 91 and 76%, respectively, by 300 ␮g/ml heparin treatment during 12 h IVM culture (Fig. 2, A and B). Conversely, two granulosa cell specific markers normally repressed by oocytes or GDF9; Cyp11a1 and Lhcgr, were significantly increased in COC matured in the presence of heparin by 20-fold and 94-fold, respectively (Fig. 2 C, and D). In each case the expression of cumulus and mural granulosa specific genes was restored to control levels by opposing heparin with GDF9 500 ng/ml cotreatment. These alterations in cumulus and granulosa cell marker gene expression showed a dose-dependent sensitivity to heparin; all four genes examined were affected more than 3-fold by heparin concentration as low as 30 ␮g/ml (Supplemental Fig. 1, C–F). GDF9 signaling is transduced through phosphorylation of SMAD2/3 transcription factors. We assessed SMAD2 phosphorylation to further verify that GDF9 signaling was disrupted in heparin-treated COC. Western blot of protein extracts from COC matured in the presence of heparin for 6 h showed a dramatic reduction in phos-

FIG. 1. Inhibition of cumulus expansion by heparin is restored by GDF9 treatment. Cumulus oocyte complexes were cultured under in vitro maturation conditions (␣MEM ⫹ 2% fetal calf serum, 10 mIU/ml FSH, 10 ng/ml Egf) without or with exogenous heparin (300 ␮g/ml) or heparin ⫹ Gdf9 (500 ng/ml). A–C, Representative images of control (A), heparin-treated (B), and heparin ⫹ GDF9 (C)-treated COC after 12 h culture. CEI was scored by a trained observer blinded to treatments. Mean ⫾ SEM of CEI was from three independent experiments (D). Values indicated by different characters are significantly different P ⬍ 0.001 by one-way ANOVA and Tukey’s post hoc test.

4548

Watson et al.



FIG. 2. Heparin blocks cumulus-specific gene expression and signal transduction mediated by GDF9. A–D, Expression levels of cumulus-specific Has2 (A) and Tnfaip6 (B) or mural granulosa cell specific; Cyp11a1 (C) and Lhcgr (D) transcripts were evaluated after 12 h culture in no treatment (control), heparin (300 ␮g/ml), or heparin ⫹ GDF9 (500 ng/ml). Expression levels are normalized to rpL19 mRNA loading control assayed in parallel and expressed relative to expression in a pool of whole-ovary calibrator RNA sample. Values indicated by asterisks are significantly different (P ⬍ 0.05) by using one-way ANOVA and Tukey’s post hoc test. E, Phosphorylated SMAD2 abundance in extracts of COC after 6 h culture in the indicated treatments. Hep, Heparin. Equivalent protein content was loaded in each lane, and ␤-actin was evaluated to confirm equal sample loading. F, Venn diagram illustrating the number of gene transcripts identified in microarray analysis with significantly up- or down-regulated after heparin or SB431542 treatment compared with controls. Numbers indicate the number of transcripts with significantly up- or down-regulated expression level in pairwise comparisons against controls. Overlapping areas represent the number of transcripts whose levels are commonly altered by both treatments. Genes with significantly altered steady-state expression were defined by P ⬍ 0.05 after ANOVA with step-up adjustment and greater than 2-fold change in expression level (n ⫽ 4 independent experiments). Mwt, Molecular weight.

pho-SMAD2 abundance, which was completely reversed with the addition of exogenous GDF9 (Fig. 2E), suggesting that heparin directly interfered with GDF9 receptor activation and signal transduction. To further confirm that heparin fully antagonizes GDF9 action in cumulus cells, we compared global gene

expression by microarray analysis in COC from control IVM or IVM treated with heparin (300 ␮g/ml) or the activin receptor-like kinase (Alk)-4/5/7 inhibitor SB431542 (2 ␮M). As expected, these two treatments both significantly reduced cumulus expansion compared with controls (CEI 3.57 vs. 2.50 or 2.65 for control, heparin, or SB431542 treatment, respectively, P ⬍ 0.01). From 35,557 unique probes detectable in the arrays, we found that SB431542 altered expression of 398 genes, and of these, 300 genes (75%) were similarly altered by heparin. This relationship between treatments was equally evident in the up- and down-regulated gene profiles after heparin or SB431542 treatment (Fig. 2F). The striking overlap between genes affected by Alk4/5/7 inhibition and heparin demonstrates the conserved action between the two treatments. Heparin significantly altered expression of 629 genes of which 300 (48%) responded similarly after Alk4/5/7 inhibition, indicating that GDF9 signaling is a major pathway but perhaps not the only pathway affected by heparin. The additional effects of heparin compared with SB41542 were likely caused by the high dose of heparin chosen for the present studies, whereas the 2-␮M dose of SB431542 used is the minimum dose that fully blocks oocyte signaling (32, 33). Consistent with our conventional quantitative PCR results, the arrays identified Has2, Tnfaip6, and Glypican 1 as genes significantly down-regulated and Cyp11a1, Has2, and Betaglycan as genes significantly up-regulated by both treatments.

