A total of 60 female Sprague-Dawley (SD) rats were provided by Shanghai Jie Esprit Experimental Animal Co., Ltd. [Shanghai, China; certificate no. SCXK (Shanghai) 2013-0006]. The animals were of specific-pathogen-free (SPF) grade and were 21 days old, with a mean body weight of 58±4 g.

The 60 21-day-old female SD rats of SPF grade were randomly divided into three groups following common feeding for 2 days; In the PCOS model (PM) group (n=20), the rats were administered with dehydroepiandrosterone (DHEA; Sigma-Aldrich; Merck KGaA, Darmstadt, Germany) at 6 mg/(100 g. d) + 0.2 ml of injectable soybean oil via hypodermic injection for 35 consecutive days (19,20); in the oil control (OC) group (n=20), the rats were administered with 0.2 ml of injectable soybean oil via hypodermic injection over the same period; in the normal control (NC) group (n=20), the rats underwent no specific treatment. All animals were housed with a normal diet every day under a 12 h light/dark cycle at a temperature of 20-25°C.

The model was successfully established when rats of the PM group showed loss of their estrous cycle. All rats were detected for body weight, ovarian weight, organ coefficient and vaginal smear, and serum was collected and analyzed via an enzyme-linked immunosorbent assay (ELISA). In addition, four ovaries of each group were randomly selected for histological examination of ovarian tissues to further evaluate the efficiency of the established model.

Following the final day of modeling, all rats were weighed and then anesthetized with 2% sodium pentobarbital (30 mg/kg) for laparotomy to collect abdominal aorta blood and ovarian tissue specimens. When the 60 ovarian tissue specimens were weighed, the four ovarian specimens of each group were rapidly fixed in 4% paraformaldehyde fixative, embedded with paraffin, and cut into sections for hematoxylin and eosin (H&E) staining and immunohistochemistry (IHC). The rest were stored in a refrigerator at −80°C, for reverse transcription-quantitative polymerase chain reaction (RT-qPCR) analysis and western blotting (WB). Serum was collected following centrifugation of all blood samples at 1000 × g for 10 min.

Following 4 h of fixation in 4% paraformaldehyde fixative, the rat ovarian tissue specimens were subjected to conventional dehydration followed by paraffin-embedding and thin sectioning (4-µm thick). H&E staining was performed to examine the pathological structures of the rat ovary. IHC staining was conducted to detect the expression of FKBP52 in rat ovaries with an optical microscope (Leica DM2500, Leica Microsystems GmbH, Wetzlar, Germany).

All serum samples were manipulated according to the manufacturer’s protocol of every ELISA kit (Elabscience, Wuhan, China) for follicle stimulating hormone (FSH), luteinizing hormone (LH), estradiol (E₂), progesterone (P) and total testosterone (T). Experiments were performed in triplicate and repeated three times.

A total of eight ovarian tissue specimens from each group were milled separately with TRIzol reagent (Beyotime Institute of Biotechnology, Haimen, China) using an IKA T 10 basic ULTRA-TURRAX disperser. A NanoDrop 2000 spectrophotometer was used to measure RNA concentration following extraction of total RNA from every specimen. Subsequently, the RNA was reverse-transcribed into cDNA using a reverse transcription kit (Takara Bio, Inc., Otsu, Japan). Eventually, 1.0 µl cDNA was used as a template for PCR amplification ChamQ™ SYBR^® qPCR Master mix (High ROX Premixed) was purchased from Vazyme (Nanjing, China). The 10 µl qPCR reaction mixture included the following: cDNA 1 µl, forward primer (10 µM) 0.5 µl, reverse primer (10 µM) 0.5 µl, SYBR qPCR mix 5 µl, ddH₂O 3 µl. The primer sequences (Sangon Biotech Co., Ltd., Shanghai, China) used were as follows: GAPDH forward, 5′-GACATGCCGCCTGGAGAAAC-3′ and reverse, 5′-AGCCCAGGATGCCCTTTAGT-3′, FKBP52 upstream, 5′-CACTACACTGGCTGGCTGCT-3′ and downstream, 5′-TGGTTGCCACAGCAATATCC-3′; AR upstream, 5′-CCTTCACAGCAGCAGTCAGC-3′ and downstream, 5′-CCTGATCTGGAGGAGCTGGT-3′. The amplification conditions were as follows: Pre-denaturation at 95°C for 10 min, followed by 40 cycles of denaturation at 95°C for 15 sec and annealing at 60°C for 1 min. Data were collected and used for relative quantitative analysis with the 2^−∆∆Cq method (21). The relative mRNA expression level was obtained by comparing data of the experimental group with those of the control group. The experiments were performed in triplicate and repeated three times.

