Intraperitoneal translocation of gut microbiota induces NETosis and promotes endometriosis

Background Endometriosis is a debilitating gynaecological disorder with an elusive pathogenesis. While gut microbiota dysbiosis has been implicated, the causal role of gut-peritoneum microbial translocation and the specific mechanisms driving disease progression remain elusive. Notably, the role of peritoneal neutrophils and neutrophil extracellular traps (NETs) in the development of endometriosis remains unknown.

This study aims to delineate the pathogenic pathway linking gut microbiota to peritoneal neutrophil activation and the development of endometriosis. Design We combined single-cell RNA sequencing of clinical peritoneal fluid immune cells with functional validation in heterologous and homologous mice models. We further adopted microbial source-tracking analysis of patient cohorts and interventional strategies, including faecal microbiota transplantation (FMT) and administration of green fluorescent protein (GFP)-tagged Pseudomonas aeruginosa.

We identified a unique membrane metalloendopeptidase (MME) positive neutrophil subset (Neu_MME) that is expanded in endometriosis and primed for NETs formation (NETosis). These Neu_MME released NETs in response to bacterial lipopolysaccharides (LPS), which directly captured endometrial cells and enhanced their proliferation and migration, driving lesion development. Accordingly, inhibiting NETosis or degrading NETs significantly suppressed endometriosis in mice. Furthermore, FMT from patients with endometriosis to mice disrupted the intestinal barrier, promoting the translocation of gut microbiota, particularly Pseudomonas, into the peritoneal cavity and the lesions. This translocated Pseudomonas was identified as a key driver of LPS-induced NETosis and disease progression.

Our findings define a gut-peritoneum axis in endometriosis, where gut-derived Pseudomonas triggers NETosis in peritoneal Neu_MME to promote disease, suggesting that targeting this bacterium or NETosis represents a viable therapeutic strategy. - Bacterial Translocation - Intestinal Barrier Function - Intestinal Bacteria - Immunology - Cell Proliferation Data availability statement Data are available upon reasonable request. The data that support the findings are available within the article and its supplementary materials. The raw 16s rRNA microbiota data have been deposited at https://www.biosino.org/node under project ID: OEP00006185. The raw single-cell RNA sequencing data of peritoneal immune cells are available at the same repository under project ID: OEP00006180. Raw data not included therein can be obtained with the consent of the corresponding author. Statistics from Altmetric.com WHAT IS ALREADY KNOWN ON THIS TOPIC Although gut microbiota has been associated with the development of endometriosis, the gut-peritoneum microbial crosstalk and its causal effects in endometriosis remain unclear. The specific role of neutrophils and neutrophil extracellular traps (NETs) in the development of endometriosis remains unknown. WHAT THIS STUDY ADDS We identified a pathogenic membrane metalloendopeptidase (MME) positive neutrophil subset (Neu_MME) in peritoneal fluid as the primary source of NETs. NETs promote lesion development by capturing endometrial cells and enhancing their proliferation and migration. We found a causal gut-peritoneum axis, where patient-derived gut microbiota disrupts the intestinal barrier, enhancing Pseudomonas translocation and lipopolysaccharide-induced NETs formation (NETosis). HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY Targeting peritoneal Pseudomonas and inhibiting NETosis are potential strategies for endometriosis prevention and treatment.