GDF9 binds to heparin with high affinity To confirm that GDF9 directly interacts with heparin, we tested the ability of immobilized heparin-sepharose to bind recombinant pro-mature human and mouse GDF9. Both mouse and human recombinant GDF9 preparations contained the approximately 60 kDa pro-mature GDF9 complex as well as the approximately 17-kDa mature peptide as detected


by Western blot targeting the mature peptide after reducing gel electrophoresis (Fig. 3, lane 2). When passed over a column of heparin-sepharose, both forms of GDF9 were depleted from the flowthrough solution. Washing the column with low ionic strength PBS did not remove the GDF9 from the heparin. Elution with 0.25 or 0.5 M NaCl eluted very little bound GDF9 from the column, whereas elution with 1 M NaCl removed approximately 90% of the GDF9 from the heparin sepharose and the remainder was eluted with 2 M NaCl (Fig. 3), indicative of a high-strength ionic interaction between both human and mouse GDF9 and heparin-sepharose similar to other growth factors (34). Proteoglycan expression in the COC and granulosa cells during oocyte maturation To determine which HSPG may participate in mediating GDF9 signaling in the follicle, we characterized the expression of the major HSPG families in COC and granulosa cells. We also compared HSPG regulation in COC throughout oocyte maturation in vivo vs. in vitro (IVM) (Fig. 4). Expression of the syndecans and Betaglycan did not differ markedly between cumulus and mural granulosa cells. Syndecan 2 and Syndecan 3 showed little notable change in expression during the periovulatory period post-hCG treatment, whereas Syndecan 1 and Syndecan 4

FIG. 3. Mouse and human GDF9 bind to heparin with high affinity. IMAC-purified murine (upper panel) or human (lower panel) recombinant GDF9 condition media were incubated with heparin sepharose then eluted in stepwise increasing NaCl concentrations as described in Materials and Methods. Samples were separated on reducing SDS-PAGE gel and GDF9 was detected by Western blot (described in Supplemental Methods) probing with mAb-53 directed against the mature peptide. M, Molecular weight marker (sizes in kilodaltons indicated on the left). SM, Starting material; FT, flowthrough. Arrowheads represent GDF9 (bands at ⬃60 kDa are the intact promature protein; bands at 17 kDa are the processed mature monomer). The lower molecular band is likely a degradation product.


4549

were strongly induced in both COC and GC during oocyte maturation. Syndecan 1 and Syndecan 4 showed comparable induction in IVM vs. in vivo COC up to 8 h after oocyte maturation stimulus, but after 16 h, both had significantly reduced expression in IVM (25.6- vs. 10.93- and 15.34- vs. 9.39-fold change, respectively) compared with in vivo. Betaglycan was the only HSPG gene to show a significant down-regulation in COC (4-fold, P ⬍ 0.05) and was inversely up-regulated in granulosa cells during the periovulatory period in vivo (Fig. 4). All glypicans except Glypican 6 were more highly expressed in the mural granulosa cell compartment of preovulatory follicles before hCG stimulation (Fig. 4). Glypican 5 was detectable only in the granulosa cell compartment and was the only glypican significantly induced in the granulosa compartment after hCG, being 9- and 18-fold higher 8 and 16 h after hCG, respectively (Fig. 4). Glypican 1 mRNA increased in COC in vivo and in vitro after oocyte maturation stimuli but was not significantly induced in mural granulosa cells. Contrary to the pattern observed in IVM for Syndecan 1 and Syndecan 4, Glypican 1 was more highly up-regulated in COC matured in vitro compared with those matured in vivo (21.2- vs. 13.47-fold induction, respectively, after 16 h maturation). Protein abundance and immunolocalization of Syndecan 1 and Glypican 1 in granulosa and COC We further examined protein abundance and localization of SYNDECAN 1 and GLYPICAN 1 by immunohistochemistry in sections of mouse ovary and oviduct. Negative controls using a nonspecific IgG in place of primary antibody resulted in no visible immunostaining (data not shown). Staining for both SYNDECAN 1 and GLYPICAN 1 protein was absent from the mural granulosa layers both 44 h after treatment with eCG and 8 h after treatment with hCG, but GLYPICAN 1 was detected in some cells of the ovarian stroma (Fig. 5A). By contrast, SYNDECAN 1 was evident on all cumulus cell membranes but less abundant in granulosa cells in ovaries 44 h after eCG and was increased in abundance in COC in follicles after 8 h hCG administration (Fig. 5A). After 16 h hCG treatment ovulated COC in oviducts possessed intense SYNDECAN 1 immunostaining on cumulus cell surfaces (Fig. 5A), matching the mRNA expression pattern (Fig. 4). GLYPICAN 1 immunostaining was detectable on cumulus and antral lining granulosa cells 44 h after treatment with pregnant mare serum gonadotropin (Fig. 5A). More intense staining was evident in COC after 8 h hCG, and COC in the oviduct after 16 h hCG had strong GLYPICAN 1 immunostaining on cumulus cell surfaces as well as in the expanded COC matrix (Fig. 5A).

