Women with male infertility or tubal factor infertility, regular menstrual cycles, serum testosterone (T) level below 2 nmol/L, and body mass index (BMI) below 27 served as controls. PCOS was diagnosed refer to the Rotterdam revised criteria [12]. PCOS patients with high LH received oral contraceptive pretreatment (OCP) to reduce the basal LH before controlled ovarian stimulation (COS). The non-obese PCOS patients enrolled had oligo- or anovulation, serum T level above 2 nmol/L, polycystic ovary morphology and BMI below 27. The exclusion criteria were abnormal serum level of prolactin, endometriosis, dysfunctional thyroid or systemic diseases. Plasma levels of luteinizing hormone (LH), anti-Müllerian hormone (AMH), T, estradiol (E2) and follicle-stimulating hormone (FSH) were collected and measured between day 3 and day 5 of the menstrual cycle before controlled ovarian stimulation.

Serum levels of AMH, LH, FSH, T, E2 and prolactin (PRL) were determined by chemiluminescence immunoassay (CLIA). The inter-assay coefficients were 5.9% for AMH, 6.8% for LH, 5.6% for FSH, 12.1% for T and 9.2% for E2. The intra-assay coefficients were 3.6% for AMH, 5.1% for LH, 4.4% for FSH, 7.9% for T, 7.0% for E2 and 11.2% for PRL.

All patients underwent ovarian stimulation using the long gonadotrophin-releasing hormone (GnRH) agonist (Ferring, Kiel, Germany) protocol according to our previous work [13]. Briefly, patients were given daily subcutaneous injections of 0.1 mg GnRH agonist from day 20 of a menstrual cycle until the day of human chorionic gonadotrophin (hCG) trigger. Ovarian stimulation was initiated with 150 international unit (IU) recombinant human FSH (rhFSH, Merck Serono, Geneva, Switzerland) per day. Recombinant hCG (Ovidrel, Merck Serono) was injected, when at least three follicles reached $\ge$17 mm diameter. Oocyte retrieval was performed 34–36 h after hCG injection.

Follicular fluids from follicles reached $\ge$17 mm diameter were collected. After centrifugation at 340 g for 8 min to pellet the GCs, the cell pellets were resuspended with 1$\times$PBS solution (Gibco, Carlsbad, CA, USA) and subsequently layered on 40%/80% gradient solution (PureCeption, SAGE) followed by centrifugation at 320 g for 20 min. The GC layers in the interface were collected and washed with 1$\times$PBS solution at 340 g for 5 min. Then, the purified GCs were cultured as previously described [13].

For the GC pellet harvested from follicular fluids and ovaries from the murine model, total RNA was extracted by the Total RNA Kit II Kit (Omega Bio-Tek). For the primary cultured GCs, total RNA was extracted by MicroElute Total RNA Kit (Omega Bio-Tek, Norcross, GA, USA). The methods were performed from a protocol in our previous study [13]. The efficiencies of the PCR reaction were above 95%. The primers were hSIRT3-F and hSIRT3-R for human SIRT3, hGAPDH-F and hGAPDH-R for human GAPDH, mAMH-F and mAMH-R for mouse AMH, mFOXO1-F and mFOXO1-R for mouse FOXO1, mFOXO3A-F and mFOXO3A-R for mouse FOXO3A, mSIRT3-F and mSIRT3-R for mouse SIRT3, mGAPDH-F and mGAPDH-R for mouse GAPDH. The sequences of the primers are listed in Table 1.

The method was performed as described previously [13]. The murine monoclonal antibody against SIRT3 (Santa Cruz Biotechnology, sc-365175, Santa Cruz Biotechnology, Dallas, TX, USA) and beta-actin (Proteintech, 60008-1-Ig, Proteintech, Wuhan, Hubei, China) were used. The signal intensities of the bands were measured by ImageJ software (Bethesda, Maryland, NIH). The protein expression of SIRT3 was normalized to $\beta$-Actin (internal control, v1.8.0, NIH, Bethesda, MD, USA).

