Breeding pairs were purchased from Jackson Laboratories (Bar Harbor, ME, United States). Female C57BL/6J (B6) mice (8-week old) and B6.Cg-Lepob/J (ob/ob) were housed in the Animal Facility of Institute of Animal Reproduction and Food Research, Polish Academy of Sciences in Olsztyn. Mice were housed with free access to food and water for the duration of the study (humidity 50 ± 10%; 23°C; 12L:12D cycle). All procedures were approved by the Local Animal Care and Use Committee for the University of Warmia and Mazury in Olsztyn. Guidelines for animal experiments followed EU Directive 2010/63/EU. Throughout the experiments, mice were monitored for any sings of welfare or disease. At 8 week of age, mice were subjected to various protocols.

For hormonal treatment protocol, the estrous cycle was monitored after vaginal cytology analysis. Cells were collected on a glass slide and stained with Diff Quik^® kit (Medion Diagnostics AG, Switzerland, DQ-ST). Estrus (E) was characterized by cornified epithelium cells, metestrus by both cornified cells and leukocytes, diestrus (D) by predominant leukocytes, and proestrus by nucleated cells, as previously described (Kyrönlahti et al., 2011). Subsequently, mice received either eCG or hCG, as previously described (Hasegawa et al., 2016). One group of female B6 mice (seven to eight mice/group) was injected in E with equine chorionic gonadotropin (eCG, G4877; 5 IU, Sigma Aldrich) followed by human chorionic gonadotropin (hCG, Chorulon, 5 IU, MSD Animal Health) 48 h after, and tissues were collected 18–20 h later in E. In the second group (seven to eight mice/group), animals were injected with hCG, and tissues were collected 16–18 h later in D. In the DIO model, mice (8–10/group) were placed on standard chow diet (CD, #5053, Picolab Rodent diet 20 with 13% of calories) or HFD (59% of calories, AIN-76A with 33% hydrogenated coconut oil; LabDiet) for 4 or 16 week. In the hyperleptinemic model (8–10 mice/group), high circulating levels of leptin were obtained through intraperitoneal administrations of leptin twice (100 μg/day injected at 09:00 and 21:00), while the control group received saline (Recombinant Mouse Leptin, GFM26, Cell Guidance Systems). Regarding the ob/ob model, mice (8–10 mice/group) were kept on CD until 12 week of age.

Both immunohistochemistry (IHC) and immunofluorescence (IF) followed the protocols previously described (Wołodko et al., 2020). Briefly, mouse ovaries were collected during the D stage, fixed in 4% paraformaldehyde, and stored at 4°C. On the following day, the tissues were dehydrated in ethanol. Paraffin embedded tissues were then cut in 5-μm sections and mounted on a glass slide. After deparaffinization in xylene, samples were rehydrated in increasing ethanol concentrations. For antigen retrieval, sections were boiled in citrate buffer (10 mM, pH = 6). Next, sections were incubated with BSA (A2153; Sigma Aldrich) for 1 h at room temperature (RT), followed by incubation with rabbit polyclonal antibody against NLRP3 (1:200, ab214185; Abcam) overnight at 4°C. In order to test the specificity of the primary antibody, sections were also incubated with rabbit polyclonal anti-immunoglobulin G (IgG, ab37415; Abcam) or without primary antibody (negative controls). On the following day, sections were washed and incubated for 1 h at RT with peroxidase-conjugated goat anti-rabbit IgG polyclonal antibody (1:100, 410972; Dako). Finally, slides were incubated in 3,3-diaminobenzidine (DAB) for 15 s before visualization. The sections were then counterstained for 2 min using hematoxylin (MHS16; Sigma Aldrich). Stained sections were dehydrated and mounted under glass coverslips.