4550

Watson et al.



FIG. 4. Regulated expression of proteoglycans in the COC and mural granulosa cells during oocyte maturation. Expression of Syndecan 1– 4, Betaglycan, and Glypican 1– 6 was evaluated in mural granulosa cells (filled triangles) or COC (filled squares) and at the indicated times after hCG treatment in vivo or after EGF (10 ng/ml) and FSH (10 mIU/ml) treatment in IVM culture (open squares). Left-hand bar graphs show relative mRNA abundance in mural granulosa cells (filled bars) compared with cumulus cells (open bars) in follicles at eCG 44 h treatment before administration of hCG. Line graphs show fold change in expression of each gene relative to the 0 h control. Each specific gene was normalized to rpL19 mRNA as a loading control. Values indicated by different characters are significantly different (P ⬍ 0.05) from the 0 h control using two-way ANOVA and Tukey’s post hoc test. ND, Not detectable.

To further verify the difference in Glypican 1 gene expression between IVM and in vivo matured COC, GLYPICAN 1 protein abundance was analyzed by Western blot. Equivalent amounts of protein extract from IVM

or in vivo-matured COC or mural granulosa cells were loaded in each lane, and equal loading was verified by ␤-actin staining (Fig. 5B). Consistent with the expression of mRNA, GLYPICAN 1 protein (64 kDa) increased



4551

To confirm the localization of GLYPICAN 1 in the expanded matrix of COC after hCG treatment, we isolated COC from oviducts after 16 h hCG and used hyaluronidase to solubilize the COC matrix. The cell pellet and extracellular matrix (ECM) fraction were then assessed by Western blot. Proteins known to be either intracellular [␤-actin and signal transducer and activator of transcription (STAT)-3] or ECM localized (VERSICAN) were examined to verify effective separation of cell and ECM fractions. As expected, ␤-actin and STAT3 were almost entirely found in the cellular compartment, whereas VERSICAN was exclusively in the ECM (Fig. 5C). GLYPICAN 1 was localized entirely to the ECM fraction, indicating that GLYPICAN 1 protein is associated with the ECM of mature COC. GDF9 regulates proteoglycan gene expression in COC To determine whether the regulated pattern of proteoglycan expression in COC is under the control of oocyte signaling factors, we assessed expression in the presence of heparin or heparin ⫹ GDF9 of selected HSPG that previously FIG. 5. Protein abundance and immunolocalization of SYNDECAN 1 and GLYPICAN 1 in exhibited temporal regulation in COC. granulosa and COC. A, Localization of SYNDECAN 1 (upper panels) and GLYPICAN 1 (lower Glypican 1 showed a trend toward sigpanels) in ovarian follicle sections from mice after eCG ⫹ hCG 0 h, 8 h hCG, or sections of COC in oviducts after 16 h hCG stimulation as indicated. B, Western blot for GLYPICAN 1 in nificant 2-fold reduction after 12 h 10 ␮g protein extract from COC, or mural granulosa cells after eCG 44 h ⫹ 0 h or 16 h hCG, IVM culture with heparin and restoraor COC matured in vitro by FSH (10 mIU/ml) ⫹ EGF (10 ng/ml) stimulation as indicated. tion to control expression levels by hepEquivalent sample loading was confirmed by ␤-actin analysis of the same samples. Intensity of arin ⫹ GDF9 (P ⫽ 0.0549, Fig. 6A), GLYPICAN 1 bands normalized to ␤-ACTIN quantitated from three independent experiments are presented in the lower bar graph, shown as percent of band intensity in granulosa cells of suggesting that Glypican 1 induction in eCG 44 h-treated mice. The asterisk indicates a significant difference from all other maturing COC normally requires treatments (P ⬍ 0.05, one-way ANOVA and Tukey’s post hoc test). C, Cumulus cells and ECM GDF9 action. Syndecan 4 was infractions of COC from eCG 44 h ⫹ hCG 16 h-treated mice were separated by incubating COC for 1 min with hyaluronidase to dissociate cells and solubilize cumulus ECM (ECM creased 4-fold and Betaglycan infraction), and then cumulus cells were pelleted and extracted in urea Triton proteoglycan creased 10-fold by heparin treatment. extraction buffer (cell extract). Cell and ECM extracts equivalent to 10 COC were separated The increase in Syndecan 4 was signifon reducing SDS-PAGE and subjected to Western blot analysis for GLYPICAN 1 and compared with known cumulus ECM protein (VERSICAN) and intracellular proteins ␤-ACTIN and STAT3. icantly reversed by cotreatment with Data are representative of three independent experiments with similar results. GDF9, although not returned to control levels (Fig. 6B). Betaglycan expresgreater than 10-fold in COC 16 h after hCG compared sion returned to levels not significantly different from conwith eCG only-treated COC. Furthermore, GLYPICAN 1 protein was 2-fold higher in COC vs. mural granulosa cells trols after heparin ⫹ GDF9 treatment (Fig. 6C), indicating 16 h after hCG treatment in vivo. Unexpectedly, the pro- that GDF9 may contribute to the repression of Betaglycan tein abundance of GLYPICAN 1 was lower after 16 h IVM mRNA levels seen in maturing COC. Syndecan 1 and compared with in vivo matured COC and showed no in- Glypicans 2, Glypicans 4, and Glypicans 6 showed no evidence of GDF9 regulation (data not shown). crease over immature COC (Fig. 5B).

