Altered Expression of GDF9 and BMP15 Genes in Granulosa Cells of Diminished Ovarian Reserve Patients: A Case-Control Study.

Diminished ovarian reserve (DOR) is one of the causes of infertility in women. Subfertility is considered to be related to DOR in older women ( 1 ). Approximately 10% of women seeking fertility treatments have DOR ( 2 ). DOR is manifest heterogeneous disorder with various etiology such as infertility, genetics, autoimmune, idiopathic, and iatrogenic. Iatrogenic causes include ovarian surgery, radiation, and chemotherapy ( 2 , 3 ). DOR is associated with infertility and poor ovarian response following controlled ovarian hyperstimulation. Delays in childbearing have led to an increase in a significant number of couplesneeding assisted reproductive technology (ART) due to older age. DOR has adverse effects on ART outcomes such as decreased ovarian response to gonadotropins, and quality of oocytes and embryos, implantation rates, and live birth rates ( 4 , 5 ). Sub-fertile women are regularly screened for DOR before starting therapy because of the relationship between DOR and poor in vitro fertilization (IVF) outcome ( 6 ). The Bologna Criteria, published in 2011, defines poor ovarian response as women having at least two of the following criteria: maternal age (≥39 years), poor ovarian reserve history (DOR; ≤3 oocytes), and abnormal ovarian reserve tests (antral follicle count [AFC] <5-7 follicles or antiMüllerian hormone [AMH]<0.5-1.1 ng/ml) ( 7 ). The functional unit of the ovary is the ovarian follicle ( 8 ). Development of the ovarian follicle requires coordination between somatic cell proliferation and differentiation with oocyte growth and its maturation ( 9 ). Granulosa cells (GCs) are one of the somatic cells of the follicular environment that are closely associated with the developing oocyte, ovulation, and fertilization ( 10 ). The ovarian function resulted from feedback of gonadotropins, follicle-stimulating hormone (FSH) and luteinizing hormone (LH), and local ovarian factors ( 11 ). In addition, the transforming growth factor-beta (TGF-β) superfamily is comprised of >35 proteins that have similar common structural motivess. These proteins include activin/inhibin, growth and differentiation factor ( GDF ), bone morphogenetic protein ( BMP ) subfamilies, and AMH ( 12 ). Among these, GDF9 and BMP15 play a critical role in ovarian function. GDF9 and BMP15 are crucial for folliculogenesiswhich involved in primary, secondary, and antral follicles except primordial follicles. However, the mRNAs level of GDF9 and BMP15 detected in all stages of folliculogenesis ( 13 ). Spontaneous mutations of either GDF9 or BMP15 affect fertility in females ( 14 ). As with other members of the TGF-β superfamily, both GDF9 and BMP15 exert their biological functions by forming heteromeric complexes with types 1 and 2 receptors on the cell surface ( 13 , 15 ). GDF9 derived from the oocyte, is necessary for folliculogenesis, cumulus expansion, and GCs proliferation ( 15 ). GDF9 knockout mice were infertile because of disruptions to folliculogenesis, ( 16 ). GDF9 motivates the expression of the FSH receptor ( FSHR ), decreases expression of the LH receptor ( LHR ), prevents follicle atresia, and affects the GCs to steroidogenesis ( 17 ). BMP15 with GDF9 plays a critical role in folliculogenesis and fertility. In contrast to GDF9 , BMP15 knockout mice had subfertility because of defects in ovulation and early-stage embryonic development ( 18 ). BMP15 , during the initial and final stages of folliculogenesis, has a positive role in follicular development ( 17 ). BMP15 and GDF9 , like other TGFβ superfamily members, are translated as pre-proteins ( 19 ). In vitro studies show that both BMP15 and GDF9 can be expressed as linked homodimers or heterodimers ( 20 , 21 ). Since studies about GDF9 and BMP15 focused on premature ovarian failure and comprehensive information in DOR is not available, we aimed to assess the fold of expression of GDF9 and BMP15 as important genes involved in folliculogenesis in the GCs of DOR patients.

