{"paper_id":"b5a7dfb8-7119-42c4-b452-9d962924f464","body_text":"Article | Article in Press https://doi.org/10.1038/s41467-026-78453-5\nHigh-resolution spatial transcriptomics\nrevealsfibroblast and neuroimmune\nmicroenvironments in endometriosis lesions\nCaroline M. Haney 1, Elaheh Alizadeh1, Joshua Lee 1,2, Meryl Sullivan1,\nJasmina Kuljancic1, William F. Flynn 1,P a u lR o b s o n1,2,3,4,B r i a nS .W h i t e1,\nDanielle E. Luciano 1,5 &E l i s eT .C o u r t o i s1,2,5\nE n d o m e t r i o s i si sac hronic, systemic, inflammatory disease characterized by\nt h ep r e s e n c eo fe n d o m e t r i u m - l i k et i s s u eg r o w i n go u t s i d eo ft h eu t e r u s .\nAffecting one in ten individuals assigned female at birth, it causes chronic pain\nand inflammation that significantly impairs quality of life. Yet the mechanisms\nthat drive pain and disease progression remain poorly defined. This study\nestablishes the precise spatial transcriptomic cartography of human ovarian\nand peritoneal lesions, two of the most commonly found lesions. We identify\nshared spatial features across lesion types, including immune cell infiltration,\nfibroblast compartments surrounding epithelial glands, and distinct distribu-\ntions of neuronal and macrophage subsets. We precisely defined sensory\nneuronal subtypes and mapped their spatial location relative to immune cells.\nWe further validate the epithelial-neuronal interactome, using an in vitro 3D\nmodel of peripheral sensory brain organoids co-culturedw i t hh u m a ne n d o -\nmetriosis epithelial andfibroblast cells. By mapping spatial cellular interac-\ntions and conserved features across lesion types linked to pain, our study\nadvances understanding of endometriosis pathophysiology and informs tar-\ngeted therapeutic development.\nEndometriosis is a chronic, systemic, in flammatory, and estrogen-\ndependent condition characterized by the growth of endometrium-\nlike tissue outside of the uterus, termed lesions. These lesions are\ndiverse and heterogeneous in morphology and are most commonly\nlocated on the pelvic peritoneum, ovaries, and invading surrounding\norgans. Endometriosis affects ~ 7– 10% of individuals assigned female at\nbirth, with symptoms including heavy menstrual bleeding, chronic\npelvic pain, infertility, and urinary or gastrointestinal dysfunction\ntypically emerging during the reproductive years. However, diagnosis\nis often delayed by 7 – 10 years, allowing the disease to progress and\ncomplications to accumulate, including the development of peripheral\nand central sensitization\n1– 6. Pain is one of the main symptoms that\ndrives patients to seek treatment. Endometriosis’ sa s s o c i a t e dp a i nc a n\nmanifest through diverse ways: while some patients experience\nheightened neuroplastic sensitivity or increased activation of central\npain pathways, others present with different pain mechanisms entirely.\nIn addition, pain perception can be synchronized or not with men-\nstrual cycle phases\n6. Emerging work in clinical phenotyping highlights\nthe importance of stratifying patients based on the nature and\nmechanisms of their pain symptoms\n2,7,8. Endometriosis is a lifelong\ndisease with substantial morbidity, beginning in adolescence. Cur-\nrently, there is no cure for endometriosis, and available pharma-\ncotherapies are limited to hormonal regulation and pain management.\nSurgical intervention is currently the gold standard for pain relief.\nReceived: 25 July 2025\nAccepted: 16 September 2026\nCheck for updates\n1The Jackson Laboratory for Genomic Medicine, Farmington Connecticut, USA.2Department of Cell Biology, University of Connecticut School of Medicine,\nFarmington, CT, USA. 3Department of Genetics and Genome Sciences, University of Connecticut School of Medicine, Farmington, CT, USA.4Institute for\nSystems Genomics, University of Connecticut, Farmington, CT, USA.5Department of Obstetrics and Gynecology, University of Connecticut School of\nMedicine, Farmington, CT, USA. e-mail: elise.courtois@jax.org\nNature Communications Article in Press\n1234567890():,;\n1234567890():,;\n\nThese treatment options are associated with signi ficant side effects\nand limited efficacy and, most importantly, fail to treat the root cause\nof the disease and address the underlying pathophysiology9.\nHistologically, endometriosis lesions are defined by the presence\nof two out of the following three markers: endometrium-like glandular\nepithelial cells,fibroblasts, and/or an abundance of hemosiderin-laden\nmacrophages. Clinical pathology assessment of endometriosis lesions\nis often difficult due to their small size nature; glandular epithelial cells\ncan disappear in subsequent sections while fibroblasts and macro-\nphages are still present. This makes it important to better understand\nthe spatial environment of lesions. Recent endometriosis single-cell\nstudies have expanded our understanding of the human complex\nlesion microenvironment, revealing critical contributions from het-\nerogeneous fibroblasts subpopulations, immune cells, and blood\nvessels\n10– 12. Collectively, these studies suggest that human endome-\ntriosis lesions drive immune cell recruitment and rely on afinely tuned\nbalance of pro-inflammatory and immunosuppressive signals– medi-\nated by diverse cell types including macrophages, dendritic cells\n(DCs), and others – to establish an immune-rich niche. Our previous\nwork identified both shared and distinct cellular phenotypes between\nperitoneal and ovarian lesions, the two most common subtypes in\nhuman disease, underscoring the importance of resolving lesion het-\nerogeneity to identify conserved therapeutic targets.