Short-chain fatty acid-butyric acid ameliorates granulosa cells inflammation through regulating METTL3-mediated N6-methyladenosine modification of FOSL2 in polycystic ovarian syndrome

A total of 18 patients, including 6 male infertility patients or patients receiving in vitro fertilization (IVF) treatment for tubal factors in the control group (control group) and 12 patients with PCOS, were recruited to participate in this study. The patients were divided into two groups; the mere obese group (FAT) and the mere hyperandrogenism group (HA) according to body mass index (BMI) and serum androgen levels. The demographic and clinical characteristics of the study participants, including age, body mass index (BMI), anti-Müllerian hormone (AMH), follicle-stimulating hormone (FSH), luteinizing hormone (LH), and testosterone (T), are presented in Table 1. Age was comparable among the three groups. However, patients with polycystic ovary syndrome (PCOS) exhibited significantly higher levels of serum AMH, LH, and LH/FSH ratio compared to the control group. No significant differences in T levels were observed between the FAT and control groups. Our findings suggest that PCOS patients have elevated plasma levels of pro-inflammatory cytokines when compared to healthy individuals. Furthermore, within the PCOS subgroups, obese patients in the FAT group showed higher levels of IL-6 (93.53 ± 12.96 pg/ml vs. 49.17 ± 9.49 pg/ml) and TNF-α (195.20 ± 27.02 pg/ml vs. 124.01 ± 11.17 pg/ml) compared to the HA group, which is consistent with previous studies [18, 19]. These results indicate a more robust inflammatory response in obese PCOS patients.

We collected the feces of three groups of patients and analyzed the distribution of microflora using 16S sequencing to explore the effect of obesity on the intestinal flora of PCOS patients. The 16 s sequencing data showed that the Operational Taxonomic Unit (OTU) numbers of the three groups were similar. The results also demonstrated that 1053 OTU were co-expressed in the three groups, as shown in Fig. 1A, B. Measures of microbial diversity (α-diversity) revealed that obesity did not affect the Simpson and Chao1 species (OTU) richness in the FAT group (Fig. 1C, D). However, Shannon’s diversity index was lower in the normal group (Fig. 1E). The β-diversity was calculated and ordinated using PCoA to compare the microbial composition between the three groups. Samples from the FAT group were differentiated from the HA and Control groups, as shown in Fig. 1F. These analyses revealed that microbial composition was significantly different between the three groups.

In this study, we analyzed the distribution of intestinal flora in the three groups of patients and found that the main types of flora in the three groups were similar at the family level. However, there were differences in their respective dominant flora (Additional file 1: Fig. S2A). For instance, Enterobacteriaceae and Bifidobacteriaceae were the predominant groups of bacteria in the intestinal microbiota in the normal group. On the other hand, Ruminococcacea and Lachnospiraceae were the dominant bacterial groups in the FAT group. The abundance of Prevotellaceae family is higher in the PCOS fecal samples compared to the control group, as shown in Fig. 2A. The dominant flora of each group at the genus level was analyzed using pairwise LEfSe analysis (LDA score > 2.0 and P < 0.05). Streptococcus, Fusobacterium, Rhizobacter, and Achromobacter were the representative flora in the FAT group, and Clostridium and Romboutsia were abundant in the HA group (Fig. 2B). The content of SCFAs in the serum of patients in each group was analyzed. The results showed that the level of butyric acid in the FAT group had decreased. However, the level of butyric acid was not significantly different between the FAT group and the HA and control groups (Fig. 2C). A combined Spearman’s rank test of the flora and SCFAs found that the increased level of Streptococcaceae and decreased Rikenellaceae affect the butyric acid in PCOS patients (Fig. 2D).

