Direct impact of gonadotropins on glucose uptake and storage

Abstract

Background: Polycystic ovary syndrome (PCOS) is often associated with higher levels of LH, and arrested ovarian follicular growth. The direct impact of high LH on FSH mediated metabolic
responses in PCOS patients is not clearly understood.

Method: In order to investigate the impact ofFSH and LH on glucose metabolism in preovulatory granulosa cells (GCs), we used [U-14C]-2 deoxyglucose, D-[U-14C]-glucose or 2-NBD glucose to
analyse glucose uptake and its incorporation into glycogen. To reproduce the high androgenic potential in PCOS patients, we administered hCG both in vitro and in vivo. The role of IRS-
2/PI3K/Akt2 pathway was studied after knockdown with specific siRNA. Immunoprecipitation and specific assays were used for the assessment of IRS-2, glycogen synthase and protein phosphatase 1. Furthermore, we examined the in vivo effects of hCG on FSH mediated glycogen increase in normal and PCOS rat model. HEK293 cells co-expressing FSHR and LHR were used to demonstrate glucose uptake and BRET change by FSH and hCG.

Results: In normal human and rat granulosa cells, FSH is more potent thanhCG in stimulating glucose uptake, however glycogen synthesis was significantly upregulated only by FSH through
increase in activity of glycogen synthase via IRS-2/PI3K/Akt2 pathway. On the contrary, an impaired FSH-stimulated glucose uptake and glycogen synthesis in granulosa cells of PCOS-patients indicated a selective defect in FSHR activation. Further, in normal human granulosa cells, and in immature rat model, the impact of hCG on FSH responses was such that it inhibited the FSH-mediated glucose uptake as well as glycogen synthesis through inhibition of FSH-stimulated IRS-2 expression. These findings were further validated in HEK293 cells overexpressing Flag-LHR and HA-FSHR, where high hCG inhibited the FSH-stimulated glucose uptake. Notably, an increased BRET change was observed in HEK293 cells expressing FSHR-Rluc8 and LHR-Venus possibly suggesting increased heteromerization of LHR and FSHR in the presence of both hCG and FSH in comparison to FSH or hCG alone.

Conclusion: Our findings confirm a selective attenuation of metabolic responses to FSH such as glucose uptake and glycogen synthesis by high activation level of LHR leading to the inhibition of IRS-2 pathway, resulting in depleted glycogen stores and follicular growth arrest in PCOS patients.

Keywords: FSH, LH, Granulosa cells, PCOS, IRS-2, Glucose, Glycogen, Metabolism.

1. Introduction

Polycystic ovary syndrome (PCOS) is one of the most common endocrinopathies leading to subfertility or infertility in women of reproductive age [1-2]. PCOS women are more susceptible to
develop metabolic immune parameters abnormalities such as insulin resistance [1,3-4]. In PCOS women, the ovarian follicular growth is disordered as it gets arrested at mid-antral stage [5]. Most PCOS women have an abnormal ratio of luteinizing hormone (LH) and follicle stimulating hormone (FSH), due to higher basal level and increased pulse frequency of LH [6-7]. Several hypotheses have emerged as to how LH hypersecretion may adversely affect follicular growth and differentiation in PCOS patients. These hypotheses were substantiated by transgenic mice model where over-expression of LH or hCG resulted in polycystic ovaries and infertility [8-9]. The most popular hypothesis states that increased steroidogenesis and hyperandrogenism due to high LH level contribute to PCOS development [2]. An elevated LH may also dysregulate follicular development by decreasing FSH sensitivity [10]. Most studies showed a normal to exaggerated serum E2-responsiveness to FSH in normal-weight PCOS women [11].

Granulosa cells (GCs) and oocytes have metabolic cooperation for the continuous supply of glucose and any alteration in this process may have deleterious effects on follicular growth and oocyte maturation. Both FSH and LH stimulate glucose uptake as demonstrated in oocyte-cumulus cell complexes [12] and GCs [13]. However, the relative difference in the potency of FSH and LH for glucose uptake in preovulatory GCs is not known. Previous studies have established that LH is more glycolytic than FSH [12]. Further, the metabolic fate of FSH-stimulated glucose uptake in preovulatory GCs is not clearly known. Earlier, accumulation of glycogen has been observed in ovarian follicles especially in GCs of antraland preovulatoryovine follicles 48hbefore ovulation [14-15]. However, little is known about the relative role of FSH and LH in the regulation of glycogen levels in preovulatory follicles. Intriguingly, the impact of high LH on the metabolic responses of FSH, which may contribute to PCOS condition, is also not clear.

Glycogen synthesis is primarily known to be stimulated by insulin in peripheral tissues as well as in ovaries [16]. The molecular mechanisms by which the enzymatic effectors involved in glycogenesis are regulated in preovulatory follicles remain poorly understood. FSH and insulin have overlapping effects on the signaling cascades to increase glucose uptake via
translocation of glucose transporter 4 (GLUT4) to the plasma membrane and the activation of hexokinase [16- 17].Both FSH and LH bind to their specific G protein-coupled receptors and activate the classical cAMP/protein kinase A (PKA) pathway in addition to several other signaling pathways such as IRS- 2/PI3K/Akt2 and MAPK pathways [17- 18]. LH receptor (LHR) binds both pituitary-derived LH and the placental hormone, chorionic gonadotropin (hCG), while the FSH receptor (FSHR) only binds the pituitary-derived FSH. LHR gets constitutively expressed on theca and interstitial cells whereas its expression is regulated by FSH in preovulatory GCs [19-20]. FSHR expression starts in GCs of early follicles, whereas LHR expression becomes prominent just prior to ovulation (preovulatory stage) and luteinizing follicles. Once acquired FSHR remains on GCs of healthy follicles until they become atretic or luteinize. In PCOS follicles, GCs have premature and higher expression of LHR whereas the expression ofFSHR is either normal or higher [21]. No change in the expression of insulin receptor (InsR) has been observed and the expression of IGF- 1 receptor (IGF- 1R) is normal in PCOS GCs [5,10, 22].Earlier, we had shown that FSH increases IRS-2 expression,which is crucial for PI3K-Akt2-mediated translocation of GLUT4 to cell membrane and consequent uptake of glucose in preovulatory rat GCs [17]. FSH-stimulated increase in IRS-2 which is an early step in the cross-talk between FSH and insulin
/IGF- 1 pathways was impaired in PCOS GCs [17]. Previous studies indicate that IRS-2-deleted mice are infertile and resistant to the exogenous gonadotropins [23-24]. However, InsR knockout mice have normal fertility and litter size [25]. Therefore, we hypothesized that FSH signaling could cross-talk with insulin signaling pathway through IRS-2 and a defect in this pathway could be responsible for impaired metabolism in PCOS patients and poor fertility outcome.Here, we assessed the impact of LH on FSH-regulated metabolic responses in physiological conditions as well as in PCOS patients. Our findings suggest that abnormal LH levels may selectively attenuate the metabolic responses of FSH such as glucose uptake and its storage as glycogen in GCs. Our findings reported here will help in exploring new therapies and therapeutic targets for women with PCOS in future.

