Nervonic acid triggered ovarian inflammation by inducing mitochondrial oxidative stress to activate NLRP3/ IL-1β pathway.

Determined using a commercially available assay kit (Cusabio) via the ELISA method. Total RNA was extracted from tissues or cell cultures using the guanidine isothiocyanate-phenol method. The concentration, purity, and quality of RNA were assessed using a Nanodrop 2000 (Thermo Fisher). Total RNA was reverse transcribed using complementary DNA synthesis reagents. SYBR-green detection was used for mRNA level assessment, and qPCR was performed on a Light Cycler 480II qPCR system. The qPCR reaction mixture consisted of 5 μL SYBR Premix Ex Taq, 0.3 μL forward primer, 0.3 μL reverse primer, 0.4 μL double-distilled water, and 4 μL cDNA. Relative gene expression data were determined using the 2 -ΔΔCt method and normalized to the reference gene β-actin. Primers were designed using online NCBI and Primer 6.0 software. The gene and reference primer sequences used in this experiment are provided in Supplementary Table 2 . The transfection of siRNA and overexpression plasmids was performed using Lipofectamine 8000 from Invitrogen. The specific procedure was as follows: Before transfection, cells were cultured to a density of 30–50%. According to the manufacturer's instructions for Lipofectamine 8000, it was diluted with Opti-MEM I Reduced Serum Medium and mixed with siRNA. The mixture was then added to the cell culture medium. Subsequently, cells were placed in a CO 2 incubator at 37°C. After 72 h of transfection, the cells were collected to assess transfection efficiency. The siRNA and overexpression plasmids used in the experiment was designed and synthesized by Shanghai Gima Pharmaceutical Technology Co., Ltd. The specific sequences are listed in Supplementary Table 3 . Take mouse ovarian tissue and quickly place it in a 1.5 mL Eppendorf tube containing RPMI1640 medium pre-cooled to 4°C with 1% FBS. Chop the tissue into small pieces and transfer it to a small culture dish containing digestion solution with a final concentration of 0.01 mg/mL DNases and 4 mg/mL collagenase. Incubate in a 37°C cell culture incubator for 20–40 min. Terminate the digestion by adding RPMI1640 medium containing 10% FBS. Place a 70 μm cell strainer on top of a 50 mL centrifuge tube and position the tissue on the strainer. Use a 5 mL syringe plunger to gently grind the tissue on the strainer, periodically washing with FACS buffer to allow cells to pass through the strainer into the centrifuge tube. Centrifuge at 1700 rpm for 10 min at 4°C, discard the supernatant. Add 3 times the volume of ACK lysis buffer to the pellet, gently pipette to mix the pellet, and let it lyse on ice for 3–5 min. Then add FACS buffer to dilute to 30 mL. Centrifuge at 1700 rpm for 10 min at 4°C, discard the supernatant. Resuspend the cells in RPMI1640 medium containing 1% FBS. Wash the cells twice with pre-chilled staining buffer (PBS containing 1% BSA) and centrifuge at 400g using a Sigma 3-30K low-speed centrifuge. Resuspend the cells in pre-chilled staining buffer to achieve a cell concentration of 1x10^7/mL. Take 0.1 mL (100 µL) of cell suspension and transfer it to a round-bottom centrifuge tube. Add the antibodies according to the instructions, whether it's dual-stained or triple-stained antibodies, and incubate in the dark at 37°C for half an hour. Wash the cells twice with 1 mL of staining buffer to remove unbound antibodies. Centrifuge at 400g for 5 min using the Sigma 3-30K centrifuge. Carefully remove the supernatant. Proceed with machine detection using a Flow Cytometer (Brand: Biodata, Model: BeamCyte-1026). All experiments were repeated at least three times. Statistical analyses and graphical presentations were performed using R.4.2.2 and GraphPad Prism (version 9.5.0). Continuous variables are presented as means ± s.e.m. Normality tests were performed before analyses. Comparisons between two groups were performed by unpaired two-sided Student’s t -test. Comparisons among multiple groups were one-way or two-way ANOVA with Tukey’s post hoc test for continuous variables. Sample size and detailed statistical information for each experiment can be found in the corresponding Fig. legends. In vitro and in vivo experiments were conducted randomly. Data from animal and cell studies were collected in a blinded fashion. No data were excluded when performing the statistical analysis. P ＜0.05 were regarded as significant. The R and GSEA programming language software was used for RNA-seq data analysis. All calculated P values are two sided, and with P  < 0.05, P  < 0.01 and P  < 0.001 indicated as *, **, and ***, respectively; ns indicated no significant differences.

In the current study, sows in their early pregnancy with different embryo survival rates were used as an ideal model for investigating the effects of metabolic syndrome on pregnancy outcomes. The lipid metabolism levels of sows with high and low embryo survival rates were analyzed, and the correlation between the embryo survival rate and fatty acid composition was further assessed ( Fig. 1 A). Through lipid metabolomics analysis, we identified 29 serum lipid metabolites that showed significant differences in their concentrations according to the embryo survival rates ( Fig. S1 A). These differences in the metabolites were associated with the following differentially regulated pathways: lipid synthesis and degradation, ovarian steroid hormone synthesis, glycerolipid metabolism, and cholesterol metabolism ( Fig. 1 B). To further confirm the association between abnormal lipid metabolism and embryo survival rates in the sows, we employed liquid chromatography-mass spectrometry (LC-MS) to analyze the fatty acid composition of the back fat ( Fig. S2 A), abdominal fat ( Fig. S2 B), liver ( Fig. S2 C), and ovaries ( Fig. 1 C) at 28 days of pregnancy. Additionally, we examined their correlation with parameters such as embryo count, number of corpora lutea, body weight, backfat thickness, serum estradiol concentration, progesterone concentration, and leptin concentration. The results revealed a significant negative correlation between the accumulation of NA (in back fat, abdominal fat, liver, and ovaries) and the number of embryos that survived at 28 days of pregnancy ( Fig. S2 A–C, Fig. 1 C). The fitted regression curves illustrating the relationship between NA content in various sow tissues and embryo counts are presented in Fig. 1 D and Fig. S1 D–F. Furthermore, we conducted a comparative analysis of NA content in various tissues between sows with different embryo survival rates, and the results showed that the levels of NA in the ovaries, liver, serum, and back fat were significantly elevated in sows exhibiting low embryo survival rates ( Fig. 1 E). Analysis of NA-related metabolites in the serum metabolic profile also revealed aberrant NA metabolism in sows with low embryo survival rates ( Fig. S1 B). Fig. 1 The impact of abnormal NA accumulation on fertility in sows, rats, cows, PCOS women. (A) An overview of serum lipid metabolism profiles and tissue fatty acid compositions in high embryonic survival rate (HRP, embryo count > 16) and low embryonic survival rate (LRP, embryo count < 11) sows on the 28th day of pregnancy. (B) Analysis of differential metabolic pathways in the serum lipid metabolism of LRP sows, based on n = 3 biological replicates. (C) Correlation analysis between the fatty acid composition of sow ovaries on the 28th day of pregnancy and parameters such as embryo count, corpus luteum count, back fat thickness (BF), and body weight (BW), based on n = 39 biological replicates, the sizes of the dots indicate the degree of the correlation, the red dots indicate a negative correlation, while the blue dots indicate a positive correlation. (D) Regression curve illustrating the fitted relationship between nervonic acid content in sow ovaries on the 28th day of pregnancy and embryo count. (E), NA content in ovaries, liver, serum, back fat, and abdominal fat (LRP, n = 11, HRP, n = 12) of HRP and LRP sows on the 28th day of pregnancy, data are represented as mean ± s.e.m..(F) NA content in serum and ovaries of rats during estrus and early pregnancy (7th day) via tail vein injection of NA (1.65 mM·mL-1/day). Data are represented as mean ± s.e.m., based on n = 5 biological replicates. (G) Number of uterine embryos on the 7th day after tail vein injection of nervonic acid in rats. Data are represented as mean ± s.e.m., based on n = 8 biological replicates. (H) NA content in follicular fluid from Holstein cows with different follicle diameters, n = 11, 12,10. Data are represented as mean ± s.e.m.. (I) NA content in follicular fluid from Holstein cows with different genetic merit for fertility (Fert + and Fert-), n = 10 (Fert-) and 16 (Fert + ). Data are represented as mean ± s.e.m.. (J) Volcano plot of differential lipid metabolites in the serum of mice with metabolic syndrome induced by high-fat diet on the fifth day of pregnancy. n = 6. (K) NA content in plasma of patients with polycystic ovary syndrome (PCOS) compared to healthy women, based on n = 25 biological replicates, and in obese individuals compared to normal and weight individual, based on n = 25 biological replicates. and in NAFLD individuals, based on n = 69 biological replicates, and in GDM individuals, based on n = 53 biological replicates. Data are represented as mean ± s.e.m.. Statistical significance was determined using Student’s two-tailed unpaired t -test (E, F, G, I, J) or one-way ANOVA with Dunnett’s post-hoc test (H). * P ≤0.05 ** P ≤0.01, *** P ≤0.001. The impact of abnormal NA accumulation on fertility in sows, rats, cows, PCOS women. (A) An overview of serum lipid metabolism profiles and tissue fatty acid compositions in high embryonic survival rate (HRP, embryo count > 16) and low embryonic survival rate (LRP, embryo count < 11) sows on the 28th day of pregnancy. (B) Analysis of differential metabolic pathways in the serum lipid metabolism of LRP sows, based on n = 3 biological replicates. (C) Correlation analysis between the fatty acid composition of sow ovaries on the 28th day of pregnancy and parameters such as embryo count, corpus luteum count, back fat thickness (BF), and body weight (BW), based on n = 39 biological replicates, the sizes of the dots indicate the degree of the correlation, the red dots indicate a negative correlation, while the blue dots indicate a positive correlation. (D) Regression curve illustrating the fitted relationship between nervonic acid content in sow ovaries on the 28th day of pregnancy and embryo count. (E), NA content in ovaries, liver, serum, back fat, and abdominal fat (LRP, n = 11, HRP, n = 12) of HRP and LRP sows on the 28th day of pregnancy, data are represented as mean ± s.e.m..