Hypothalamic microglial cells were isolated from newborn female rats (age, <24 h postpartum) as previously described (14). Briefly, each newborn female animal was injected subcutaneously with 300 mg/kg pentobarbital sodium (15), cardiac arrest was used to confirm death, and then the hypothalamus tissue was excised and digested into a single-cell suspension using DMEM/F12 (Hyclone; Cytiva) containing 0.25% trypsin (Hyclone; Cytiva) for 30 min at 37°C in a 5% carbon dioxide cell incubator. The hypothalamic cell suspension was then filtrated and gradient centrifugation at 300 × g for 10 min at 37°C was performed to obtain presumptive microglia. Finally, microglial cells were sorted for anti-CD11b-FITC (1:100; cat. no. FITC-65229; Proteintech Group, Inc.) using a FACScan Flow Cytometer (BD Biosciences). The cells were divided into five groups and treated with 0, 10⁻¹², 10⁻¹⁰, 10⁻⁸ and 10⁻⁶ mol/l leuprolide acetate (LA, MedChemExpress) at 37°C in a 5% carbon dioxide cell incubator for 6, 12 and 24 h.

Thirty-two Sprague-Dawley rats were purchased from Wushi Experimental Animal Supply Co. Ltd. The animals were maintained under a 14 h light/10 h dark schedule at 24±2°C with humidity of 50±10% and free access to chow and water. The experimental protocol was in accordance with the Guide for the Care and Use of Laboratory Animals, by the Institutional Animal Care and Use Committee, Fujian Normal University (Fuzhou, China; approval no. IACUC-20180011). Each animal was anesthetized using 0.05 mg/kg atropine (Sigma-Aldrich; Merck KGaA) administered subcutaneously and 2.5 mg/kg diazepam (Sigma-Aldrich; Merck KGaA) administered intraperitoneally for deep anesthesia. The abdomen was opened, then the ovaries were removed for subsequent analysis and blood was drawn for the assessment of serum hormone concentrations. All animals were then sacrificed by cervical dislocation while still anesthetized. All efforts were made to minimize animal discomfort and to reduce the number of animals used.

Six-week-old female rats with two consecutive 4-day estrous cycles were randomly divided into 2 groups as follows: The carboxymethyl cellulose (CMC) group as a vehicle group (n=16) and a PCOS group (n=16). PCOS was induced in the PCOS group by the intragastric administration of 1.0 mg/kg/day letrozole dissolved in 1.0% CMC (2.0 ml/kg) for 21 days, while the CMC group was injected with equal volume of CMC, as previously described (16). The estrous cycles of all rats were assessed using a vaginal smear method as previously reported (1). PCOS rats were diagnosed using the Rotterdam diagnostic criteria (1), which are based on the exclusion other diseases that cause hyperandrogenism and the meeting of two of the following three points: i) Oligo-ovulation or anovulation, ii) clinical hyperandrogenism and/or hyperandrogenism, and iii) polycystic ovary.

Following modeling, the CMC and PCOS groups were further divided into CMC + quiet (CQ, n=8), CMC + exercise (CE, n=8), PCOS + quiet (PQ, n=8) and PCOS + exercise (PE, n=8) groups. After 1 week of acclimatization, all rats in the CE and PE groups were trained on the treadmill at 75% VO_2max for 60 min per day, 6 days a week for 4 consecutive weeks; during that time, all the rats in the CQ and PQ groups were allowed free movement around the cage. The treadmill training was performed between 8:00 and 10:00 daily.