\nLocal tissue remodeling and inflammation are central features of\nendometriosis-driven microenvironmental changes\n10.R e c e n ts t u d i e s\nhave explored the role of neuro-immune signaling in the generation of\nendometriosis-associated pain, with a particular focus on macrophage-\nneuron crosstalk in humans 13,14 and mice 15– 21.I nm i c e ,m a c r o p h a g e s\nsecrete a range of pro-in flammatory cytokines, such as IL-1 β,T N F -α,\nIL6 and prostaglandins, that can act on nearby nociceptive neurons to\namplify pain signaling\n16. In addition, macrophages contribute to the\ngrowth of endometriosis lesions in mouse models by increasing\nangiogenesis, matrix remodeling, and tissue invasion\n15. Moreover, in\nhumans, macrophages secrete neurotrophic factors such as Brain-\nDerived Neurotrophic Factor (BDNF), Insulin Growth Factor−1 (IGF-1),\nand Nerve Growth Factor (NGF), which can stimulate neuronal\nsprouting\n22. Conversely, sensory neurons secrete neuropeptides such\nas Substance P (SP) and Calcitonin Gene-Related Peptide (CGRP),\nwhich signal back to macrophages and other immune cells to further\nenhance inflammation\n23.R e c e n tw o r ki nam o u s em o d e lf o re n d o m e -\ntriosis showed that blocking CGRP-RAMP1 signaling in macrophages\nreduced lesion size and pain symptoms\n17. Together, these findings\nsuggest a model where reciprocal neuropeptide-receptor interactions\nlead to a feedforward loop of inflammation and pain, but this needs to\nbe further validated in humans.\nHowever, translating these neuro-immune insights into human\ndisease remains a major challenge. Mouse models, while useful, fail to\nfully recapitulate human lesions and their diversity. In addition,\nimmunohistochemistry-based studies have con firmed that nerve\nfibers are present near human endometriosis lesions in a broad sen-\nsory vs. sympathetic manner, but this approach lacks the resolution to\ndefine their molecular identity or spatial organization/interactions\n11– 13.\nAs a result, the neuronal landscape and full complexity of neuro-\nimmune interactions in human endometriosis remain poorly under-\nstood. This is significant, as several studies have emphasized the dis-\nconnect between lesion burden and pain severity\n24. This indicates the\npresence of uncovered mechanisms driving pain transmission. One\nleading hypothesis is that the neuro-immune signaling driven by local\nneuro-immune interactions underlies the increased nociceptive phe-\nnotypes and sustains chronic inflammation.\nDespite the powerful insights gained from single-cell approaches,\nthese dissociative methods inherently disrupt tissue architecture,\nlimiting our ability to study sensitive cell types or spatially integrated\ncell populations such as neurons and their projections. Even single-\nnuclei RNAseq fails to capture neurons whose cell bodies reside\noutside lesion boundaries, while low-resolution spatial transcriptomics\n(ex. Visium) lack the continuous capture area to uncover the small\nneurites found in lesion areas. The absence of high-resolution spatial\ninformation precludes direct inference of cellular neighborhoods, cell\ncommunities and cell-cell interactions within lesions. Thus, integrating\nsingle-cell resolution spatial information with single-cell data is\nessential to resolve the multicellular architecture and signaling\ndynamics driving endometriosis.\nOur previous work generated a comprehensive single-cell atlas of\nendometriosis lesions, revealing coordinated immune phenotypes,\nincreased angiogenesis, and similar lesion microenvironments\n10.\nBuilding on this foundation, we now extend the analysis to subcellular\nspatial resolution— placing cell types within their spatial context and\ndefining spatial interaction networks that, critically, incorporate neu-\nrons as key components implicated in pain. This study: (1) creates a\nsubcellular and continuous map of the cell type spatial architecture of\nhuman ovarian and peritoneal endometriosis lesions. (2) Identi fies\nspatially defined cellular layers common across ovarian and peritoneal\nlesions, explained by their specific fibroblast enrichment, their relative\ndistance to epithelial glands and their neuronal enrichment. (3)\nUncovers the spatially organized ensemble of sensory neurons and\nassociated cell types— fibroblasts, immune cells, vascular components\n— likely engaged in bidirectional signaling that contributes to pain\ngeneration and sustained in flammation. Finally, (4) functionally vali-\ndates the role of epithelial andMME + fibroblasts in promoting sensory\nneurons, using in vitro 3D co-culture cellular models.