Endometriosis is a chronic gynaecological disorder affecting about 10% of reproductive-aged women worldwide,1 characterised by ectopic endometrial-like tissue growth that causes pelvic pain and infertility.2 Its pathogenesis remains incompletely understood.3 Sampson’s theory of retrograde menstruation proposes that shed endometrial cells implant in the peritoneal cavity,4 while subsequent hypotheses involving hormonal, genetic and immune factors highlight the role of microenvironmental modulators.5 Recent evidence implicates the microbiome, particularly gut and peritoneal microbiota, in the development and progression of endometriosis through inflammatory pathways.6 The gut microbiota, a key regulator of systemic immunity and inflammation,7 exhibits dysbiosis in patients with endometriosis,8 with increased Proteobacteria and decreased Firmicutes.9 Similarly, the peritoneal cavity, once considered sterile, harbours distinct microbiota profiles in patients with endometriosis,10 marked by increased pathogenic bacteria (eg, Ruminococcus and Pseudomonas)9 and elevated endotoxin levels (eg, lipopolysaccharide (LPS)).11 These microbial shifts may promote chronic inflammation via LPS-Toll-Like Receptor 4 (TLR4) signalling.12 Shared microbial genera between gut and peritoneal fluid suggest potential gut-peritoneum microbial crosstalk.9 However, the evidence for gut-peritoneum bacterial translocation remains limited. Neutrophils, the most prevalent immune cells,13 are recruited to the peritoneal cavity in endometriosis and contribute to inflammation.14 Activated neutrophils release bioactive substances, including cytokines, chemokines, proteolytic enzymes, reactive oxygen species and neutrophil extracellular traps (NETs). NETs play important roles in innate immunity15 but can also exacerbate tissue damage.16 Microbial products like LPS can trigger NETs formation (NETosis),17 yet the pathways linking gut-peritoneum microbiota to NETosis in endometriosis are unclear. This study investigates how gut microbiota translocation induces NETosis via peritoneal Pseudomonas to promote endometriosis development.

and materials Detailed methods are provided in Supplementary file 1 (online supplemental file 1). 18–36 All reagents and antibodies used in this study are listed in key resources table (online supplemental key source table). Supplemental material Supplemental material Single-cell sequencing Peritoneal fluid CD45+ immune cells from six patients with endometriosis and three controls were subjected to single-cell RNA sequencing (scRNA-seq) using the Becton, Dickinson and Company (BD) Rhapsody system. ‘NETs-endometrial cells’ capturing assay A custom ‘NETs-endometrial cells’ capturing assay was designed to study NETs-endometrial cells interactions. F-actin was labelled with phalloidin- while NETs and endometrial cells were identified using antibodies against citrullinated histone H3 (citH3) and estrogen receptor 1 (ESR1) (online supplemental key source table), respectively. Mice models for endometriosis Homologous (C57BL/6) and heterologous (nonobese diabetic (NOD)/ShiLtJGpt-Prkdc em26 Il2rg em26/Gpt, NCG) endometriosis murine models were established using murine or human endometrial tissues.26 27 Derived intervention models include NETosis inhibition, neutrophil depletion, faecal microbiota transplantation (FMT) and GFP-tagged Pseudomonas aeruginosa (PAO1) treatment. Bacterial DNA sequencing and source tracking The V3-V4 region of the 16S rRNA gene was sequenced via the Illumina MiSeq platform.29 Amplicon sequence variants (ASVs) were taxonomically classified using the naive Bayes classifier with the SILVA 138–99 reference sequences.33 Microbial source-tracking was performed with the Fast Expectation-Maximisation for Microbial Source Tracking (FEAST) algorithm.34 To quantify the translocation of specific bacterial genera, all corresponding ASVs were used as input for the FEAST analysis. Statistical analyses All the data were analysed using GraphPad Prism V.8.0.2 for Windows (www.graphpad.com) and are presented as the mean±SD. For comparisons between two groups, Student’s t-test was used. For comparisons among more than two groups, one-way analysis of variance was performed, followed by an appropriate post hoc test (Tukey’s or Dunnett’s) for specific group comparisons. Correlations between variables were assessed using Pearson’s or Spearman’s correlation analysis. A two-sided value of p<0.05 was considered statistically significant.