Concerning the immunofluorescence protocol, after deparaffinization in xylene, samples were in increasing ethanol concentrations. Sections were then permeabilized in 0.3% Triton X-100 (T8787; Sigma Aldrich) two times for 5 min at RT. Antigen retrieval was done after heating the sections in citrate buffer (10 mM, pH = 6). Blocking was done with BSA (A2153; Sigma Aldrich) with 0.3 M glycine (G8898; Sigma Aldrich) in phosphatase-buffered saline (PBS)–0.1% Tween 20 (P7949; Sigma Aldrich) (PBST) solution for 2 h, at RT. In order to minimize the autofluorescence, tissues were treated with 0.3% Sudan Black (199664; Sigma Aldrich) (in 70% ethanol) for 10 min at RT. Subsequently, sections were washed eight times for 5 min in PBST (0.1% Tween20). The sections were then incubated with rabbit polyclonal antibody against NLRP3 (1:100, NBP2-12446; Novus Biologicals) overnight at 4°C. The negative controls were after incubating the slides with rabbit polyclonal anti-immunoglobulin G (IgG, ab37415; Abcam) or without primary antibody (negative controls). On the following day, the sections were washed six times for 5 min in PBST at RT, followed by incubation with the secondary antibody cyanine 3 (Cy3)-AffiniPure donkey polyclonal anti-rabbit IgG (H+L) (1:400, 711-165-152; Jackson ImmunoResearch) for 2 h at RT. The sections were washed six times for 5 min in PBST at RT. In order to visualize the nucleus, DAPI (dilution 1:100 with PBS, D9542-10MG; Sigma Aldrich) was added for 30 min at RT. Finally, the slides were covered with Vectashield medium (H-1000; Vector Laboratories) and stored at −20°C. Images were captured using Zeiss Axio Observer System (Carl Zeiss, Germany) with × 20/0.4 NA, and confocal microscope (Zeiss LSM 800) with × 40/1.2 or × 63/1.4 NA oil immersion objectives. For examination of the slides, an Axio Observer Systems Z1 microscope (Carl Zeiss Microscopy GmbH, Germany) and the Zeiss ZEN 2.5 lite Microscope Software (Carl Zeiss, Germany) were used.

Protein expression in mouse ovary and liver was assessed by Western blotting. Ovaries and livers were homogenized with RIPA buffer (R0278; Sigma Aldrich) after adding protease inhibitors (phenylmethylsulfonyl fluoride, PMSF and Protease Inhibitor Cocktail, P8340; Sigma-Aldrich) and phosphatase inhibitors (Pierce Phosphatase Inhibitor Mini Tablets 88667; Thermo Fisher Scientific) and incubating on ice for 1 h. After centrifugation (20,000 × g, 15 min, 4°C), the supernatants were collected, and protein concentration was determined with the Smith et al. (1985) copper/bicinchoninic assay [Copper (II) Sulfate, C2284; Sigma and Bicinchoninic Acid Solution, B9643; Sigma Aldrich]. Samples were run (40 μg of protein) on 10–18% polyacrylamide gels. After transfer, the membranes were blocked in PBS solution containing 3% powdered milk for 1 h. Immunoblotting was performed using the primary antibodies NLRP3 (AG-20B-0014-C100; Adipogen), CASP1 (ab108362; Abcam), IL-18 (ab71495; Abcam), β-actin (A2228; Sigma Aldrich) and glyceraldehyde 3-phosphate dehydrogenase (GAPDH, ab9485; Abcam) on nitrocellulose (10600009; GE Healthcare Life Science) or polyvinylidene fluoride (PVDF) membrane (IPVH00010; Merck Millipore). Primary antibodies were incubated overnight at 4°C. The following day, proteins were detected by incubating the membranes with polyclonal anti-mouse horseradish peroxidase (HRP)-conjugated secondary (1:10,000, 31430; Thermo Fisher Scientific), polyclonal anti-rabbit HRP-conjugated secondary (1:20,000, 31460; Thermo Fisher Scientific), polyclonal anti-mouse alkaline phosphatase-conjugated secondary (1:10,000, 31321; Thermo Fisher Scientific), and polyclonal anti-rabbit alkaline phosphatase-conjugated secondary (1:10,000, A3687; Sigma Aldrich) antibodies, for 1.5 h in the chemiluminescence method or 2.5 h in the colorimetric method at RT. All antibody specifications are summarized in Table 1. Immunocomplexes were visualized subsequently using chemiluminescence detection reagent (SuperSignal West Femto kit, 34095; Thermo Fisher Scientific) or chromogenic substrate NBT/BCIP diluted 1:50 (11681451001; Roche) in alkaline phosphate buffer. Band density for each of the target protein was normalized against β-actin for NLRP3 and IL-18, while GAPDH was used for CASP1 as a reference protein. Finally, bands were quantified using the ChemiDoc or VersaDoc MP 4000 imaging system (Bio-Rad). Quantitative measurements of blot intensity were performed using ImageLab software.