\nResults\nSpatial niches in ovarian and peritoneal lesions\nWe performed whole transcriptome spatial analysis using Visium v2\n(10x Genomics) on three tissue types: endometrium biopsy (EMB),\nperitoneal superficial lesion, and ovarian lesion. The Visium assay is a\nlow-resolution spatial assay that employs 55 µm spots to capture\nprotein-coding genes. These spots are spread out across the tissue\nwith a 100 µm distance between the center of each spot, leaving sub-\nstantial tissue space uncaptured (Fig.1a). To identify epithelial glands–\none of the key components of both EMB and lesions – we performed\nnuclei segmentation on H&E-counterstained tissues using QuPath and\ntrained a learning model for gland detection. This model was inte-\ngrated with spatial expression data for the epithelial markerEPCAM to\ndefine the spatial Epi cluster (Fig. 1a ) .I na d d i t i o nt ot h ek n o w np r e -\nsence of CD10+ (encoded by MME) fibroblasts, our previous endo-\nmetriosis single-cell study led us to define differentfibroblast subtypes\nin lesions. We speci fically observed an increased osteoglycin ( OGN)\nexpressing fibroblast subpopulation in endometriosis tissues (EMB\nand lesions)\n10. Therefore, we investigated the spatial distribution of\nMME and OGN expression. Unsupervised clustering of the low-\nresolution spatial gene signatures failed to resolve the cellular com-\nplexity surrounding the glands, likely due to the heterogenous cell\ntype composition within each spot, where we captured ~ 25 cells per\nspot (Fig. 1b). We therefore manually annotated two stromal regions\nbased on spatial proximity to epithelial glands and gene expression:\nthe ‘ Close Stroma’ , an immediately adjacent region, marked by ele-\nvated MME expression, and the ‘ Far Stroma’ ; and a more distal region\nenriched for OGN expression (Fig. 1c, d). This spatial strati fication\nrevealed a clear, layered pattern of MME and OGN expression in both\novarian and peritoneal lesions (Fig.1c, d) but not in EMB tissue (Sup-\nplementary Fig. 1a– f). These findings suggest the presence of lesion-\nspecific stromal organization around epithelial structures.\nTo evaluate the molecular differences between stromal niches, we\nperformed pseudo-bulking of the measured gene expression in Close\nand Far Stroma. We then conducted differential gene expression\nanalysis (DEGs) within the peritoneal (Fig. 1e) and ovarian (Supple-\nmentary Fig. 1c) lesions. In peritoneal lesion, the Far Stroma niche was\nenriched for genes associated with WNT signaling (such asSFRP4)a n d\nArticle https://doi.org/10.1038/s41467-026-78453-5\nNature Communications Article in Press\n\nvascular processes signaling (CNN1 and THBS1) while the Close Stroma\ndisplays elevated expression of immune cell, particularly NK cell,\nrelated genes, includingGZMA, GZMB, GNLY,a n dNKG7 (Fig. 1e, f). We\nobserved enrichment of nervous system-related pathways in Far\nStroma of the peritoneal lesion (Fig. 1f). Though these findings were\ninconsistent in the ovarian lesion (Supplementary Fig. 1d), it became\nclear on further visual inspection of neuronal related genes that the\nplatform lacked the sensitivity to reliably detect fine neuronal pro-\ncesses such as small axons, as reported by Liu et al.\n12.T oo v e r c o m et h i s\nlimitation and enhance spatial precision, we transitioned to high-\nresolution spatial transcriptomics using the Visium HD (10x Genomics)\nplatform.\na\nbc\nd\nEPCAM MME OGN\nNES\nEPCAM MME OGN\nEpi Close Far\nGene Expression Level / Spot\nEpi Close Far Epi Close Far\nPeritoneal\n-1       0      1       2  \nlog/g3676 FC\n-log/g3675/g3674 p.adj\ne\nf\nFig. 1 | Low-resolution spatial transcriptomics of peritoneal endometriosis.\na Schematic of the low-resolution spatial transcriptomics (Visium V2) workflow\nperformed on peritoneal superficial lesion (n = 1), ovarian lesion (n = 1) and endo-\nmetrium (n = 1). The H&E was used for nuclei segmentation and gland identifica-\ntion, the spatial transcriptomics data was used to identify EPCAM +, MME + , and\nOGN + spots. The working model is as pictured with Close Stroma enriched in\nMME + fibroblasts directly surrounding the epithelial glands and Far Stroma enri-\nched in OGN + fibroblasts outside of the Close Stroma cluster. Created in BioR-\nender. Haney, C.63. https://BioRender.com/e64ttk0is licensed under CC BY 4.0.\nb Peritoneal lesion unsupervised clustering (sample P5).c Zoom of identified gland\narea showing spots with detectableEPCAM, MME,a n dOGN expression.d Violin plot\nof EPCAM, MME,a n dOGN expression levels in peritoneal Epithelial (Epi), Close\nStroma (Close) and Far Stroma (Far) clusters. e Volcano plot of the differentially\nexpressed genes (DEGs) between the Close Stroma (blue, upregulated above zero)\nand Far Stroma (purple, downregulated below zero). The top 15 genes for each\ngroup are highlighted. DEGs were calculated using DEseq2 which uses a Wald Test.\nThe horizontal gray dotted line represents the significance threshold of Benjamini-\nHochberg-adjustedp-value (FDR) < 0.05, and the two vertical lines represent the\nLog2 fold change (FC) <− 0.5 or > 0.5.MME (Log\n2 FC = 0.611, FDR = 0.604) andOGN\n(Log2 FC = − 2.064, FDR = 0.048) are labeled in red. f Bar plot of significant GO:\nBiological Process pathways summarized by representative pathways in the Close\nStroma [blue, positive normalized enrichment score (NES)] compared to the Far\nStroma (purple, negative NES). A complete list of significant pathways and what the\nsummarized processes represent can be found in Supplementary Data 5, 6.