2. Materials and Methods
2.1 Materials:

The materials used in this study can be found in supplemental materials and in our previous publications [3, 17].

2.2 Subjects:

Following the approval by institutional ethics committees (IEC/NP-293/2012-RF- Ixabepilone 16/2013; IHEC/DU/NP-2/2012; IHEC/DU/NP- 1/2018), we enrolled PCOS women with or without insulin resistance as per Rotterdam criteria as described earlier [3,17, 26]. The details of the subjects contributing to this study are given in Supplemental Table 1. General inclusion criteria for all participants were age less than 35 years, normal prolactin levels, and normal thyroid function [3, 17]. The selection criteria for control women were as follows: regular menstrual cycles occurring every 25-35 d, no clinical or biochemical evidence of hyperandrogenism, no polycystic ovaries, and without insulin resistance. These women were receiving assisted reproduction for non-ovarian indications, such as male or tubal factor infertility. PCOS patients did not receive clomiphene citrate or antidiabetic drugs during stimulation cycles. All methods performed are in accordance with the relevant guidelines and regulations. We collected the ovarian aspirates of 41 non-hirsute ovulatory women and 61 PCOS women with or without insulin resistance, after the gonadotropin therapy for in vitro fertilization (IVF). Insulin resistance in PCOS was assessed by calculating the homeostasis model assessment (HOMA-IR) index and 2.5 was selected as a cutoff point [3, 17].

2.3 Granulosa cell culture and treatments

We isolated the human GCs from the follicular fluid aspirates obtained after the IVF therapy of normal and PCOS women as described earlier [17]. The cells were suspended incomplete media,Dulbecco’s Modified Eagle’s medium (DMEM) containing 5% foetal bovine serum (FBS, cat # RM9955, South American origin, EU approved), antibiotic and antimycotic solution (Himedia
Biosciences, India), and cultured in 5% CO2 at 37 °C (Thermo Scientific, USA).The animal experiments were performed under the guidance of institutional animal ethics committee (DU/2001/IAEC-R/2013/37). Twenty-six day old immature female rats (Holtzman strain, Total=42) were primed with pregnant mare serum gonadotropin (PMSG, 10U/d for 3 d) and GCs were isolated
from preovulatory follicles described earlier [17]. We used GCs cultured in DMEM supplemented with 10% FBS, antibiotic solution and grown to 70% confluency in 5% CO2 at 37 °C.Human or rat granulosa cells were cultured and serum starved on day 3 of culture for 16 h overnight and then treated with different concentrations of recombinant human FSH (10,000 U/mg protein,
Recagon, Organon, Ireland), human chorionic gonadotropin (Urinary origin hCG, CG- 10, 10,000 U/mg protein, Sigma-Aldrich, see details in Supplemental materials) or in combination for 1 h before addition of [U-14C]-2 deoxyglucose (1 µCi) or 2-(N-(7-Nitrobenz-2-oxa- 1,3-diazol-4-yl) amino)-2-deoxyglucose (2-NBDG, 200µM) for 30 min for glucose uptake and 3h for glycogen synthesis. To reproduce the high androgenic potential in PCOS patients, we administered hCG both in vitro and in vivo. To compare the effect of FSH and hCG on glucose uptake, rat preovulatory GCs were treated with FSH alone or hCG alone and in combination for 1h before the addition of 2-NBDG for 30 min, in vitro.

To study FSH-mediated glycogen synthesis, GCs were also treated with LY294002 (PI3K inhibitor,10 μM), Akt inhibitor IV (7 μM), phosphatase inhibitors- okadaic acid (OA, 50 nM) and calyculin A (Cal A, 5 nM), and D-[U-14C]-glucose (specific activity, 150-250 mCi/mmole) 30 min before the FSH treatment.To find out if the role of FSH in stimulating the glycogen synthesis in preovulatory GCs is Estradiol (E2) mediated, we examined the glucose uptake and glycogen synthesis in rat preovulatory GCs in the presence of 3.3 nM FSH or 50 nM E2 [27] or 4µg/ml anti-E2, a non-steroidal compound, 2-[piperdinoethoxyphenyl]-3-[4-methoxyphenyl]-2H benzopyran (K-7) [28] or in combination of these for 1h before addition of 2-NBDG for 3h. Glycogen was extracted by KOH method and fluorescence was measured as described in the section 2.4.The siRNA specific to Akt (Akt2), PI3K (p85), IRS-2 and scrambled siRNA (Santa Cruz Biotechnology, USA), were transfected into GCs by using RNAiFect (Qiagen, Germany). The efficiency and specificity of each siRNA-mediated knockdown was monitored as described earlier [17].

2.4 Glycogen assay

Serum starved GCs were treated with FSH for 1h, [U-14C]-2 deoxyglucose (1 µCi) or 2-NBDG (200 µM) was added to each well and further incubated for 3h [29]. The cells were homogenized in 30% KOH saturated with sodium sulphate, and the extracts were boiled for 30 min after addition of 2 mg carrier glycogen. Then, 2 volumes of 95% ethanol was added and glycogen was precipitated overnight at -20ºC. In these conditions, free glucose did not precipitate. Radioactivity was measured in the pellets with a liquid scintillation counter (Wallac 1450 MicroBeta® TriLux scintillation counter). The 2-NBDG-glycogen fluorescence in samples was measured in black 96-well plate (Greiner) using fluorescence plate reader (Fluostar® Optima, BMG Labtech GmbH (Ortenberg, Germany) at 480 nm excitation and 535 nm emission wavelengths. Protein content was measured in cell lysates by Bradford assay [17].

For confocal imaging of glycogen granules, control and FSH-treated GCs in 4-well slides (BD Biosciences) were fixed in 4% paraformaldehyde in 1X PBS for 10 min. After washing, the cells were mounted using Ultra Cruz mounting medium. Fluorescent images were observed with a 63x oil immersion objective of Leica TCS SP5 confocal microscope. The images were processed using LAS
AF Lite software (Leica Microsystems Inc. Germany). The quantification of relative mean optical intensity (ROI) was done using Image J software.