(F) NA content in serum and ovaries of rats during estrus and early pregnancy (7th day) via tail vein injection of NA (1.65 mM·mL-1/day). Data are represented as mean ± s.e.m., based on n = 5 biological replicates. (G) Number of uterine embryos on the 7th day after tail vein injection of nervonic acid in rats. Data are represented as mean ± s.e.m., based on n = 8 biological replicates. (H) NA content in follicular fluid from Holstein cows with different follicle diameters, n = 11, 12,10. Data are represented as mean ± s.e.m.. (I) NA content in follicular fluid from Holstein cows with different genetic merit for fertility (Fert + and Fert-), n = 10 (Fert-) and 16 (Fert + ). Data are represented as mean ± s.e.m.. (J) Volcano plot of differential lipid metabolites in the serum of mice with metabolic syndrome induced by high-fat diet on the fifth day of pregnancy. n = 6. (K) NA content in plasma of patients with polycystic ovary syndrome (PCOS) compared to healthy women, based on n = 25 biological replicates, and in obese individuals compared to normal and weight individual, based on n = 25 biological replicates. and in NAFLD individuals, based on n = 69 biological replicates, and in GDM individuals, based on n = 53 biological replicates. Data are represented as mean ± s.e.m.. Statistical significance was determined using Student’s two-tailed unpaired t -test (E, F, G, I, J) or one-way ANOVA with Dunnett’s post-hoc test (H). * P ≤0.05 ** P ≤0.01, *** P ≤0.001. To validate these findings, we established an in vivo rat model of abnormal NA accumulation via tail vein injections of NA (1.65 mmol/L) during estrus and early pregnancy. Through this model, we aim to simulate the abnormal accumulation of nervonic acid in the sows’ serum of LRP compared to HRP. Our results demonstrated that the increased accumulation of NA in both serum and ovaries ( Fig. 1 F) significantly impaired embryo survival rates during early pregnancy in the rats ( Fig. 1 G). Furthermore, we conducted a correlation analysis on the levels of NA in back fat and serum and reproductive performance in sows at 28 days of pregnancy and sows during estrus. We observed a significant negative correlation of the NA content in both back fat and serum with reproductive performance ( Fig. S2 A–H). These findings suggest that abnormal NA accumulation significantly diminishes embryo survival rates in both rats and sows. Next, we tried to further validate whether abnormal NA accumulation resulted in similar reproductive dysfunction in humans and other mammals. We first collected follicular fluid from Holstein cows with varying follicle diameters. Through fatty acid analysis, we observed that the NA content in dominant follicles was significantly lower than that in small follicles ( Fig. 1 H). In addition, we conducted data mining analysis, which revealed a significant decrease in NA levels in the follicular fluid of Holstein cows with high reproductive performance, in comparison with cows with low reproductive performance ( Fig. 1 I) [46] . By feeding a high-fat diet, we established an early pregnancy metabolic syndrome mouse model. We observed obvious lipid metabolism disorders in mice with HFD ( Fig. S1 C&D), as well as significantly high levels of serum NA ( Fig. 1 J). Further, through analysis of the published data, it was found that individuals with PCOS, obesity, non-alcoholic fatty liver (NAFL) or gestational diabetes mellitus (GDM), all of which are closely linked to metabolic syndrome, showed significantly heightened levels of plasma NA and its metabolites ( Fig. 1 K) [47] , [48] , [49] , [50] . Thus, abnormal NA accumulation appears under metabolic syndrome condition, which ultimately decreased fertility. The corpus luteum is a crucial functional cell in the ovary that originates from thecal and granulosa cells [51] . When stimulated by luteinizing hormone, granulosa cells undergo luteinization to form corpus lutea, which have the capability to secrete progesterone, a vital hormone for the maintenance of pregnancy [52] . Hence, we hypothesized that abnormal NA accumulation might reduce sow embryo survival rates by affecting the formation of the corpus luteum. Accordingly, correlation analysis revealed that abnormal NA accumulation was significantly negatively correlated with a decrease in the number of corpus lutea in sows with low embryo survival rates ( Fig. 2 A). Further, abnormal NA accumulation induced via tail vein injection or feed supplementation in rats caused a significant decrease in the number of corpus lutea in the ovaries ( Fig. 2 B, Fig. S4 E), and this was accompanied by a reduction in the serum levels of progesterone ( Fig. 2 C). Fig. 2 NA caused ovarian dysfunction, leading to follicular atresia. (A) Regression curve of NA content in sow ovaries vs. corpus luteum count. Based on n = 40 biological replicates. (B，C) Number of corpora lutea in rat ovaries on the 7th day of pregnancy after intravenous injection of nervonic acid (n = 8) and serum progesterone levels (n = 6). (D) Impact of NA on the mRNA expression levels of ovarian granulosa cell luteinization-related genes induced by LH and FSH as determined by qPCR. Based on n = 4 biological replicates. (E) Effect of NA on the mRNA levels of genes related to steroid hormone synthesis in porcine ovarian granulosa cells as determined by qPCR. Based on n = 12 biological replicates.(F) Measurement of estrogen levels in the supernatant of granulosa cells after NA treatment by Elisa. Based on n = 6 biological replicates. (G) Western blot analysis of LHCGR, FSHR, GDF9, StAR, CYP19A1 protein in granulosa cells after NA treatment, with GAPDH as a loading control. Based on n = 3 biological replicates ( Fig. S4 A). (H) Impact of NA treatment on the mRNA levels of genes related to follicular atresia in granulosa cells. Based on n = 12 biological replicates. (I) Western blot analysis of BAX protein in granulosa cells after NA treatment, with GAPDH as a loading control. Based on n = 3 biological replicates ( Fig. S4 F). (J) Statistical chart of changes in the diameter of rat follicles in vitro after NA treatment (left), and microscopic images of follicles on the fourth day of NA and control groups, magnified at 20 × . Based on n = 17 (Con) and 19 (NA) biological replicates (day 1). Data are represented as mean ± s.e.m.. Statistical significance was determined using Student’s two-tailed unpaired t-test (B, C, E, F, H, J) or one-way ANOVA with Dunnett’s post-hoc test (D). *P≤0.05, **P≤0.01, ***P≤0.001. NA caused ovarian dysfunction, leading to follicular atresia. (A) Regression curve of NA content in sow ovaries vs. corpus luteum count. Based on n = 40 biological replicates. (B，C) Number of corpora lutea in rat ovaries on the 7th day of pregnancy after intravenous injection of nervonic acid (n = 8) and serum progesterone levels (n = 6). (D) Impact of NA on the mRNA expression levels of ovarian granulosa cell luteinization-related genes induced by LH and FSH as determined by qPCR. Based on n = 4 biological replicates. (E) Effect of NA on the mRNA levels of genes related to steroid hormone synthesis in porcine ovarian granulosa cells as determined by qPCR. Based on n = 12 biological replicates.(F) Measurement of estrogen levels in the supernatant of granulosa cells after NA treatment by Elisa. Based on n = 6 biological replicates. (G) Western blot analysis of LHCGR, FSHR, GDF9, StAR, CYP19A1 protein in granulosa cells after NA treatment, with GAPDH as a loading control. Based on n = 3 biological replicates ( Fig. S4 A). (H) Impact of NA treatment on the mRNA levels of genes related to follicular atresia in granulosa cells. Based on n = 12 biological replicates. (I) Western blot analysis of BAX protein in granulosa cells after NA treatment, with GAPDH as a loading control. Based on n = 3 biological replicates ( Fig. S4 F). (J) Statistical chart of changes in the diameter of rat follicles in vitro after NA treatment (left), and microscopic images of follicles on the fourth day of NA and control groups, magnified at 20 × . Based on n = 17 (Con) and 19 (NA) biological replicates (day 1). Data are represented as mean ± s.e.m.. Statistical significance was determined using Student’s two-tailed unpaired t-test (B, C, E, F, H, J) or one-way ANOVA with Dunnett’s post-hoc test (D). *P≤0.05, **P≤0.01, ***P≤0.001. To further validate the hypothesis, we established a porcine granulosa cell luteinization model in which significant upregulation of the luteinization-related genes hormone/choriogonadotropin receptor (lhcgr) , follicle stimulating hormone receptor ( fshr ), cytochrome P450 family 11 subfamily a member 1 ( cyp11a1 ), cytochrome P450 family 17 subfamily a member 1( cyp17a1 ), 3β-Hydroxysteroid dehydrogenase type 1 ( hsd3β1 ), and steroidogenic acute regulatory protein ( star ) was induced in granulosa cells through luteinizing hormone (LH) and follicle-stimulating hormone (FSH) treatment. When these cells were treated with a large dose of NA, the expression of these genes was significantly decreased ( Fig. 2 D). Next, we assessed the impact of NA treatment on estrogen synthesis in granulosa cells. The results showed that elevated NA levels also significantly suppressed the expression of the key estrogen synthesis genes lhcgr , fshr , growth differentiation factor 9 ( gdf9 ), star , and cytochrome P450 family 19 subfamily a member 1 ( cyp19a1 ) ( Fig. 2 E & G, Fig. S4 A) and reduced the secretion of estrogen in the cell culture supernatant ( Fig. 2 F). These findings were corroborated by our observations in mouse granulosa cells (KK1) treated with NA and in the ovaries of rats administered NA via tail vein injection ( Fig. S4 B–D). In addition to the secretion of steroid hormones and corpus luteum formation, follicular development in the ovary is also crucial for the success of pregnancy, as it directly determines the quantity and quality of mature oocytes [53] . Therefore, we tried to determine whether heightened NA levels also affect follicular atresia. On treatment of a granulosa cell model with NA, we observed significant upregulation in the expression of key genes associated with follicular atresia (Bcl-2-associated X protein ( bax ), BH3-interacting domain death agonist ( bid ) , and caspase-6 ) ( Fig. 2 H). Analysis of protein expression also indicated a significant increase in Bax protein expression in response to NA treatment ( Fig. 2 I, Fig. S4 F). Accordingly, in an in vitro rat follicle culture model, aberrant NA accumulation was found to significantly promote follicular atresia, which in turn, impeded the development of rat follicles ( Fig. 2 J). Taken together, these data suggest that abnormal NA accumulation may lead to ovarian