An MME+ neutrophil subset is expanded in the peritoneal fluid of endometriosis and prone to NETosis To delineate the immune landscape, we performed scRNA-seq of peritoneal fluid CD45+ cells (figure 1a; online supplemental table S2). Based on the marker genes of immune cell clusters (online supplemental figure S1a-d), eight distinct clusters were identified as follows: B cells and plasma cells, dendritic cells, eosinophils and basophils, macrophages and monocytes (Ma-Mono), mast cells, neutrophils, T cells and natural killer cells (T-NK), and proliferative cells (figure 1b). Among these clusters, Ma-Mono was the predominant one, accounting for 53.7% in the control group and 67.5% in the endometriosis group (figure 1c). Notably, the proportions of Ma-Mono and neutrophils were significantly higher in the endometriosis group, whereas those of T-NK and mast cells were significantly lower (figure 1d, online supplemental figure S1h). Moreover, a comparative analysis of patients with endometriosis with different revised American Society for Reproductive Medicine (rASRM) Scores showed that the proportion of neutrophils was significantly higher in patients at rASRM stages III-IV (figure 1e). Supplemental material Supplemental material Neutrophils were further classified into subclusters as previously reported37 (online supplemental figure S1e). Five distinct subclusters were identified: Neu_CD83, Neu_IFIT1, Neu_PTGS2, Neu_LCN2 and membrane metalloendopeptidase (MME) positive neutrophil subset (Neu_MME). Neu_MME refers to a distinct subset of neutrophils characterised by high expression of the MME, S100 calcium binding protein A12 (S100A12), matrix metallopeptidase 9 (MMP9) and peptidyl arginine deiminase 4 (PADI4) (online supplemental figure S1e). Remarkably, the proportion of the Neu_MME subcluster showed a significant increase in the endometriosis group (figure 1f,g). There was prominent communication among all subtypes, specifically between Neu_IFIT1, Neu_PTGS2 and Neu_MME, mediated by the CXCL8-CXCR2 signalling axis (online supplemental figure S3). In a clinical cohort (EM n=20, Con n=20; online supplemental table S4), the proportions of neutrophils and Neu_MME (CD45+ CD11b+ CD16b+ MME+) in the peritoneal fluid were significantly higher in the endometriosis group than those in the control group. Moreover, the proportions of neutrophils and Neu_MME were also higher in patients at rASRM stages III-IV (figure 1h,i, online supplemental figure 2a–c), indicating that an increased abundance of Neu_MME in the peritoneal fluid might be related to the progression of endometriosis. Supplemental material Supplemental material Supplemental material According to the pseudotime ordering of neutrophils, the cluster with the lowest CytoTRACE Score (Neu_CD83) was designated as the starting point, and Neu_LCN2 and Neu_MME were identified as the endpoints of the trajectory (figure 1j,k). Gene ontology biological process and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analyses showed that the Neu_MME was enriched in pathways including neutrophil degranulation, chemotaxis, response to LPS, NETosis, regulation of actin cytoskeleton and leucocyte transendothelial migration (figure 1l), suggesting their potential to undergo NETosis in response to LPS within the endometriosis microenvironment. In summary, the recruitment and activation of Neu_MME in the peritoneal fluid are associated with the development and progression of endometriosis, potentially affected by bacterial components in the peritoneal microenvironment. Bacterial LPS induces NETosis in Neu_MME Consistent with their NETosis-prone signature, we found elevated levels of myeloperoxidase (MPO) and citH3, markers of activated neutrophils and NETs, in patient lesions and peritoneal fluid (figure 2a,b). Co-expression of MME, CD16b and citH3 suggested Neu_MME as a source of NETs (figure 2c). The peritoneal fluid supernatant (PF_SN) from patients significantly upregulated MPO and citH3 in neutrophils, indicating the presence of proinflammatory stimuli (figure 2d). Consistent with this, levels of LPS in peritoneal fluid and LPS binding protein (LBP) in serum were significantly elevated in patients (figure 2e). LPS treatment markedly increased the proportion of Neu_MME (figure 2f). Moreover, an analysis of differentially expressed genes in Neu_MME between patients with endometriosis and control subjects revealed a significant upregulation of FosB proto-oncogene (FOSB), Fibronectin 1 (FN1) and Cathepsin L (CSTL) in the endometriosis group (online