The microglial cells attached to the cell slides at 37°C in a 5% carbon dioxide cell incubator for 2 days. Following washing, all slides were fixed in 4% paraformaldehyde at room temperature for 10 min followed by incubation with 0.3% H₂O₂ at room temperature for 15 min and blocked using 10% bovine serum albumin (Sigma-Aldrich; Merck KGaA) at room temperature for 45 min. The slides were incubated with anti-transmembrane protein (TMEM)119 (1:200; cat. no. NBP2-30551; Novus Biologicals, LLC), anti-ionized calcium binding adaptor molecule 1 (Iba1; 1:200; cat. no. ab178846; Abcam) and anti-GnRHR (1:50; cat. no. 19950-1-AP; Proteintech Group, Inc.) primary antibodies separately, overnight at 4°C. Following washing with PBS, the slides were incubated with CoraLite 594-conjugated goat anti-rabbit IgG secondary antibody (1:200; cat. no. SA00013-4, Proteintech Group, Inc.) at room temperature for 1 h. The nuclei were stained using DAPI at room temperature for 10 min. All slides were imaged using a fluorescence microscope.

Total mRNA was extracted from the primary cultured rat microglia using TRIzol^® solution (Life Technologies; Thermo Fisher Scientific. Inc.). The extracted mRNA samples were then reverse-transcribed using an iScript™ Select cDNA synthesis kit (Bio-Rad Laboratories, Inc.). The reverse-transcribed products were amplified using a FasQuant RT Kit (Tiangen Biotech Co., Ltd.) and the mRNA expression levels of GnRHR were assessed using 1% agarose gel electrophoresis, which was stained with ethidium bromide at room temperature for 30 min and imaged using an ultraviolet lamp. The primer sequences used were as follows: GnRHR forward (F), 5′-AGGACCCACGCAAACTACAG-3′ and reverse (R), 5′-TCCAGCAGATGACAAAGGAG-3′; and β-actin F, 5′-CGTAAAGACCTCTATGCCAACA-3′ and R, 5′-AGCCACCAATCCACACAGAG-3′. β-actin was used as a loading control. All procedures were performed according to the manufacturer's protocols.

The aforementioned rat ovaries were fixed in 4% paraformaldehyde at 4°C for 48 h and embedded using paraffin. Sections (5 µm) were cut and mounted on slides. The sections were stained at room temperature with hematoxylin for 8 min and eosin for 10 sec, and then imaged using a BX51 light microscope (Olympus Corporation).

Blood was centrifuged at 1,000 × g for 20 min to obtain serum at 4°C. The serum concentrations of testosterone (T), luteinizing hormone (LH), follicle-stimulating hormone (FSH) and GnRH were assessed using T ELISA kit (H090-1-1, Nanjing Jiancheng Bioengineering Institute), LH ELISA kit (H206-1-2, Nanjing Jiancheng Bioengineering Institute), FSH ELISA kit (H101-1-2, Nanjing Jiancheng Bioengineering Institute) and GnRH ELISA kit (H297, Nanjing Jiancheng Bioengineering Institute) according to the manufacturer's protocols.

Total protein from cells and hypothalamic tissue was extracted using RIPA lysis buffer (Beyotime Institute of Biotechnology). A total of 20 µg of protein samples were subjected to 10% SDS-PAGE gel electrophoresis and then transferred onto a PVDF membrane. These membranes were washed with TBST (0.1% Tween-20) and probed with primary antibodies (Table SI) overnight at 4°C. Following washing with TBST (0.1% Tween-20), the membrane was incubated with the corresponding secondary antibodies at 37°C for 1 h. The immunoblotting signals were assessed using enhanced chemiluminescence BeyoECL Plus buffer (P0018S, Beyotime Institute of Biotechnology) and semi-quantified by densitometry using ImageJ (version 1.44p; National Institutes of Health).

Data were presented as the mean ± SE. Differences in mean values among multiple groups were evaluated using one-way ANOVA, followed by Tukey's post-hoc test. P<0.05 was considered to indicate a statistically significant difference.

In vitro, rat microglia were identified using the marker of microglia, TMEM119 (Fig. 1A). The expression of GnRHR in microglia was demonstrated using immunocytochemistry (Fig. 1B) and reverse transcription PCR (Fig. 1C), which indicated the expression of GnRHR in rat microglia.