2.5 Estimation of glycogen in ovaries

Immature female rats (26 d old, total=45) were divided into seven groups based on the hormonal treatments. In the first experiment, these were: group 1. FSH (8 IU/d) treatment for one day (n=5),group 2. FSH (8 IU/d) treatment for two days (n=5), and group 3. FSH (8 IU/d) treatment for three days (n=10). group 4. Treatment with FSH (8 IU/day) for 3 days followed by hCG (10 IU/day) on 3rd day (n=5). group 5. Treatment with hCG alone (10 IU/day) for 3 days (n=5), group 6. A control group of immature female rats was treated with saline for respective duration (n=10). The animals were sacrificed at the end of treatments and ovaries were processed for glycogen extraction by KOH treatment, and estimation by phenol-sulphuric acid method [30].

2.6 Glycogen synthase (GS) activity

For determining the GS activity, serum starved GCs were treated with FSH for varying time periods.After washing with PBS, the cells were scraped and then sonicated in NaF-EDTA solution (2.5 mM EDTA, 10 mM NaF) for 10 secs and 100 μl of this extract was added to GS assay buffer (200 μl). To measure the activity of GS (in the absence of glucose-6-phosphate), abuffer
containing 3 mM Uridine diphosphate glucose and UDP-[U-14C] glucose (specific activity: 200 mCi/mmol) was added and then incubated for 20 min at 37°C. After overnight precipitation with ethanol (95%) at -20°C, samples were centrifuged and radioactivity was measured in the pellets using a liquid scintillation counter. The activity of GS in this assay was represented as nmoles of UDP-glucose incorporated per mg protein per hour.

2.7 Animal model of PCOS

Adult rats (3 months old) received subcutaneous (s.c.) injections (4 mg/0.2ml olive oil/d, n=6/group, Total=18) of RU486 (Sigma Chemicals, St. Louis, MO, USA) daily for 18 days beginning on the day of proestrus (Day 1 of the experiment) as described earlier [17]. After the treatment, the ovaries were processed for glycogen extraction and estimation as described above. The control ovaries were taken from rats that were in proestrus or oestrus stage after olive oil injection for 18 days.

2.8 Immunoprecipitation

Control, and FSH-treated GCs were harvested in 20 mM Tris-HCl pH 7.5, 150 mM NaCl, 0.5% sodium deoxycholate, 1% Triton X- 100, 2 mM EDTA and the extracts were prepared by homogenization. Equal amount of protein was taken for each sample and to the precleared supernatant, 10 μlofIRS-2 antibody (Santa Cruz Biotechnology, USA) and protein-A agarose were added, and the mixture was kept for 15h at 4°C. Immune complexes were eluted from protein-A agarose and were subjected to Western blotting.

2.9 Western Blotting

Proteins extracted from GCs (50 µg), and the immune complexes obtained with IRS-2 antibody were fractionated on 10% and 7.5% SDS-PAGE respectively. The proteins were transferred onto nitrocellulose membranes which were then incubated for 2h at 25°C with either of the primary antibodies to p-GSK3β, GSK3β, p-Tyr, p-Ser or β-actin (Santa Cruz Biotechnology, USA). The Western blots were further incubated with appropriate horseradish peroxidase-conjugated secondary antibodies, anti-goat IgG-Cy3 and anti-rabbit IgG-Cy3 antibodies (Sigma Aldrich Chemicals Pvt Ltd, USA). Presence of bound antibodies was detected by enhanced chemiluminescence (ECL) reaction using the ECL Pluskit (Millipore, USA).

2.10 Immunofluorescence

Cultured cells were fixed with 4% paraformaldehyde for 10 min and permeabilized with 0.1% Triton-X 100. For studying the colocalization of GS with 2-NBDG, monolayer cultures of GCs in culture slides (BD Bioscience) were incubated with 2-NBDG (500 μM) after 1h incubation with FSH at a concentration of 3.3 nM equivalent approximately to EC50 of glycogen synthesis for 3h. After blocking with 5% BSA, the slides were incubated with 1:50 dilution of anti-GS- 1 (mouse) antibody (Santa Cruz Biotechnology, USA) at 4°C overnight. GS exhibited red fluorescence from Cy3-conjugated anti-mouse secondary antibody (Santa Cruz Biotechnology, USA) and 2-NBDG, a green fluorescence while nuclei were counterstained with DAPI (4′, 6-diamidino-2-phenylindole). The images were processed as described earlier. The quantification of signal overlap was determined using Pearson Correlation Coefficient (Leica LAS AF software).

2.11 RNA isolation and reverse-transcription quantitative PCR (qRT-PCR)

Total RNA was extracted using Tri-reagent (Sigma) from rat GCs treated with FSH, hCG and FSH + hCG and untreated cells that were used as control. IRS-2 RNA was quantified by qPCR as described earlier [17]. The primers for IRS-2 and β2M were as follows:IRS-2 (F) 5′-TCGGACAGCTTCTTCTTCA-3′, (R) 5′-ATGGTCTCGTGGATGTTCT-3′,β2M (F) 5′-TGCTCGCGCTACTCTCTCTT-3′, (R) 5′-TCAACTTCAATGTCGGATGG-3′.

2.12 Protein phosphatase 1 (PP1) assay

PP1 activity was determined by measuring the formation of DiFMU (6,8-Difluoro-7-Hydroxy-4-Methylcoumarin) from the substrate DiFMUP (6,8-Difluoro-4-Methylumbelliferyl Phosphate) at 37°C
[31]. The cells were lysedin homogenization buffer and supernatants were incubated with PP1 antibody (Santa Cruz Biotechnology, USA) and the immune complexes were immobilized on protein A-agarose. After washing, the immune complexes immobilized on protein A-agarose were incubated at 37 °C for 1 hwith PP1 assay buffer at pH 7.5. The reaction was initiated by the addition of 50 µM DiFMUP substrate. To study the signaling pathways involved in FSH-mediated increase in PP1 activity, GCs were also treated with PI3K inhibitor (LY294002, 10 μM), Akt inhibitor IV (7 μM),Phosphatase inhibitors-okadaic acid (OA, 20 nM) and calyculin A (Cal A, 0.5 nM), 30 min before the FSH treatment. Thereafter, the fluorescence was measured in the fluorescence plate reader (Fluostar Optima) at the excitation/emission wavelengths of 360 nm / 460 nm. For the measurement of total PP1 enzyme activity, a linear regression analysis was performed and its slope indicates phosphatase activity in relative fluorescence units per hour (RFU/h). The phosphates released were quantified by comparison of the measured RFUs with the fluorescence of DiFMU. A standard curve derived from the provided reference standards was used to convert the fluorescence units obtained in the assay into nanomoles (nmoles) of phosphate. One nmole of the included reference standard, DiFMU, is equivalent to one nmole of phosphate released by the cleavage of the substrate, DiFMUP. The results are represented as nmoles of phosphate
released per mg protein per hour.