supplemental figure S4a). Neutrophils treated with LPS showed a significant upregulation of TLR4, tumour necrosis factor alpha, PADI4, MME, FOSB and citH3, further suggesting their involvement in NETosis (figure 2g). Collectively, these findings indicate that the levels of NETs derived from Neu_MME are significantly increased in the peritoneal fluid of patients, with LPS serving as a key trigger. Supplemental material The expression of MPO and citH3 was significantly higher in clinical endometriotic lesions compared with normal or eutopic endometria (figure 2h; online supplemental table S5). Receiver operating characteristics curve analysis demonstrated high diagnostic accuracy for MPO and citH3, with area under the curve values of 0.924 and 0.935, respectively, suggesting their potential as diagnostic biomarkers for detecting endometriosis lesions (figure 2i). Higher levels of MPO and citH3 were observed in ESR1-positive regions. ESR1 is a marker for endometrial glandular cells in endometriosis. These findings indicate that the recruitment and activation of neutrophils predominantly occur in the glandular cell layer adjacent to endometriosis lesions, suggesting that NETosis and NETs may play a role in the pathogenesis and progression of endometriosis (figure 2h,j). Supplemental material Furthermore, fluorescence in situ hybridisation (FISH) with Eubacteria probe 338 (EUB338) detected bacterial 16S rRNA in endometriosis lesions, along with positive LPS staining but low or undetectable LTA (figure 2k; online supplemental table S1). The distributions of EUB338 and LPS were enriched in the glandular cell layer adjacent to the lesions, consistent with the distribution patterns of MPO and citH3 (figure 2h,k). Patients with rASRM stages III-IV showed significantly higher levels of MPO, citH3, EUB338 and LPS than those with stages I-II (figure 2l). Correlation analysis indicated significant positive associations between MPO, citH3, EUB338, LPS and rASRM Scores (figure 2m,n and online supplemental figure S4b). These data establish bacterial LPS as a key inducer of NETosis from Neu_MME in endometriosis. Supplemental material NETs capture endometrial cells and drive lesion development We observed ‘neutrophil-NETs-endometrial cell’ aggregates in patients’ peritoneal fluid, which implies a potential interaction between NETs and endometrial cells in the peritoneal fluid (figure 3a). A custom-designed assay was developed to examine this interaction. Neutrophils and primary endometrial cells were successfully isolated and their characteristics were determined (online supplemental figure S5). Fluorescence co-localisation analysis revealed that endometrial cells tended to accumulate in NETs-containing regions. Even when F-actin inhibitor (cytochalasin D) was used to exclude the effect of cell autonomous migration, endometrial cell aggregation in NETs-containing regions could still be observed, which supported the hypothesis that NETs facilitate the capture of endometrial cells (figure 3b). Supplemental material In vitro experiments demonstrated that NETs enhanced the migratory and proliferative abilities of primary endometrial epithelial cells and stromal cells. This effect was eliminated by treatment with DNase I, which specifically degrades NETs (figure 3c–e). In an in vivo model of endometriosis using immunodeficient NCG mice, intraperitoneal administration of NETs significantly promoted the proliferation of endometrial cells and increased the volume of lesions (online supplemental figure S6a–e; online supplemental table S6). These findings emphasise the crucial role of NETs in capturing endometrial cells, promoting their migration and proliferation, and facilitating lesion formation. Supplemental material Supplemental material To further explore the role of NETs in endometriosis, endometriosis models in C57BL/6 mice were established.27 We observed that robust Ly6G+ neutrophils infiltrated into the nascent lesions at this early stage (day 3 after endometrial plantation, online supplemental figure S7). Mice were then treated intraperitoneally with LPS, GSK484, DNase I or phosphate-buffered saline (figure 3f). We found that LPS exacerbated endometriosis, increasing lesion size, stromal thickness, proliferation (Ki67) and angiogenesis (CD31). These effects were reversed through inhibition of NETosis using GSK484 to target PADI4, or through degradation of NETs using DNase I to target the DNA backbone of NETs (NET-DNA) (figure 3f–l, online supplemental figure S8a). LPS enhanced neutrophil infiltration (Ly6G) and NETosis (citH3) in lesions and peritoneal fluid, which was similarly suppressed by GSK484 and DNase I (figure 3m–p). Furthermore, the proportions of myeloid cells, macrophages, neutrophils and Neu_MME in the peritoneal fluid were significantly increased in the LPS group, while they were reduced in the GSK484 and DNaseI groups (figure 3q, online supplemental figure S8b). Moreover, neutrophil depletion mediated by anti-Ly6G antibody reduced the number of lesions, cell proliferation and angiogenesis (online supplemental figure S9a–j). Together, these results demonstrate that LPS-induced NETosis critically promotes endometriosis development. Inhibition of NETosis or elimination of NET-DNA could be a promising therapeutic approach for endometriosis. Supplemental material Supplemental material Supplemental material FMT from patients with endometriosis alters gut barrier and induces peritoneal NETosis We next investigated the sources of peritoneal bacteria and LPS. Mice were colonised with human faecal microbiota through patient-derived FMT (figure 4a). FMT from patients with endometriosis (FMT_EM) increased the number, size and stromal layer thickness of the lesions compared with the FMT from controls (FMT_Con), indicating that FMT from patients facilitated the formation and development of endometriosis lesions (figure 4b–f). The lesions in FMT_EM mice exhibited higher Ki67 expression and a greater number of CD31-positive capillaries, suggesting increased cell proliferation and angiogenesis (figure 4c,g,h, online supplemental figure S10a). Moreover, FMT_EM mice had higher levels of Ly6G and citH3 in the lesions, suggesting enhanced neutrophil recruitment and NETosis (figure 4i,j). Supplemental material FMT_EM led to an increase in LPS levels in both ileum and colon (figure 4k, online supplemental figure S10g). Moreover, the levels of LPS in the peritoneal fluid and LBP in serum were higher in the FMT_EM group (figure 4l). The alteration of intestinal permeability was further corroborated by the increased level of 70 KDa dextran in the serum of the FMT_EM group (figure 4m). Morphological analysis showed shorter villi and shallower crypts (figure 4n), as well as decreased expression of tight junction protein 1 (ZO-1) and Claudin4, suggesting disrupted tight junctions and an altered gut morphology (figure 4o, online supplemental figure S10b–g). Furthermore, bacterial infiltration into the intestinal wall was observed, with clear co-localisation of the universal bacterial probes (EUB338 mix) and the Pseudomonas-specific probe signals (figure 6d–f; online supplemental table S1), confirming the colonisation and invasion of the gut barrier by Pseudomonas. Pseudomonas signals were also detected within the ectopic lesions (figure 6d-e). Thus, endometriosis-associated microbiota disrupts the gut barrier, facilitating bacterial translocation and systemic inflammation that drives NETosis and disease progression. Gut microbiota is a source of peritoneal Pseudomonas The clinical faeces (EM_F group, n=105; Con_F group, n=100) and peritoneal fluid (EM_P group, n=91; Con_P group, n=101) samples were analysed using 16S rRNA gene sequencing (figure 5a, online supplemental figure S11a,b; online supplemental table S3). We observed that Proteobacteria, Firmicutes and Bacteroidetes were the dominant phyla across all sample types, and the proportion of Proteobacteria was higher in the endometriosis group than that in the control group, especially in the peritoneal fluid samples (figure 5b). Twelve differential genera at the genus level were shared between EM_F and Con_F, as well as between EM_P and Con_P, including Pseudomonas, Methylobacterium-Methylorubrum, Staphylococcus and Comamonas. Notably, Pseudomonas was the most abundant genus in the peritoneal fluid, and its abundance was significantly higher in the endometriosis group than that in the control group (figure 5c and online supplemental figure S11c). Supplemental material Supplemental material We found that the microbiota at the phylum level in peritoneal fluid was correlated with those in faecal samples in both the control and the endometriosis groups (online supplemental figure S11d). Based on source-tracking analysis, we observed that the proportion of microbiota in peritoneal fluid potentially originating from faecal microbiota was significantly higher than that from unknown sources in both the control and the endometriosis groups (figure 5d). On further analysis at the genus level of Pseudomonas, we observed that the proportion of