Total RNA was extracted from whole ovary and 10 mg of liver, using TRI reagent (T9424; Sigma Aldrich) following the instructions of the manufacturer. RNA samples were stored at −80°C. Concentration and quality of RNA was determined spectrophotometrically, and the ratio of absorbance at 260 and 280 (A_260/280) was analyzed confirming good RNA quality. Subsequently, 2 μg of RNA was reverse transcribed into cDNA using Maxima First Strand cDNA Synthesis Kit for RT-qPCR (K1642; Thermo Fisher Scientific) (Galvão et al., 2012).

Real-time PCR assays were performed in a 7900 Real-time System (Applied Biosystems), using a default thermocycler program for all genes: a 10-min preincubation at 95°C was followed by 45 cycles of 15 s at 95°C and 1 min at 60°C. A further dissociation step (15 s at 95°C, 15 s at 60°C, and 15 s at 95°C) ensured the presence of a single product. Ribosomal protein L37 (Rpl37) was chosen as a housekeeping gene and quantified in each real-time assay together with the target gene. Based on gene sequences in GenBank (National Center for Biotechnology Information), the primers for Nlrp3, Casp1, Il-1β, Il-18, Asc, Il-10, and Tnf, which sequences are presented in Table 2, were designed using Primer Express 3.0 software (Applied Biosystems). All reactions were carried out in duplicates in 384-well plate (4309849; Applied Biosystems) in 12 μl of total solution volume (Galvão et al., 2014). The data were analyzed using the real-time PCR Miner algorithm (Zhao and Fernald, 2005).

The concentrations of IL-1β in tissue extracts of ovaries and livers were determined using an IL-1 beta Pro-form Mouse Uncoated ELISA kit (88-8014-22; Thermo Fisher Scientific) following the instructions of the manufacturer. The standard curve concentrations ranged from 25 to 3,000 ng/m, and interassay coefficient variation (CV) was 7.27%.

We used RNA-seq data from CCs previously produced by Wołodko et al. (2020). The dataset is publicly available under the GEO accession number GSE180300¹. Library generation was previously described (Wołodko et al., 2020). Briefly, we collected approximately 50 CCs per animal, after superovulation, and RNA-seq libraries were generated using a Smart-seq2 oligo-dT method. Read counts were normalized by size factor and log-2 transformed. Partial correlations were used to determine coexpressing genes (cor > 0.9) using the corpcor package (Schäfer and Strimmer, 2005). Network statistics were calculated using the igraph package (Csárdi and Nepusz, 2006). Network visualization was done with Cytoscape (Shannon et al., 2003).

Statistical analyses were performed using the GraphPad Prism Software (Version 9.01, GraphPad Software, Inc.; La Jolla, CA, United States). Sample normal distribution was determined using the D’Agostino–Pearson omnibus test. Mann–Whitney test, simple t-test, or multiple unpaired t-test were used to analyze the data, and statistical significance was calculated with Bonferroni–Sidak corrections for multiple comparison, depending on the experiment (details in figure legend). Results were presented as means with standard deviation. Differences between means for all tests were considered statistically significant if p < 0.05.