\nArticle https://doi.org/10.1038/s41467-026-78453-5\nNature Communications Article in Press\n\nHigh-resolution spatial mapping of lesion microenvironment\nWe performed high-resolution spatial transcriptomics using Visium\nHD on eight endometriosis lesions: four superficial peritoneal and four\novarian lesions (Fig.2a and Supplementary Fig. 2). All lesions used were\nselected based on the presence of epithelial glands surrounded by\nfibroblasts. We focused on these lesion types as they represent the two\nmost prevalent clinical subtypes – peritoneal lesions are observed\nacross all stages of disease, while ovarian lesions affect ~50% of\npatients undergoing treatment for fertility\n25.V i s i u mH De n a b l e sh i g h -\nresolution spatial mapping by capturing transcriptomic data across a\nab\nc\nd\nf\nOvarian Lesion\nPeritoneal Lesion\nOvarian Lesions\nNES\nPeritoneal Lesions\nNES\nEPCAM MME OGNH&E\nCD10 OGN Pan-CK/EpCAM\n100 μm\nSpatial Cluster\nEpi\nClose\nFar\nP13\nP12\nP11\nO9\nZ-score\nSample\nSpatial \nCluster\nSample\ng\nh\ni\nSpatial1\nSpatial 2\nSpatial1\nSpatial 2\nSpatial1\nSpatial 2\nSpatial1\nSpatial 2\nEPCAM\nMME\nOGN\nEpi\nClose\nFar\nZ-score Sample Spatial \nCluster\nj\ne\nOvarian LesionPeritoneal Lesion\nOvarian LesionPeritoneal Lesion\nASPN\nCPXM2\nKCNMB1\nJPH2\nFILIP1L\nSYNM\nAOC3\nSGCA\nMME\nMMP11\nDIO2\nCOL7A1\nWNT2\nHOXD10\nCD276\nNTN1\nCALB2\nUPK3B\nMST1\nHPSE\nPANX2\nEPCAM\nCDH1\n100 μm\nDistance from PanCK \nboundary (μm)\nZ-score across 8μm bins\nFig. 2 | High-resolution spatial transcriptomics of ovarian and peritoneal\nendometriosis. a Schematic of high-resolution spatial transcriptomics and\nantibody-based platform validating spatial clusters on three independent cohorts\nof samples. Visium HD was performed on peritoneal superficial (n = 4) and ovarian\n(n = 4) lesions. Xenium spatial transcriptomic on peritoneal superficial (n =3 )a n d\novarian (n = 1) lesions. Imaging Mass Cyometry (IMC) on peritoneal superficial\n(n =4 )a n do v a r i a n(n = 4) lesions. Created in BioRender. Courtois, E. (2026)https://\nBioRender.com/2ufvq8cis licensed under CC BY 4.0. b Zoom of identified gland\narea showing bins with > 0 expression ofEPCAM, MME,a n dOGN (n = 4 ovarian and\n4 peritoneal lesions). c (left) Representative image of ovarian lesion (sample O1a)\nand peritoneal lesion (sample P1) spatial clusters labeled as Epi, Close Stroma, Far\nStroma or Other. The other six tissues can be found in Supplementary Fig. 2. Scale\nbar 1 mm. d Heatmap of z-scoredEPCAM, MME and OGN expression per VisiumHD\nsample defining the Epi, Close and Far spatial clusters. e, f Bar plot of selected\nsignificantly enriched GO: Biological Process pathways in the Close Stroma [blue,\npositive normalized enrichment score (NES)] compared to the Far Stroma (purple,\nnegative NES) for all ovarian lesions (e) and all peritoneal lesions (f). A full list of\npathways is in Supplementary Data 9 and 10. g Representative images of Xenium\nspatial transcriptomics data from (top) ovarian lesion (sample O9) and (bottom)\nperitoneal lesion (sample P11), colored the same as in (c). Scale bars 1 mm.\nh Heatmap of Xenium spatial transcriptomic samples showing z-score expression\nof genes used to define spatial clusters based on VisiumHD differential gene ana-\nlysis (Supplementary Fig. 3 and Supplementary Data 11 and 12).OGN is not present\nin the predesigned human 5000 gene panel. i IMC protein validation of CD10\n(encoded byMME) surrounding epithelial gland (Pan-cytokeratin/EpCAM collapsed\ninto one color) as Close Stroma and OGN farther away as Far Stroma. Scale bar is\n100 µm. Ovarian lesion (sample O7) and peritoneal lesion (sample P8). Additional\nimages in Supplementary Fig. 3c (n = 4 ovarian and 4 peritoneal lesions). j Kernel-\nsmoothed z-scored profiles of CD10 and OGN antibody expression plotted against\nthe distance from the Pan-CK expressing boundary (epithelial gland), highlighting\nrelative enrichment patterns. Source data are provided as a Source Data file.\nArticle https://doi.org/10.1038/s41467-026-78453-5\nNature Communications Article in Press\n\ncontinuous tissue area using 2 × 2 µm bins, leaving no tissue unsam-\npled. We aggregated 4 × 4 squares of these 2 × 2µmb i n si n t o8×8µm\nbins. This approach yielded a dataset comprising 2.62 million 8 × 8µm\nbins, with an average of 327,000 bins per tissue section, providing an\nunprecedented subcellular spatial map of lesions (Supplemen-\ntary Data 2).\nWe identified epithelial glands using glandular feature detection\non H&E images combined with spatial EPCAM expression across the\nbins. Unlike the low-resolution spatial analysis, unsupervised cluster-\ning of the transcriptomic pro files of 8 × 8 µm bins outlined distinct\nspatial niches. For each lesion, we labeled the unsupervised clusters\nb a s e do nm R N Ae x p r e s s i o no fMME – for Close Stroma, OGN – for Far\nStroma and EPCAM – for Epi (Fig. 2b– d). The Far Stroma included\nspatial clusters enriched in fibroblast-associated markers ( PDGFRA,\nLUM, SFRP4) in addition to OGN. Clusters labeled ‘ Other’ lacked fibro-\nblast markers or represented low-quality bins. This allowed for com-\nparison across multiple samples and lesion types.