Pseudomonas in peritoneal fluid originating from faecal microbiota was significantly higher than that from unknown sources in both the control and the endometriosis groups (figure 5e, online supplemental figure S11e). These findings imply that gut microbiota may act as a potential source of peritoneal Pseudomonas in endometriosis. Regarding microbe-derived functional profiles, including KEGG Orthology (KO) genes and pathways, several pathways were enriched in the endometriosis groups, including UDP−2,3−diacetamido−2,3−dideoxy−α−D−mannuronate biosynthesis, the superpathway of N−acetylneuraminate degradation and L−rhamnose degradation I (online supplemental figure S11f). Several KO genes associated with LPS were significantly more abundant in the peritoneal fluid microbiota and faecal microbiota, such as pagL, amT, waaP, Waag and lpxO (figure 5f). Spearman’s correlation analysis indicated that the expression levels of these LPS-associated KO genes were positively correlated with the abundance of Pseudomonas, especially in the peritoneal fluid samples (figure 5g). Among the top 20 genera that functionally contribute to the LPS-associated KO genes, Pseudomonas was identified as a potentially crucial genus in both faecal and peritoneal fluid samples (figure 5h). In clinical endometriotic lesions, we found distinct co-localisation of EUB338 and Pseudomonas (figure 6a,b), confirming Pseudomonas as an authentic component of the microbiota in human endometriotic lesions. At the species level, P. aeruginosa was also positive in these lesions (figure 6c). In the FMT model, co-localisation of EUB338 and Pseudomonas in the intestinal wall and positive Pseudomonas within the ectopic lesions were also observed (figure 6d–f), suggesting the translocation of Pseudomonas from the gut into the lesions. These findings position the gut as a source of peritoneal Pseudomonas in endometriosis. P. aeruginosa translocation promotes endometriosis via LPS-induced NETosis To investigate the role of Pseudomonas in endometriosis, PAO1 was used in in vitro and in vivo experiments. PAO1 is the most used strain for research on this ubiquitous and metabolically versatile opportunistic pathogen.38 Initially, PAO1 was cultured and its characteristics were determined in vitro (online supplemental figure S12a–g). We found that PAO1 effectively recruited neutrophils, significantly increased the proportion of Neu_MME and promoted NETosis (online supplemental figure S12h–k). In a murine model, PAO1 was administered intraperitoneally or intragastrically (figure 7a). Both intraperitoneal and intragastric administration of PAO1 significantly increased the number and size of lesions, as well as cell proliferation and angiogenesis, when compared with the blank and antibiotics mixture with vancomycin, neomycin, metronidazole and ampicillin groups (figure 7b,c, online supplemental figure S13a). These findings suggest that P. aeruginosa promotes the development of endometriosis. Supplemental material Supplemental material Administration of PAO1 significantly increased the proportion of Neu_MME in the peritoneal fluid and enhanced the activation of neutrophils and NETosis in lesions (figure 7c,d, online supplemental figure S13b). PAO1 was detected in the lesions of both the intraperitoneal_PAO1 and intragastric_PAO1 groups (figure 7c,d). Notably, PAO1 was cultured and identified in the peritoneal fluid and detected in the intestinal walls of the ileum and the colon in the intragastric_PAO1 group (figure 7e,g and online supplemental figure S14a–e). The expression level of LPS was significantly higher in the ileum and the colon of the intragastric_PAO1 group (figure 7f,h). The levels of LPS and LBP were elevated in the peritoneal fluid and serum of intragastric_PAO1 group mice (figure 7i). These findings suggest that P. aeruginosa translocates from the gastrointestinal tract into the peritoneal cavity, contributing to an LPS-induced inflammatory response. Supplemental material Intestinal permeability was increased in the intragastric_PAO1 group, as evidenced by the elevated levels of 70 KDa dextran in the peritoneal fluid and serum (figure 7j). Morphological analysis showed shorter villi and shallower crypts, as well as decreased expression of ZO-1 and Claudin4 in both the ileum and the colon, suggesting disrupted tight junctions and an altered intestinal epithelial structure (figure 7k–n, online supplemental figure S13c). Therefore, P. aeruginosa translocates from the gut, inducing NETosis via LPS to fuel endometriosis progression.