LPS was used to stimulate human ovarian granulosa tumor (KGN) cells to construct the inflammatory state of PCOS patients. The KGN cells were treated with LPS for 12 h. After LPS treatment, the cells were treated with butyric acid of different concentrations for 12 h. The effect of butyric acid on inflammatory KGN cells was observed and recorded. Subsequently, KGN cells were divided into a low-dose group (L-BA) and a high-dose group (H-BA) according to the dosage of butyric acid. The Edu method was used to detect the level of cell proliferation. The results showed that the proliferation level of KGN decreased after LPS treatment and recovered after butyric acid treatment (Fig. 3A, B). Caspase3 detection showed that the apoptotic cells increased significantly after adding LPS but decreased after adding butyric acid (Fig. 3C, D). Since several studies have shown that butyric acid can improve mitochondrial function, JC-1 was used to detect mitochondrial membrane potential. The fluorescence results showed that butyric acid could improve mitochondrial membrane potential, as shown in Fig. 3E, F. Studies have reported that butyric acid can activate PPAR-γ, which increases the expression of GLUT4, improving the glucose uptake ability of cells. Therefore, we used the 2-NBDG fluorescence tracer method to detect the glucose uptake capacity of KGN cells. The results showed that LPS decreased and butyric acid improved the glucose uptake capacity of KGN cells (Fig. 4A, B). The expression and localization of GLUT4 in KGN cells in each group were detected using an immunofluorescence assay. The immunofluorescence staining showed that the expression of GLUT4 protein was higher in butyric acid-treated groups than LPS treated groups (Fig. 4C, D). In addition, QRT-PCR and western blot showed that PPAR-γ, GLUT4 mRNA, and protein were more highly expressed in butyric acid-treated KGN cells than in LPS-treated KGN cells (Fig. 4E–G). The results indicated that butyric acid could improve the glucose metabolism of KGN cells in an inflammatory state.

A METTL3 overexpression plasmid was constructed and transferred into KGN cells, and we used WB and QPCR to detect the overexpression efficiency of METTL3 and found that METTL3 mRNA and protein were highly expressed in the METTL3- overexpression (METTL3-OE) group than in the METTL3-normal control (METTL3-NC) group (Fig. 5A–C). The m6A ratio of the METTL3-NC and METTL3-OE groups was analyzed. An m6A kit was used to detect the content of m6A in the total RNA of KGN cells in two groups. The results showed that the level of m6A modification in the METTL3-OE group was significantly higher than in the METTL3-NC group (Fig. 5D). The MERIP sequencing results showed that the enriched sequence was AAGGACC (Additional file 1: Fig. S2B). The sequence met the requirements of GGACU of the m6A modified site. Moreover, the modification site of m6A is mainly located in the 3'UTR region, as shown in Fig. 5E, F. Analyzing the MERIP sequencing results showed that FOSL2 might be a potential METTL3 regulatory target (Fig. 5G). The m6A content of FOSL2 was detected using the MERIP kit. The findings demonstrated that the m6A-modified mRNA expression of FOSL2 was increased in the METTL3 overexpression group (Fig. 5H). Two m6A modified sites were found when searching the location of GGACU in the 3’UTR section of FOSL2 in the database. There were significant differences between the two groups in motif 1, suggesting that METTL3 acts on motif 1 (Fig. 6A, B). We also detected the expression of FOSL2 mRNA and protein in KGN cells overexpressing METTL3. The results showed that the expression of FOSL2 in KGN cells overexpressing METTL3 was increased. The increase in FOSL2 was more obvious after adding LPS (Fig. 6C, D). The protein expression of NLRP3 was also detected. The findings showed that the expression trend of NLRP3 was consistent with FOSL2, as shown in Fig. 6C, D. Since studies have shown that FOSL2 can promote the expression of inflammatory factors, inflammatory factors IL-6 and TNF-α were tested. It was found that the expression trend of these two inflammatory factors was consistent with FOSL2 (Fig. 6E, F). To verify the regulatory role of FOSL2 on IL-6 and TNF-a, we used siRNA to knockdown FOSL2 in KGN cells and measured the expression of IL-6 and TNF-α (Additional file 1: Fig. S2C, S2D, S2E). We found that the protein levels of IL-6 and TNF-α were decreased when FOSL2 expression was downregulated in KGN cells (Additional file 1: Fig. S2F, S2G).

