TGFß Gene Members and Their Regulatory Factors in Granulosa Compared to Cumulus Cells in PCOS: A Case-Control Study.

Polycystic ovary syndrome (PCOS), one of the most common causes of an-ovulatory, is a heterogeneous endocrine disorder in infertile women, affecting up to 6-20% of women of reproductive age ( 1 ). PCOS is confirmed by existing at least two of three special features based on the Rotterdam criteria: clinical/ biochemical evidence of hyperandrogenism, oligomenorrhea/ anovulation, and polycystic ovaries ( 2 ). A wide variety of genetic alterations in differential tissues and cell types impact normal follicular development in this syndrome. Imbalanced follicle-stimulating hormone (FSH), luteinizing hormone (LH), and some key factors that are involved in the inappropriate progression of folliculogenesis are the main molecular features associated with impaired and arrested development of the germinal vesicle stage of oocytes toward the dominant and mature follicles in polycystic ovaries ( 3 , 4 ). The delicate communication process between different types of cells in the mammalian follicles, namely oocyte and its surrounding somatic cells, granulosa (GCs), and cumulus cells (CCs), has a pivotal impact on the oogenesis process, oocyte maturation, cumulus expansion, and ovulation ( 5 , 6 ). Paracrine signals of the oocyte are the potent stimulators in GCs/CCs, which regulate the above physiological processes and consequently affect female fertility ( 7 ). Among the secreted factors by the oocyte or follicle somatic cells, transforming growth factor-beta (TGF-β) superfamily and its members, including anti-mullerian hormone ( AMH ), growth and differentiation factor 9 ( GDF9 ), and bone morphogenetic protein 15 ( BMP15 ) which are expressed in follicular stage-related manner, have been implicated as key regulators in various aspects of intra-ovarian development ( 8 ). On the other hand, alteration in expression of these factors, especially AMH , is related to pathological ovarian conditions, such as PCOS and Endometriosis ( 9 ). AMH is now a quantitative marker of ovarian reserve as well as a diagnostic factor for PCOS ( 10 ). TGF-β family members bind to their receptors in GCs/CCs which induces a cascade of several downstream protein-coding genes. Even though GCs are differentiated into CCs and both of them have a common progenitor during folliculogenesis, the mentioned genes show different expression and even regulation patterns in those specialized cell types ( 11 , 12 ). Based on previous studies, alterations in the gene expression pattern of granulosa-cumulus cells play a definite role in the impaired final maturation of antral follicles to later-stage growth, which is one of the main contributing factors and underlying causes of polycystic ovaries ( 13 ). AMH is secreted from ovarian cells in the growing follicles and can regulate the FSH-dependent follicle development. Previous studies showed that the expression and secretion of AMH from granulosa-cumulus cells are increased in preovulatory follicles, which contain atretic oocytes ( 14 , 15 ). GDF-9 and BMP-15, as other members of the TGF-β superfamily, are known as growth promoters during normal folliculogenesis ( 16 ). A study of intraovarian interactions, regulatory mechanisms, and possible abnormalities which interfere with the balance between these molecular factors in patients suffering PCOS compared to women with normal oogenesis can provide a better overview of controlling and monitoring ovarian function that may be resulted in improving the fertility of such patients. This study aimed to investigate the mRNA expression profile of BMP15, GDF9, AMH genes, and their respective receptors, which include: BMPR1A, BMPR1B , and BMPR2 , in GCs and CCs of PCOS patients, compared to women with normal oogenesis undergoing controlled ovarian stimulation (COS) treatment cycles.