In this study, we demonstrated that bacterial LPS induces NETosis in Neu_MME within peritoneal fluid, promoting cell proliferation, angiogenesis and lesion development in endometriosis. We further identified that Pseudomonas translocates from gut microbiota into the peritoneal cavity and lesions, triggering LPS-induced NETosis and facilitating endometriosis progression. These findings provide new insights into the role of peritoneal microbiota in endometriosis and suggest potential novel therapeutic strategies. The peritoneal immune microenvironment in endometriosis is recognised as a ‘pro-endometriotic niche’,39 yet research has predominantly focused on macrophages40 41 in peritoneal fluid and lesions.42 43 Our work shifts the spotlight to neutrophils, confirming their significant recruitment in patients, particularly at advanced stages. We extend beyond mere quantification by identifying Neu_MME, a distinct subpopulation poised for NETosis. The high expression of MME, S100A12, MMP9 and PADI4 in this subset not only provides a specific cellular marker but also points to enhanced proteolytic and inflammatory potential, positioning Neu_MME as a key effector cell in the peritoneal milieu. Previous studies suggest that neutrophil infiltration is an early event in endometriotic lesion formation.14 Similarly, our data confirmed that neutrophils infiltrated into the nascent lesions at their early stage, and neutrophil depletion effectively inhibits lesion development. The efficacy of neutrophil depletion in curbing lesion development underscores their functional importance beyond early infiltration. The role of NETs is well established in cancer and inflammatory diseases,44 45 where they can facilitate metastasis by capturing circulating tumour cells.46 Our data robustly translate this concept to endometriosis. We provide direct evidence that NETs are not merely present but are functional: they form aggregates with endometrial cells in the peritoneal fluid and actively promote cellular proliferation and migration. This ‘capture and growth’ mechanism may explain how refluxed endometrial fragments are retained and nurtured within the peritoneal cavity. The therapeutic success of NETosis inhibition (GSK484) and NETs degradation (DNase I) in our models solidifies their pathogenic role. Our finding that NET-DNA is a primary active component opens avenues for therapeutic intervention, for instance, by targeting its receptor, CCDC25, which has been identified as a key mediator of NET-DNA signalling in other pathologies.47 48 The involvement of the microbiome in endometriosis is gaining support.6 Recent evidence and hypotheses have proposed the active involvement of subclinical genital microbial infections in the development and clinical progression of endometriosis.12 49 Comparative studies that emphasise the role of gut microbiota rather than cervical microbiota in disease progression9 highlight the potential routes of microbial translocation and their impact on local immune responses.50 In specific disease conditions, such as a compromised gut barrier or alterations in immune function,51 microorganisms can translocate into the peritoneal cavity, resulting in an imbalance in the microbial community. In peritoneal fibrosis, the potential interaction between gut and peritoneal microbiota has been highlighted, and the concept of the gut-peritoneum axis has been proposed.52 While debates continue regarding the relative contributions of the genital versus gut tract, our data provide compelling evidence for a gut-peritoneum axis. We move from correlation to causation by demonstrating a direct chain of events: endometriosis-associated gut microbiota compromises intestinal barrier integrity, facilitating the translocation of viable bacteria. Through sophisticated source-tracking and definitive FISH imaging, we identify Pseudomonas as a key translocating genus. The use of PAO1 was pivotal, allowing us to visually confirm its journey from the gut lumen to the peritoneal cavity and ultimately into the ectopic lesions. This translocation drives a feed-forward loop of inflammation via LPS-induced NETosis. Importantly, the core features of this pathway: neutrophil infiltration, NETosis and Pseudomonas presence, were consistent across peritoneal, ovarian and deep infiltrating endometriosis (figures 2h–n and 6a–c). This suggests we have identified a