Table 2 demonstrates the demographic data and clinical characteristics of DOR and the NOR at baseline. There was no significant difference between the two groups in age and age at menarche. The result of IVF revealed achieved oocytes, MII, two PN were higher in the NOR group compared to DOR patients these differences were statistically significant (P<0.05, Table 2 ). Quantitative expression of these genes in GCs of DOR patients and the control group was performed by RNA extraction and cDNA synthesis with primers designed for each gene. The GAPDH primer was used as the internal control for real-time PCR. Demographic characteristics, oocyte and embryo parameters between DOR and NOR groups Data are presented as mean ± SD. *; Obtained by independent sample t test. Statistically significant level at 0.05. MI; Metaphase I, MII; Metaphase II, GV; Germinal vesicle, PN; Pronuclei, DOR; Diminished ovarian reserve, and NOR; Normal ovarian reserve. The standard curve was used to evaluate the efficiency of primers and relative quantification. For normalization of the GDF9 and BMP15 genes and drawing a standard curve, was investigated in gene database sites such as NCBI and Gene Cards. The highest expression of these genes was found in testicular tissue. Therefore, standard curves were prepared by serial dilution to evaluate the efficiency of primers and real time PCR procedures. All amplified RT-PCR products were at the expected size for GDF9 and GAPDH (housekeeping) genes. GDF9 expression in patients was 0.23 times lower than the control group,which was significant (P<0.0001, Fig .1 ). All amplified RT-PCR products were at the expected size for BMP15 and GAPDH (housekeeping) genes. BMP15 expression in patients was 0.32 times lower than the control group, which was significant (P<0.0001, Fig .1 ). Comparison of GDF9 and BMP15 expressions between DOR and NOR groups. *; Statistical significant level at 0.05, DOR; Diminished ovarian reserve, and NOR; Normal ovarian reserve. Although the serum level of AMH was significantly decreased in the case group, the mean serum level of FSH and LH on the third day of menstruation was significantly increased in cases in comparison with control participants (P<0.0001, Fig .2 ). The serum thyroid-stimulating hormone (TSH) level was not different between the two groups (P≥0.05, Fig .2 ). Mean serum hormones level between DOR and NOR groups. *; Statistical significant level at 0.05, FSH; Follicle-stimulating hormone, LH; Luteinizing hormone, AMH; Anti-Müllerian hormone, TSH; Thyroidstimulating hormone, DOR; Diminished ovarian reserve, and NOR; Normal ovarian reserve.

In our study, we assessed the expression of GDF9 and BMP15 in GCs of DOR patients with bologna criteria and 28-40 years old. In comparison with male factor patients as a control group, we observed that the mean relative expressions of GDF9 and BMP15 were significantly lower in GCs of the infertile DOR patients compared with the control group. Recent studies of genetic mutations in sheep, goats, and mice highlight the importance of oocyte-secreted factors in regulating ovarian follicular development and ovulation ( 13 , 19 , 26 ). Although some genetic causes of DOR are established, little is known about definitive gene mutations associated with DOR ( 2 ). During the procedure of folliculogenesis, interplays between the oocyte and the somatic cells that surround it (GCs) are recognized that are the enduring effects of this interaction on the potential for the stages after fertilization for embryonic development ( 23 ). The TGF-β superfamily, in particular GDF9 and BMP15 , oocyte-secreted growth factors play a critical role in ovarian organization and fertility. These factors are essential for the growth, development, and function of GCs according to the results from studies carried out on animals. Knockout mouse technology has been used over the past decade to define the essential role of ovarian gene expression and the discovery of genetic interactions. The distribution of the GDF9 , BMP15 mRNA,proteins, and BMP receptor mRNA,were evaluated in goat and bovine ovaries ( 13 , 27 , 28 ) to determine if these TGF-β members may play an important role in follicular development in goats ( 13 ). In humans, there is a lack of expression level of these genes in DOR and only polymorphisms and mutations of GDF9 are associated with DOR. In a study of Chinese women, 3 out of 139 (2.2%) women with DOR had a specific mutation (p.R146C), whereas this mutation was not present in the control group (n=159) ( 15 ). In another study, GDF9 was associated with DOR. A study compared 103 Chinese women with DOR to 123 agematched women with normal ovarian reserve. The women were analyzed for three single nucleotide polymorphisms (SNPs) of GDF9 . A higher prevalence of the GA/AA genotype was found in those with poor ovarian response (32%) during IVF cycles compared with those with the control group (19.5%) ( 29 ). Recently, Gong et al. ( 30 ) had evaluated GDF9 and BMP15 expression in GCs of poor patients according to bologna criteria. They subdivided patients via age including 40 years. Also they observed the expression of these genes decreased with the age of poor patients especially after 40 years which related to lower oocyte quality and pregnancy outcome. In the current study, the number of achieved oocytes, MII, and 2PN embryos were significantly lower in DOR patients. Since, these genes involved in folliculogenesis it seems lower expression of these genes involved in poor oocyte achievements in DOR patients. Because achieve oocyte from DOR patients is difficultanalysis was performed in limited sample size and further studies with more samples is essential. BMP15 and GDF9 are oocyte-secreted factors that we assessed expression of these genes in the targeted tissue (granolosa cells) because of limited number of oocyte. A better suggestion is to check their receptor or target genes in GCs. In future, we collect enough samples and are going to study these receptor or target genes.