We first sought to characterize the expression of NLRP3-induced inflammasome components in the ovaries of mice throughout the estrous cycle. Fifteen female 8-week-old C57BL/6 (B6) mice were treated with hormones in order to synchronize the estrous cycle (Figure 1A). Ovaries were collected in the E and D stage and further processed for mRNA or protein expression analysis, respectively. Real-time PCR analysis (n = 6–7/group) revealed increased levels of Casp1, Il-1β, and Il-18 mRNA in the D stage (Figure 1B, p < 0.05). Moreover, Western blotting (n = 7–8/group) revealed increased NLRP3 protein expression in D (Figure 1C, p < 0.05), as well as the pro-peptide (p24) and mature form (p18) of IL-18 (Figures 1F,G; p < 0.01). Regarding CASP1, the long form (p45) was decreased in D (Figure 1D; p < 0.05), but no significant changes were observed for the active CASP1 (p20) (Figure 1E). These results suggest the activation of NLRP3 inflammasome in the D stage, through upregulation of NLRP3 and its downstream mediator IL-18. Next, we characterized the cellular distribution of NLRP3 protein in the ovaries collected from mice in D, using IHC and IF (n = 2–3/group). We confirmed that NLRP3 protein was visible in ovarian follicles, corpora lutea, and interstitial space (Figure 1I). On the other hand, a closer observation of IHC sections revealed staining in granulosa cells (GC) and theca cells (TC), as well as in oocytes, in all developmental follicular stages (Figures 1K–M). Negative controls stained exclusively with secondary antibodies did not reveal any brown staining (Figures 1H,J). The specificity of our IHC staining was corroborated by IF, in which a clear yellow staining was observed in GC, TC, and oocytes (Figures 1O–Q). Negative control stained with rabbit immunoglobulin type G (IgG) confirmed no staining (Figure 1N). Our results not only confirmed the presence of NLRP3 protein in GC, TC, and oocytes by IHC and IF, but also confirmed the upregulation of NLRP3 in the ovaries of eCG-treated mice. Therefore, in subsequent experiments, collections were consistently performed in D.

In the following experiment, we tested the effects of short- (4 week) vs. long-term (16 week) HFD treatment on NLRP3-induced inflammasome activation in the ovary of mice (Figure 2A). Throughout the mouse DIO protocol, the average body weight (BW) was recorded every 4 week (Table 3). After collection, ovaries were processed for mRNA and protein expression analysis. Real-time PCR analysis (n = 6–8/group) revealed increased mRNA of Nlrp3 after 4-week HFD (Figure 2B, p = 0.06), whereas Il-1β levels were increased after both 4 and 16 week of HFD (Figure 2B, p < 0.05). Regarding Western blotting analysis (n = 7–8/group), we found that the expression of NLRP3, CASP1 p45, and pro IL-18 p24 were increased in the 4-week HFD group, compared with the control group (Figures 2C,D,F, p < 0.05, respectively). The opposite pattern was seen after 16-week HFD, with the downregulation of NLRP3 expression, the mature form of CASP1 p20 and both forms of IL-18 (p24 and p18) (Figures 2C,E–G, p < 0.05). Finally, we also confirmed that IL-1β protein level was upregulated in 16-week HFD by enzyme-linked immunosorbent assay (ELISA) (n = 4–6) (Figure 2H, p = 0.082). We presently showed that despite the upregulation of NLRP3, the pro-proteins IL-18 (p18, p24) and CASP1 (p45) after 4-week HFD treatment, the expression profile of NLRP3, CASP1 (p20), and both forms of IL-18 (p18, p24) were consistently downregulated after a 16-week HFD treatment.