2.13 Measurement of glucose uptake in HEK293 cells expressing both LHR and FSHR

For transfections, plasmid DNA (50 ng/well) of HA-FSHR or FLAG-LHR was diluted in 25 µl/well DMEM without phenol red and FBS. Both plasmids and HEK293 cells were provided by Dr. Aylin
Hanyaloglu (Imperial College London, London, UK). Transfection Reagent (ViaFect™ Promega, 0.5 µL/well) was mixed with 25 µl/ well of DMEM and the two solutions were mixed and incubated at
room temperature for 20 min. HEK293 cells (80,000 cells/well) were added to the above solution in a 96-well plate and incubated at 37°C in 5% CO2. After 48 hours, cells were serum-starved overnight.HEK293 cells expressing both FSHR and LHR were treated with different concentrations of FSH alone, hCG alone and FSH + hCG for 1h, followed by addition of 2-NBDG (200µM) for 30 min. The cells were washed with PBS and transferred to Greiner 96-well black plate. Fluorescence due to 2-NBDG uptake was measured using fluorescence plate reader (FLUOstar OPTIMA) at 480 nm excitation and 535 nm emission wavelength.

2.14 Bioluminescence Resonance Energy Transfer (BRET) measurement in HEK293 cells after co-transfection with FSHR-Luc8 and LHR-Venus

HEK293 cells were cultured for forty-eight hours before BRET measurement. HEK293 cells were co-transfected with the FSHR fused with the Renilla luciferase 8 BRET donor (FSHR-Luc8) and with
LHR fused with the Venus BRET acceptor (LHR-Venus) using Metafectene Pro (Biontex Laboratories GmbH, Munich, Germany). Both plasmids were provided by Dr. Aylin Hanyaloglu (Imperial College London, London, UK). The two transfection mixes (Mix A and B) were prepared and incubated separately for 5 min. Mix A contained 50 ng/well ofFSHR-luc8 plus 120 ng/well of LHR-Venus diluted in 25 µl/well in DMEM. Mix B contained 0.5 µL/well of Metafectene Pro and 25 µl/ well of DMEM. The two solutions were mixed and incubated at room temperature for 20 min.,then cells (80,000 cells/well) were added and cultured in opaque 96-well plates (Greiner Bio One International GmbH, Kremsmünster, Austria). After 48 hours, cells were starved overnight in a serum-free medium. These cells in 30 µl/well of PBS without Ca2+ and Mg2+ were stimulated with increasing concentrations of FSH or hCG in 10 µl (FSH was a kind gift from Merck KGaA, Darmstadt,
Germany; hCG was a kind gift from Dr. Y. Combarnous, CNRS, Nouzilly, France) and a fixed concentration of hCG or FSH in10 µl, respectively. Finally, 10 µl/well of Coelenterazine H, the
Renilla luciferase substrate (Interchim, Montluçon, USA) was added to each well to a final concentration of 5 µM and the BRET signal was immediately detected and registered for 30 min by
using a Mithras LB 943 plate reader (Berthold Technologies GmbH & Co.,Wildbad, Germany).Results are expressed as area under the curve from 5 experiments and were analyzed by one way ANOVA.

2.15 Statistical analysis

Statistical significance of differences between two groups was determined by unpaired Student’s t-test. One way ANOVA was applied to compare the data between the groups. [14C]-Glycogen levels in GCs were converted from cpm topmoles of glucose units incorporated per mg protein per hour.Incorporation of 2-NBDG into glycogen was expressed as relative fluorescence units (RFUs).Statistical analyses were performed using Prism Version 7.0a (GraphPad software Inc. USA). The results arepresented as mean ± SEM and statistical significance determined asp < 0.05. 3. Results
3.1 Both FSH and hCG increase glucose uptake but only FSH upregulates glycogen synthesis