\nThe Close and Far stroma gene expression pro files were subse-\nquently pseudo-bulked for DEG and pathway analysis. Pathway analysis\nrevealed distinct, cell type speci fic biological processes associated\nwith each spatial cluster, consistent across lesion types. Pathways\nrelated to lymphatic vessels in peritoneal lesions and B cells in ovarian\nlesions were significantly enriched in the Far Stroma (Fig. 2e, f). Strik-\ningly, sensory perception signaling pathways were enriched in the\nClose Stroma compartment across both ovarian and peritoneal lesions\n(Fig. 2e, f). Although classically associated with olfactory tissues, these\npathways encompass a family of ectopically expressing G-protein-\ncoupled receptors and chemosensory receptors that are known to\nregulate inflammatory and neuroimmune signaling in peripheral tis-\nsues. Their enrichment in endometriosis lesions likely reflects altered\nchemosensory and neuroin flammatory activity within the Close\nStroma microenvironment. This is in sharp contrast withfindings from\nlow-resolution spatial results (Fig.1f), underscoring the need for high-\nresolution spatial profiling to resolve neuronal signatures accurately.\nTo further validate thesefindings and increase the sample size, we\nperformed single-cell spatial transcriptomics on an additional subset\nof peritoneal lesions ( n = 3) and ovarian lesions ( n =1 ) u s i n g t h e\nXenium (10x Genomics) platform and predesigned human 5000 gene\npanel. To define the spatial clusters, we used unsupervised clustering\nof the transcripts (Points2Regions26), then evaluated the top differen-\ntially expressed genes from the VisiumHD data (Supplementary Fig. 3).\nFor the Close Stroma, this includedMME, MMP11, DIO2, COL7A1, WNT2,\nHOXD10, CD276, NTN1, CALB2, UPK3B, MST1 ,a n d HPSE.F o rt h eF a r\nStroma, this included A S P N ,C P X M 2 ,K C N M B 1 ,J P H 2 ,F I L I P 1 L ,S Y N M ,\nAOC3, SGCA,a n d CCN5 – note that OGN is not included in the prede-\nsigned Xenium 5k probe set. The results show well defined Close and\nFar clustering across all tissues, further validating our initial observa-\ntions using an independent high-de finition spatial transcriptomic\nplatform and biological replicates (Fig. 2g, h). Consistent with these\nfindings, quantitative protein expression relative to epithelial glands\nwas con firmed by imaging mass cytometry (IMC) in independent\nperitoneal lesions (n = 4) and ovarian lesions ( n =4 ) ( F i g .2ia n dS u p -\nplementary Fig. 3c). The pixel intensity analysis averaged across all\ntissues shows that CD10 expression is highest in proximity to epithelial\nglands, whereas OGN increases as CD10 signal declines when moving\nfurther away from epithelial glands (Fig.2j).\nTogether, across two spatial transcriptomic platforms and one\nprotein expression platform, all ovarian and peritoneal lesions formed\nspatial clusters surrounding epithelial glands, with MME, MMP11 and\nWNT2 highly expressed in the Close Stroma andOGN, ASPN and FILIP1L\nhighly expressed in the Far Stroma.\nStromal compartments are conserved across lesions\nTo assess the full cellular complexity of the lesion microenvironment,\nwe predicted the most likely cell type within each bin using marker\ngenes from in-house and published single-cell datasets10– 12 (Fig. 3a, b,\nMethods section and Supplementary Data 3 and 4).\nShifts in predicted cell type proportions between Epi, Close and\nFar Stromal spatial clusters showed similarities and differences\nbetween ovarian and peritoneal lesions (Supplementary Fig. 4b, linear\nmixed effect model). As a good control, epithelial cells were sig-\nnificantly enriched in Epi spatial clusters. Across both lesion types,\nvascular cells (which comprises endothelial and perivascular markers)\nare signi ficantly enriched in the Far Stroma, and there is a trend\nt o w a r d si n c r e a s e dm o n o c y t e sa n dm a c r o p h a g e si nt h eC l o s eS t r o m a .\nDifferences across lesion types were manifest in the signi ficant\nenrichment of neurons within the Close Stroma of peritoneal lesions,\nwhile neurons were evenly distributed across Close and Far Stroma in\novarian lesions. NK cells are significantly enriched in the Close Stroma\ncluster of ovarian lesions (Fig. 3a, b and Supplementary Fig. 4a, b).\nB cells have been associated with endometriosis and autoimmune\ndisorders involving B cells have also been linked to the disease\n27.Bc e l l\ncounts are increased in the peritoneal fluid of patients 28– 31,a n dw e\npreviously observed tertiary-like structure involving B cells in lesions10.\nHowever, their role remains unclear. Interestingly, we noted that B\ncells form aggregates throughout the lesion, and more abundantly in\nthe Far Stroma in both lesion subtypes, across all tissues analyzed\n(Fig. 3a– d and Supplementary Fig. 5a– f).\nTo evaluate globally the similarities across lesion types, we inte-\ngrated all tissues into one dataset and performed unsupervised hier-\narchal clustering on the pairwise similarities between all bins. The\nresults indicate similar transcriptomic signatures across lesion types,\nmostly aligning with the spatial clusters initially defined independently\nwithin each sample by the stromal compartments (Fig. 3e, black\nboxes). UMAP projection of spatial-clustered bins shows discrete\nClose, Far and Epi clusters, with a little overlap, while lesion and sample\nU M A Pp l o t si n d i c a t en os a m p l eb i a s( F i g .3f). This further confirms our\nstrategy to define spatial compartments and that the identified spatial\nfeatures apply across both ovarian and peritoneal lesions.