After temporally characterizing the expression profile of NLRP3-induced inflammasome components in the ovary of DIO mice, we further interrogated whether the activation of NLRP3 inflammasome was regulated by leptin signaling. Therefore, we analyzed the levels of NLRP3 inflammasome components in the ovaries of a previously validated mouse model of pharmacological hyperleptinemia, which presented increased systemic levels of leptin and increased leptin signaling in the ovary without obesity (Wołodko et al., 2020), and a genetically obese mouse ob/ob characterized by extreme obesity and lack of circulating leptin. In the pharmacological hyperleptinemic model (n = 8–10/group), B6 female mice were treated with leptin intraperitoneally 16 days (16 L), whereas controls were administered saline (16 C) (Wołodko et al., 2020). Moreover, in the ob/ob model (n = 8–10/group), control females (+/?) and homozygous mutant females (−/−) were kept on CD for 4 week (Figure 3A). Body weights of the mice were registered periodically (Table 3). Ovaries from all groups were collected and processed for mRNA and protein expression analysis. Real-time PCR analysis (n = 6–8/group) revealed an increase in Nlrp3 and Casp1 in 16 L, but a decrease in ob/ob −/− mice (Figure 3B, p < 0.05). Furthermore, the mRNA of Il-1β was upregulated in 16 L (Figure 3B, p < 0.05). Finally, the mRNA of Il-18 was significantly downregulated in the ob/ob −/− group (Figure 3B, p < 0.05). With regard to protein expression (n = 6–8/group), we found that the 16 L group presented increased levels of NLRP3 (Figure 3C, p < 0.05), whereas the opposite pattern was observed in ob/ob −/− mice, compared with control groups (Figure 3C, p < 0.05). Accordingly, both pro-peptides CASP1 (p45) and CASP1 (p20) showed increased levels in 16 L (Figures 3D,E; p = 0.07 and p < 0.05, respectively); nonetheless, no significant changes were found in the ob/ob model. Importantly, IL-1β protein levels measured by ELISA (n = 4–5/group) were increased in 16 L, but decreased in ob/ob −/− (Figure 3F, p < 0.05). In this experiment, we revealed the functional link between leptin signaling and NLRP3 inflammasome component regulation in the ovary, with leptin treatment inducing the activation of NLRP3 and CASP1 with subsequent secretion of IL-1β. Furthermore, we confirmed the downregulation of NLRP3 expression in the ovaries of ob/ob −/−.

In this experiment, we examined whether the association between leptin signaling and NLRP3 inflammasome activation previously observed in whole ovaries holds true at the cellular level. We, therefore, analyzed particularly the somatic companions of the female gamete, the CCs. We started reanalyzing the RNA sequencing (RNA-seq) datasets from CCs from 4- and 16-week DIO and pharmacological hyperleptinemic model (Figure 4A; Wołodko et al., 2020). We confirmed the expression levels of leptin and NLRP3 pathway components for 16 L and 4-week HFD, and despite no changes in Nlrp3 in CCs after 4-week HFD, the gene was upregulated in 16 L (Figure 4B). The low coverage of the samples (an average of 5.5 million reads) and the weak expression level of Nlrp3 in CCs may account for the lack of changes in 4-week HFD. Nonetheless, the consistent upregulation of various components of the NLRP3 inflammasome, like Nlrp3, Il-18, Casp1, Il-1β, and Asc in 16 L is suggestive of the stimulatory effects of leptin on the expression level and activation of NLRP3 inflammasome genes also in CCs (Figure 4B). As previously shown, the DESeq analysis of 4-week DIO protocol revealed 997 differentially expressed genes (DEGs) in 4-week HFD, whereas for pharmacological hyperleptinemic model, a total of 2,026 DEGs were found in 16 L (Wołodko et al., 2020), in comparison with their control groups (p < 0.05; Wołodko et al., 2020). In the present analysis, we overlapped the DEGs from 4-week HFD and 16 L and identified seven genes either up- or downregulated in both conditions (Figure 4C). Subsequently, we integrated the 14 DEGs with the main components of NLRP3 and leptin signaling pathways (Wołodko et al., 2020), based on the correlation between their expression levels (p > 0.90), and obtaining five clusters (Figure 4D). Of note, one of the clusters revealed the gene interaction between Casp1, phosphatase and tensin homolog (Pten) and signal transducer and activator of transcription 5a (Stat5a), while others showed a link between Socs3 and Il-1β, known as an important axis involved in the mediation of immune response (Chaves de Souza et al., 2013). Importantly, other genes were highlighted in the network, as solute carrier family 22 member 15 (Slc22a15), a cell membrane transporter and metabolic gene (Nigam, 2018), or stress-associated endoplasmic reticulum protein 1 (Serp1) involved in protein unfolding and stress response (Yamaguchi et al., 1999). Indeed, metabolic performance in the preovulatory follicle is tightly regulated and involves the crosstalk between GC and oocyte (Figure 4D; Wołodko et al., 2021). The gene ontology analysis for the presented network revealed three main events: negative regulation of glucose transport, positive regulation of cytokine biosynthesis, and response to ATP (Figure 4E, p < 0.05). Finally, we plotted a subset of genes known to directly activate the NLRP3 inflammasome signaling pathway (Weber et al., 2020), particularly regarding the regulation of glutathione, major mediator of NLRP3 signaling (Hughes et al., 2019), as well as other genes involved in the pathway regulation (Zhou et al., 2010; Barlan et al., 2011; Iyer et al., 2013; Mitoma et al., 2013; Wang et al., 2014; Kim et al., 2015; He et al., 2016b; Jo et al., 2016; Li et al., 2016; Wolf et al., 2016; Guglielmo et al., 2017; Hughes and O’Neill, 2018; Palazón-Riquelme et al., 2018; Shuvarikov et al., 2018; Billon et al., 2019; Hughes et al., 2019; Martine et al., 2019; Zhang et al., 2021). We confirmed a similar behavior between 16 L and 4-week HFD for those genes, in opposition to 16-week HFD (Figure 4F). Thus, systemic administration of leptin-activated genes from the NLRP3 inflammasome pathway in CCs, corroborating the functional link between leptin signaling and NLRP3 inflammasome activation in CCs of DIO mice.