The uptake of 2-NBDG was significantly increased by FSH alone and hCG alone, but FSH was more efficient (Fig. 1A, Table 1, p < 0.001, n=5) as well as potent (Fig. 1A, Table 1, p < 0.001, n=5) than hCG (Table 1, n=6).We also examined 2-NBDG incorporation into glycogen in GCs and FSH alone robustly increased the incorporation of 2-NBDG into glycogen in a concentration-dependent manner (Fig. 1B, Table 1, p < 0.001, n=6) whereashCG alone had no significant effect (Fig. 1B, Table 1). Maximum incorporation of D-[U-14C]-glucose into glycogen was observed after 4h of incubation of rat GCs with FSH (Supplemental Fig. 1).FSH increased the uptake of glucose in GCs with EC50 = 0.92 ± 0.02 nM,whereas it increased the glycogen level with EC50 = 3.18 ± 0.07 nM. FSH concentration (1.5 nM) approximately equivalent to EC60 was used in subsequent experiments so as to get adequate glucose uptake for studying the inhibitory effect of hCG [32]. A concentration of FSH (3.3 nM) approximately equivalent to EC50 was used for experiments on glycogen synthesis in GCs.FSH increased the glycogen granules in preovulatory GCs in vitro in a time dependent manner up to 4h as it can be seen in the images taken under confocal microscope (Fig. 1C). But after 6h, there was no significant difference in comparison to untreated controls (Fig. 1C). To understand the physiological importance of the observations, we examined the glycogen content in the ovaries of immature rats treated with FSH (8IU/d) for three days. As shown in Fig.1D, a robust increase in ovarian glycogen content was observed on day 1 to 3 as compared to controls (p < 0.001, n=5).To examine whether FSH increased the activity of glycogen synthase (GS) which is a rate limiting enzyme in glycogen synthesis step, we measured the incorporation of UDP-[U-14C] glucose into glycogen. Incorporation of UDP-[U-14C] glucose into glycogen was stimulated by FSH with a significant increase in GS activity after 1h of treatment (Fig. 1E, p < 0.05, n=3), in a G6P-independent way. FSH also increased the protein expression of GS in preovulatory rat GCs (Supplemental Fig. 2, p < 0.001, n=4).Increased 2-NBDG (green) co-localization with GS (red) after FSH treatment was observed by immunocytochemistry with antibodies specific to GS (Fig. 1F, lower panel; Fig. 1G, R2 = 0.81, p < 0.0001, 10- 12 cells) as compared to untreated control cells (Fig.1H, upper panel, R2 = 0.135, p = 0.425).To understand the effect of hCG on glycogen levels built up by FSH, GCs were treated with FSH and 2-NBDG for 4 hand then divided into 2 groups, one set of GCs were treated with hCG for another 4 h and the other set was not treated any further and was taken as control. Glycogen content of GCs measured at the end of 4 h of incubation with FSH was taken as 100%. There was a significant decrease in the glycogen content in GCs treated with hCG for 4 h in comparison to the controls (Fig.1I, *p < 0.05 vs at time 0, #p < 0.05 vs. control, n=3). To ascertain the direct role of FSH in modulation of glucose metabolism, we investigated the role of E2 in FSH-stimulated increase in glucose uptake and glycogen synthesis in rat preovulatory GCs.FSH-stimulated glycogen was not significantly decreased in the presence of E2 inhibitor (Fig. 1J).There was significant increase in glycogen in response to E2 (*p < 0.01, n=5 Fig. 1J) which was inhibited by anti-E2 : K-7 (*p < 0.001, n=5 Fig. 1J). A significant difference between the glycogenic response ofFSH and E2 was observed (E2 : FC=1.35 0.01 vs FSH: FC=2.51 0.08, *p < 0.001, n=5 Fig. 1J). The glycogen content was significantly decreased in GCs co-treated with FSH and E2 in comparison to the cells treated with FSH alone (*p < 0.001, n=5 Fig. 1J). There was no effect of E2-inhibitor alone. Similar pattern of changes were observed in FSH-stimulated glucose uptake in the presence of E2 (Supplemental Fig. 3). 3.2 FSH-stimulated glycogen synthesis in rat preovulatory GCs is dependent on IRS-2/PI3K/AKT pathway To confirm the role of IRS-2/PI3K/Akt pathway in FSH-stimulated glycogen synthesis, firstly we checked the phosphorylation status of IRS-2. Phosphorylation of tyrosine or serine residues in IRS-2 was examined in GCs treated with FSH for 45 min. IRS-2 immunoprecipitates were analysed by Western blotting with phospho-tyrosine (p-Tyr) and phospho-serine (p-Ser)-specific monoclonal antibodies. There was a significant increase in the tyrosine phosphorylation of IRS-2 (Fig. 2A, p<0.05,n=3) and a decrease in serine phosphorylation of IRS-2 (Fig. 2B, p<0.05, n=3) in GCs treated with FSH.Further, D-[U-14C]-glucose incorporated into glycogen was measured in GCs after transfection with specific siRNA and control siRNA. A significant decrease in FSH-stimulated incorporation of D-[U- 14C]-glucose into glycogen was observed after knockdown of IRS-2, PI3K and Akt2 with siRNA specific to IRS-2, PI3K (P85) and Akt2 (Fig. 2C, p<0.001, n=3), or after treatment with LY294002 and Akt inhibitor IV (Fig. 2D, p<0.001, n=3). The serine/threonine protein kinase GSK-3β inhibits GS by phosphorylation and is considered to be a major regulator of GS. We therefore investigated the effect of FSH on the phosphorylation and deactivation of GSK3β. FSH significantly increased the phosphorylation of GSK3β (5- 10 fold) in GCs at Ser 9 site (Fig. 2E, F, p<0.001, n=3). GCs transfected with siRNA specific to IRS-2, PI3K (P85) and Akt2 showed significant reduction in FSH-stimulated phosphorylation of GSK3β at Ser 9 (Fig. 2E, p<0.001, n=3) while the cells treated with scrambled siRNA sequences maintained the basal levels of phosphorylation (Fig. 2E). The cells treated with inhibitors of PI3K (LY294002) and Akt (Akt inhibitor IV) showed marked decrease in the FSH-stimulated phosphorylation of GSK3β (Fig. 2F, p<0.001, n=3). 3.3 FSH activates Protein phosphatase 1 (PP1) The activity of GS isreversibly controlled diversity in medical practice by phosphorylation/dephosphorylation mechanism,therefore, the role of phosphatases in FSH-stimulated glycogen synthesis was examined. Significant inhibition of FSH-stimulated incorporation of D-[U-14C]-glucose into glycogen was observed after treatment of GCs with okadaic acid or calyculin A (Fig. 3A, p<0.001, n=3).PP1 is the primary phosphatase for the activation of GS by dephosphorylation of the key phosphorylated sites of GS. Therefore, the phosphatase activity was estimated in the immune complexes of PP1 isolated from control and FSH-treated GCs. Fluorescent product (DiFMU) thus formed in the presence of PP1 was measured. After 1h of treatment, FSH had significantly increased PP1 activity in rat preovulatory GCs (Fig. 3B, p<0.05, n=3). A decrease in the FSH-stimulated PP1 activity was observed after treatment of GCs with PI3K inhibitor, LY294002 or Akt inhibitor IV (Fig.
3C, p<0.05, n=3). Both okadaic acid and calyculin A abolish PP1 activity (Fig. 3C). Further, these observations were confirmed after siRNA-mediated knockdown of IRS-2, PI3K, or AKT2, where
FSH-stimulated PP1 activity was significantly inhibited in GCs (Fig. 3D, p<0.05, n=3). 3.4 Impairment of FSH-mediated glucose uptake and glycogen synthesis in GCs of PCOS patients with or without insulin resistance To understand the physio-pathological significance of the regulation of glucose uptake and glycogen content in GCs by FSH, we checked both parameters in GCs of PCOS women with and without