The intracellular ROS of GCs were examined by the general oxidative stress indicator (CM-H2DCFDA, Invitrogen). The CM-H2DCFDA was prepared in DMSO immediately before use. The cultured GCs were incubated with 10 $\mu$M CM-H2DCFDA for 30 minutes at 37 ℃ and then observed under a confocal microscope (TCS SP8, Leica, Wetzlar, Germany) with an excitation wavelength of 490 nm and an emission wavelength of 525 nm. The fluorescence intensity was analyzed and quantified by ImageJ software according to a previous study [11].

All procedures were reviewed and approved by the Animal Ethical Committee of the Laboratory Animals Center of Chongqing Medical University. Female C57BL/6J mice were purchased from Shanghai slake experimental animal Co., Ltd (Shanghai, China). Mice were bred and housed in a temperature-control room (20–22 °C, on a 12/12 h light/dark cycle. The protocol of generating a PCOS mouse model was modified from the previous study [14]. After acclimating to the laboratory conditions with a standard diet and water for 3 days, mice (19 days of age) arranging from 8.6 to 10.5 g weight were randomly divided into two groups (n = 8 per group). The dihydrotestosterone (DHT) group was implanted with a DHT release pellet (Innovative Research of America, Sarasota, FL, USA). The pellets contained 2.5 mg of DHT with a 90-day continuous release. The control group was treated with a placebo pellet. After 90 days of treatment, mice were killed by cervical dislocation, and the ovaries were collected.

Vaginal smear analysis was performed to assess the estrous cycle for 10 consecutive days before the mice were killed. Briefly, vaginal cells were collected by saline lavage and placed on a glass slide. After air-drying, the slides were stained with 0.1% methylene blue solution (Sigma-aldrich, St. Louis, MO, USA). The stages of the estrous cycle were identified by the relative ratio of cell types observed in the slides as described previously [15].

Total serum T was assayed using RIA kit with 125I-labelled ligands (Beijing North institute of Biological Technology) according to the manufacturer’s instruction. The radioactivity was measured by a Wallac Wizard 1470 automatic gamma counter (PerkinElmer). The detection efficiency was 75%. The extraction efficiency was 92%. The detection limit was 0.02 ng/mL.

Data are presented as the means $\pm$ SD. The normality and homogeneity of variance of the data were assessed. Differences between groups were determined by Student’s t-test (for only two groups) or one-way analysis of variance (ANOVA) followed by the Tukey multiple comparison test (for comparing all groups) using SPSS 16.0 software package (IBM Corp., Chicago, IL, USA). Significance was set at p 0.05.

The clinical parameters of patients were summarized in Table 2. Compared with the normovulatory controls, basal levels of AMH and T as well as antral follicle count (AFC) were significantly increased in non-obese PCOS patients (p 0.05). The insulin-glucose parameters showed the PCOS patients had insulin resistance. Moreover, there was a positive correlation between T and HOMA-IR levels. SIRT3 expression in controls and PCOS patients was examined using qRT-PCR and western blot. SIRT3 expression was significantly elevated at both mRNA and protein levels in PCOS patients compared with the controls (Fig. 1A,B). Western blot quantification revealed that SIRT3 protein expression was approximately 4-fold higher in GC of PCOS patients than in the controls (Fig. 1C). Additionally, the Pearson’s correlation was performed to evaluate the relationship between the mRNA levels of SIRT3 and total antral follicle counts, testosterone levels. We found mRNA expression of SIRT3 was correlated with testosterone levels, but not total antral follicle counts.

Fig. 1.