common pathogenic driver that transcends anatomical subtypes, potentially explaining shared inflammatory symptomatology and offering a unified target for intervention. This study has several limitations that chart the course for future research. While we established the translocation and pathogenicity of P. aeruginosa in our models, studies in germ-free mice colonised with defined microbiota would be the gold standard to confirm its sufficiency in initiating endometriosis and to delineate its interactions with other members of the microbial community. Furthermore, although targeting NETosis was therapeutically effective, the translational potential of modulating the gut microbiome itself remains a tantalising prospect. Future studies should explore interventions such as probiotics, prebiotics or targeted antibiotics to restore eubiosis and assess whether this strategy can prevent or reverse disease progression by shutting off the source of inflammatory triggers. In conclusion, our study delineates an integrated pathogenic mechanism in endometriosis: translocation of gut-derived Pseudomonas to the peritoneal cavity, where its LPS component activates Neu_MME neutrophils to release NETs. These NETs subsequently capture and stimulate endometrial cells, thereby driving the formation and development of lesions. Our findings highlight the therapeutic potential of targeting peritoneal Pseudomonas and inhibiting NETosis for the prevention and treatment of this debilitating condition. Data availability statement Data are available upon reasonable request. The data that support the findings are available within the article and its supplementary materials. The raw 16s rRNA microbiota data have been deposited at https://www.biosino.org/node under project ID: OEP00006185. The raw single-cell RNA sequencing data of peritoneal immune cells are available at the same repository under project ID: OEP00006180. Raw data not included therein can be obtained with the consent of the corresponding author. Ethics statements Patient consent for publication Ethics approval This study involves human participants. This study was conducted in accordance with the principles of the Declaration of Helsinki. The study protocol involving human participants was reviewed and approved by the Institutional Research Ethics Committee of the First Affiliated Hospital of Guangzhou Medical University (approval number: 2020074). Participants gave informed consent to participate in the study before taking part. All animal experiments were approved by the Animal Ethics Committee of Guangzhou Medical University (approval numbers: 2021036, 2021143, 2021191, 2021228, 2021549, 2022045, 2022062 and 20240008). Acknowledgments The authors thank Xiao-xia Zhou, Chuang-hao Lin, Ying-ying Yang and Xue Jiang for their assistance in collecting clinical samples. The authors also thank all enrolled participants for providing the clinical specimens for this study.

Footnotes XW, MW, HL, YY and HS contributed equally. RZ, LF and WW contributed equally. Contributors WW, LF and XW conceptualised and designed the project. XW, HL, HS, YY and MW were involved in concept development, detailed design, manuscript drafting, data collection, analysis, interpretation and critical revision of the manuscript for important intellectual content. RZ, WZ and LT assisted with analysing the single-cell RNA sequencing data and 16S rRNA sequencing data. SG, JC, YW, XZ, SL and BL assisted in collecting clinical data and preparing clinical specimens. WW, LF and RZ supervised the project and contributed to the writing and revision of the manuscript. All authors read and approved the final manuscript. WW is the guarantor of the work and is responsible for the overall content. Funding This work was supported by the National Natural Science Foundation of China (grant number 81971341 to WW, 92251307 and 82170542 to RZ), the Postdoctoral Fellowship Programme of CPSF (grant number GZC20230611 to XW) and the Hospital Technology Transfer Project (grant number ZH201905 to WW). The funders had no role in the study design, data collection and analysis, decision to publish or preparation of the manuscript. Competing interests None declared. Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research. Provenance and peer review Not commissioned; externally peer reviewed. Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.