insulin resistance as well as normal ovulatory women. Patient biochemical and clinical features are given in Supplemental Table 1. FSH treatment for 1h resulted in significant increase in 2-NBDG uptake in 30 minin the GCs of normal women (Fig. 4A, p<0.001, n=10), but there was a significant decrease in 2-NBDG uptake in the GCs of PCOS women with and without IR (Fig. 4A, p<0.001,n=7). FSH increased the glycogen content of GCs in the normal group (Fig. 4B, 4C, 4D-upper panel,p<0.001, n=6-9), much less in GCs of non-IR women with a significant increase only at highest concentration of FSH (4 nM) (Fig. 4B, 4C, 4D-middle panel p<0.001, n=6-9), but no significant increase in PCOS IR (Fig. 4B, 4C, 4D-lower panel, n=9). The results were comparable in both experiments where D-[U-14C]-glucose (2 nM FSH, p<0.001, n=9) or 2-NBDG (2 nM FSH, n=6,p<0.001) were used as a probe.However, rate of the basal glucose uptake and glycogen synthesis were not significantly different in the GCs of normal as well as PCOS women as measured by incorporation of D-[U-14C]-glucose (Fig. 4B) and 2-NBDG (Fig. 4C).A significant number of PCOS patients with or without IR had abnormal FSH-mediated glucose uptake and glycogen synthesis (Table. 2). 3.5 Decreased glycogen content in ovaries of rat PCOS model To validate the decrease in FSH-mediated glycogen levels in PCOS GCs, we examined the glycogen content of cystic ovaries of rats (developed by treating with antiprogestin RU486, 4mg/d, for 18 days,[17]. FSH levels were normal in PCOS rats but LH levels were 2.8±0.25 fold higher than in control rats. Glycogen content in the cystic ovaries was significantly decreased in comparison to the control ovaries of the rats in the proestrus stage (Fig. 4E, p < 0.001, n=6). Since the PCOS rats showed oestrus vaginal cell morphology [33], glycogen content was also compared with that found at oestrus stage of normal rats. The glycogen content in the ovaries at oestrous stage was significantly lower than the ovaries from the proestrous stage (Fig. 4Ep < 0.001, n=6). 3.6 Inhibition of FSH-stimulated glucose uptake and glycogen synthesis by hCG To understand the role of high LH levels in FSH-regulated glucose uptake and glycogen synthesis, GCs isolated from ovarian aspirates of normal ovulatory women after gonadotropin therapy for IVF were treated with FSH alone, hCG alone, and FSH + hCG for 1h followed by incubation with 2-NBDG (200 μM) for 30 min for glucose uptake and 3h for the estimation of glycogen content. In normal human GCs, FSH stimulated the uptake of glucose (Fig. 5A, p < 0.001, n=11) as well as the synthesis of glycogen in human GCs (Fig. 5B, p < 0.001, n=10). In comparison to the response to FSH alone, glucose uptake was significantly lower in GCs co-incubated with FSH and hCG (Fig. 5A, p < 0.001, n=8). FSH-stimulated increase in glycogen content was attenuated in the presence of hCG (Fig. 5B, p < 0.001, n=9).Consistently, in normal rat preovulatory GCs, there was a robust inhibition of FSH-stimulated glucose uptake in the presence of hCG, when FSH was kept constant (Fig. 5C, IC50= 13.05 ± 1. 15 nM, p < 0.001, n=6). FSH-mediated increase in 2-NBDG incorporation in glycogen was also decreased in a dose-dependent manner in the presence of hCG (Fig. 5D, IC50= 8.97 ± 1.12 nM, p < 0.001, n=6).Further, we examined the effect of hCG in vivo, on the glycogen content in FSH-stimulated immature rat ovaries.Immature rats were treated with FSH alone, hCG alone and FSH + hCG for one day. Glycogen per mg ovarian weight was significantly more in FSH-treated rats (Fig. 5E, p < 0.01, n= 5) than in hCG-treated rats (Fig. 5E, ns, n=5) in comparison to untreated control animals. Rats treated with FSH + hCG had significantly lower glycogen content than FSH-treated immature rat ovaries (Fig.5E, p < 0.01, n= 5).Next, we examined the incorporation of D-[U-14C]-glucose into glycogen in preovulatory rat GCs in response to insulin (10 nM) for 4h and how the presence of hCG can impact it. Insulin increased the glycogen synthesis in GCs compared to untreated cells (Fig. 5E, p < 0.01, n= 3) and insulin-mediated response was significantly inhibited in the presence of hCG (5-20 nM) (Fig. 5E, p < 0.05, n=3). 3.7 Downregulation of FSH-stimulated IRS-2 expression by hCG In our earlier publication, we showed that the expression of IRS-2 was increased by FSH but not by hCG alone after 3h of treatment in vitro [17]. Therefore, we examined the effect of hCG on FSH-stimulated IRS-2 expression. In the presence of hCG, there was a significant decrease in FSH-stimulated IRS-2 mRNA expression in rat preovulatory GCs (Fig. 5G). However, hCG alone had no significant effect on IRS-2 mRNA expression as reported earlier [17]. 3.8 Attenuation of FSH-stimulated glucose uptake by hCG in HEK293 cells expressing both LHR and FSHR To confirm the impact of hCG on FSH-stimulated glucose uptake, we examined the uptake of 2-NBDG in HEK293 cells expressing both FLAG-LHR and HA-FSHR. When 2-NBDG uptake was examined in the presence of constant FSH (3.3 nM) and increasing hCG concentrations, the uptake of 2-NBDG was significantly inhibited at higher concentrations of hCG (Fig. 6A, Table 3, p<0.001,n=6). At low concentrations (0.01-0.5 nM), hCG enhanced 2-NBDG uptake by FSH (Fig. 6A, p<0.05, n=6). When the HEK293 cells expressing both FLAG-LHR and HA-FSHR were treated with constant hCG (3.3 nM) and increasing FSH concentrations, the inhibition of glucose uptake was at a much lower IC50 of FSH than IC50 of hCG in the last experiment (Fig. 6B, Table 3A, p<0.001, n=6). Basal
FSH alone or hCG alone data was obtained from the same transfections as the hCG or FSH co-treatments respectively. The concentration of hCG (3.3 nM) approximately equivalent to EC70 was
chosen so as to get adequate glucose uptake required for inhibition studies in the presence of increasing concentrations of FSH [32]. In HEK293 cells expressing both FLAG-LHR and HA-FSHR, the uptake of 2-NBDG was significantly increased by FSH alone (Fig. 6C, Table 4, p<0.001, n=6) and hCG alone (Fig. 6D, Table 3B, p<0.001, n=6). Impact of FSH and hCG on BRET change in HEK293 cells expressing both LHR and FSHR Heteromerization ofFSHR and LHR has been indicated as a probable mechanism that may attenuate FSH actions [34]. Therefore, we studied the change in BRET signals in HEK293 cells expressing both Renilla luciferase 8 (Rluc8) tagged hFSHR and mVenus tagged hLHR. The changes in intermolecular BRET signals between sensors of hFSHR-Rluc8 and hLHR-Venus were monitored upon stimulation of these cells with increasing concentrations (10 pM to 100 nM) of FSH or hCG in the presence of a constant concentration (3.3 nM) of hCG or FSH respectively. BRET signal change w.r.t untreated cells was measured. A robust increase in BRET signals was elicited in the presence of increasing concentrations of FSH in the presence of a fixed concentration of hCG (Fig. 6E, EC50 =1.12 ± 0.02 nM, p<0.0001, n=5). Although, the BRET signals increased significantly in response to varying concentrations of hCG in the presence of constant FSH (Fig. 6F, p<0.021, n=5), the concentration dependence of changes in BRET signals was not as robust as seen with varying concentration of FSH and constant hCG. The BRET signals did not increase in the presence of FSH alone (Fig. 6G) or hCG alone (Fig. 6H). Taken together, real-time kinetic analysis of BRET signals in HEK293 cells expressing hFSHR-Rluc8 and hLHR-Venus and stimulated with both hCG and FSH showed a significant increase in the proximity ofFSHR and LHR receptors leading to heteromerization. In other words, the difference in the BRET signals elicited by different ratios of FSH:hCG suggests the existence of hormone-specific regulation of FSHR-LHR heteromerization. 4. Discussion This study demonstrates for the first time the upregulation of glycogen synthesis in preovulatory GCs by FSH through IRS-2/PI3K/Akt2 pathway. Interestingly, hCG had no robust effect on glycogen synthesis, rather in preovulatory rat GCs, it was more glycolytic than FSH. These findings confirm that FSH is the key driver for glycogen synthesis through increase in glycogen synthase activity in preovulatory GCs. It is noteworthy that FSH not only increases the activity of glycogen synthase but also its protein expression in preovulatory GCs. Our earlier and the present study provide evidence that FSH stimulates the expression as well as the activity of IRS-2, which is involved in signaling cascades of insulin, IGF- 1, interleukin, IFN, growth hormone and integrins, and maybe an important component in the complex cross-talk between their receptors [22-24, 35]. This is corroborated by the fact that the absence of InsR in GCs does not adversely affect the fertility of mice in terms of pups per litter or the number of oocytes ovulated [25]. The novel mechanisms of FSH action are now becoming clear and add to our understanding about the complex processes involved in metabolic homeostasis during folliculogenesis. Apparently, signaling through a limited type of effectors may not be sufficient for FSHR and it may cross-regulate the signaling pathways of other GPCRs or RTKs, thereby leading to diverse physiological responses during follicular maturation [17, 36-38]. In preovulatory GCs, FSH-mediated glycogen stores may either have a basic role of energy reserve (as glucose or lactate) in the growing follicles, or it maybe important for other metabolic pathways required for the complex processes offolliculogenesis, oocyte maturation and ovulation. Glycogen maybe important for generation of nucleotides required for DNA repair,
proliferation of GCs and neutralizing the oxidative stress in the growing follicles. Glycogen may help in cell survival in the hypoxic environment of the preovulatory follicles [39]. In addition, glycogen may work like a metabolic sensor in preovulatory follicles due to its tight metabolic coupling with that of lipids [40-41].