In the last experiment, we compared the temporal activation of NRLP3 inflammasome between the ovary and another metabolic organ as the liver, during obesity progression. Thus, we analyzed the expression profile of NLRP3 inflammasome genes in the liver of DIO mice. Furthermore, we tested once more the functional link between leptin signaling and activation of NLRP3 inflammasome in the liver, using pharmacological hyperleptinemic and ob/ob mouse models. Liver samples were collected from DIO, leptin treated, and ob/ob female mice, for mRNA transcription and protein expression analysis (Figure 5A). Real-time PCR analysis (n = 6–9/group) showed no significant changes in expression of all inflammasome components, except for the increase in Nlrp3 and Il-1β mRNA in 16-week HFD, but downregulation of Nlrp3 in ob/ob−/− (Figure 5B, p < 0.05). Furthermore, protein analysis determined by Western blotting (n = 4–8/group) showed an increase in protein levels of NLRP3, CASP1 (p20), and IL-18 (p18) after 16-week HFD treatment, compared with control (Figures 5C,E,G, p < 0.05, p = 0.07, p = 0.08). Regarding the pro-peptide of CASP1 (p45), its protein was upregulated in 16 L, but downregulated in ob/ob−/−, compared with controls (Figure 5D, p < 0.05 and p < 0.01, respectively). Finally, CASP1 (p20) and IL-18 (p18) proteins were decreased in ob/ob−/− (Figures 5E,G, p < 0.05 and p = 0.08, respectively). No significant changes were observed for pro IL-18 (p24) (Figure 5F). The present results of our analysis in the liver indicate a site-dependent NLRP3 inflammasome regulation throughout obesity, since overexpression of NLRP3, CASP1 (p20), and IL-18 (p18) took place only at 16 week of DIO. Thus, we observed drastic differences in NLRP3 inflammasome profile between liver and ovary (Figure 5H), which showed the upregulation of NLRP inflammasome components after 4-week DIO (Figure 2). Another important observation was the downregulation of NLRP3 and CASP1 (p20) in livers from ob/ob −/− mice, suggesting that leptin directly modulates NLRP3 inflammasome activation at hepatic level. Hence, we confirmed the latency of NLRP3 inflammasome activation in the liver of DIO female mice, which showed signs of upregulation only after 16- week HFD treatment. Furthermore, we have confirmed the functional link between leptin and NLRP3 inflammasome activation in the liver.