A total of eight ovarian tissue specimens from each group were milled separately with RIPA lysis buffer using IKA T 10 basic ULTRA-TURRAX. The detection of protein concentration was performed according to the manufacturer’s protocol of the Enhanced BCA Protein Assay kit (Beyotime Institute of Biotechnology) following total protein extraction from every specimen. The electrophoretic separation of known antigenic proteins were run using the sodium dodecyl sulfate polyacrylamide gel electrophoresis method; 50 µg protein were separated by 10% SDS-PAGE and then transferred onto a polyvinylidene fluoride membrane. Subsequently, the membranes were blocked in 5% bovine serum albumin (2 h at room temperature) and then incubated with primary (4°C overnight) and secondary antibodies (2 h at room temperature). Anti-FKBP52 antibody (EPR6618; cat. no. ab129097; 1:1,000), anti-androgen receptor antibody (EP670Y; cat. no. ab52615; 1:1,000) and anti-ERK1 + ERK2 antibody (EPR17526; cat. no. ab184699; 1:10,000) were purchased from Abcam (Cambridge, UK). Phospho-p44/42 MAPK (Erk1/2) (Thr202/Tyr204) (D13.14.4E) XP^® rabbit monoclonal antibody (cat. no. 4370; 1:2,000), GAPDH (14C10) rabbit monoclonal antibody (cat. no. 2118; 1:1,000) and anti-rabbit IgG and HRP-linked antibody (cat. no. 7074; 1:50,000) were purchased from Cell Signaling Technology, Inc. (Danvers, MA, USA). The blots were developed with Immobilon Western Chemiluminescent HRP Substrate (Merck KGaA). GAPDH served as an internal control. Quantitative analysis of protein expression was conducted by using ImageJ software 1.8.0 (National Institutes of Health, Bethesda, MD, USA).

A total of 10 female SD rats were provided by Shanghai Jie Esprit Experimental Animal Co., Ltd. [certificate no. SCXK (Shanghai) 2013-0006]. The animals were of SPF grade and were 21 days old, with a mean body weight of 58±4 g. All animals were housed with a normal diet every day. Every rat was administrated with pregnant mare serum gonadotropin (20 IU; Sansheng Pharmaceutical Industry Co., Ltd., Ningbo, China) through intraperitoneal injection following 2 days of common feeding. After 48 h, the animals were anesthetized with 2% sodium pentobarbital (30 mg/kg), sacrificed by cervical dislocation, and soaked in 75% alcohol for 30 min. All ovaries, which were obtained by laparotomy, were placed into PBS and transferred into DMEM-F12 medium (Thermo Fisher Scientific, Inc., Waltham, MA, USA) following washing once again with PBS.

The GCs were harvested by puncturing the follicles with microsurgical forceps, and ~95% of the GCs were negatively stained using a Trypan blue assay. The GCs were then centrifuged at 300 × g, 4°C for 5 min, filtered with a cell sieve, and centrifuged again. The pellet was resuspended at 5×10⁵/ml with DMEM-F12 medium for the RT-qPCR and WB methods, and at 5×10⁴/ml for H&E, immunocytochemistry (ICC), and immunofluorescence (IF) techniques in 6-well plates. The GCs were cultured in the incubator (37°C, 5% CO₂) and, after 72 h, the cell adherence rate was ~80%.