Table 2 presents all the demographic data. Age, BMI, and duration of infertility were similar between the PCOS and control groups (P>0.05). However, there were significant differences among them considering the main PCOS features and characteristics, such as the ratio of LH/ FSH, LH and AMH levels ( Table 2 ). The PCOS subjects showed higher LH level than normo-ovulatory females (control group) (P<0.01). Demographic information The results are reported as mean ± SD. AMH; Anti-mullerian hormone, BMI; Body mass index, FSH; Follicle-stimulating hormone, TSH; Thyroid stimulating hormone, and *; P<0.05. As shown in Figure 1 , all amplified RT-PCR products were at the expected size for GDF9, BMP15, AMH, BMPR1B, BMPR1A , and BMPR2 genes. In GCs isolated from PCOS subjects, an overexpression of AMH (P<0.001) and BMPR1A (P0.001), BMP15, BMPR1B, and BMPR2 gene expression were significantly decreased (P<0.001, Fig .2A ). On the other hand, CCs from PCOS patients display the same patterns as observed in GCs genes expression increased expression of AMH (P<0.001) and BMPR1A (P<0.038); decreased expression of GDF9 (P<0.001), BMP15 (P=0.005), BMPR1B (P<0.004), and BMPR2 (P<0.001) mRNA significantly ( Fig .2B ). Results of RT-PCR for mRNA expression of GDF9, BMP15, AMH, BMPR1B, BMPR1A , and BMPR2 genes (DNA ladder 50 base pair). RT-PCR; Reverse transcription polymerase chain reaction. Expression level of GDF9, BMP15, AMH and their receptors in polycystic ovarian syndrome (PCOS) women and control group. A. Gene expression of granulosa cell from PCOS women compared to control group. B. Gene expression of cumulus cells from PCOS women compared to control group. Differential gene expression profile in granulosa compared to cumulus cells. A. Gene expression profile in granulosa cells compared to cumulus cells in polycystic ovarian syndrome (PCOS) group. B. Gene expression profile in granulosa cells compared to cumulus cells in control groups. Comparing CCs and GCs from PCOS patients revealed GDF9, BMP15 (P<0.05) and AMH (P<0.001) have higher expression in cumulus cells; in contrast, BMPR1A (P<0.05) expression level is higher in GCs of such patients ( Fig .3A ). On the other hand, BMP15 (P<0.001) and BMPR1B (P<0.03) are effe∆ively expressed in normal granulosa cells compared to normal CCs, and AMH (P<0.001) and GDF9 (P<0.001) have higher mRNA expression in CCs of control than PCOS patients ( Fig .3B ).

The differential gene expression levels of AMH and its regulatory factors that include BMP15 and GDF9 , as well as their primary receptors: BMPR2, BMPR1B , and BMPR1A , were detected in granulosa and cumulus cells in PCOS women compared to the control group. In addition, these genes showed different expression patterns in two ovarian cells (GCs and CCs). Granulosa cells show phenotypic heterogeneity in the developing ovarian Follicle. During folliculogenesis, the heterogeneous GC populations differentiate and divide into mural GCs and CCs. Their function is associated with their position in the follicles and proximity to the oocyte resulting in different properties and gene expression patterns ( 19 , 20 ). The mural GCs have a great steroidogenic function, a high level of LH receptors, and mainly are involved in the acquisition of signals from outside Follicle; however, the cumulus cells provide essential regulatory factors for oocytes to promote final cytoplasmic and nuclear maturation ( 21 , 22 ). The resumption of oocyte meiosis is regulated by activating some genes in the oocyte itself and a broad signaling network in other follicular cells ( 23 ). The increased expression of BMP15 and BMPR1B receptors in the periovulatory phase of ovarian cells showed their implication in the ovulatory process. On the other hand, these BMP proteins, the largest subfamily belonging to the TGFβ superfamily, regulate numerous main biological processes in the ovarian follicle, including cell differentiation, proliferation, and apoptosis ( 24 ). Also, the expression of BMPR1B is correlated with serum oestradiol (E2) level in normal folliculogenesis ( 25 ). Alterations in BMPs signaling are involved in some folliculogenesis disorders such as PCOS ( 24 ). In Bmp15 knockout mice, uni- and bi-lateral cysts developed, ovulation rate was decreased and female mice were subfertile due to reduced developmental potential of oocytes and ovulation ( 26 ). The reduced expression of this ligand and its receptor in PCOS patients, as shown in the present study, has revealed the association mechanism of PCOS-related anovulation and impaired oocyte competency and maturation. The study performed by Kedem-Dickman et al. ( 27 ) showed a reverse association between GC -cumulus AMH expression and oocyte maturity in preovulatory follicles. It means that corresponding GC-cumulus cells of the Germinal Vesicle oocyte stage express higher AMH mRNA levels than the Metaphase stage ones. They suggested that AMH meiosis inhibition, which had been confirmed previously ( 28 ), was the mechanism by which the oocytes may fail to resume meiosis. Some recent studies showed that high levels of AMH decrease the FSH receptors