In the current study, we observed the significant reductions of GDF9 and BMP15 gene expressions in GCs of DOR patients compared to the control group. The decreased oocyte numbers and 2PN embryos in the DOR group might be secondary to the decrease in the expressions of the above genes. Based on the above findings and the information we have regarding the function of these two genes, it seems that there is a role for these paracrine factors in the folliculogenesis process as well as the interaction of GCs with oocytes which further studies with larger sample size is essential for its confirmation.

This case-control study approved by the Reproductive Biomedicine Research Center Ethics Committee at Royan Institute (IR.ACECR.ROYAN.REC.1397.067). Participants were included patients admitted for infertility treatment at Royan Institute, Tehran, from 2017 to 2020. Twenty-six women (14 in the study and 12 in the control group),between 28-40 years of age undergoing IVF/intracytoplasmic sperm injection (ICSI) treatment were enrolled in this study. All patients gave their consent for collection and use of their discarded follicular fluid for research purposes. Participants were divided into two groups, DOR and control (NOR). The Bologna criteria were used to select DOR patients. Patients were considered to have DOR if they had serum AMH levels <1.1 ng/ml and/or an AFC 10 IU/l. The NOR participants were women of normal ovarian reserves with male factor infertility cause. All women had normal karyotype and those with fragile X mental retardation 1 ( FMR1 ) gene permutation, polycystic ovary syndrome (PCOS), endometriosis as well as autoimmune disorders were excluded. The antagonist protocol is increasingly used in the management of women with a DOR who undergo ARTs. All DOR patients and the control group underwent pituitary down-regulation with a gonadotropin realising hormon (GnRH) antagonist protocol. Patients received exogenous gonadotropins on the second or third day of menstruation while the leading follicle reached 13 mm in diameter, followed by a GnRH antagonist. Human chorionic gonadotropin (hCG) was prescribed when the follicles were dominant in terms of size. When follicles reached ≥ 18 mm diameters, follicles puncture was performed and GCs from follicular fluid were collected. Following follicles puncture oocytes number, quality, and embryos grading were assessed ( 22 ). The aspirated follicular fluid was centrifuged at 2000 g for 10 min, then 4 ml of salt tayrod added to the pellet, and it was slowly layered on a 50% sill select gradient and centrifuged at 3000 g for 30 min. The GCs collected and removed using a sterile transfer pipette and placed into a 15 ml tube that contained 3 mL of cold DMEM/F-12 media supplemented with 1x penicillin/streptomycin and 10% fetal bovine serum, followed by centrifugation. Enzymatic digestion with hyaluronidase enzyme was performed to disperse the GCs ( 23 ). Multinucleated giant cells were then washed in DMEM/F-12. The tube was centrifuged at 1500 rpm for 5 minutes at 21°C, and then 5 ml RBC lysing buffer (RLB) added to the pellets. The RLB solution consisted of ammonium chloride, potassium bicarbonate, and EDTA. The diluted solution was kept at room temperature for 2-5 minutes and centrifuged at 1500 rpm for 3 minutes at 21°C. Then, it was washed once with DMEM/F12. Cell counts and viability assessments were performed before the second centrifugation. The GCs were washed, centrifuged at 1500 rpm for 5 minutes, pelleted, and frozen at -80°C until RNA extraction ( 24 , 25 ). RNeasy Mini-kit (Qiagen, Valencia, CA, USA, cat. no: 74004) was used according to the manufacturer’s protocol to extract the RNA. The amount of RNA was measured by using a Nano Drop ND-1000 spectrophotometer (Thermo Scientific, Nano Drop spectrophotometer). For cDNA synthesis, we used 30 ng total RNA according to the manufacturer’s instructions in the QuantiTect Whole Transcriptome Kit (Qiagen, cat. no: 207045). Polymerase chain reaction (PCR) targets were created from the template RNA using the manufacturer’s protocols. Gene-specific primers were designed by using Perl Primer software. Then, the primers examined by Primer-BLAST to ensure that they were not linked to non-specific sites of the genome. The Primers used for real-time PCR are shown in the Table 1. Primer sets used for real-time RT-PCR mRNA quantification was performed by quantitative reverse transcription polymerase chain reaction (qRTPCR) using a Step-One RT-PCR system (Applied Biosystems, USA). Amplification of the genes of interest and GAPDH were performed in duplicate wells. GAPDH was used as the endogenous control for normalization. Statistical calculations were performed using the IBM SPSS statistic 22 software (IBM, United states). The student’s t test with a two-tailed distribution for equality of variances was used. P<0.05 indicated statistical significance. Gene expression data were analyzed using the 2 −ΔΔCt algorithm to calculate the GDF9 and BMP15 mRNAs level relative to the level of GAPDH .