We had earlier demonstrated that upregulation of IRS-2 expression by FSH was crucial for activation of PI3K/Akt pathway and glucose uptake in preovulatory rat GCs [17]. Here,we demonstrate an insulin-independent increase in tyrosine phosphorylation, decrease in Ser/Thr phosphorylation and activation of IRS-2 by FSH,however further studies are needed to understand the underlying mechanisms.IRS-2 is activated by phosphorylation of its tyrosine residues, thereby initiating the signaling cascades [42]. It is also regulated by phosphorylation status of specific serine residues leading to either decrease or increase in its activity [43]. Additionally, FSH-stimulated PKA may increase IRS-2 protein stability by phosphorylation of certain Ser/Thr residues [43]. We also need to find the specific tyrosines ofIRS-2 that are phosphorylated and whether these tyrosine residues are different from the ones phosphorylated in response to insulin [42, 44].

The physiological importance of the upregulation of glycogen synthesis by FSH was substantiated by our findings, where there were defects in FSH-stimulated glucose uptake and glycogen levels in GCs of both insulin resistant as well as non-insulin resistant PCOS patients [17]. Interestingly, insulin-stimulated glycogen synthesis through its receptor was found to be normal in PCOS patients but at higher concentrations of insulin it could not elicit an appropriate response [4, 22, 45]. The tyrosine kinase domain of InsR gene is normal in women with hyperinsulinaemia and PCOS [46]. An increase in mitogenic activity by IGF- 1 has been reported in PCOS [22]. A decrease in FSH-stimulated IRS-2 levels would limit its availability to insulin receptors. This is corroborated by our earlier findings which demonstrated a defect in FSH-stimulated expression of IRS-2 in PCOS GCs [17]. In view of normal or hyper-steroidogenic responses of insulin as well as FSH in PCOS women, these novel findings indicate that the impairment of metabolic pathways in PCOS GCs are due to defective FSH signaling caused by higher than normal levels of LH. The findings here establish the deleterious effects of high LHR activity on FSH-mediated glucose uptake and glycogen synthesis in PCOS GCs.This study provides an important model system for understanding the mechanism of selective FSH resistance in PCOS patients which may increase their susceptibility to develop insulin resistance later in life.

One of the biochemical consequences of high LH is hyperandrogenism, predominantly of ovarian origin, in PCOS patients. Several lines of evidence have linked elevated androgen levels with insulin resistance, but most studies have been inconclusive while conferring a direct role upon androgens [27, 47-50]. This is further supported by the fact that the androgen receptor (AR) antagonists are not able to reverse high testosterone-linked insulin resistance [47-48, 51]. Suppression of hypothalamic-pituitary axis with GnRH analogues did show improvement in insulin sensitivity in PCOS women with hyperandrogenism but not in all studies [52-55]. Mechanisms by which androgens may adversely affect insulin sensitivity in women with PCOS may include indirect androgenic actions or non- androgenic mechanisms including defective lipolysis in adipocytes, oxidative stress, beta cells dysfunction and increase in the secretion of insulin [48, 56-57].Not with standing, several other studies have supported the association of high testosterone with insulin sensitivity in males, but later it was elucidated to be through its conversion to E2 and its action via ER [27, 56-61]. Both physiological and genetic evidences favored the role of E2 in insulin sensitivity, but supra-physiological levels of E2 or T were found to increase insulin-stimulated inhibitory phosphorylation of IRS- 1Ser636 but these studies could not be confirmed by other groups [48-49, 62]. A robust evidence is still lacking on the direct effect of E2 and T on glucose metabolism especially in ovary or in GCs and would bean interesting aspect for further study. In a study by Gibbs and colleagues [63], healthy men treated with aromatase inhibitor showed reduction in insulin sensitivity. Intriguingly, in addition to low estrogen levels in their plasma, they had significantly high levels of LH.