dysfunction and a decrease in embryo survival rates by blocking the formation of the corpus luteum via inhibition of the expression of genes related to luteinization and estrogen synthesis in granulosa cells. Next, we attempted to assess whether ovarian corpora luteum dysfunction induced by abnormal NA accumulation (observed in the previous set of experiments) was associated with increased inflammatory reactions. To this end, we conducted hematoxylin-eosin (HE) staining analysis to assess the levels of inflammation in the ovaries of sows with low embryo survival rates. As shown in Fig. 3 A, the ovaries of sows with low embryo survival rates exhibited abnormal follicular structure and loose arrangement of granulosa cell layers. In addition, we observed partial degeneration and a significant reduction in the number of granulosa cells, deep nuclear condensation, and infiltration of a small number of lymphocytes and neutrophils. Next, immunohistochemical staining analysis of macrophage-1 antigen (MAC-1) and myeloperoxidase (MPO) demonstrated that there was a significant increase in the number of macrophages and neutrophils in the ovaries of sows with low embryo survival rates ( Fig. 3 B). We further measured the expression levels of inflammatory cytokines in the ovaries of sows with different embryo survival rates and found that tumor necrosis factor-alpha (TNF-α), IL-1β, and nuclear factor-kappa B (NF-κB) were significantly upregulated in the ovaries of sows with low embryo survival rates ( Fig. 3 C). In line with these results, a higher level of NA in the mouse model, achieved via tail vein injection of NA in pregnant mice for 1 week ( Fig. S5 A), caused a significant increase in the number of B cells in the ovaries ( Fig. 3 D) and an significant increase in the serum IL-1β levels ( Fig. S5 A). Fig. 3 NA triggered ovarian inflammation, enhanced the expression of IL-1β. (A) Histological examination of ovarian slices stained with HE to assess the survival rate of high- and low-embryo survival rate sows. The scale bar is 50 μm. Based on n = 3 biological replicates. (B) Immunohistochemical staining (left) and normalized average light density statistics (right) of ovarian MAC-1 (up) and MPO (down) in high- and low-embryo survival rate sows. The scale bar is 20 μm. Based on n = 3 biological replicates. (C) qPCR analysis of the mRNA levels of tnfα, il-1β, il-6, and nf-kb in the ovaries of high- and low-embryo survival rate sows. Based on n = 10 biological replicates. (D) Representative flow cytometry plots showing mouse ovarian B cells on the 7th day of pregnancy after tail vein injection (left), and statistical analysis of the proportion of other ovarian white blood cells (right). Based on n = 3 biological replicates. (E) GSEA plot comparing control with NA treated granulosa cells for the nod-like receptor signaling pathway. NES, normalized enrichment score. (F) qPCR analysis of the mRNA levels of nlrp3, caspase1, gsdmd, il-18, and il-1β in NA-treated granulosa cells. Based on n = 6 biological replicates. (G) Western blot analysis of IL-1β protein in NA-treated granulosa cells, with GAPDH used as a loading control. Based on n = 3 biological replicates ( Fig. S5 B). (H) Western blot analysis of NLRP3/caspase1/IL-1β axi protein with NA (0.5 mM, 3 h) or LPS (1 μg/mL, 3 h) + ATP (5 mM, 45 min) or LPS (1 μg/mL, 3 h) + nigericin (10 μM, 45 min) in granulosa cells, with GAPDH used as a loading control. Based on n = 3 biological replicates. (I) Western blot analysis of NLRP3/caspase1/IL-1β axi protein with NA (0.5 mM, 3 h) or IFN39 (10 μm, 24 h) + NA (0.5 mM, 3 h) or CY-09 (10μmL, 24 h) + NA (0.5 mM, 3 h) in granulosa cells, with GAPDH used as a loading control. Based on n = 3 biological replicates. (J) qPCR analysis of il-1β mRNA levels in the corpora lutea of high- and low-embryo survival rate sows. Based on n = 10 biological replicates. (K) Elisa measurement of IL-1β content in the serum of mice after tail vein injection of NA. Based on n = 4 biological replicates. Data are represented as mean ± s.e.m.. Statistical significance was determined using Student’s two-tailed unpaired t-test. *P≤0.05, **P≤0.01, ***P≤0.001. Nominal P values and false-discovery rates (FDRs) were calculated with default method of the GSEA software (E). NA triggered ovarian inflammation, enhanced the expression of IL-1β. (A) Histological examination of ovarian slices stained with HE to assess the survival rate of high- and low-embryo survival rate sows. The scale bar is 50 μm. Based on n = 3 biological replicates. (B) Immunohistochemical staining (left) and normalized average light density statistics (right) of ovarian MAC-1 (up) and MPO (down) in high- and low-embryo survival rate sows. The scale bar is 20 μm. Based on n = 3 biological replicates. (C) qPCR analysis of the mRNA levels of tnfα, il-1β, il-6, and nf-kb in the ovaries of high- and low-embryo survival rate sows. Based on n = 10 biological replicates. (D) Representative flow cytometry plots showing mouse ovarian B cells on the 7th day of pregnancy after tail vein injection (left), and statistical analysis of the proportion of other ovarian white blood cells (right). Based on n = 3 biological replicates. (E) GSEA plot comparing control with NA treated granulosa cells for the nod-like receptor signaling pathway. NES, normalized enrichment score. (F) qPCR analysis of the mRNA levels of nlrp3, caspase1, gsdmd, il-18, and il-1β in NA-treated granulosa cells. Based on n = 6 biological replicates. (G) Western blot analysis of IL-1β protein in NA-treated granulosa cells, with GAPDH used as a loading control. Based on n = 3 biological replicates ( Fig. S5 B). (H) Western blot analysis of NLRP3/caspase1/IL-1β axi protein with NA (0.5 mM, 3 h) or LPS (1 μg/mL, 3 h) + ATP (5 mM, 45 min) or LPS (1 μg/mL, 3 h) + nigericin (10 μM, 45 min) in granulosa cells, with GAPDH used as a loading control. Based on n = 3 biological replicates. (I) Western blot analysis of NLRP3/caspase1/IL-1β axi protein with NA (0.5 mM, 3 h) or IFN39 (10 μm, 24 h) + NA (0.5 mM, 3 h) or CY-09 (10μmL, 24 h) + NA (0.5 mM, 3 h) in granulosa cells, with GAPDH used as a loading control. Based on n = 3 biological replicates. (J) qPCR analysis of il-1β mRNA levels in the corpora lutea of high- and low-embryo survival rate sows. Based on n = 10 biological replicates. (K) Elisa measurement of IL-1β content in the serum of mice after tail vein injection of NA. Based on n = 4 biological replicates. Data are represented as mean ± s.e.m.. Statistical significance was determined using Student’s two-tailed unpaired t-test. *P≤0.05, **P≤0.01, ***P≤0.001. Nominal P values and false-discovery rates (FDRs) were calculated with default method of the GSEA software (E). To further investigate the potential relationship between NA and ovarian inflammation, we conducted RNA-seq to analyze the impact of aberrant NA accumulation on granulosa cell transcription. The sequencing results revealed that NA treatment significantly upregulated 103 genes and downregulated 452 genes in porcine ovarian granulosa cells ( Fig. S6 A). GO analysis showed that the significantly upregulated genes were involved in biological processes related to oxidative stress, interleukins, immunity, inflammation, and cytokines, among others. In contrast, the significantly downregulated genes were involved in processes related to steroid biosynthesis, hormonal regulation, and oxidative phosphorylation enzyme activity ( Fig. S6 B). Gene set enrichment analysis (GSEA) revealed that NA treatment significantly upregulated the NLRP signaling ( Fig. 3 E), oxidative phosphorylation, and lipid metabolism pathways ( Fig. S6 C), while it significantly downregulated the JAK-STAT, TGF-β, cytokine-cytokine receptor interaction, neuroactive ligand-receptor interaction, and natural killer cell-mediated cytotoxicity pathways ( Fig. S6 D). Thus, NA treatment affected the transcription of granulosa cell genes related to inflammation, among other pathways. Previous studies have indicated that NLRP3 is a classic inflammasome in the NLRP signaling pathway [27] , [54] . Therefore, using transcriptomic data, we analyzed the activation status of NLRP3-related signaling pathways in porcine ovarian granulosa cells following NA treatment. The results showed that NA treatment significantly activated the gene expression levels of nlrp3 , caspase1 , gasdermin d ( gsdmd ), il-18, and il-1β ( Fig. 3 F) and promoted the expression of il-1β ( Fig. 3 G and Fig. S5 B). By employing activators of NLRP3 (LPS + APT or Nigericin [32] ) and inhibitors (IFN39 [55] or CY-09 [56] ), we determined that NA induces elevated levels of IL-1β in porcine ovarian granulosa cells through the NLRP3/caspase/IL-1β axis ( Fig. 3 H and I). Consistent with these results, IL-1β expression was significantly increased in the corpus luteum and liver of sows with low embryonic survival rates ( Fig. 3 J and Fig. S5 C), and this increase in serum IL-1β levels ( Fig. 3 K), IL-18 and Mip-1α ( Fig. S5 D) was also observed in rats that were administered NA via tail vein injection. Collectively, these results suggest that abnormal accumulation of NA in the ovaries may induce ovarian inflammation by promoting the activation of NLRP3 and the expression of IL-1β. Previous studies have indicated that NLRP3 activation can be induced by mitochondrial oxidative stress [26] . Imbalance in intracellular lipid homeostasis also constitutes a significant contributor to mitochondrial oxidative stress [57] . Therefore, we further investigated whether NA could induce an increase in ROS levels in granulosa cells. Utilizing ROS staining, we observed a significant elevation of KK1 mitochondrial ROS upon NA treatment ( Fig. 4 A), accompanied by a marked reduction in mtDNA content, ATP levels, and mitochondrial membrane potential ( Fig. 4 B-D). By employing the ROS scavenger Mito-Tempo, we confirmed that reducing ROS production could alleviate NA-induced damage to KK1 mitochondrial function ( Fig. 4 E-H). Moreover, inhibiting the increase in ROS effectively suppressed NLRP3 activation induced by NA. Additionally, our findings revealed that elevated NA significantly inhibited the expression of NDUFB8 protein from mitochondrial respiratory chain complex I and Cytb protein from mitochondrial respiratory chain complex III. It also significantly downregulated the protein abundances of ATP5A and UQCRC2, which are two key factors for ATP synthesis ( Fig. 4 I). Thus, our results demonstrated that NA induced oxidative stress, through destroying mitochondrial respiratory chain integrity, to stimulate NLRP3 activation in the granulosa cells. Fig. 4 NA induced mitochondrial ROS to mediate the activation of NLRP3. (A) Mitochondrial ROS staining was observed by