H&E staining was performed to observe the morphology of the GCs with an Olympus inverted microscope (Olympus Corporation, Tokyo, Japan). As FSH receptor (FSHR) is only expressed in GCs, ICC and IF were conducted to identify its purity. It met the requirements of subsequent trials when the positive rate was >95%.

The adenovirus vectors Ad-Oe-FKBP4-EGFP (2×10¹⁰ PFU/ml), Oe negative control virus (10¹¹ PFU/ml), Ad-siRNA-FKBP4-EGFP (2×10¹⁰ PFU/ml) and RNAi negative control virus (5×10¹⁰ PFU/ml), provided by Shanghai GeneChem Co., Ltd. (Shanghai, China) were constructed to transfect GCs. First, the multiplicity of infection (MOI) was ascertained by using different diluted concentrations of adenovirus vectors (MOI = virus titer x virus volume/cell numbers). The adenovirus vectors were diluted according to a concentration gradient between 10⁻² and 10⁻⁷. When cell adherence rate reached ~80% in 96-well plates, the medium was discarded. The GCs were then cultured with medium containing different concentrations of virus vectors. After 12 h, the normal medium was replaced. The expression of green fluorescence in the GCs was observed following 48 h of infection (Cell transfection rate = number of green fluorescent cells/total number of cells ×100%). When the rate reached ~80%, the virus concentration was considered the optimal dilution concentration.

The GCs were divided into three groups: Normal control group (CO group), EGFP-Oe group (EO group), and FKBP4-EGFP-Oe group (FO group). When the cell attachment rate reached ~80%, the medium was discarded. Based on previous pre-experiment results, DMEM-F12 medium was added to the CO group, DMEM-F12 medium with Oe negative control virus (2 µl) was added to the EO group, and DMEM-F12 medium with FKBP4-Oe virus (20 µl) was added to the FO group. After 12 h, the medium was replaced with normal DMEM-F12 medium, and RNA and protein were obtained following 48 h of infection.

The GCs were divided into three groups: Normal control group (CR group), EGFP-RNAi group (ER group), and FKBP4-EGFP-RNAi group (FR group). When the cell attachment rate reached ~80%, the medium was discarded. Based on previous pre-experiment results, DMEM-F12 medium was added to the CR group, DMEM-F12 medium with RNAi negative control virus (2 µl) was added to the ER group, and DMEM-F12 medium with FKBP4-RNAi virus (20 µl) was added to the FR group. After 12 h, the medium was then replaced with normal DMEM-F12 medium, and RNA and protein were obtained following 48 h of infection.

The statistical analysis was performed using GraphPad Prism 7 software (GraphPad Software, Inc., La Jolla, CA, USA). The measurement data are expressed as the mean ± standard deviation. A normality test and homogeneity of variance test were performed prior to making comparison between the groups. Comparisons of three samples were performed using one-way analysis of variance at the 0.05 level and multiple comparison between the groups was performed using Tukey method. Variables that did not meet a normal distribution were analyzed using a Kruskal-Wallis test.

At 35 days following DHEA treatment, the rats in the PM group had lost their regular estrous cycles and all of remained in the estrus period. Microscopy of stained smears of vaginal secretions showed the presence of large numbers of keratinized cells, suggesting anovulation. By contrast, the rats in the NC and OC group had a regular estrous cycle (Fig. 1A).

The body weights of the rats at 21 and 58 days among the three groups showed no statistically significant difference. However, the ovarian weight and organ coefficient of the PM group were significantly lower than those in the other two groups (Fig. 1B).