number and reduces the sensitivity of GC to FSH, which consequently prevents the provoke activation of critical genes in steroidogenesis processes such as P450scc and CYP19A1 genes ( 29 ). PCOS is a disease in which AMH has putative roles in its pathophysiology ( 30 ). Usually, it is the first cause of anovulation infertility in women ( 31 ). It seems that AMH contributes to the mechanisms leading to antral follicle arrest, anovulation, and oligomenorrhoea/ amenorrhoea ( 32 , 33 ). Taken together, these studies suggest that increased AMH levels in PCOS women create an AMHdominated microenvironment, which counteracts the action of FSH on follicles selection ( 34 ). The increased pre-antral follicle number, which is the primary source of AMH, is not merely contributing factor, but also elevated AMH concentration in PCOS is largely due to increased AMH production by individual follicles ( 35 ). These studies have shown indirect confirmation of AMH’s role in PCOS. Further observations elucidated the mechanism by which AMH expression is up-regulated in GCs. A limited number of BMPs such as BMP15 can bind to BMPR1B and BMPR2 receptors, and the activated receptors recognize SMAD1, which is translocated to the AMH promoter sequence. This mechanism significantly causes the AMH expression induction in human GCs ( 36 , 37 ). In our study, mRNA expression of BMP15 in cumulus cells of PCOS patients is higher than in granulosa ones, and AMH expression was also higher in these cells. However, when CCs/GCs of PCOS are compared to normal oogenesis ones, this pattern was not observed. The high level of AMH expression in PCOS may be due to an additional source of BMP15 secretion, especially from the corresponding oocyte, which could induce AMH expression in CCs/GCs of such patients. In PCOS, however, BMPR1A in GCs and GDF9, BMP15 and AMH in CCs have a substantial expression. These distinct alterations in the follicular microenvironment of PCOS patients may negatively interfere with intra-follicular signals such as the follicle sensitivity to FSH, compromise the dominant follicle selection and the other functions of the GC/CC cells from the viewpoint of the oocyte maturation aspect and reproductive processes ( 38 , 39 ). On the other hand, in ovarian cells, the most important receptor for AMH is BMPR-1A. Therefore some members of the TGF-β superfamily share their signaling pathway with AMH, and target genes involved in folliculogenesis and GCs differentiation could also be regulated by AMH ( 40 ). To explore whether gene expression pattern is diverse between cumulus cells and granulosa cells, we analyzed the expression profile. Data showed a differential overexpression of GDF9, BMP15 , and AMH in infertile PCOS cumulus cells in comparison to PCOS granulosa cells. However, AMH and GDF9 have the same pattern in normal cyclic cumulus cells, but BMP15 and BMPR1B significantly decreased in granulosa cells of normal patients. Therefore, the gene expression profile in both GC and CC cells showed interesting diversity. Since the same antagonist protocol for ovarian stimulation was used in both PCOS and control groups, the results were independent of the protocols.

To conclude, based on the previous results, increased gene expression of AMH and BMPR1A and decreased gene expression of GDF9, BMP15 , and BMPR1B may contribute to alterations in the downstream pathway, which are involved in ovulatory function, follicular growth, and oocyte maturation. The results of the current study are important for understanding the mechanisms of follicular growth arresting and anovulation cycles in patients suffering from PCOS. It will be considered to screen the epigenetic regulatory mechanism of the promoter regions of the TGFβ family members and corresponding receptors in future studies.

The case-control research was endorsed by the Royan Institute Ethics Boards (IR.ACECR.ROYAN. REC.1394.86). Thirty patients (15 PCOS patients with infertility and 15 normo-ovulatory patients, 22≤age≤38 years old) were recruited to the study between November 2014 and April 2016 who underwent COS for IVF/ICSI at the Infertility Clinic (Royan Institute, Iran). Informed consent were obtained from all subjects. Patients who met the following criteria were included in the control group: absence of clinical/ biochemical signs of hyperandrogenism or other endocrine diseases, polycystic ovaries, diabetes, with regular menstrual cycles (26<cycle length<32), having normal hormone profile (TSH, FSH, and LH). All controls were undergoing ovarian stimulation treatment due to their tubal factor infertility. Rotterdam diagnostic criteria are routinely used for PCOS phenotyping which includes: chronic oligo and/ or anovulation (35 days of cycle length), biochemical or clinical evidence of hyperandrogenism (total testosterone concentration >0.5 ng/ml, hirsutism, acne, and alopecia), and the presence of 12 or more ovarian cysts with 2-10 mm diameter per cyst and/or ovarian volume ≥10 cm 3 on ultrasound (polycystic ovaries). Inclusion criteria for all women (PCOS and control) were including the first IVF/ICSI cycle, ovarian