Expression of SIRT3 in human GCs between normovulatory women and non-obese PCOS patients. (A) Differential expression of SIRT3 mRNA in GCs between the two groups. Data were expressed as fold changes relative to the control. Bars represent means $\pm$ SD, n = 32. *p 0.05 vs the control. (B) SIRT3 protein in GCs of the two groups determined by Western blot analysis. The $\beta$-Actin was served as the internal control in both groups. (C) The signal intensities of the western blot bands were assessed by the ImageJ software (NIH). The protein expression of SIRT3 was normalized to $\beta$-Actin. Bars represent means $\pm$ SD, n = 3. *p 0.05 vs the control.

To assess the effects of hyperandrogenism, GCs from the controls and PCOS patients were treated with vehicle or DHT, a nonaromatizable androgen, and the intracellular ROS levels were measured by CM-H2DCFDA (Fig. 2A–D). Then, the fluorescence intensity was quantified. GCs of the controls showed scant ROS generation, while ROS was significantly increased in GCs from PCOS patients (Fig. 2E). And DHT treatment significantly stimulates ROS generation in GCs from both the controls and PCOS patients. Next, we performed quantitative RT-PCR analysis to assess the effects of DHT on the expression of SIRT3 (Fig. 2F). The result revealed that DHT up-regulated the expression of SIRT3.

Fig. 2.

Effects of DHT on intracellular ROS and SIRT3 expression in the luteinized GCs of the controls and PCOS patients. The primary cultured GCs were treated with vehicle or DHT (10 nM) for 24 h before further analysis. Intracellular ROS levels were measured by CM-H2DCFDA. Representative images are shown as following: (A) GCs of the controls treated with vehicle. (B) GCs of PCOS patients treated with vehicle. (C) GCs of the controls treated with DHT. (D) GCs of PCOS patients treated with DHT. The fluorescence intensities were assessed by the ImageJ software (E). Bars represent means $\pm$ SD, n = 5. Alphabets a–c indicate significant differences between the groups (p 0.05). The mRNA levels of SIRT3 were quantified with qRT-PCR analysis (F). Data were expressed as fold changes relative to the ethanol vehicle control. Bars represent means $\pm$ SD, n = 5. *p 0.05 vs the control.

For in vivo experiments, a DHT-induced PCOS murine model was established. Menstrual irregularities are a key feature of PCOS patients. We performed vaginal smears experiments to determine the estrous cycles of the mice. The control mice showed regular estrous cycles, while the DHT-treated mice exhibited disrupted estrous cycles (Fig. 3A,B). No cystic follicles were detected in the control mice. In contrast, such follicles were observed in DHT-treated mice (Fig. 3C,D). The number of antral follicles was obviously increased in DHT-treated mice compared with the controls (Fig. 3D). To confirm that the DHT treatment caused hyperandrognism, the serum T level was measured. As expected, the serum T level was dramatically raised in DHT-treated mice (Fig. 3E). The expression of specific genes such as AMH, FOXO1 and FOXO3A were associated with PCOS in ovaries [16, 17]. Our results showed that the expression of these genes was significantly enhanced in the ovaries of the DHT-treated mice compared with the controls (Fig. 3F). Consistent with the results in primary culture cells, the expression of SIRT3 was also significantly elevated in DHT-treated mice (Fig. 3F).

Fig. 3.

Ovarian morphology, serum testosterone levels and ovarian gene expression in a murine model of PCOS. (A) Representative estrous cycle of the control mice. (B) Reprentative estrous cycle of the DHT-treated mice. (C) Representative HE staining of ovarian sections from the controls and DHT-treated mice. Asterisk indicates cyst-like follicles. Triangle indicates antral follicles. # indicates corpora lutea. Scale bar = 50 $\mu$m. (D) Effects of DHT treatment on the number of antral follicles and cyst-like follicles. (E) Serum testosterone levels in the two groups. (F) Expression of AMH, FOXO1, FOXO3A and SIRT3 mRNA in ovaries determined by qRT-PCR. Bars represent means $\pm$ SD, n = 8. *p 0.05 vs the control.