Morphological changes in the rat ovarian tissue specimens were examined by light microscopy. In the NC and OC groups, microscopic examination revealed the presence of follicles of different developmental stages and a few corpora lutea; arrangement of the GCs was orderly with an intact form, mostly in 6-8 layers. In the PM group, the number of follicles with saccular dilatation increased whereas few follicles of different developmental stages and corpora lutea were observed; the GCs were arranged loosely in ~2-3 layers, and with atresia of certain follicles. This result was consistent with PCOS characteristics (Fig. 1C).

There were no statistically significant differences in the levels of FSH, LH or LH/FSH in the rats among the three groups. However, the levels of E₂ and T in the PM group were significantly higher than those in the other two groups, and E₂/T was significantly lower. The experiments were performed in triplicate and repeated three times (Fig. 1D).

FKBP52-positive staining (yellow) was present not only in the nucleus, but also in the cytoplasm among all types of cell in the rat ovary. In the GCs, the expression of FKBP52 in the PM group was higher than in the other two groups (Fig. 2).

The mRNA expression levels of FKBP52 and AR in the rat ovarian tissues of the PM group were significantly higher than those in the NC and OC groups. There was no statistically significant difference between the NC and OC groups. The experiments were performed in triplicate and repeated three times (Fig. 3A).

The protein expression levels of FKBP52, AR and P-ERK1/2 in the rat ovarian tissues of the PM group were significantly higher than those of the NC and OC groups. There was no statistically significant difference between the NC and OC groups. The protein expression of ERK1/2 did not differ significantly among the three groups. The experiments were performed in triplicate and repeated three times (Fig. 3B).

The GCs exhibited a polygonal or cuboidal appearance under inverted phase contrast microscopy on adhering to the culture surface, proliferating, and spreading to form a monolayer (Fig. 4A). FSHR-positive staining (yellow) was present in the nucleus and the cytoplasm of the GCs (Fig. 4B). Therefore, IF staining was performed. The positive rate was >95%, therefore, the GCs extracted met the requirements of subsequent trials (Fig. 4C).

Based on the preliminary trial, appropriate results were obtained following calculation according to the formula of MOI: For the Oe negative control virus, the optimal dilution concentration was 10⁻³, MOI=200; the optimal dilution concentration of the FKBP4-Oe virus was 10⁻², MOI=400; the optimal dilution concentration of RNAi negative control virus was 10⁻³, MOI=100; the optimal dilution concentration of FKBP4-RNAi virus was 10⁻², MOI=400. The cell transfection rate was >80% (Fig. 5A-D).

The mRNA expression levels of FKBP52 and AR in the FO group were significantly higher than those in the CO group and EO group. There was no statistically significant difference between the CO group and EO group. The experiments were performed in triplicate and repeated three times (Fig. 6A).

The protein expression levels of FKBP52, AR and p-ERK1/2 in the FO group were significantly higher than those in the CO group and EO group. There was no statistically significant difference between the CO group and EO group. The protein expression of ERK1/2 did not differ significantly among the three groups. The experiment was repeated three times (Fig. 6B).

The mRNA expression levels of FKBP52 and AR in the FR group were significantly lower than those in the CR group and ER group. There was no statistically significant difference between the CR group and ER group. The experiments were performed in triplicate and repeated three times (Fig. 6C).

The protein expression levels of FKBP52, AR and ERK1/2 in the FR group were significantly lower than those in the CR group and ER group, whereas p-ERK showed the opposite result. There was no statistically significant difference between the CR group and ER group. The experiment was repeated three times (Fig. 6D).

At 35 days post-DHEA treatment, the vaginal smear (Fig. 1A), H&E staining (Fig. 1C), and ELISA results (Fig. 1D) revealed that the PCOS model was successfully established. Differences between rats and humans in the mechanism of ovulation may explain why the ovarian weights and ovarian coefficients of the PM group rats were statistically lower in than the control groups (Fig. 1B).