A mice model of obese PCOS was created to investigate the impact of butyric acid on ovary function. The mice were divided into four groups, namely the control group, the obese PCOS model group, the low-dose butyric acid group, and the high-dose butyric acid group. No significant difference in mice quality was observed before the experiment commenced, and no weight loss was observed after one week of intraperitoneal injection of butyric acid (Additional file 1: Fig. S2H). The ovarian histomorphology of mice in each group was different. The ovaries of mice in the control group were bright red, without expanding follicles. Histomorphological analysis of the ovaries revealed significant differences among the groups, with the ovaries of the obese PCOS model mice showing multiple follicles of varying sizes, while the control group exhibited bright red ovaries without expanding follicles. The low-dose butyric acid treatment did not significantly alter ovarian morphology, but the high-dose butyric acid treatment reduced the number of ovarian follicles, as shown in Fig. 7A. ELISA was used to detect the serum concentration of FSH, LH, T, E2, and insulin of mice in each group. It was observed that low doses of butyric acid could improve the level of hormones in PCOS mice. The level of related hormones changed significantly after taking a high concentration of butyric acid (Fig. 7B). The expression levels of PPAR-γ and GLUT4 were also detected in the ovaries of mice in each group. The protein expression levels of PPAR-γ and GLUT4 decreased in PCOS model mice and increased after taking butyric acid (Fig. 7C, D). The expression of NLRP3, IL-6 and TNF-α in the ovaries of mice decreased after taking butyric acid (Fig. 7E, F).

This study was approved by the Ethics Committee of the Affiliated Hospital of Nanjing University of Chinese Medicine (Ethics No. 2019NL-194-02). The participants signed the informed consent. The Rotterdam criteria was used to diagnose polycystic ovarian syndrome (PCOS) [54]. The exclusion criteria include: hypertension, cardiovascular disease, hyperprolactinemia, diabetes, liver or renal disease, endometriosis, ovarian tumors, ovarian insufficiency, and other conditions affecting follicular development. The standard of obese PCOS is BMI ≥ 25 kg/m [55]. In this study, the participants were not treated with Glucocorticoids, oral contraceptives, immunomodulatory drugs, insulin sensitizers, and antibiotics.

A total of 18 patients, including 6 male infertility patients or patients receiving IVF treatment for tubal factors in the control group (Control group) and 12 patients with PCOS, were recruited to participate in this study. The patients were divided into two groups; the mere obese group (FAT) and the mere hyperandrogenism group (HA) according to body mass index (BMI) and serum androgen levels. Participants were included in the HA group if their androgen levels were ≥ 70 ng/dl with a clear diagnosis of PCOS. Blood was drawn from the fasting vein 2–3 days after the onset of menstruation. Subsequently, the levels of sex hormones, including follicle-stimulating hormone (FSH), luteinizing hormone (LH), estrogen (E2), testosterone (T), AMH, IL-6, and TNF-α, were detected using chemiluminescence enzyme-linked immunosorbent assay.

To analyze the intestinal flora, 16S ribosomal ribonucleic acid (rRNA) sequencing was performed on stool samples. DNA was extracted from the stool samples of the three groups and pre-amplified and sequenced by Feng Zi Biomedical Technology (Nanjing, China). The Qiime2 diversity plug-in was used to analyze the alpha and beta diversity of the formed OTU representative sequences. The Kruskal–Wallis test and LEfSe analysis were used to analyze the significant differences between species at various classification levels.

The researchers used serum samples to identify SCFAs through gas chromatography-mass spectrometry (GC–MS). They employed an Agilent GC–MS/MS (7890B-7000D) gas chromatography-mass spectrometer, known for its sensitivity and accuracy, to carry out the analysis.

Total RNA was extracted from the KGN cells and mice ovary using the TRIzol method (Invitrogen, USA) as described in the product manual. The extracted RNA was quantified using NanoDrop (Thermo Scientific, USA). The total RNA was reverse transcribed according to the instructions on the reverse transcription kit HiScript III 1st Strand cDNA Synthesis Kit (Vazyme, China). The total RNA was reverse transcribed on ice and stored at − 80 °C. The primer was synthesized by Qing Ke Biotechnology Service Company (Additional file 2: Table S1). The cDNA obtained through reverse transcription was used as the template for the qRT-PCR reaction according to SYBR qPCR Master Mix instructions (Vazyme, China) of cDNA synthesis. After the reaction, the amplification and melting curves of Real-time PCR were confirmed, and the data were analyzed using the 2^−ΔΔCT method.