4552

Watson et al.


FIG. 6. GDF9 regulates proteoglycan gene expression in COC. Expression of proteoglycans, Glypican 1, Syndecan 4, and Betaglycan, was evaluated in COC after 12 h IVM culture [␣MEM ⫹ EGF (10 ng/ml) and FSH (10 mIU/ml)] either with no treatment (control) or treated with heparin (300 ␮g/ml) or heparin ⫹ GDF9 (500 ng/ml). Expression levels are normalized to rpL19 mRNA loading control assayed in parallel and presented as expression relative to a calibrator sample of whole-ovary RNA. Values are mean ⫾ SEM of three independent experiments. Those with different characters are significantly different (P ⬍ 0.05). #, P ⫽ 0.0549, one-way ANOVA and Tukey’s post hoc test.

Discussion The ovarian follicle is a highly organized structure, with several specialized cell layers surrounding the oocyte exerting specific roles in endocrine signaling and oocyte maturation. Differential cell fate determination involves distinct responses to paracrine signaling factors produced by oocytes, predominantly GDF9 in the mouse. Both mural and cumulus cell types express the cognate receptors and respond readily to GDF9 (13, 14); yet because oocytes specify the cumulus lineage (3, 35), their signaling must be restricted to the cell layers closest to the oocyte, allowing the mural granulosa cell lineage to diverge in more distal cell layers. Most members of the TGF␤ and FGF family of ligands interact with heparan sulfate proteoglycans, which act to sequester and promote local signaling actions of these ligands. Our findings that exogenous heparin potently blocks GDF9-mediated gene expression indicate that GDF9 signaling involves an interaction with endogenous HS, which may participate in regulating the pattern and intensity of response to oocytes during mouse COC expansion and oocyte maturation.


We found that COCs treated with heparin during in vitro maturation culture with EGF and FSH, had reduced expansion but importantly also changed gene expression pattern indicative of a transition from cumulus to mural granulosa cell phenotype. Highly significant reductions in cumulus specific ECM genes Has2 and Tnfaip6 demonstrates a loss of cumulus cell function while equally striking elevation of Lhcgr and Cyp11a1 expression indicates activation of mural granulosa-like function in heparin treated COCs. These changes mimic the reported effects of oocytectomy or pharmacological Alk4/5/7 inhibition on cumulus cell function (3, 32). We found a striking overlap in the consequent gene expression changes produced by Alk4/5/7 inhibitor or heparin treatment. Thus the results indicate that GDF9 signaling is comprehensively and efficiently disrupted by heparin. Furthermore, we found SMAD2 phosphorylation was strikingly reduced in COCs cultured in the presence of heparin, supporting the idea that the endogenous signal transduction from the GDF9 receptor was blocked by heparin. Soluble heparin can either promote or inhibit HSPG-dependent growth factor signaling; depending whether it mimics endogenous heparan sulfate mediation of receptor-ligand interaction or competitively binds the ligand antagonizing the endogenous HSPG co-receptor function (36). In our results the action of OSF in cumulus cells was competitively inhibited by heparin indicating endogenous OSF signaling involves an interaction with HSPG on cumulus cells. In mouse COCs, GDF9 is the dominant active oocyte signaling factor required for cumulus lineage maintenance (37). When we added recombinant mature GDF9 at a standard concentration used in COC IVM culture (6, 7) it fully reversed the effect of heparin, strongly suggesting that heparin predominantly blocked the action of endogenous GDF9. Native GDF9 in all species is secreted from oocytes as a ternary complex of the mature protein dimer linked to a pro-domain dimer. The pro-domain of many TGF␤ family members mediates ECM binding and sequestration (38). Based on protein sequence analysis we predicted that the GDF9 pro-domain may interact with HSPG via several basic amino acid rich motifs that are highly conserved among vertebrate species. Similar associations have been recently reported for activin (18) and other members of the TGF␤ family including TGF␤1 and ␤2 (17). In support of this concept we found that human pro-mature GDF9 complex bound avidly to immobilized heparin Sepharose. In addition, the fact that recombinant fully processed mature GDF9 was able to overcome the action of exogenous heparin at a standard IVM dose suggests that the mature protein did not interact strongly with heparin. The mechanism of direct interaction of GDF9 pro-domain with heparan sulfate is the subject of ongoing investigation.