\nNeurons and immune cells spatially organize within lesion\nNeuro-immune signaling has been hypothesized to play an important\nrole in the cell-to-cell interactions that occur in endometriosis lesions\ndue to the association between increased in flammation and pain\nsignaling4,19. Thus, we evaluated the enrichment of immune cells in the\nimmediate neuron periphery (within 16µm distance) and compared it\nto all other bins (Fig. 4a). Results indicated increased NK- and T-cell\nenrichment around neurons for the Epi cluster, while B cells appear to\nbe less present within the immediate neuron periphery in different\nspatial clusters. We also noted an enrichment in mast cells in proximity\nto neurons within both the Close and Far Stromal clusters and across\nperitoneal and ovarian lesions. This observation is consistent with\nprior studies reporting mast cells in direct apposition to nervefibers in\nlesions, where they are believed to contribute to nociceptive signaling\nthrough degranulation and release of pro-inflammatory mediators\n32,33\n(Fig. 4a). Interestingly, macrophages appear to be mostly enriched\nwithin the immediate neuron periphery in the Epi cluster. Our analysis\nis based on a grid of 8 × 8 µm bins rather than true single-cell resolu-\ntion; therefore, individual bins may contain transcripts from multiple\ncell types. To assess classi fication robustness, we examined the ‘ sec-\nond-best’ cell type assignment for bins classified as neuron, across all\ntissues. Although fibroblasts consistently emerged as the next most\nlikely identity, the confidence score for neuronal assignment was sig-\nnificantly higher than that of any alternative cell type in these bins\nacross all tissues, supporting the robustness of the neuron bin\nassignments (Supplementary Fig. 6).\nCrosstalk between macrophages and neurons is critical in endo-\nmetriosis, as we and others have previously shown based on functional\nstudies and in single-cell analysis\n10,17,19. As such, we compared the dis-\ntribution of distances between macrophages and neurons to that of\nArticle https://doi.org/10.1038/s41467-026-78453-5\nNature Communications Article in Press\n\ndistances between macrophages and non-neurons. Within spatial\nclusters, we showed a signi ficant correlation between macrophage –\nneuron distance, normalized to all other cell types, indicating that\nmacrophages and neurons position closer to each other, while not\nestablishing immediate contact. This was particularly emphasized\nwithin the Epi and Far Stroma for ovarian and peritoneal lesions,\nrespectively (Fig. 4b). These results were further con firmed through\nIMC as we saw CD68+ macrophages located close to βIIITub+ neur-\nites (Fig. 4c).\nTo further evaluate the communication between neurons and\nmacrophages, we performed ligand-receptor analysis on the neuron\nand macrophage bins across the Close Stroma and Far Stroma. Spatial\nligand– receptor analysis revealed pronounced regional differences in\nneuroimmune signaling within endometriosis lesions. In the Close\nStroma, we observed extensive bidirectional communication between\nmacrophages and neurons. Both macrophage to neuron and neuron to\nmacrophage signaling were represented by a broad set of\nligand– receptor families, including PDGF, SEMA –NRP/PLXNA, MDK-\nLRP1 and multiple COL–ITG extracellular matrix interactions\nac\ne\nSpatial Cluster\nSamples\nLesion Type\nSpatial Cluster\nSamples\nLesion Type\nSpatial Cell Type Assignment\nOvarian Lesion\nf\nPeritoneal Lesion\nUMAP 1\nUMAP 2\nUMAP 1\nUMAP 2\nUMAP 1\nUMAP 2\nInfered Cell Type Proportion Neighborhood Analysis\nLesions\nbd\nSpatial Cluster\nB cells\nDCs\nEpithelial Cells\nFibroblasts\nMast cells\nMonocytes\nNK cells\nNeurons\nOther\nT cells\nVascular cells\nB cells\nDCs\nEpithelial Cells\nFibroblasts\nMast cells\nMonocytes\nNK cells\nNeurons\nOther\nT cells\nVascular cells\nB cells\nDCs\nEpithelial Cells\nFibroblasts\nMast cells\nMonocytes\nNK cells\nNeurons\nOther\nT cells\nVascular cells\nB cells\nDCs\nEpithelial Cells\nFibroblasts\nMast cells\nMonocytes\nNK cells\nNeurons\nOther\nT cells\nVascular cells\nSamples\nMME\nUMAP 1\nUMAP 2\nOGN\nUMAP 1\nUMAP 2\nSpatial1\nSpatial 2\nSpatial1\nSpatial 2 Epi\nEpi\nClose\nClose\nFar\nFar\nOther\nOther\nProportion of Predicted Cell TypesProportion of Predicted Cell Types\nFig. 3 | Global cellular map of endometriosis lesions. a, b (left) Spatial plot of\nrepresentative ovarian (sample O2) (a) and peritoneal lesion (sample P2) (b)\nshowing bin-level cell type assignment for Epithelial Cells, Fibroblasts, Vascular\nCells, Neurons, Monocytes/Macrophages, T Cells, NK Cells, B Cells, DCs and Mast\nCells (Supplementary Data 4). Arrows indicate B-cell clusters and the scale bar\nrepresents 1 mm in (a, b). (right) Bar plot represents the average proportion for\neach cell type in the Epi, Close Stroma, Far Stroma and Other spatially defined\nclusters for ovarian (a) and peritoneal lesions (b). Proportions for each tissue\nindividually are in Supplementary Fig. 4a, and larger images are in Supplementary\nFig. 11. A linear mixed-effect model was used to compare cell type proportion\ndifferences across the stroma clusters in each lesion type; significant comparisons\nare shown in Supplementary Fig. 4b.c, d Neighborhood analysis heatmap of global\ncell types for lesions represented in panel (a)a n d(b). Per tissue analysis is found in\nSupplementary Fig. 4. e Heatmap of unsupervised clustering across all merged\nsamples comparing spatial cell cluster, individual samples, and lesion type. Unsu-\npervised clustering comparing peritoneal lesions and, separately, ovarian lesions is\nin Supplementary Fig. 4c. Black boxes represent similar transcriptomic signatures\nin the spatially defined clusters. f Uniform manifold approximation and projection\n(UMAP) of bins across all merged samples, colored by lesion type (top left), spatial\ncell cluster (top right), and samples (bottom left) and MME and OGN expression\n(bottom right).