stimulation by antagonist protocol, 22≤age≤38 years, and body mass index (BMI) of 18-28 kg/m 2 . The severe male factor, including oligo-astheno-teratozoospermia, retrieved sperm from testis, and semen freeze was excluded. The semen was analyzed in accordance with the criteria of the 5 th edition of the World Health Organization (WHO) in 2010. All hormone tests (FSH and LH levels) were carried out on days 2 to 3 of the menstrual cycle using electrochemiluminescence immunoassay kits (ECLIA kits, Roche Diagnostics GmbH, Germany). Further evaluation of total testosterone concentrations in PCOS patients was performed by using ELISA kits (Monobind Inc., USA), and serum AMH levels were assessed using an electrochemiluminescence immunoassay by Elecsys ® and Cobas Immunoassay Analyzers (Roche GmbH). Down-regulation of the pituitary gland with GnRH antagonist protocol [the administration of 75–150 IU recombinant FSH (Gonal F; Merck Serono, or Puregon; MSD, the Netherlands)] was done for all women up to the day of human chorionic gonadotropin (hCG, Ovitrelle®, Merck-Serono) injection. The COS was performed considering the age of the women and the antral follicle count (AFC). The following steps were taken to isolate the follicular fluid and extract cumulus and granulosa cells from collected cumulus-oocyte complex (COC) as described previously ( 17 , 18 ). Briefly, follicular fluid (FF) was obtained after hCG injection (34-36 hours) during ovarian puncture. COCs were collected from FF and repeatedly washed in culture medium G1V5™ (Vitrolife AB, Sweden), then incubated at 37°C and 5% CO 2 . Just before ICSI, the COC was denuded with hyaluronidase enzyme, then CCs washed in free enzyme medium, and cold phosphate buffered saline (PBS), then the cells were pelleted by centrifugation twice for 8 minutes at 3000 rpm; the pellet was snap-frozen in liquid nitrogen then stored at -80°C ( 17 ). After retrieval of COCs, aspirated FF was centrifuged for 10 minutes at 2000 rpm. The clear supernatant (without blood) was isolated, the pellet was solved in Tyrode’s solution, and the suspension was added on Sill select gradient and centrifuged at 3000 rpm for 13 minutes. Granulosa cells (The layer formed between Sill select gradient and supernatant) were removed and washed with 3 ml of DMEM / HamF12 medium 10%. In the next step, the suspensions were centrifuged at 3000 rpm for 13 minutes. Afterward, the diluted hyaluronidase enzyme was added, and the suspension was incubated for 3 minutes; a culture medium was added to neutralize the effect of the enzyme, then the suspension was centrifuged for 5 minutes at 1500 rpm; Red blood cell lysing buffer was added to the cell suspension and incubated for 5 minutes, the solution was removed; granulosa cells were taken for extraction of RNA. Total RNA was extracted, then purified by removing DNA genomic from the samples using the RNeasy Micro Kit (Qiagen, cat. no: 74004) as per the ‘ ‘manufacturer’s protocol which is also previously described ( 18 ). After assessment of RNA quantity and quality by using the NanoDrop spectrophotometer (NanoDrop 2000 spectrophotometer), cDNA was synthesized using 20 nanograms of total RNA according to QuantiTect Whole Transcriptome Kit (Qiagen, Cat.No:207045) instruction and stored at -20∘ C until molecular analysis. For granulosa cells’ RNA extraction, TRIzol Reagent (Sigma- Aldrich) was used following the ‘ ‘manufacturer’s protocol. Genomic DNA contaminations were removed from all extracted RNA samples by treating them by DNaseI (Fermentas, Thermo Scientific, Germany). DNase I enzyme was inactivated by incubation for 7 minutes at 65∘ C with EDTA (50 mM, Fermentas, Germany). The purity of RNA was detected by NanoDrop spectrophotometer to determine A260/A280 ratio. RNA (1 µg) was reverse transcribed to cDNA using firststrand cDNA synthesis kit (K1632- Fermentas, Thermo Scientific, Germany) instruction, then stored at -20∘ C until molecular analysis. A quantitative mRNA analysis was performed using the Step-One RT-PCR system for doing qRT-PCR (Applied Biosystems, USA). All reactions were run in duplicate. Evaluation of mRNA gene expression of GDF9, BMP15, AMH, BMPR1A, BMPR1B, BMPR2 between the patient and control groups was done by calculation of ∆C t and 2 -∆Ct . To perform this analysis, cDNAs were synthesized from granulosa and cumulus cells of patients and healthy individuals. The qPCR was performed with humanspecific primers. The product size and primer sets are listed in Table 1 . GAPDH primer was used as a control gene expression. PCR products were visualized using Gel Red ZellBio, Germany) staining and electrophoreses (1.7%-2% agarose, Paya pazhoh pars, Iran) under UV light by Molecular Imager ® Gel Doc™ XR+ (BioRad, USA). Sequence of the primers used for quantitative polymerase chain reaction (qPCR) experiments Data obtained were analyzed using the Independent Samples t test, using IBM SPSS software version 22. The one-sample Kolmogorov-Smirnov (K-S) test was used to determine whether the variables come from the normal distribution in each group. Data are presented as mean ± SD at the level of P<0.05.