A number of independent HSPG with reported ability to modulate growth factor activity exist, of which only CD44 has been widely studied, showing up to 100-fold induction in periovulatory COC (39, 40). We investigated in detail the level and pattern of expression of the other main families of cell surface HSPG, the syndecans, glypicans and Betaglycan in the mural granulosa cells and COC. Syndecans and Betaglycan showed similar relative mRNA expression in mural granulosa and cumulus compartment of preovulatory follicles. Higher abundance of the glypicans in mural granulosa cells than COC of preovulatory follicles, suggests a distinct role for glypicans in the mural compartment before LH stimulation of ovulation and oocyte maturation. The role of glypicans, particularly Glypican 5 in folliculogenesis warrants further study and may participate in Wnt/␤-catenin (41), hedgehog (42) and/or TGF␤ family signaling during folliculogenesis and luteinization. However, notwithstanding the higher Glypican 1 mRNA in granulosa, our Western blot and immunohistochemistry indicated that GLYPICAN 1 protein abundance was in fact comparable in mural granulosa and cumulus cells before the LH surge. During the period of COC expansion and ovulation induced by hCG in vivo or by FSH and EGF in IVM culture, highly dynamic regulation of certain HSPG occurred. Betaglycan was rapidly and significantly down regulated specifically in the COCs and was the only HSPG to show a significant down-regulation during this period. Betaglycan is a repressor of cell motility, and its down-regulation promotes motility in metastatic cancers (43). We recently reported that cumulus cells selectively acquire invasive motility during the periovulatory window (44) and this down-regulation of Betaglycan would be expected to facilitate this acquisition of motility in ovulating COC. Heparin treatment of COCs significantly increased Betaglycan expression which was reversed by GDF9 co-treatment providing the first evidence that Betaglycan mRNA is repressed in COC and this is GDF9 dependent. This is consistent with other reports showing that TGF␤1 also represses Betaglycan expression in Sertoli and adrenal cells (45). A significant up-regulation of Betaglycan in posthCG granulosa cells indicates an alternate role during luteinization possibly involving its known role in inhibin or TGF␤ action. Syndecan 4, another HSPG widely associated with cell motility regulation (46), was up-regulated in COC 16 h after hCG in vivo, but the increase was not significant in IVM COCs. The patterns of Betaglycan and Syndecan 4 expression were not consistent with a role of these HSPG as key mediators of OSF signaling during COC maturation. Syndecan 1 and Glypican 1 were strongly up-regulated in COC by both hCG in vivo and EGF⫹FSH in IVM.


4553

These HSPG are both widely reported potentiators of growth factor signaling (47, 48). In the Drosophila germ cell niche Glypican 1 homolog Dally both restricts diffusion and enhances the localized signal of the TGF␤ homolog decapentaplegic (49, 50). Likewise in mammalian pancreatic cells GLYPICAN 1 mediates localized TGF␤1 action (51, 52). Interestingly, Glypican 1 mRNA and protein was strongly up-regulated in COC but not granulosa cells suggestive of a specific role in the COC during the periovulatory period. In IVM COC Glypican 1 mRNA was elevated significantly more than in vivo matured COC, yet protein abundance was reduced. We found that GLYPICAN 1 protein was associated with the hyaluronan matrix, hence this loss of GLYPICAN 1 from IVM matrix is consistent with a deficiency in ECM components of IVM COC we have reported previously (53). Heparin and SB431542 repressed Glypican 1 mRNA and GDF9 upregulated it suggesting the Glypican 1 induction seen in maturing COC is oocyte dependent, a pattern similar to other genes critical to COC expansion and/or oocyte maturation (3). Together our observations suggest a model for precise positional responses to oocyte signaling during COC expansion and oocyte maturation whereby secreted GDF9 interacts with heparan sulphate to form a morphogenic gradient. Striking regulation of HSPG expression may subsequently influence the pattern of GDF9 signaling during folliculogenesis. Further studies are required to determine the specific HSPG involved, but the current results show that GDF9 action during periovulatory oocyte maturation is required for induction of Glypican 1 and repression of Syndecan 4 and Betaglycan.