Design and synthesize siRNA targeting FOSL2, cultivate KGN cells to a transfectable state and transfect with siRNA according to the transfection instructions. After 48 h of cultivation, extract RNA and protein from KGN cells in advance and examine the mRNA and protein expression levels of FOSL2, the sequence of siRNA is provided in Additional file: Table S1. 2

Ovarian tissues were fixed with 4% paraformaldehyde (PFA) (Beyotime, China) at room temperature for 24 h and embedded in paraffin. Subsequently, the ovarian tissues were cut into 4 μm thick sections, dewaxed, hydrated, and stained with hematoxylin for 5–10 min. The ovarian histomorphological changes of mice in each group were observed under the microscope.

The cell creeper was put into a 6-well plate. When the fusion rate of KGN cells on the creeper reached 70%—80%, the cell creeper was fixed with 4% PFA for 15 min. Subsequently, the cells were permeated in 0.5% Triton X-100 for 20 min, washed thrice with phosphate buffer (PBS), and incubated with BSA at room temperature for 30 min. The cells were hybridized overnight at 4 °C with primary antibody to the anti-GLUT4 antibody (1:100) (Proteintch, China). The fluorescent-labeled secondary antibody was incubated at room temperature for 2 h and washed thrice using PBS. Then, DAPI was incubated at room temperature in the dark for 1 min and observed under a fluorescent microscope.

The total protein sample concentration was detected using the BCA method. In this study, 40 μg of each total protein sample was used for Western blotting. Proteins separated on the gel were transferred to a polyvinylidene fluoride (PVDF) membrane (250 mA, 80 min) and the membrane incubated in 5% skim milk powder at room temperature for one hour. The membranes were incubated overnight at 4 °C with PPAR-γ, GLUT4, METTL3, FOSL2, IL-6, TNF-α, NLRP3 antibodies. The next day, the membrane was washed three times with TBST and incubated with anti-horseradish peroxidase (HRP)-labeled sheep anti-rabbit IgG (1:5000) (Fudebio, China) for 2 h at room temperature. After incubation, the membrane was washed three times using TBST. The grayscale values of the protein bands were detected using chemiluminescence. The relative expression of the detected protein was expressed as the ratio of the grayscale values of the target protein to the Tubulin bands.

The human ovarian granulosa cell line, designated KGN, was cultured in DMEM/F12 medium (Invitrogen, USA) containing 10% FBS (Invitrogen, USA) and 1% penicillin/streptomycin (Invitrogen, USA) and incubated at 37 °C with 5% CO₂. The 100ug/ml of LPS in DMEM/F12 media was used in KGN cells for 24 h in 6-well cell plates. The granulosa cells were treated with several doses of butyric acid (BA) (1.1 mg/ml and 11 mg/ml) in DMEM/F12 media for 24 h in 6-well cell plates.

The ability of KGN cells to uptake glucose was evaluated by measuring the fluorescent glucose 2-NBDG (APEXBIO, USA). The KGN cells seeded in 96-well plates were cultured in DMEM/F12 medium. The cells were gently rinsed with DPBS and incubated with 10 μM 2-NBDG at 37 °C for 10 min. The fluorescent intensity of the cells was measured using fluorescence microscopy (Olympus, Japan) with a 450 nm excitation. The level of intracellular fluorescence was quantitatively analyzed using ImageJ software.

The mitochondrial membrane potential of KGN cells was measured using fluorescent probe JC-1 (Beyotime, China). The cells were incubated with JC-1 (10 μm) at 37 °C for 20 min in the dark. Subsequently, the cells were rinsed with DPBS twice. Fluorescent images of the cells were taken under a fluorescent microscope (Olympus, Japan). When the mitochondrial membrane potential is high, JC-1 aggregates in the mitochondrial matrix, forming a polymer that produces red fluorescence. When the mitochondrial membrane potential is low, JC-1 cannot accumulate in the mitochondrial matrix and exists as a monomer, resulting in green fluorescence. The level of intracellular fluorescence was quantitatively analyzed using ImageJ software. The red-to-green fluorescence ratio is often used to compare the mitochondrial membrane potential between each group of cells.