fluorescence microscope after KK1 was treated with 2.5 μM rotenone or 0.5 mM NA for 6 h. Based on n = 3 biological replicates. (B) The fluorescence value of KK1 ATP level treated with 2.5 μM rotenone or 0.5 mM NA for 6 h was detected by microplate reader. Based on n = 6 biological replicates. (C) mtDNA content after 2.5 μM rotenone or 0.5 mM NA treatment 6 h in KK1. Based on n = 3 biological replicates. (D) TMRM staining of mitochondrial membrane potential by fluorescence microscopy after KK1 was treated with 2.5 μM rotenone or 0.5 mM NA for 2 h. Based on n = 3 biological replicates. (E) The fluorescence value of KK1 mitochondrial membrane potential treated with 2.5 μM rotenone or 0.5 mM NA for 2 h was detected by microplate reader. Based on n = 6 biological replicates. (F) Mitochondrial ROS staining was observed by fluorescence microscope after KK1 was treated with 10 μM Mito-Tempo 24 h + 0.5 mM NA 6 h or 0.5 mM NA for 6 h. Based on n = 3 biological replicates. (G) The fluorescence value of KK1 ATP level treated with 10 μM Mito-Tempo 24 h + 0.5 mM NA 6 h or 0.5 mM NA for 6 h was detected by microplate reader. Based on n = 6 biological replicates. (H) mtDNA content after 10 μM Mito-Tempo 24 h + 0.5 mM NA 6 h or NA treatment 6 h in KK1. Based on n = 3 biological replicates. (I) TMRM staining of mitochondrial membrane potential by fluorescence microscopy after KK1 was treated with 10 μM Mito-Tempo 24 h + 0.5 mM NA 2 h or 0.5 mM NA for 2 h. Based on n = 3 biological replicates. (J) The fluorescence value of KK1 mitochondrial membrane potential treated with 10 μM Mito-Tempo 24 h + NA 2 h or 0.5 mM NA for 2 h was detected by microplate reader. Based on n = 6 biological replicates. (K) Western blot analysis of NLRP3, Cytb, ATP5A, UQCRC2, NDUFB8 protein with NA (0.5 mM, 6 h) or 10 μM Mito-Tempo 24 h + NA 6 h in KK1, with GAPDH used as a loading control. Based on n = 3 biological replicates. Data are represented as mean ± s.e.m.. Statistical significance was determined using Student’s two-tailed unpaired t -test. * P ≤0.05, ** P ≤0.01, *** P ≤0.001. NA induced mitochondrial ROS to mediate the activation of NLRP3. (A) Mitochondrial ROS staining was observed by fluorescence microscope after KK1 was treated with 2.5 μM rotenone or 0.5 mM NA for 6 h. Based on n = 3 biological replicates. (B) The fluorescence value of KK1 ATP level treated with 2.5 μM rotenone or 0.5 mM NA for 6 h was detected by microplate reader. Based on n = 6 biological replicates. (C) mtDNA content after 2.5 μM rotenone or 0.5 mM NA treatment 6 h in KK1. Based on n = 3 biological replicates. (D) TMRM staining of mitochondrial membrane potential by fluorescence microscopy after KK1 was treated with 2.5 μM rotenone or 0.5 mM NA for 2 h. Based on n = 3 biological replicates. (E) The fluorescence value of KK1 mitochondrial membrane potential treated with 2.5 μM rotenone or 0.5 mM NA for 2 h was detected by microplate reader. Based on n = 6 biological replicates. (F) Mitochondrial ROS staining was observed by fluorescence microscope after KK1 was treated with 10 μM Mito-Tempo 24 h + 0.5 mM NA 6 h or 0.5 mM NA for 6 h. Based on n = 3 biological replicates. (G) The fluorescence value of KK1 ATP level treated with 10 μM Mito-Tempo 24 h + 0.5 mM NA 6 h or 0.5 mM NA for 6 h was detected by microplate reader. Based on n = 6 biological replicates. (H) mtDNA content after 10 μM Mito-Tempo 24 h + 0.5 mM NA 6 h or NA treatment 6 h in KK1. Based on n = 3 biological replicates. (I) TMRM staining of mitochondrial membrane potential by fluorescence microscopy after KK1 was treated with 10 μM Mito-Tempo 24 h + 0.5 mM NA 2 h or 0.5 mM NA for 2 h. Based on n = 3 biological replicates. (J) The fluorescence value of KK1 mitochondrial membrane potential treated with 10 μM Mito-Tempo 24 h + NA 2 h or 0.5 mM NA for 2 h was detected by microplate reader. Based on n = 6 biological replicates. (K) Western blot analysis of NLRP3, Cytb, ATP5A, UQCRC2, NDUFB8 protein with NA (0.5 mM, 6 h) or 10 μM Mito-Tempo 24 h + NA 6 h in KK1, with GAPDH used as a loading control. Based on n = 3 biological replicates. Data are represented as mean ± s.e.m.. Statistical significance was determined using Student’s two-tailed unpaired t -test. * P ≤0.05, ** P ≤0.01, *** P ≤0.001. Transcriptome analysis demonstrated that NA intervention not only remarkably upregulated the inflammatory signaling pathway but induced alterations in the expression of various histone deacetylases ( Fig. 5 A). Accordingly, previous studies have demonstrated that changes in epigenetic modifications are associated with the activation of inflammatory signaling pathways [32] . Notably, histone acetylation plays a crucial role in regulating gene expression at the chromatin level. [37] , [58] . Based on these findings, we hypothesized that NA-induced alterations in granulosa cell epigenetic modifications contribute to inflammation. In order to explore this hypothesis, we investigated three levels of histone modifications that are strongly affected by fatty acids, namely, histone lactylation, crotonylation, and histone acetylation. The results showed that abnormal accumulation of NA significantly altered histone acetylation levels in granulosa cells, but it did not have a significant effect on histone lactylation or crotonylation ( Fig. S7 A–C). Fig. 5 NA reduces H3K9ac modification in granulosa cells, inhibits steroid hormone synthesis gene expression, and promotes IL-1β expression. (A) Analysis of NA-treated granulosa cells using RNA-seq. The heatmap shows the relative abundance of histone deacetylase-related genes. Based on n = 3 biological replicates. (B) Western blot analysis of H3K9ac, H3K14ac, H4K5ac, H4k8ac, H4K12ac, and H4K16ac proteins in NA-treated granulosa cells, with H3 and H4 used as loading controls. Based on n = 3 biological replicates ( Fig. S7 D). (C) Western blot analysis of H3K9ac protein in the ovaries of rats injected with NA through the tail vein on the 7th day of pregnancy. Based on n = 3 biological replicates. (D) Immunofluorescence analysis of H3K9ac modification levels in granulosa cells after NA treatment. Scale bar: 40 μm. Based on n = 3 biological replicates. (E) qPCR analysis of mRNA levels of hdac1, hdac2, sirt1, and sirt6 in NA-treated granulosa cells. Based on n = 12 biological replicates. (F) Western blot analysis of SIRT6 protein in NA-treated granulosa cells, with GAPDH used as a loading control. Based on n = 3 biological replicates. (G) Distribution of H3K9ac binding peaks in the genome of NA-treated granulosa cells. (H) Distribution of the relative positions of H3K9ac binding peak summits to gene locations in NA-treated granulosa cells. The x-axis represents the distance from the peak summit to the TSS, and the y-axis represents the number of peak summits. (I) Genome-wide distribution heatmap of H3K9ac signal peak centers in NA-treated granulosa cells using CUT&Tag-seq technology. (J) Browser tracks showing H3K9ac at LHCGR, FSHR, GDF9, CYP19A1, StAR, and IL-1β. (K) TF binding motifs enriched at sites of NA treatment differential peaks enrichment in H3K9ac. (L) CUT&Tag-qPCR analysis revealed the interactions between H3K9ac and promoter regions of jun and fos genes induced by NA. (M) Western blot analysis of NLRP3、IL-1β、JUN protein in NA-treated or NA+si_jun treated granulosa cells, with GAPDH used as a loading control. Based on n = 3 biological replicates (G-M). Data are represented as mean ± s.e.m.. Statistical significance was determined using Student’s two-tailed unpaired t-test. *P≤0.05, **P≤0.01, ***P≤0.001. NA reduces H3K9ac modification in granulosa cells, inhibits steroid hormone synthesis gene expression, and promotes IL-1β expression. (A) Analysis of NA-treated granulosa cells using RNA-seq. The heatmap shows the relative abundance of histone deacetylase-related genes. Based on n = 3 biological replicates. (B) Western blot analysis of H3K9ac, H3K14ac, H4K5ac, H4k8ac, H4K12ac, and H4K16ac proteins in NA-treated granulosa cells, with H3 and H4 used as loading controls. Based on n = 3 biological replicates ( Fig. S7 D). (C) Western blot analysis of H3K9ac protein in the ovaries of rats injected with NA through the tail vein on the 7th day of pregnancy. Based on n = 3 biological replicates. (D) Immunofluorescence analysis of H3K9ac modification levels in granulosa cells after NA treatment. Scale bar: 40 μm. Based on n = 3 biological replicates. (E) qPCR analysis of mRNA levels of hdac1, hdac2, sirt1, and sirt6 in NA-treated granulosa cells. Based on n = 12 biological replicates. (F) Western blot analysis of SIRT6 protein in NA-treated granulosa cells, with GAPDH used as a loading control. Based on n = 3 biological replicates. (G) Distribution of H3K9ac binding peaks in the genome of NA-treated granulosa cells. (H) Distribution of the relative positions of H3K9ac binding peak summits to gene locations in NA-treated granulosa cells. The x-axis represents the distance from the peak summit to the TSS, and the y-axis represents the number of peak summits. (I) Genome-wide distribution heatmap of H3K9ac signal peak centers in NA-treated granulosa cells using CUT&Tag-seq technology. (J) Browser tracks showing H3K9ac at LHCGR, FSHR, GDF9, CYP19A1, StAR, and IL-1β. (K) TF binding motifs enriched at sites of NA treatment differential peaks enrichment in H3K9ac. (L) CUT&Tag-qPCR analysis revealed the interactions between H3K9ac and promoter regions of jun and fos genes induced by NA. (M) Western blot analysis of NLRP3、IL-1β、JUN protein in NA-treated or NA+si_jun treated granulosa cells, with GAPDH used as a loading control. Based on n = 3 biological replicates (G-M). Data are represented as mean ± s.e.m.. Statistical significance was determined using Student’s two-tailed unpaired t-test. *P≤0.05, **P≤0.01, ***P≤0.001. We further investigated the levels of multiple proteins modified by histone acetylation in granulosa cells after NA treatment. The results indicated a prominent decrease in H3K9ac modification levels ( Fig. 5 B & D, Fig. S7 D); in contrast, the H4K8ac and H4K12ac modification levels were significantly increased ( Fig. 5 B). Consistent with these results, we found that abnormal accumulation of NA in rats induced via tail vein injection caused a reduction in ovarian H3K9ac modification levels ( Fig. 5 C). Next, we examined the expression levels of deacetylases involved in regulating H3K9ac expression. The results indicated that NA treatment sharply upregulated the protein expression of the deacetylase SIRT6, thus leading to inhibition of H3K9ac modification levels ( Fig. 5 E & F). However, the mechanism by which H3K9ac regulates inflammation needs to be studied further. We next investigated the mechanisms underlying the impact of changes in H3K9ac modification