To understand the complexity of FSH-mediated glucose metabolism in preovulatory GCs, we checked the effect of E2 on FSH-stimulated glycogen synthesis. E2 alone increased the glycogen synthesis but it was not as robust as observed with FSH alone. It was intriguing to find that the same concentration of E2 caused a slight, though significant, decrease in the FSH-stimulated synthesis of glycogen. Similar effect of E2 was seen on glucose uptake in GCs. Overall, our findings here show a major role of FSH in glucose uptake and glycogen synthesis in preovulatory GCs, which is impeded by abnormally high LHR activity.

Several studies have elaborated on the role of FSH, LH and insulin in the maintenance of energy substrates in the oocytes of preantral/antral stages or cumulus-oocyte complexes,however our
understanding of the regulation of glucose metabolism in the preovulatory stage is limited to the production of lactate and pyruvate [12, 64-66]. The significance of glycolysis and lactate production in gonadotropin-induced follicle maturation is still controversial [67]. Also, the role of insulin during folliculogenesis is not clear and future studies will be required to understand the selective modulation of insulin signaling by FSH and LH. Previously, Ma etal., 2015 [68] reported that high LH attenuates insulin sensitivity in adipocytes, and we found it to be true in GCs as well.

There has been a controversy on the optimal doses of LH to be included in the ovarian stimulation protocol of IVF for PCOS patients, an aspect that has confounded the investigators over the last two decades [69-70]. Exposure to high LH during early follicular phase has been shown to be associated with poor maturation of oocytes and reduced fertilization rate in PCOS women [5, 69]. To address the issues arising due to high LH, we studied the concentration-dependent effects of hCG on the FSH responses such as glucose uptake and glycogen synthesis.
Interestingly, a synergistic effect of very low concentrations of hCG on FSH-stimulated glucose uptake was observed. However, at higher concentrations, it had adverse effects on FSH-mediated glucose uptake and glycogen synthesis in human and rat GCs. A selective heterologous desensitization ofFSHR is observed in the presence of high hCG such that stimulation of IRS-2 expression by FSH was inhibited. Our findings support the relative importance of FSH and hCG in the modulation of glucose uptake and glycogen storage in GCs over the other intra-ovarian growth factors like insulin, IGF- 1, EGF, and TGF-β1 [17, 71].

Both LH and hCG bind and signal through the same receptor designated as LHR or LHCGR. Casarini and colleagues [72-74] have demonstrated that hCG is 5-fold more potent than LH in binding LHR and activating cAMP/PKA pathway in heterologous cell lines expressing LHR and gonadal cells.However, there is no significant difference in the maximal testosterone response produced by LH and hCG [73]. But, LH and hCG differentially modulate progesterone and proliferative responses in granulosa-lutein cells in vitro in the presence of FSH [75-76]. Studies are required to confirm whether FSH has different mechanisms of cross-talk with hCG and LH to modulate glucose uptake and glycogen synthesis, which maybe a limitation of the present study.

In HEK293 cells co-expressing FSHR and LHR, we observed an increase in the proximity ofFSHR and LHR and a concomitant decrease in FSH-stimulated glucose uptake, when exposed to both hCG
and FSH. A larger picture is emerging herewith a positive cross-talk ofFSHR with InsR signaling pathways through IRS-2 and a negative one with higher concentration of hCG or LH as in PCOS
patients. The glycoprotein receptors (FSHR, LHR, TSHR) have been found to make dimers or trimers [32]. While LHR and FSHR form heteromers, heteromerization is known to attenuate the hormone-dependent signaling by reducing the cAMP production [34] or prolonging the Ca2+ response of the LHR [77]. Interestingly, this is the first report of a functional correlation of the FSHR-LHR heteromerization with reduction in FSH-stimulated glucose uptake. However, further studies are required to elucidate the mechanism (s) of attenuation of FSH-stimulated glucose uptake caused by high concentrations of LH through FSHR-LHR heteromerization.

Most of the earlier studies have demonstrated the regulation of GPCRs by RTKs [34, 78]. The transactivation of heterologous receptors by a GPCR may have different consequences [79-80].
Angiotensin II increased the phosphorylation of IRS- 1 and IRS-2 through its GPCR, but attenuated the insulin-stimulated PI3K activity [80]. In contrast, insulin induced desensitization of β2Adrenergic receptor (β2AR) was in an IRS- 1/IRS-2 dependent manner [78]. Our earlier findings and the ones reported herepresent compelling evidence for an insulin-independent regulation of glucose metabolism by FSH which happens through increase in expression as well as tyrosine phosphorylation of IRS-2 leading to the upregulation of glucose uptake and glycogen synthesis in preovulatory GCs.Additionally, it is pertinent to accept the transactivation of post-receptor signaling mechanisms of insulin by FSH and its impact on insulin responses.In summary, our findings demonstrate a crucial role of FSH in glucose metabolism in preovulatory follicles. FSH is more efficient as well as potent in stimulating glucose uptake in GCs thanhCG.Storage of glucose as glycogen is regulated by FSH through IRS-2/PI3K/Akt2 pathway in preovulatory GCs. Whereas, hCG not only increases glycogen depletion but also inhibits the FSH-stimulated IRS-2 expression, glucose uptake and glycogen synthesis (Fig. 7). FSH-stimulated glucose uptake and storage are impaired in women with PCOS, indicating a selective defect in FSHR activation. Interference of the FSH-stimulated glucose uptake and storage by high LH would lead to intra-ovarian glycogen deficit in PCOS patients causing follicle growth arrest and anovulation in the PCOS patients (Fig. 7). The low levels of IRS-2 in GCs may contribute to the intra-ovarian insulin resistance in PCOS patients. Together these data identify a novel mechanism of cross-talk between the FSH, LH and insulin signaling pathways to maintain metabolic homeostasis in GCs.In conclusion, this is the first report on the FSH- and LH-mediated integrated regulation of glucose and glycogen levels in preovulatory GCs and any defect in this cross-talk may have a greater impact in the pathogenesis of PCOS (Fig. 7). These data suggest a therapeutic potential of LH antagonists in the management of metabolic syndrome in PCOS patients.