Acknowledgments We thank Dr. Joy McIntosh and Professor Ken McNatty (School of Biological Sciences, Victoria University of Wellington, Wellington, New Zealand) for the mouse GDF9 expression construct. Address all correspondence and requests for reprints to: Darryl Russell, Robinson Institute, Research Centre for Reproductive Health, School of Paediatrics and Reproductive Health, The University of Adelaide, Adelaide, South Australia, 5005, Australia. E-mail: [email protected]. (http://www. adelaide.edu.au/directory/darryl.russell). This work was supported by National Health and Medical Research Council Project Grants 519235 and 1011297. D.L.R. is supported by Fellowship funding from the Australian Research Council. R.B.G. is supported by a National Health and Medical Research Council Fellowship.

4554

Watson et al.


Disclosure Summary: I certify that neither I nor my coauthors have a conflict of interest as described above that is relevant to the subject matter or materials included in this work.

19.

20.

References 1. Park JY, Su YQ, Ariga M, Law E, Jin SL, Conti M 2004 EGF-like growth factors as mediators of LH action in the ovulatory follicle. Science 303:682– 684 2. Russell DL, Robker RL 2007 Molecular mechanisms of ovulation: co-ordination through the cumulus complex. Hum Reprod Update 13:289 –312 3. Diaz FJ, Wigglesworth K, Eppig JJ 2007 Oocytes determine cumulus cell lineage in mouse ovarian follicles. J Cell Sci 120:1330 –1340 4. Li R, Norman RJ, Armstrong DT, Gilchrist RB 2000 Oocyte-secreted factor(s) determine functional differences between bovine mural granulosa cells and cumulus cells. Biol Reprod 63:839 – 845 5. Elvin JA, Yan C, Wang P, Nishimori K, Matzuk MM 1999 Molecular characterization of the follicle defects in the growth differentiation factor 9-deficient ovary. Mol Endocrinol 13:1018 –1034 6. Sugiura K, Su YQ, Diaz FJ, Pangas SA, Sharma S, Wigglesworth K, O’Brien MJ, Matzuk MM, Shimasaki S, Eppig JJ 2007 Oocytederived BMP15 and FGFs cooperate to promote glycolysis in cumulus cells. Development 134:2593–2603 7. Sugiura K, Su YQ, Li Q, Wigglesworth K, Matzuk MM, Eppig JJ 2010 Estrogen promotes the development of mouse cumulus cells in coordination with oocyte-derived GDF9 and BMP15. Mol Endocrinol 24:2303–2314 8. Elvin JA, Clark AT, Wang P, Wolfman NM, Matzuk MM 1999 Paracrine actions of growth differentiation factor-9 in the mammalian ovary. Mol Endocrinol 13:1035–1048 9. Zhang M, Su YQ, Sugiura K, Xia G, Eppig JJ 2010 Granulosa cell ligand NPPC and its receptor NPR2 maintain meiotic arrest in mouse oocytes. Science 330:366 –369 10. Sugiura K, Pendola FL, Eppig JJ 2005 Oocyte control of metabolic cooperativity between oocytes and companion granulosa cells: energy metabolism. Dev Biol 279:20 –30 11. Eppig JJ, Pendola FL, Wigglesworth K 1998 Mouse oocytes suppress cAMP-induced expression of LH receptor mRNA by granulosa cells in vitro. Mol Reprod Dev 49:327–332 12. Diaz FJ, Sugiura K, Eppig JJ 2008 Regulation of Pcsk6 expression during the preantral to antral follicle transition in mice: opposing roles of FSH and oocytes. Biol Reprod 78:176 –183 13. Vanderhyden BC, Telfer EE, Eppig JJ 1992 Mouse oocytes promote proliferation of granulosa cells from preantral and antral follicles in vitro. Biol Reprod 46:1196 –1204 14. Gilchrist RB, Ritter LJ, Myllymaa S, Kaivo-Oja N, Dragovic RA, Hickey TE, Ritvos O, Mottershead DG 2006 Molecular basis of oocyte-paracrine signalling that promotes granulosa cell proliferation. J Cell Sci 119:3811–3821 15. Häcker U, Nybakken K, Perrimon N 2005 Heparan sulphate proteoglycans: the sweet side of development. Nat Rev Mol Cell Biol 6:530 –541 16. Belenkaya TY, Han C, Yan D, Opoka RJ, Khodoun M, Liu H, Lin X 2004 Drosophila Dpp morphogen movement is independent of dynamin-mediated endocytosis but regulated by the glypican members of heparan sulfate proteoglycans. Cell 119:231–244 17. Lyon M, Rushton G, Gallagher JT 1997 The interaction of the transforming growth factor-␤s with heparin/heparan sulfate is isoform-specific. J Biol Chem 272:18000 –18006 18. Li S, Shimono C, Norioka N, Nakano I, Okubo T, Yagi Y, Hayashi M, Sato Y, Fujisaki H, Hattori S, Sugiura N, Kimata K, Sekiguchi K 2010 Activin A binds to perlecan through its pro-region that has