\nArticle https://doi.org/10.1038/s41467-026-78453-5\nNature Communications Article in Press\n\n(Supplementary Fig. 7c). In contrast, within the Far Stromal regions,\nthe communication network was markedly reduced (Supplementary\nFig. 7d), with weaker and scarce signals, suggesting that Far Stroma\nmaintains only a baseline level of neuroimmune activity rather than the\namplified, lesion associated signaling observed near glands. Taken\ntogether, these results demonstrate that endometriosis lesions harbor\na specialized stromal neuroimmune microenvironment characterized\nby intensified macrophage– neuron cross talk within the Close Stroma\nmicroenvironment.\nNext, we examined the spatial localization of macrophage sub-\ntypes (Fig. 4d, e), which we previously identified in our endometriosis\nsingle-cell atlas\n10. We observed a trend in increased presence of acti-\nvated macrophages within the Epi and Close Stroma clusters for both\nlesions (Fig. 4f and Supplementary Fig. 7b). Ovarian lesions show a\nab\nd f\nc\nOvarian Lesion Peritoneal Lesion\nOvarian Lesion Peritoneal Lesion\nSpatial Monocytes/Macrophages Cell Type Assignment\nT cells\nNK cells\nMonocyte/\nMacrophages\nMast cells\nDCs\nB cells\nCell Type Enrichment in Neuron periphery vs \nNon-Neuron periphery Bins (Log/g3676 FC)\nLesion Type\n      Ovarian\n      Peritoneal\nOvarian Lesion Peritoneal Lesion\nNeuron Label\n      Neuron bin\n      Neuron       periphery\n      Non-neuron       periphery\nProportion of Predicted Macrophage\n Cell Types\nNearest  Macrophage Distance from \nNeuron vs Non-Neuron (Log2 FC)\nLesion Type\n      Ovarian\n      Peritoneal\nCD10 OGN \nPan-CK/EpCAM\nBIIItub CD68\nEpi Close Stroma Far Stroma\nEpi EpiClose CloseFar FarOther Other\ng h\nOvarian Lesion Peritoneal Lesion\nVisium HDXenium\nEpi Close Far\nXenium Macrophage Transcript Counts\nEpi\nClose\nFar\nOther\nMacrophage\nEpi Close Far\nSpatial1\nSpatial 2\nSpatial1\nSpatial 2\nSpatial1\nSpatial 2\nSpatial1\nSpatial 2\ne\nFig. 4 | Spatial association of neurons to immune cells and macrophage sub-\npopulations. a Representation of bins annotated to neurons (orange) and two\nlayers of bins (16 µm) surrounding a neuron bin labeled as neuron periphery (left).\nPer sample distribution of Log2 fold changes (FC) of immune cell types in neuron\nperiphery compared to non-neuron periphery. Log2 FC > 0 indicates enrichment in\nneuron periphery <0 in non-neuron periphery.b Box plot of distance calculation\nbetween a neuron and macrophage bin compared to a neuron and all other cell\ntype bins. The box represents the span of the interquartile range (IQR), the middle\nline in the box is the median, the whiskers extend to the smallest and largest value\nof 1.5 x IQR, the points represent the outliers. Log\n2 FC is plotted, normalized to non-\nneuron bin distances to macrophages. Log2 FC < 0 (orange dotted line) have\nshorter distances between neurons and macrophages.N = 4 ovarian and 4 perito-\nneal lesions. Two-sided Wilcoxon test with Benjamini-Hochberg adjustedp-values\n(FDR) are plotted in order from left to right FDR = 1.41 × 10−2,1 . 7 9×1 0−10,6 . 1 7×1 0−5,\n1.25 × 10−10,1 . 9 5×1 0−10,<2 . 2×1 0 −16. c Antibody staining (IMC) indicates that\nneurons (βIII-tubulin + ) are found located close to macrophages (CD68 + ) in\nproximity to glands, indicated by white arrows. Peritoneal lesion (sample P9,n =4 )\nand ovarian lesion (sample O8,n = 4). Scale bar in (c) represents 100µm. d, e Spatial\nplots of monocyte/macrophage subpopulations and epithelial bins. Representative\nimage of an ovarian (sample O2) (d) and a peritoneal lesion (sample P2) (e). f Bar\nplot of average proportion of monocyte/macrophage subpopulations in ovarian\n(n = 1) and peritoneal lesions (n = 3). Per sample proportions and linear mixed-effect\nmodel heatmap comparing average proportions in each lesion type are in Sup-\nplementary Fig. 7.g Spatial scatter plots of representative ovarian (sample O9) and\nperitoneal (sample P12) lesions colored by Epi, Close and Far stroma with an overlay\nof macrophage transcripts in dark purple (CD68, CD163, MRC1, CD14,a n dFCGR3A).\nh Box plot (same as b) of Xenium macrophage transcript counts in each spatial\nregion (n = 4 samples). Comparisons between regions performed using two-sided\nWilcoxon rank-sum test with Benjamini– Hochberg false discovery rate correction (*\nFDR = 0.029). Source data are provided as a Source Data file.\nArticle https://doi.org/10.1038/s41467-026-78453-5\nNature Communications Article in Press\n\nsignificant increase in LYVE1+ and peritoneal macrophages within the\nClose Stroma cluster (Fig. 4f and Supplementary Fig. 7b). Altogether\nthese results suggest a specific pattern of neurons and macrophages\nco-localizing in the endometriosis lesion microenvironment, where\nactivated macrophages, expressingSPP1, FN1 and SPARC, are speci fi-\ncally positioned near epithelial glands. Notably,SPP1-expressing mac-\nrophages were reported to contribute to extracellular matrix (ECM)\nand tissue remodeling in endometriosis and other fibrotic diseases\n34.\nThese results were further con firmed on independent samples pro-\ncessed on an independent spatial transcriptomic platform, showing\nincreased macrophage transcripts ( C D 6 8 ,C D 1 6 3 ,M R C 1 ,C D 1 4,a n d\nFCGR3A) within the Close Stroma (Fig. 4g, h).