To the best of our knowledge there are no previous reports of the effect of GnRH agonist LA on rat hypothalamic microglia. Furthermore, due to the different concentrations of certain factors, LA may have different effects in different microglial polarization states (17); therefore, different concentrations of LA were used to treat the rat primary cultured microglia to evaluate the effects of GnRH on hypothalamic microglia. The results demonstrated that there were no significant differences in the protein expression levels of TMEM119 and GnRHR among all 5 groups following 6 h of treatment (Fig. 2A and C). The markers of microglia polarization were assessed and a significant increase in CD86 protein expression levels was demonstrated in the 10⁻⁸ and 10⁻⁶ mol/l groups compared with the 0, 10⁻¹², 10⁻¹⁰ mol/l groups (P<0.05, Fig. 2B and D) and a significant decrease in CD206 expression was identified in the 10⁻¹⁰, 10⁻⁸ and 10⁻⁶ mol/l groups compared with the control (0 mol/l) group (P<0.05, Fig. 2B and D) following 6 h of treatment. Moreover, no significant differences in the protein expression levels of TMEM119 and GnRHR were demonstrated among groups following 12 (Fig. S1A and C) and 24 (Fig. S2A and C) h of treatment, which was consistent with the results shown by each group when treated for 6 h. Further analysis also demonstrated that the CD86 and CD206 protein expression level changes among the groups when treated for 12 h (Fig. S1B and D) and 24 h treatment (Fig. S2B and D) were similar to those demonstrated following treatment for 6 h.

A PCOS rat model was diagnosed based on the Rotterdam diagnostic criteria (1). The ovarian histology results demonstrated a normal structure follicle with multi-layered (mostly 8–9 layers) granulosa cells, a normal sized follicular antrum and oocyte-corona cumulus complex in the CQ, CE and PE groups (Fig. 3A), the PQ group presented cystic follicles with fewer layers (mostly 2–4 layers) of granulosa cells, a bigger follicular antrum and a lack of oocyte-corona cumulus complex in the ovaries (Fig. 3A). Further analysis demonstrated that, compared with the CQ group, the levels of serum T (P<0.05, Fig. 3B), LH (P<0.05, Fig. 3C) and GnRH (P<0.05, Fig. 3E) were significantly higher in the PQ group, while the level of FSH (P<0.05, Fig. 3D) was significantly lower. Moreover aerobic exercise significantly decreased the levels of serum T, LH and GnRH, and significantly increased the serum FSH level in the PE group compared with PCOS rats without the aerobic exercise intervention (PQ, P<0.05; Fig. 3).

As the level of serum C reactive protein (CRP) is one of the most important properties of chronic low-grade inflammation (1,5), the present study assessed CRP protein expression levels and demonstrated it was significantly increased in the hypothalamic tissue of PQ rats compared with the CQ group (P<0.05; Fig. 4A and C). Furthermore, it was significantly decreased following aerobic exercise (PE) compared with the PQ group (P<0.05; Fig. 4A and C). Moreover, the protein expression levels of Iba1, a marker of microglia activity, was also assessed, which demonstrated that the changes in Iba1 protein expression levels were similar to the changes in the protein expression levels of CRP (Fig. 4A and C), which indicated that microglia activity was increased on the hypothalamic inflammation of PCOS rats.

To further evaluate the involvement of microglia-mediated inflammation in the hypothalamic inflammation of PCOS rats, the polarization of hypothalamic microglia was also assessed. The present study demonstrated that CD86 protein expression levels, as a marker of M1 polarized microglia, was significantly increased in the hypothalamic tissue of PQ rats compared with the CQ group (P<0.05; Fig. 4B and D), and then significantly decreased following aerobic exercise (PE) compared with the PQ group (P<0.05; Fig. 4A and C). Furthermore, the protein expression levels of CD206, a marker of M2 polarized microglia, were also examined, which demonstrated that CD206 expression was significantly decreased in the hypothalamic tissue of PQ rats compared with the CQ) group (P<0.05; Fig. 4B and D) and that CD206 protein expression levels were significantly increased following aerobic exercise (PE) compared with the PQ group (P<0.05; Fig. 4B and D). These results demonstrated that microglia polarization was involved in the hypothalamic inflammation of PCOS rats, while aerobic exercise may promote microglia to transform from M1 to M2 polarization.