21.

22.

23.

24.

25. 26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.


heparin/heparan sulfate binding activity. J Biol Chem 285:36645– 36655 Ohkawara B, Iemura S, ten Dijke P, Ueno N 2002 Action range of BMP is defined by its N-terminal basic amino acid core. Curr Biol 12:205–209 Bernfield M, Götte M, Park PW, Reizes O, Fitzgerald ML, Lincecum J, Zako M 1999 Functions of cell surface heparan sulfate proteoglycans. Annu Rev Biochem 68:729 –777 Shah MM, Sakurai H, Gallegos TF, Sweeney DE, Bush KT, Esko JD, Nigam SK 2011 Growth factor-dependent branching of the ureteric bud is modulated by selective 6-O sulfation of heparan sulfate. Dev Biol 356:19 –27 Shah MM, Sakurai H, Sweeney DE, Gallegos TF, Bush KT, Esko JD, Nigam SK 2010 Hs2st mediated kidney mesenchyme induction regulates early ureteric bud branching. Dev Biol 339:354 –365 Patel VN, Likar KM, Zisman-Rozen S, Cowherd SN, Lassiter KS, Sher I, Yates EA, Turnbull JE, Ron D, Hoffman MP 2008 Specific heparan sulfate structures modulate FGF10-mediated submandibular gland epithelial morphogenesis and differentiation. J Biol Chem 283:9308 –9317 Casu B, Naggi A, Torri G 2010 Heparin-derived heparan sulfate mimics to modulate heparan sulfate-protein interaction in inflammation and cancer. Matrix Biol 29:442– 452 Iozzo RV 1998 Matrix proteoglycans: from molecular design to cellular function. Annu Rev Biochem 67:609 – 652 Eppig JJ 1981 Ovarian glycosaminoglycans: evidence for a role in regulating the response of the oocyte-cumulus cell complex to FSH. Endocrinology 108:1992–1994 Downs SM 1989 Specificity of epidermal growth factor action on maturation of the murine oocyte and cumulus oophorus in vitro. Biol Reprod 41:371–379 Mottershead DG, Pulkki MM, Muggalla P, Pasternack A, Tolonen M, Myllymaa S, Korchynskyi O, Nishi Y, Yanase T, Lun S, Juengel JL, Laitinen M, Ritvos O 2008 Characterization of recombinant human growth differentiation factor-9 signaling in ovarian granulosa cells. Mol Cell Endocrinol 283:58 – 67 Pulkki MM, Myllymaa S, Pasternack A, Lun S, Ludlow H, AlQahtani A, Korchynskyi O, Groome N, Juengel JL, Kalkkinen N, Laitinen M, Ritvos O, Mottershead DG 2011 The bioactivity of human bone morphogenetic protein-15 is sensitive to C-terminal modification: characterization of the purified untagged processed mature region. Mol Cell Endocrinol 332:106 –115 Hussein TS, Thompson JG, Gilchrist RB 2006 Oocyte-secreted factors enhance oocyte developmental competence. Dev Biol 296:514 – 521 Yeo CX, Gilchrist RB, Thompson JG, Lane M 2008 Exogenous growth differentiation factor 9 in oocyte maturation media enhances subsequent embryo development and fetal viability in mice. Hum Reprod 23:67–73 Dragovic RA, Ritter LJ, Schulz SJ, Amato F, Thompson JG, Armstrong DT, Gilchrist RB 2007 Oocyte-secreted factor activation of SMAD 2/3 signaling enables initiation of mouse cumulus cell expansion. Biol Reprod 76:848 – 857 Mottershead DG, Ritter LJ, Gilchrist RB 2012 Signalling pathways mediating specific synergistic interactions between GDF9 and BMP15. Mol Hum Reprod 18:121–128 Asada M, Shinomiya M, Suzuki M, Honda E, Sugimoto R, Ikekita M, Imamura T 2009 Glycosaminoglycan affinity of the complete fibroblast growth factor family. Biochim Biophys Acta 1790:40 – 48 Eppig JJ, Wigglesworth K, Pendola F, Hirao Y 1997 Murine oocytes suppress expression of luteinizing hormone receptor messenger ribonucleic acid by granulosa cells. Biol Reprod 56:976 –984 Fannon M, Forsten KE, Nugent MA 2000 Potentiation and inhibition of bFGF binding by heparin: a model for regulation of cellular response. Biochemistry 39:1434 –1445 Yan C, Wang P, DeMayo J, DeMayo FJ, Elvin JA, Carino C, Prasad SV, Skinner SS, Dunbar BS, Dube JL, Celeste AJ, Matzuk MM 2001


38.