Previous studies on two independently derived FKBP52-deficient (52KO) mouse lines served a vital role in the reproductive development of mammals. Male 52KO mice exhibit phenotypes in accordance with partial androgen insensitivity, incorporating dysgenic prostate and seminal vesicles, ambiguous external genitalia covering hypospadias and retention of nipples into adulthood (9,13). They had reduced epididymal sperm counts and motility (22), and the sperm showed abnormal morphology. Therefore, gene knock-out experiments in male mice have revealed FKBP52 as a crucial promoter of physiological AR activity (9,13).

Female 52KO mice were yield as they exhibited normal morphology, ovulation and fertilization. The dysgenesis was the consequence of embryonic implantation and decidualization failure on account of progesterone insensitivity and uterine defects (23-25). In addition, FKBP52 may result in endometriotic lesions, with decreased expression levels of FKBP52 in patients with endometriosis, and increased cell proliferation, inflammation and angiogenesis resulting from progesterone resistance in 52KO mice (26). According to the above-mentioned findings, FKBP52 is important in reproductive development.

Previous studies have provided evidence to support a direct pathological role for AR-mediated signaling in the development of PCOS (27,28). It is an important extraovarian mediator (29). Certain findings have indicated how hyperandrogenism modulates AR ubiquitination, resulting in antral follicle growth arrest in a chronically androgenized PCOS rat model (30).

Few investigations have been performed on the impact of FKBP52 on PCOS. In a comparison of normal women of childbearing age with patients with PCOS using gene chip technology, FKBP52 protein-coding genes were different, and it may be closely associated with abnormal androgen PCOS and obesity phenotype (14). High-fat fed diabetic NONcNZO mice exhibited higher rates of peri- and post-implantation resorption, and had aberrant expression of uterine interferon-γ and progesterone receptor (PGR) and its immunophilin co-chaperone FKBP52 at nidation (30). The aberrant PGR-targeted gene expression in PCOS-like rats pre- and post-implantation overlapped with dysregulated expression of FKBP52, which was linked to endometrial dysfunction and infertility (31).

The results in the present experiments revealed that the PM group had higher mRNA and protein levels of FKBP52 (Figs. 2 and 3) and AR (Fig. 3). This suggested that the overexpression of FKBP52 and AR in the ovaries of PCOS rats may be attributed to the stimulation of hyperandrogenism.

The H&E staining showed that cells were polygonal or cuboidal under inverted phase contrast microscopy. As FSHR is a unique marker of GCs, its positive rate in IHC and IF staining was >95%. Therefore, the above results indicated that the GCs extracted met the requirements of subsequent trials (Fig. 4).

GCs have different MOIs when they meet different viral titers. An appropriate MOI not only guarantees the transfection rate, but also conserves virus usage. In the preliminary trial, when the cell transfection rate reached ~80%, the MOI was calculated using the formula described above. This met the requirements of subsequent trials (Fig. 5).

Activated AR in the cytoplasm may interact with several signaling molecules, which in turn converge on MAPK/ERK activation (15,16). In the present study, the PM rats had higher expression levels of FKBP52, AR and p-ERK1/2 in the ovary. Therefore, it was hypothesized that FKBP52 mediates AR through the MAPK/ERK pathway. The Ad-siRNA-FKBP4-EGFP and Ad-Oe-FKBP4-EGFP adenovirus vectors were constructed to verify the association of FKBP52, AR and MAPK/ERK by transfecting GCs.

When FKBP4 was overexpressed, the mRNA and expression levels of FKBP52 and AR increased and that of p-ERK1/2 increased, whereas ERK1/2 showed no change (Fig. 6A and B). Following FKBP4 silencing, the mRNA and expression levels of FKBP52 and AR decreased, that of p-ERK1/2 increased and that of ERK1/2 decreased (Fig. 6C and D). This may be interpreted as the existence of a compensatory response and this requires verification in the future. To the best of our knowledge, this is the first time the above-mentioned results have been reported. To a certain extent, these results explain the association of FKBP52, AR and MAPK/ERK, and further verification is required to confirm this.