levels on IL-1β expression and steroid hormone synthesis by using the CUT&Tag technique to generate a comprehensive genome-wide map of H3K9ac histone modifications in porcine ovarian granulosa cells treated with NA ( Fig. S7 E). The CUT&Tag data revealed that NA treatment led to a decrease in the enrichment of H3K9ac near the transcription start site (TSS) ( Fig. 5 I). Peak analysis of H3k9ac demonstrated that NA significantly reduced the peaks located in the promoter regions ( Fig. 4 G and H). Moreover, differential peak analysis indicated that NA treatment upregulated the AMPK signaling, Rap1 signaling, and Toll-like receptor signaling pathways, while it downregulated the glycerolipid metabolism, phosphatidylinositol signaling, glycerophospholipid metabolism, and P53 signaling pathways ( Fig. S7 F & G). Gene expression analysis revealed that NA significantly decreased the modification levels of the lhcgr, fshr, gdf9, cyp19a1, and star genes, which are involved in steroid synthesis. Notably, NA treatment increased H3K9ac modification levels at the IL-1β promoter region ( Fig. 5 J). Motif analysis of differentially modified H3K9ac peaks identified AP-1, NFR, and IRF8 as the top-ranked motifs ( Fig. 5 K). AP-1 is a positive regulator of IL-1β, and its subunits Jun and Fos promote secretion of IL-1β within cells [59] , [60] . In line with this, CUT&Tag qPCR showed that both the jun and fos subunits were remarkable upregulated after NA treatment ( Fig. 5 L). Furthermore, through the interference of jun expression using small RNA, we observed that while NA treatment enhanced the expression of NLRP3 protein, there was a significant reduction in the level of bioactive IL-1β protein was detected upon downregulation of JUN protein ( Fig. 5 M). Collectively, the data here demonstrated that NA selectively impacted H3k9ac modification of genes in granulosa cells, thus resulting in downregulation of steroid hormone synthesis and increase in IL-1β expression. Further investigation into the abnormal accumulation of nervonic acid within the ovaries holds paramount significance in mitigating ovarian dysfunction attributed to lipid buildup. In this set of experiments, we investigated the mechanisms contributing to the aberrant accumulation of NA from the metabolic perspective. Previous studies have identified ELOVL1 and ELOVL3 as pivotal upstream metabolizing enzymes of NA, facilitating the elongation of C22:1 fatty acids to C24:1 by adding two carbon units [61] , [62] , while CerS2 and CerS4 serve as crucial downstream metabolizing enzymes by catalyzing the attachment of fatty acids to sphingoid bases (such as sphingosine) [40] , [63] . Our examination of these NA metabolizing enzymes in sow ovaries and liver revealed that there was a significant reduction in Cers2 expression among sows exhibiting low embryo survival rates ( Fig. 6 A & B, Fig. S8 A-C). Furthermore, through elovl1- and elovl3-overexpression experiments, we confirmed that augmenting upstream metabolizing enzyme activity had no discernible impact on granulosa cell function ( Fig. 6 C, Fig. S8 D & E), which was probably owing to these enzymes involved in multiple fatty acid metabolic pathways. However, inhibition of the downstream enzyme CerS2 resulted in inhibitory effects on granulosa cells akin to those observed following NA treatment ( Fig. S8 F, Fig. 6 D & E). Fig. 6 CerS2 downregulation aggravated ovarian abnormal NA accumulation. (A) qPCR analysis of mRNA levels of ovarian NA metabolic enzymes elovl1, elovl3, cers2, and cers4 in relation to high and low embryo survival rates in mother pigs. Based on n = 10 biological replicates. (B) Western blot analysis of CerS2 protein in the ovaries of sows with high and low embryo survival rates, with GAPDH as a loading control. Based on n = 3 biological replicates ( Fig. S8 A). (C) qPCR analysis of mRNA levels of granulosa cell lhcgr, fshr, gdf9, cyp19a1, and star following overexpression of elovl1 and elovl3. Based on n = 12 biological replicates. (D) qPCR analysis of mRNA levels of granulosa cell lhcgr, fshr, gdf9, cyp19a1, and star by small RNA interference of CerS2. Based on n = 12 biological replicates. (E) Western blot analysis of LHCGR, FSHR, and CYP19A1 protein by small RNA interference of CerS2, with GAPDH as a loading control. Based on n = 3 biological replicates ( Fig. S8 H). (F) qPCR analysis of granulosa cell lhcgr, fshr, gdf9, star, and cyp19a1 following overexpression or interference of CerS2 with NA treated. Based on n = 12 biological replicates. (G) GSEA plot comparing control with CerS2 interference treated granulosa cells for the nod-like receptor signaling pathway. NES, normalized enrichment score. (H) Analysis of RNA-seq in small RNA interference-treated granulosa cells. The heatmap shows the relative abundance of inflammatory cytokine genes. Based on n = 3 biological replicates. (I) qPCR analysis of cxcl14 mRNA levels in the ovaries of mother pigs with high and low embryo survival rates. Based on n = 10 biological replicates. (J) ELISA measurement of serum cxcl14 levels in mother pigs with high and low embryo survival rates. Based on n = 10 biological replicates. (K) qPCR analysis of lhcgr, fshr, and cyp19a1 mRNA levels in granulosa cells following combined small RNA interference and NA treatment. Based on n = 8 biological replicates. (L) Intravenous injection of recombinant CXCL14 protein via an osmotic minipump of mice during estrus and pregnancy, and measurement of CXCL14 (n = 5), NA (n = 3), and IL-1β (n = 3) levels in serum on the 7th day at pregnancy. Data are represented as mean ± s.e.m.. Statistical significance was determined using Student’s two-tailed unpaired t-test (A, C, D, I, J, L) or one-way ANOVA with Dunnett’s post-hoc test (F, K). *P≤0.05, **P≤0.01, ***P≤0.001. CerS2 downregulation aggravated ovarian abnormal NA accumulation. (A) qPCR analysis of mRNA levels of ovarian NA metabolic enzymes elovl1, elovl3, cers2, and cers4 in relation to high and low embryo survival rates in mother pigs. Based on n = 10 biological replicates. (B) Western blot analysis of CerS2 protein in the ovaries of sows with high and low embryo survival rates, with GAPDH as a loading control. Based on n = 3 biological replicates ( Fig. S8 A). (C) qPCR analysis of mRNA levels of granulosa cell lhcgr, fshr, gdf9, cyp19a1, and star following overexpression of elovl1 and elovl3. Based on n = 12 biological replicates. (D) qPCR analysis of mRNA levels of granulosa cell lhcgr, fshr, gdf9, cyp19a1, and star by small RNA interference of CerS2. Based on n = 12 biological replicates. (E) Western blot analysis of LHCGR, FSHR, and CYP19A1 protein by small RNA interference of CerS2, with GAPDH as a loading control. Based on n = 3 biological replicates ( Fig. S8 H). (F) qPCR analysis of granulosa cell lhcgr, fshr, gdf9, star, and cyp19a1 following overexpression or interference of CerS2 with NA treated. Based on n = 12 biological replicates. (G) GSEA plot comparing control with CerS2 interference treated granulosa cells for the nod-like receptor signaling pathway. NES, normalized enrichment score. (H) Analysis of RNA-seq in small RNA interference-treated granulosa cells. The heatmap shows the relative abundance of inflammatory cytokine genes. Based on n = 3 biological replicates. (I) qPCR analysis of cxcl14 mRNA levels in the ovaries of mother pigs with high and low embryo survival rates. Based on n = 10 biological replicates. (J) ELISA measurement of serum cxcl14 levels in mother pigs with high and low embryo survival rates. Based on n = 10 biological replicates. (K) qPCR analysis of lhcgr, fshr, and cyp19a1 mRNA levels in granulosa cells following combined small RNA interference and NA treatment. Based on n = 8 biological replicates. (L) Intravenous injection of recombinant CXCL14 protein via an osmotic minipump of mice during estrus and pregnancy, and measurement of CXCL14 (n = 5), NA (n = 3), and IL-1β (n = 3) levels in serum on the 7th day at pregnancy. Data are represented as mean ± s.e.m.. Statistical significance was determined using Student’s two-tailed unpaired t-test (A, C, D, I, J, L) or one-way ANOVA with Dunnett’s post-hoc test (F, K). *P≤0.05, **P≤0.01, ***P≤0.001. To further validate the impact of CerS2 on abnormal NA accumulation, we induced silencing or overexpression of CerS2 in the presence of NA. The data demonstrated that CerS2 overexpression could rescue the detrimental effects of NA on granulosa cells ( Fig. 6 F, Fig. S8 G). Next, we examined whether blocking CerS2 expression could affect the gene transcription levels of granulosa cells. RNA-seq results indicated that blocking CerS2 gene expression significantly upregulated 2646 genes and downregulated 2471 genes in porcine ovarian granulosa cells ( Fig. S8 I). To our surprise, we also observed significant upregulation of the NLRP signaling pathway ( Fig. 6 G). Furthermore, GO term enrichment analysis revealed that interfering with CerS2 expression upregulated biological processes related to interleukins, cytokines, lipid metabolism, immunity, and inflammation, while it significantly downregulated steroid hormone metabolism, hormone response, receptor signaling, and oxidoreductase activity pathways ( Fig. S8 J). GSEA results showed that interfering with CerS2 expression significantly upregulated oxidative phosphorylation, the B-cell receptor signaling pathway, the T-cell receptor signaling pathway, and the TLR signaling pathway ( Fig. S8 K). By contrast, it downregulated steroid hormone biosynthesis, progesterone-mediated oocyte maturation, linoleic acid metabolism, the TGF-β signaling pathway, and steroid biosynthesis ( Fig. S8 L). These data suggest that the downregulation of cers2 expression exerts a significant impact on lipid metabolism and the synthesis and regulation of steroid hormones in granulosa cells. To further investigate whether NA-induced metabolic abnormalities driven by Cers2 are involved in inflammation, an enrichment analysis of inflammatory cytokines was conducted using transcriptome data. The results revealed that interference with CerS2 expression significantly upregulated the expression of IL-1A, PDGFB, and VEGFA and downregulated the expression of IL33, IL15, and IL34 ( Fig. 6 H). To our surprise, interference with CerS2 expression also led to a significant increase in the transcription levels of CXCL14 ( Fig. 6 H), which is a critical chemokine closely linked to lipid metabolism and inflammation [63] . To confirm whether increased CXCL14 expression is related to abnormal NA accumulation and inflammation, we analyzed the gene expression levels of CXCL14 in the liver and ovaries of sows with different embryonic survival rates. Decrease in embryonic survival was associated with a significant increase