Total RNA was isolated using the TRIzol reagent (Invitrogen, USA), and 100 ng of the total RNA was used for RNA m6A quantification. The m6A methylation levels in total RNA were evaluated using the EpiQuik m6A RNA Methylation Quantification Kit (Epigentek, USA). A standard curve was constructed according to the manufacturer's instructions. The RNA samples were coated on the strip wells and incubated with capture antibodies. After incubation, the strip wells were washed with washing buffer, and detection antibody and enhanced solution were added. The detection solution created the signals for RNA m6 quantification. The m6A content was quantified by reading the absorbance of samples at 450 nm and calculated according to the standard curve.

The m6A RNA high-throughput sequencing technology service was provided by Hangzhou Lianchuang Biotechnology Co., Ltd. Methylated RNA Immunoprecipitation (MeRIP) was performed using EpiQuik MeRIP m6A Kit (Epogentek, USA) following the manufacturer's instructions. To perform MeRIP, total RNA was extracted using Trizol reagent (Invitrogen, CA, USA). Subsequently, Agilent 2100 bioanalyzer (Agilent, CA, USA) was used to analyze the quality and quantity of total RNA. The Epicentre Ribo-Zero Gold Kit (Illumina, San Diego, USA) was used to purify RNA and remove rRNA. On the other hand, m6A antibody (Synaptic Systems, Germany) was used to bind related RNA fragments, and enrich them with protein-A beans. The eluted fragment containing m6A (IP) and the untreated input control fragment were converted into the final cDNA library. Finally, sequencing was performed using an Illumina Novaseq™ 6000 platform according to the manufacturer’s instructions.

Real-time quantitative RT-PCR analysis (qRT-PCR) was performed using samples with anti-m6A antibody, normal IgG, and input samples in triplicates. The threshold cycle (CT) value of the samples with anti-m6A antibodies were normalized to input by subtracting the Ct of input from the Ct of IP samples: ∆Ct = Ct_IP − (Ct_input − Log₂[Input Dilution Factor]). Then, the percent of input for each IP sample was calculated: % Input = 2 − ΔCt _{(normalized IP).} The primers for MeRIP-qPCR are listed in Additional file 2: Table S1.

Specific-pathogen-free (SPF) grade C57/B6 mice (6–8 weeks old) were purchased from Weitong Lihua Co. (Beijing, China). The mice were housed at constant temperature and humidity, and 12 h light/dark cycle a day and provided with adequate food and water. The experimental animal ethics committee of the Affiliated Hospital of Nanjing University of Chinese Medicine approved the animal experiment.

The mice were anesthetized, and blood was drawn from the abdominal aorta and allowed to stand for 2 h. Subsequently, the blood was centrifuged to obtain serum. The serum sex hormone indexes and concentrations of FSH, LH, E2, T, and insulin were determined using ELISA kits purchased from Bioswamp (Wuhan, China). The ELISA test was performed by diluting the samples to 100μL (1:20) and incubating them with specific capture and detection antibodies according to the manufacturer’s instructions. All samples were detected at 450 nm optical density.

All data analyses were conducted using SPSS 20.0 statistical software. All data were expressed as mean ± standard deviation (SD). One-way ANOVA was used to determine the significant difference. A P value of less than 0.05 was considered significant. After finding significant differences, a post hoc analysis was performed using a Tukey honest significant difference test. The Kruskal–Wallis test was used for not normally distributed values. The Mothur software was used to analyze α Diversity and visually display species annotation results. On the other hand, β-Diversity analysis was calculated using Qiime software (v1.7.0) to evaluate differences in the species complexity between samples. The LEfSe difference analysis of species was performed to find the communities or species that significantly impact the differences between groups. The results demonstrated that the LDA threshold was 2.0.

This study was approved by the Ethics Committee of the Affiliated Hospital of Nanjing University of Chinese Medicine (Ethics No. 2019NL-194-02). The experimental animal ethics committee of the Affiliated Hospital of Nanjing University of Chinese Medicine approved the animal experiment. All animal experiments were in line with the Guide for the Care and Use of Laboratory Animal by International Committees.

Not applicable.

The authors declare that they have no competing interests.

The datasets presented in this study can be found in online repositories. The name of the repository and accession number can be found below. National Center for Biotechnology Information; PRJNA934844 and PRJNA934689. (https://www.ncbi.nlm.nih.gov/bioproject/PRJNA934844↗, https://www.ncbi.nlm.nih.gov/bioproject/PRJNA934689↗).