39.

40. 41.

42.

43.

44.

45.

Synergistic roles of bone morphogenetic protein 15 and growth differentiation factor 9 in ovarian function. Mol Endocrinol 15:854 – 866 Harrison CA, Al-Musawi SL, Walton KL 2011 Prodomains regulate the synthesis, extracellular localisation and activity of TGF-␤ superfamily ligands. Growth Factors 29:174 –186 Borg N, Holland M 2008 The effect of glycosaminoglycans on rat gametes in vitro and the associated signal pathway. Reproduction 135:311–319 Yokoo M, Kimura N, Sato E 2010 Induction of oocyte maturation by hyaluronan-CD44 interaction in pigs. J Reprod Dev 56:15–19 Tanaka SS, Kojima Y, Yamaguchi YL, Nishinakamura R, Tam PP 2011 Impact of WNT signaling on tissue lineage differentiation in the early mouse embryo. Dev Growth Differ 53:843– 856 Li F, Shi W, Capurro M, Filmus J 2011 Glypican-5 stimulates rhabdomyosarcoma cell proliferation by activating Hedgehog signaling. J Cell Biol 192:691–704 Hempel N, How T, Dong M, Murphy SK, Fields TA, Blobe GC 2007 Loss of betaglycan expression in ovarian cancer: role in motility and invasion. Cancer Res 67:5231–5238 Akison LK, Alvino ER, Dunning KR, Robker RL, Russell DL 2012 Transient invasive migration in mouse cumulus oocyte complexes induced at ovulation by luteinizing hormone. Biol Reprod 86:125 Farnworth PG, Wang Y, Escalona R, Leembruggen P, Ooi GT, Findlay JK 2007 Transforming growth factor-␤ blocks inhibin binding to different target cell types in a context-dependent manner through dual mechanisms involving betaglycan. Endocrinology 148:5355– 5368


4555

46. Averbeck M, Gebhardt C, Anderegg U, Termeer C, Sleeman JP, Simon JC 2007 Switch in syndecan-1 and syndecan-4 expression controls maturation associated dendritic cell motility. Exp Dermatol 16:580 –589 47. Derksen PW, Keehnen RM, Evers LM, van Oers MH, Spaargaren M, Pals ST 2002 Cell surface proteoglycan syndecan-1 mediates hepatocyte growth factor binding and promotes Met signaling in multiple myeloma. Blood 99:1405–1410 48. Matsuda K, Maruyama H, Guo F, Kleeff J, Itakura J, Matsumoto Y, Lander AD, Korc M 2001 Glypican-1 is overexpressed in human breast cancer and modulates the mitogenic effects of multiple heparin-binding growth factors in breast cancer cells. Cancer Res 61: 5562–5569 49. Guo Z, Wang Z 2009 The glypican Dally is required in the niche for the maintenance of germline stem cells and short-range BMP signaling in the Drosophila ovary. Development 136:3627–3635 50. Hayashi Y, Kobayashi S, Nakato H 2009 Drosophila glypicans regulate the germline stem cell niche. J Cell Biol 187:473– 480 51. Li J, Kleeff J, Kayed H, Felix K, Penzel R, Büchler MW, Korc M, Friess H 2004 Glypican-1 antisense transfection modulates TGF␤-dependent signaling in Colo-357 pancreatic cancer cells. Biochem Biophys Res Commun 320:1148 –1155 52. Kayed H, Kleeff J, Keleg S, Jiang X, Penzel R, Giese T, Zentgraf H, Büchler MW, Korc M, Friess H 2006 Correlation of glypican-1 expression with TGF-␤, BMP, and activin receptors in pancreatic ductal adenocarcinoma. Int J Oncol 29:1139 –1148 53. Dunning KR, Lane M, Brown HM, Yeo C, Robker RL, Russell DL 2007 Altered composition of the cumulus-oocyte complex matrix during in vitro maturation of oocytes. Hum Reprod 22:2842–2850

Join The Endocrine Society and network with endocrine thought leaders from around the world. www.endo-society.org/join