in CXCL14 expression ( Fig. 6 I, Fig. S8 M). Consistent with this finding, the serum concentration of CXCL14 also exhibited an increase as the embryo survival rate decreased (Fig. 66J). Furthermore, interference with CXCL14 expression promoted the expression of genes related to steroid hormone synthesis and luteinization in porcine ovarian granulosa cells ( Fig. S8 N). Co-treatment with NA after CXCL14 expression interference rescued the inhibitory effects of NA on steroid hormone synthesis and luteinization in these cells ( Fig. 6 K). Next, we tried to further validate these findings by delivering recombinant CXCL14 via an osmotic minipump to pregnant mice for one week. The serum CXCL14 levels in the mice significantly increased, and it was accompanied by an increase in serum NA and IL-1β levels ( Fig. 6 L). These findings imply that the effects of abnormal NA accumulation on ovarian inflammation may be mediated by downregulation of CerS2. Notably, the results also indicated that this dysregulation of NA metabolism was exacerbated by the upregulation of CXCL14. In the following experiments, we aimed to preliminarily verify the effects of abnormal NA accumulation on human ovarian function, as well as the roles of CerS2 and CXCL14 in the ovary, as observed in the different animal models so far. We examined the potential adverse effects of NA on human ovarian function by using human granulosa cells as an in vitro model. First, we induced luteinization of human ovarian granulosa cells by LH and FSH stimulation. The inflammatory genes induced by NA treatment in porcine ovarian granulosa cells were validated in luteinizing human granulosa cells ( Fig. 7 A). In addition, consistent with the observation in the porcine cells, treatment with NA in human ovarian granulosa cells caused inhibition of the key genes (that is, lhcgr , fshr , gdf9 , cyp19a1 , and star ) involved in steroid synthesis ( Fig. 7 B & C). Importantly, NA significantly reduced H3K9ac modification levels and promoted IL-1β expression in the human ovarian granulosa cells ( Fig. 7 D & E). These findings collectively suggest that abnormal accumulation of NA may also trigger human ovarian inflammation and inhibit steroid synthesis in granulosa cells. Fig. 7 Molecular alterations in NA metabolism within human ovarian. (A) qPCR analysis of LH and FSH-induced lhcgr , fshr , hsd3β1 , cyp11a1 , and star mRNA levels in human ovarian granulosa cells after NA treatment. Based on n = 12 biological replicates. (B) qPCR analysis of lhcgr , fshr , gdf9 , star , and cyp19a1 mRNA levels in human ovarian granulosa cells after NA treatment. Based on n = 12 biological replicates. (C) Western blot analysis of LHCGR and GDF9 protein in human ovarian granulosa cells treated with NA, with GAPDH as a loading control. Based on n = 3 biological replicates. (D) Western blot analysis of IL-1β protein in human ovarian granulosa cells treated with NA, with GAPDH as a loading control. Based on n = 3 biological replicates. (E) Western blot analysis of H3K9ac protein in human ovarian granulosa cells treated with NA, with GAPDH as a loading control. Based on n = 3 biological replicates. (F) IHC staining showing CerS2 expression in ovarian granulosa cells of PCOS women. Scale bar 2 mm, 100 μm. Based on n = 6 biological replicates. (G) IHC staining showing CXCL14 expression in ovarian granulosa cells of PCOS women. Scale bar 2 mm, 100 μm. Based on n = 6 biological replicates. Data are represented as mean ± s.e.m.. Statistical significance was determined using Student’s two-tailed unpaired t -test (B) or one-way ANOVA with Dunnett’s post-hoc test (A). * P ≤0.05, ** P ≤0.01, *** P ≤0.001. Molecular alterations in NA metabolism within human ovarian. (A) qPCR analysis of LH and FSH-induced lhcgr , fshr , hsd3β1 , cyp11a1 , and star mRNA levels in human ovarian granulosa cells after NA treatment. Based on n = 12 biological replicates. (B) qPCR analysis of lhcgr , fshr , gdf9 , star , and cyp19a1 mRNA levels in human ovarian granulosa cells after NA treatment. Based on n = 12 biological replicates. (C) Western blot analysis of LHCGR and GDF9 protein in human ovarian granulosa cells treated with NA, with GAPDH as a loading control. Based on n = 3 biological replicates. (D) Western blot analysis of IL-1β protein in human ovarian granulosa cells treated with NA, with GAPDH as a loading control. Based on n = 3 biological replicates. (E) Western blot analysis of H3K9ac protein in human ovarian granulosa cells treated with NA, with GAPDH as a loading control. Based on n = 3 biological replicates. (F) IHC staining showing CerS2 expression in ovarian granulosa cells of PCOS women. Scale bar 2 mm, 100 μm. Based on n = 6 biological replicates. (G) IHC staining showing CXCL14 expression in ovarian granulosa cells of PCOS women. Scale bar 2 mm, 100 μm. Based on n = 6 biological replicates. Data are represented as mean ± s.e.m.. Statistical significance was determined using Student’s two-tailed unpaired t -test (B) or one-way ANOVA with Dunnett’s post-hoc test (A). * P ≤0.05, ** P ≤0.01, *** P ≤0.001. Abnormal accumulation of NA was observed in the plasma samples of patients with PCOS ( Fig. 1 K). Immunohistochemical analysis showed that the CerS2 and CXCL14 proteins exhibit low expression and high expression, respectively, in the ovaries of patients with PCOS ( Fig. 7 F A and G). These findings are also in line with the observations in the earlier models. In summary, these results indicate that abnormal NA accumulation may promoted ovarian dysfunction in human, as observed in sows and the mouse models, and the related mechanisms also appear to be similar. However, these findings provide only preliminary evidence, and further research utilizing more precise methodologies is necessary to definitively establish the impact of NA on female ovarian function.

The animal procedures used in the experiments strictly adhere to the Beijing Municipal Laboratory Animal Welfare and Ethics Guidelines and are conducted in accordance with the regulations established by the China Agricultural University Animal Welfare and Animal Experiment Ethics Committee. All experimental animal usage is approved by the China Agricultural University Animal Welfare and Ethics Committee (AW81703202-1–1). For human specimens (n = 12, of which 6 were women with PCOs and 6 were healthy women), all PCOS patients met the Rotterdam criteria for diagnosis [39] . The inclusion criteria for nonPCOS controls were (1) age between 20 and 37 years; (2) regular menstrual cycles ranging from 25 to 37 days; (3) body mass index (BMI) ≤ 28 kg/m2; and (3) normal basal serum FSH (≤ 10 mIU/mL) and estradiol (E2) (≤ 75 pg/mL). Exclusion criteria for controls were any diagnosis of chronic conditions, metabolism-related disorders, diminished ovarian reserve, and endometriosis. Informed consent was provided by each subject before sample collection (according to the Helsinki Declaration), and the study was approved by the Ethics Committee of Peking Union Medical College Hospital (K23C0561). Materials/reagents are listed with catalogue numbers and vendors in Supplementary Table 1 . The experimental sows (n = 39, 2–3 parities) used in the study were all crossbred Landrace × Yorkshire sows and were obtained from the Fengning Experimental Station of China Agricultural University in Chengde City, Hebei Province (Academician Workstation of Chengde Jiuyun Agriculture and Animal Husbandry Co., Ltd.). These sows were raised and managed according to the standards set by the experimental station. Sample collection was conducted at two time points: during the estrous period and on the 28th day of pregnancy. All sows were randomly selected, and their parity, reproductive performance in each parity, backfat, and body weight were recorded. Blood serum samples were collected from fasting sows via the anterior vena cava and centrifuged at 200 × g for five minutes. The supernatant was then stored at −80°C in a freezer. Tissue sample collection included both slaughter sampling and in vivo biopsy gun sampling. Slaughter sampling was performed at a slaughterhouse in Chengde City, Hebei Province. Some ovarian tissues were fixed in tissue fixation solution, while the remaining ovarian tissues, liver, corpus luteum, back fat, and abdominal fat tissues were preserved in liquid nitrogen. The number of corpus luteum, total embryo count in the uterus, and the number of live embryos in the 28-day pregnant sows were recorded. In vivo sampling was conducted using Bard biopsy guns. Fat tissue samples were collected at the fixed location on the back of the sows, just before the second last rib. These samples were then frozen and stored in liquid nitrogen. The rats used in the experiments were SD rats (n = 30, 7–8 weeks), and the mice were C57 BL/6J mice (n = 60, 8 weeks), all purchased from Beijing Huafukang Biotechnology Co., Ltd. All rodents were housed at the China Agricultural University Animal Experimental Platform and were kept under specific pathogen-free conditions. Euthanasia was performed using carbon dioxide asphyxiation, followed by cervical dislocation. To create a model of intracellular NA accumulation, rats were subjected to two methods: tail vein injection of 1.65 mM NA (based on total blood volume of rats) and feeding with a diet containing 1.0% NA. Both methods were carried out for one estrous cycle (4 days) and the early pregnancy period (7 days). In the control group, rats underwent the same tail vein injection procedure with equal saline (0.9% of NaCl), or fed basal diet without NA. The specific procedure for tail vein injection involved anesthetizing the rats each morning at 8 a.m., cleaning the rat's tail with alcohol swabs, and then injecting 1 mL of 1.65 mM NA into the tail vein using a sterile 1 mL syringe. The diet for the rats was prepared by Beijing Huafukang Biotechnology Co., Ltd. For mice, the tail vein injection procedure was the same as described above, but the dose was 0.5 mM. Where indicated, mice were anesthetized with 1.5% isoflurane, and implanted with subcutaneous Alzet osmotic mini-pumps (Cupertino, CA, USA) containing CXCL14 (R&D Systems, Minneapolis, MN, USA) calibrated to release the chemokine at a rate of 4.5 ng·g −1 per day for 7 days [40] . Preparation of NA Stock Solution: Weigh a specific mass of NA and dissolve them in a centrifuge tube containing sodium hydroxide solution of the same concentration as the stock solution. After thorough shaking, place the centrifuge tube in a 70°C water bath for 30 min, periodically shaking the solution. Preparation of NA Working Solution: Depending on the required concentration for the experiment, prepare a certain volume of 10% mass fraction bovine serum albumin (BSA) solution (using culture medium or physiological saline as the solvent if needed). Withdraw the required volume of NA stock solution into the BSA solution and place it in a 55°C water bath for 40 min, periodically shaking the solution. The human ovarian granulosa cell line (KGN) and the mouse ovarian granulosa cell line (KK1) were cultured in RPMI 1640 medium supplemented with 10% FBS (fetal bovine serum). The culture conditions were maintained at 5% CO 2 and 37°C. Porcine ovaries were purchased from a local slaughterhouse (Hebei, China) and transported to the laboratory at 37℃ in 0.9% NaCl within 30 min. The detailed method for isolating porcine primary granulosa cells had been reported in our previous studies [41] . Cultivate granulosa cells to 60% confluence in the dish, wash the cell surface twice with PBS, add luteinization inducers (2.5 IU/mL FSH and LH, purchased from the Jiangsu Ningbo Second Hormone Factory), and incubate for 3 days to complete luteinization. The process of immunohistochemistry involves dehydrating, embedding, and sectioning ovarian tissue specimens fixed in formalin, typically into thin sections of 4 to 6 µm thickness. These sections are then rehydrated, subjected to antigen retrieval, and blocked to reduce nonspecific protein binding. Primary antibodies specific to target proteins (MAC-1, MPO, CerS2, CXCL14) are applied, followed by washing to remove unbound antibodies. Secondary antibodies labeled with detectable markers are applied, and after further washing, images of the immunohistochemistry-stained sections are captured using a microscope or imaging system. Image analysis software is used to quantitatively analyze the staining results, allowing for the visualization and assessment of the presence and localization of the target proteins in the ovarian tissue specimens. In this study, transcriptome sequencing was conducted by outsourcing to Beijing Ovisen Gene Technology Co., Ltd. Differential gene expression analysis was performed, and the differentially expressed genes were annotated using databases such as NCBI, UniProt, GO (Gene Ontology), and KEGG (Kyoto Encyclopedia of Genes and Genomes). This annotation process provided detailed descriptive information for the differentially expressed genes. Protein samples were quantified by the BCA protein assay kit (Thermo) using BSA as a standard. In a brief, the procedure involves fixing fresh tissue in 4% paraformaldehyde for at least 24 h, followed by trimming, dehydrating through a graded ethanol and xylene series, embedding in paraffin, and sectioning at 4 µm thickness. The sections are then floated on a water bath, transferred to slides, and dried at 60°C. For staining, the sections are deparaffinized, stained with Harris hematoxylin, differentiated with acid alcohol, counterstained with ammonia water, and stained with eosin. The sections are then dehydrated, cleared with xylene, mounted, and examined under a microscope for image capture and analysis. Cell or tissue lysis was performed in RIPA buffer with the addition of a protease inhibitor cocktail. Equal amounts of protein were separated via SDS-PAGE gels, transferred to PVDF membranes (Millipore, MA, USA) and probed with corresponding primary antibodies as specified. β-actin or GAPDH was applied as the loading control. To prepare for cell permeabilization, dissolve Triton 100 in PBS to a final concentration of 0.5% and incubate at 37°C for 2–3 h with shaking. Make a 2% BSA blocking solution by dissolving 1g of BSA in 50mL of PBS at 50°C with shaking. In a six-well plate, wash cells three times with PBS, add 600μL of 4% paraformaldehyde to fix the cells, and incubate on a shaker at room temperature for 20 min. Wash the cells three times with PBS, add 500μL of 0.5% Triton 100 for permeabilization, and incubate on a shaker at room temperature for 20 min. Wash the cells with PBS again, add 500μL of 2% BSA, and incubate on a shaker at room temperature for 30 min to block the cells. After blocking, add 500μL of primary antibody diluted 1:3000 in 2% BSA to each well and incubate on a shaker at 4°C overnight. Wash the cells three times with PBS, add 500μL of secondary antibody diluted 1:500 in 2% BSA, and incubate on a shaker at room temperature for 1 h in the dark. Wash the cells three times with PBS, add DAPI solution (1:1000 dilution in PBS, 200μL per well) for 5–10 min, wash with PBS, and observe the cells under a fluorescence microscope. Euthanasia of animals was performed using carbon dioxide asphyxiation, followed by cervical dislocation. The abdomen was disinfected with alcohol, and the abdominal cavity was opened. The ovaries were located on the ventral side, below the kidneys, and placed in Dulbecco's Modified Eagle Medium (DM). Clean ovaries were added to equilibrated Eagle's Minimum Essential Medium (EM) and torn into 2–3 pieces, which were then placed in an incubator for enzyme treatment for 40 min. The ovaries in EM were transferred to DM, and mechanical pipetting was performed for 5 min using a 200 μL pipette. A portion of the ovarian tissue was retained, while the rest was transferred to maturation medium (MM) and placed in an incubator. After a 30-minute incubation, the follicle separation process began (the separation time for each piece of tissue outside the incubator should not exceed 30 min). Using the left hand to hold the needle, the tissue was fixed to the bottom of the culture dish, and with the right hand, the follicles were gently separated. Care was taken to remove most of the surrounding extracellular matrix (ECM) without puncturing the follicles. As many follicles as possible were separated within the 30-minute timeframe. Separated follicles were transferred into DM using a pipette. During the transfer, care was taken to minimize the uptake of DM, and to prevent the follicles from adhering to the needle, no more than <10 follicles were transferred at a time. To avoid adhesion between follicles, a portion of liquid was aspirated before taking another follicle. When placing the follicles into MM, they were positioned at different locations. This process was repeated until the desired number of follicles was separated. Microcapsules were created using alginate solution, and the follicles were aspirated into the alginate droplets. Cross-linking was performed for 2 min, and the gel droplets were then transferred into MM, equilibrated for 30 min in an incubator, and subsequently moved into a 96-well plate. Encapsulated follicles were cultured for 8 days under conditions of 37°C and 5% CO 2 . The culture medium was changed every other day, replacing 50% of the medium each time. Mitochondrial dysfunction is considered a key factor in the activation of NLRP3, therefore, we measured mitochondrial membrane potential, mitochondrial reactive oxygen species (mtROS), and ATP levels [42] . 2.5 μM rotenone was used as a positive control (6 h or 2 h) to induce mitochondrial oxidative stress in KK1 [43] . 10μM Mito-Tempo was used as mitochondria-targeted superoxide dismutase mimetic with superoxide and alkyl radical scavenging properties to pretreat KK1 for 24 h [44] . Also, 0.5mM NA was used to treat KK1 6 h or 2 h. Mitochondrial membrane potential (Ψm) and mtROS were measured as previously described [42] . In brief, Using TMRM (tetramethylrhodamine methyl ester, Sigma, NO. T5428) to measure mitochondrial membrane potential. KK1 were incubated with 200nM TMRM at 37℃ for 30 min. After two washes, the fluorescence intensity was measured using a FilterMax F5 multimode plate reader (Molecular Devices). mtROS were measured using Mito-SOX (Thermo, NO. M36008 ). KK1 were treated with 4 mM Mito-SOX for 20 min, washed twice, suspended in PBS, and counted. Equal numbers of cells from different treatment groups were then plated onto a 96-well plate to reduce variation between treatment groups. Fluorescence intensity was measured at 510/580nm using a FilterMax F5 plate reader (Molecular Devices), and images were captured using fluorescence microscopy. Cellular ATP levels were measured using the CellTiter-Glo luminescent assay kit (Promega, NO. G7570) [45] . The measurement of total mtDNA was described by Zhong et al., 2018 [42] . In brief, Total DNA was extracted using the Allprep DNA/RNA Mini Kit (Catalogue No. 80204, QIAGEN). Quantitative PCR (qPCR) was performed using mtDNA-specific primers targeting the mitochondrial D-loop region for mtDNA quantification. Nuclear DNA normalization was achieved using primers specific for Tert. D-loop F: 5′-CTTCAGGGCCATCAAATGCG-3′; D-loop R: 5′-GGTGATTGGGTTTTGCGGAC-3′; Tert F: 5′-CTGTGCCTACCAGGGGAGAT-3′; Tert R: 5′-GGCCTTGAGCCCAGAAAGAT-3′. Approximately 1×10 5 cells of each sample were collected and analyzed with Hyperactive Universal CUT&Tag Assay Kit for Illumina (Vazyme, TD904) following the recommended protocol. Briefly, harvest fresh cultures and count 1×10 5 cells for each sample. Then bind cells to Concanavalin A beads at room temperature for 10 min. Next, remove liquid and add primary antibody with icecold Antibody Buffer and incubate overnight at 4 ℃. The next day, remove liquid and add H3K9ac/IGg secondary antibody with Dig-Wash Buffer and incubate at room temperature for 1h. Then, wash sufficiently and bind CUT&Tag pA/G-Tn5 Transposomes at room temperature for 1h. After gently but throughly wash, tagmentation at 37 ℃ for 1h and then performed DNA extraction with DNA extraction beads. Finally, combinate i7 and i5 Indexed Primer for PCR amplification and perform cleanup for quantification and Illumina sequencing. For peak calling, we used the BEDTools' bamtobed function to convert sequence alignments in BAM format into BEDPE records. Peak calling for CUT&Tag data was performed on merged replicate samples and normalized to the input using MACS2 v2.1.1.20160309 software. Peak calls with a false discovery rate (FDR) of ≤ 5% were retained for downstream analysis. Peak annotation was carried out using the R package ChIPseeker with default parameters. Peak comparisons and overlaps were assessed using the BEDTools suite for autosomal chromosomes. For quantitative analysis, we normalized read counts by calculating the reads per kilobase of transcript per million mapped reads (RPKM) for each peak. RPKM values were computed using the merged replicate BAM files and the bamCoverage tool from deepTools software. To minimize the effects of batch and cell type variation, RPKM values were further subjected to Z-score transformation, calculated within 100-bp bins across the entire genome, excluding outlier regions. Visualization of H3K9ac CUT&Tag data was performed using the washU epigenome browser ( https://epigenomegateway.wustl.edu/browser/ ). Porcine serum lipid-targeted metabolomics were detected by Beijing Allgwegene Health Co.,Ltd.

This study explores the influences of metabolic syndrome on fertility and the corresponding mechanisms through experiments on various animal models, and importantly, confirms the results in human cell models and plasma samples. The findings presented herein are important from the viewpoint of preserving and promoting the reproductive health of human beings and mammalian animals. The findings in experimental sows demonstrate abnormal ovarian accumulation of NA in sows with low embryo survival rates that had abnormal lipid metabolism. Similarly, the level of NA was abnormally high in cows with low fertility rates and women with PCOS. Thus, these findings indicate that abnormal NA accumulation may be associated with fertility dysfunctions. NA is a type of single unsaturated very-long-chain fatty acid with a chain length of 24 carbons that is abundant in the brain, liver, and kidneys and plays important roles in cell membrane formation, apoptosis, and neural transmission [64] , [65] in the free form or in the form of sphingolipids. It has been reported that the accumulation of long-chain fatty acids in non-adipose cells can induce lipotoxicity, triggering endoplasmic reticulum stress and inflammation [66] , [67] . Accordingly, in the present study, we found that the ovaries of sows with low embryo survival rates and abnormal ovarian accumulation exhibited obvious neutrophil infiltration, as well as higher expression of the pro-inflammatory cytokines TNF-α, IL-1β, and NF-kB. In line with these findings, a large dose of NA injected via the tail vein in mice caused a significant increase in the numbers of B cells and T cells in the ovaries. These data indicate that abnormal NA accumulation induced ovarian inflammation. Previously, ovarian inflammation has been found to inhibit steroid hormone synthesis, corpus luteum formation, and the folliculogenesis of oocytes, thus resulting in follicular atresia and a reduction in the quality and quantity of oocytes [68] , [69] , [70] , [71] . In line with these studies, the present experiments also demonstrated that abnormally high levels of NA significantly suppress the expression of key genes and proteins involved in steroid hormone synthesis and luteinization under in vitro and in vivo conditions, as well as promote follicular atresia. Thus, ovarian inflammation induced by abnormal NA accumulation may lead to ovarian dysfunction via inhibition of steroid hormone synthesis and oocyte folliculogenesis. Accordingly, it has been previously reported that CerS2-knockout mice exhibited significantly elevated levels of NA [72] . Interestingly, in the current study, the results revealed significant downregulation of CerS2, a downstream metabolic enzyme of NA, in sows with low embryo survival rates. In addition, partial functional deficiency of CerS2 was found to lead to diet-induced fatty liver inflammation and insulin resistance [73] . In line with these results, our data uncovered that interfering with CerS2 expression in granulosa cells caused an increase in the expression levels of the inflammatory cytokines IL-1A, PDGFB, and VEGFA activated the NLRP signaling pathway, and also inhibited steroid hormone synthesis and luteinization. These effects were rescued by CerS2 overexpression. Further, these results were confirmed in an in vitro mouse model in which a large dose of NA was injected via the tail vein. Notably, interference with CerS2 expression resulted in upregulation of the chemokine CXCL14, and high mRNA expression of CXCL14 in both the liver and ovaries and its serum content were significantly associated with low embryo survival rates in sows. CXCL14 is a novel non-ELR (glutamine-leucine-arginine) chemokine with broad biological activity that mainly regulates immune cell migration [33] . CXCL14 is typically considered to have widespread chemotactic effects on monocytes, immature dendritic cells, neutrophils, and NK cells [63] . Accordingly, our data indicated that injection of recombinant CXCL14 via the tail vein during early pregnancy enhanced serum IL-1β levels in mice. Thus, the abnormal metabolism of NA and its inflammatory effects may be mediated by CerS2 and CXCL14. Notably, both NA administration and silencing of CerS2 expression led to upregulation of the NLRP pathway in granulosa cells. Further, both our in vivo and in vitro experimental results demonstrated that NA activated intracellular NLRP3, which is an intracellular sensor that can detect a wide range of microbial motifs, endogenous danger signals, and environmental irritants, resulting in the formation and activation of the NLRP3 inflammasome [27] . In line with our results, it has been reported that various fatty acids can activate or inhibit intracellular NLRP3 signaling, thereby influencing the occurrence of inflammation [26] , [28] , [29] . Our results indicated that NA-induced activation of NLRP3 ultimately led to the promotion of IL-1β secretion by induced mitochondrial ROS production. There is already ample evidence indicating a close association between the increase in mitochondrial ROS and the activation of inflammation, with the activation of the NLRP3 inflammasome also being regulated by ROS [74] . We have found that NA can induce mitochondrial oxidative stress by inhibiting the mitochondrial respiratory chain complex Ⅰ protein NDUFB8 and complex Ⅲ protein Cytb [57] . Mature IL-1β is a potent proinflammatory mediator in various immune responses [75] , and in the ovarian context, elevated IL-1β levels are closely associated with ovarian cancer and PCOS [76] , [77] . Furthermore, research suggests that IL-1β can inhibit the secretion of estradiol by granulosa cells [78] and, thereby, impede FSH-induced granulosa cell differentiation, which is detrimental to follicular development [79] . These data demonstrated that NA induced ovarian inflammation by activating the NLRP3 pathway, which subsequently led to an increase in IL-1β. Notably, in studies investigating PCOS, heightened levels of IL-1β have also been observed to detrimentally impact hormone synthesis and follicular development [80] . A recent study also indicates that elevated levels of IL-1β in the ovary inhibit the cAMP-PKA pathway in granulosa cells, leading to granulosa cell apoptosis and reduced estradiol synthesis, which in turn results in the occurrence of premature ovarian insufficiency (POI) [81] . Therefore, the abnormal accumulation of NA leading to NLRP3/IL-1β activation may severely disrupt ovarian function. The present study also investigated the important question of how NA enhanced IL-1β expression. Through a comparison of three types of fatty acid-influenced epigenetic modifications (that is, histone butyrylation, histone lactylation, and histone acetylation), we found that high levels of NA significantly inhibit the histone H3K9ac modification level in granulosa cells, resulting in a decrease in the expression of steroid hormone synthesis-related genes such as LHCGR and STAR . Consistent with our findings, it has been previously reported that an increase in H3K9ac modification levels induced by butyric acid is favorable for the synthesis of granulosa cell steroid hormones [82] . The alterations in these genes indicate that the disruption of ovarian steroid hormone synthesis and follicle development is attributed to the abnormal accumulation of NA. Additionally, butyric acid-induced increase in the H3K9ac modification level of the PPARα promoter has been found to inhibit the NF-ĸB-mediated inflammatory signaling pathway [83] . Of note, the acetylation levels of H3K9ac in the IL-1β promoter region and the expression of AP-1 were significantly enhanced in response to high levels of NA. Published researches reported that activated AP-1 binds to the promoter region of the IL-1β gene to promote its transcription and expression [59] , [60] . siRNA interference experiment targeting the core AP-1 gene, jun , further substantiates that diminished AP-1 expression impedes the secretion of bioactive IL-1β in response to NA stimulation, despite NLRP3 expression being activated. Based on these findings, it can be deduced that NA promotes IL-1β expression by enhancing H3K9ac modification of the IL-1β promoter region and upregulating AP-1 expression. To summarize, the present findings indicate that the effects of metabolic syndrome on reproductive function in humans and other mammals may be related to abnormal NA accumulation that leads to follicular atresia, a decrease steroid hormone synthesis by granulosa cells, and blocking of corpus luteum formation. The mechanistic basis of this effect is likely to be NA-induced ovarian inflammation caused by an increase mitochondrial ROS production which activate NLRP3/IL-1β pathway. Further, the increase in IL-1β expression was found to be a result of enhanced H3K9ac modification of the IL-1β promoter region and upregulation of AP-1 expression. In addition, the findings indicate that the abnormal NA accumulation in the ovaries may be induced by downregulation of its downstream metabolizing enzyme CerS2. However, our study only provides preliminary evidence suggesting a close association between CerS2 and CXCL14 with abnormal NA accumulation. Further comprehensive investigations are warranted to elucidate the impact of reduced CerS2 expression on ovarian lipid homeostasis, the influence of elevated CXCL14 on ovarian immune homeostasis, and the consequential effects toward mammalian ovarian function and reproductive performance. These findings imply that NA could serve as a potential biomarker for ovarian dysfunction induced by metabolic syndrome in mammals and women; Further, the NA-related metabolic pathways identified could also provide targets for strategies to improve fertility.

Metabolic syndrome is a serious public health issue across the world that is prevalent in children, adolescents, and adults. In recent years, the prevalence of metabolic syndrome has been reported to be as high as 20%–30% [1] , [2] , [3] . According to estimates reported by the World Health Organization, by the year 2025, 300 million people worldwide will probably suffer from metabolic syndrome and its related disorders [4] . Metabolic syndrome is primarily characterized by central obesity, insulin resistance, dyslipidemia, and hypertension, and it can increase the risk of type 2 diabetes and cardiovascular diseases [5] , [6] . In addition, obesity and polycystic ovary syndrome (PCOS) are typical female reproductive complications of metabolic syndrome that result in infertility and poor pregnancy outcomes [7] . Recently, there has been a lot of interest in how metabolites associated with metabolic syndrome trigger reproductive system dysfunction [8] , [9] . The pathophysiology of metabolic syndrome is largely attributed to the excessive accumulation of fatty acids. The proinflammatory state associated with excessive fatty acid accumulation may further contribute to the development of this syndrome [10] , [11] , [12] . The abnormal fatty acid accumulation and the proinflammatory state potentially lead to the inhibition of reproductive functions [9] , [13] , [14] . During pregnancy, the abnormal accumulation of fatty acids, especially saturated and monounsaturated long-chain fatty acids, has been reported to induce placental oxidative stress and high levels of inflammation and, thereby, increase the risk of blockage of embryonic development and miscarriage [15] , [16] , [17] , [18] , [19] . The inflammatory pathways involved mainly include the toll-like receptor (TLR) pathway [20] , [21] , [22] , nuclear factor-κB (NF-ĸB) pathway [23] , cytochrome P450 pathway [24] , and NOD-like receptor family pyrin domain-containing 3 (NLRP3) inflammasome pathway [25] , among others. However, it is not clear how the abnormal accumulation of fatty acids and the induction of proinflammatory pathways observed in metabolic syndrome are mechanistically linked with loss of ovarian function. Fatty acids can mediate the activation of various inflammatory signaling pathways and the release of inflammatory cytokines [22] , [26] . NLRP3, a cellular sensor, detects a wide range of endogenous danger signals and environmental stimuli, and in response, initiates the formation and activation of the NLRP3 inflammasome [27] . Multiple fatty acids can activate or inhibit intracellular NLRP3 signaling, thereby influencing the occurrence of inflammation [26] , [28] , [29] , [30] , [31] . For instance, palmitic acid activates NLRP3 by inhibiting the adenosine monophosphate-activated protein kinase pathway, which leads to an increase in reactive oxygen species production [29] . Additionally, changes in the levels of epigenetic modifications have been found to directly impact NLRP3 activation. For instance, acetylation at the K21 and K22 sites of histones enhanced NLRP3 activation [32] , while increase in the expression of histone deacetylase sirtuin 2 (SIRT2) inhibited NLRP3 activation [32] . Further, histone deacetylase 6 was found to induce the autophagic degradation of NLRP3 [33] . In addition, there is ample evidence to indicate that fatty acids can alter intracellular histone acetylation modifications [34] , [35] , [36] , [37] . Thus, the mechanism via which fatty acids activate inflammatory pathways may involve histone modifications of the NLPR3 inflammasome. However, it is not clear whether the abnormal accumulation of fatty acids impacts histone acetylation and mediates NLRP3 activation in the context of metabolic syndrome. In the present study, we have tried to fill in the currents gaps in the literature by investigating the role of fatty acid accumulation and pro-inflammatory pathways associated with metabolic syndrome in abnormal reproductive function and the related mechanisms. We have used sows in the early pregnancy period as an experimental model, and as expected, sows with low embryo survival rates exhibited abnormal lipid metabolism and ovarian inflammation. Lipid metabolomics analysis revealed that the ovarian inflammation and dysfunction in the sows with low embryo survival rates were attributable to the abnormal accumulation of nervonic acid (NA) [38] , a type of very-long-chain monounsaturated fatty acid closely linked to endocrine function. These findings were corroborated by similar findings in humans, cows, and mice. Based on this finding, our in vivo and in vitro mechanistic experiments focused on how the abnormal accumulation of NA affects the NLPR3 inflammasome and its related inflammatory pathways, as well as their epigenetic modifications, in ovarian cells.

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.