\nWhile this spatial profiling does not provide mechanistic insights,\nit does confirm that macrophages and neurons are key partners in the\nendometriosis microenvironment. Our data further supports the\nconcept of speci fic immune cell-neuron interactions that are not\nincidental but represent a conserv ed feature across endometriosis\nlesion types.\nNociceptors sensory neurons spatially distribute close to glands\nWe next examined the spatial distribution of neuronal subtypes, by\nsubclassifying neuron bins (Fig. 3), expressing pan-neuronal markers\n(Fig. 5a), based on subtype-specific markers (Fig. 5b). We verified the\nspecificity of the selected neuronal markers by analyzing their\nexpression in a dissociative single-cell endometriosis atlas dataset,\nwhich excludes the neuronal population\n10. As expected, we found\nminimal or undetectable expression of neuronal markers in major\ncanonical cell types found in lesions (Fig. 5c). Prior studies have\nreported increase of sensory neurons near endometriosis\nlesions\n4,5,23,35– 38, prompting us to evaluate speci fic sensory neurons\nsubtypes: nociceptor (TRPV1, NTRK1) which respond to in flammation\nand transmit pain signals17, proprioceptor (RUNX3, NTRK3)i n v o l v e di n\nbody position sensing39, and mechanoreceptor (PIEZO2, NTRK2)w h i c h\nrespond to mechanical stimuli40. We also included sympathetic neu-\nrons (TH, SCL6A2), down regulators of stress response41,a n dn e u r o n a l\nbins increased in immune-signaling related genes ( IL6R, TNFRSF1A),\nlabeled neuroimmune (Fig.5a, b, d, e). 17% of the neuronal bins could\nnot be subclassi fied, though they expressed pan-neuronal mar-\nkers (Fig. 5a).\nAmong neuronal subpopulations, we identi fied a signi ficant\nenrichment of nociceptor neurons, marked by expression of pain\nreceptors such asSCN9A and TRPV1, that localize within the Epi cluster\nin both ovarian and peritoneal lesions (Fig. 5d, e and Supplementary\nFig. 8b). On average, nociceptor neurons account for 18% of the neu-\nronal bins across lesions. Ovarian lesions show a trend towards\nincreased neuroimmune neurons – which express inflammatory che-\nmokines such as TNFRSF1A and RAMP1 – within the Close Stroma.\nMechanoreceptor neurons, marked by NTRK2 expression, show a\ntrend toward enrichment in the Close and Far clusters across both\nlesion types (Fig.5d– f, Supplementary Fig. 8b). Notably, they represent\nthe predominant neuronal subset accounting for an average of 40% of\nthe neuronal bins. In contrast, s ympathetic neurons are the least\nrepresented, comprising only ~ 0.1% of the neuronal compartment.\nWhile we report average expression across lesions, individual tissues\nexhibit patient-specific heterogeneity (Supplementary Fig. 8a). Con-\nsistent with the gene expression data showing that TUBB3, the gene\nencoding βIII‑tubulin, is expressed at very low levels in all other cell\ntypes (Fig. 5c), representative IHC images demonstrate minimal colo-\ncalization between βIII‑tubulin and CD10 (Fig. 5f). These findings\nconfirm that βIII‑tubulin reliably marks neuronal cells. Applying this\nmarker to an independent IMC sample cohort, we observed increased\nneuronal βIII-tubulin expression the Close Stroma of ovarian and\nperitoneal lesions (Fig.5g). In addition, nociceptor transcripts (SCN9A,\nTRPV1, TRPA1, NTRK1, CALCA ,a n d TAC1)a r ei n c r e a s e di nt h eC l o s e\nStroma in the Xenium cohort of samples (Fig. 5h, i).\nBeyond nociceptive neurons, other sensory neurons, such as\nproprioceptors and mechanoreceptors, remain poorly characterized\nin endometriosis. Our data establishes a comprehensive overview of\nneuronal diversity within endometriosis lesions. Collectively, these\nresults also reveal distinct spatial localization of neuronal subtypes,\nwith sensory neurons – particularly nociceptors – enriched in the\nepithelial compartment.\nIn vitro validation of neuron-lesion dynamic interactions\nGiven the preferential localization of the sensory neurons near the\nepithelial glands and adjacent to the Close Stroma compartment, we\nsought to validate the dynamic interaction between epithelial, stroma\ncells – particularly MME expressing fibroblasts – and neurons. We\ngenerated 3D in vitro models recapitulating each of these components\nusing: (1) peripheral sensory brain organoids (PSBOs) derived from\ntwo female iPSC lines; (2) primary epithelial organoids derived from\nEMB and peritoneal and ovarian lesions; and 3) CD10+ fibroblasts\nderived from EMB and lesions, co-cultured in Matrigel domes (see\nMethods) (Fig. 6a). PSBOs were differentiated for 86 days. Differ-\nentiation was validated by observing a decrease in proliferation mar-\nker, Ki67, and the increase in βIII-tubulin and TRKA protein detection\nover the course of day 35 to day 95 (Fig.6b). It takes considerable time\nfor neurons to reach maturity in in vitro cultures, as others have\nreported\n42,43. We evaluated the gene expression of mature neuronal\nmarkers (MAP2, PENK,a n d RBFOX3) and nociceptor markers ( CALCA,\nCALCB, NTRK1, NTRK2, NTRK3 ,a n d TAC1) in day 95 brain organoids\n(Supplementary Fig. 9a). All markers showed clear expression in both\nPSBO iPSC lines, supporting the emergence of mature neuronal and\nnociceptor‑like identities.\nCo-cultures were maintained for 10 days, and RNA was extracted\nfor bulk RNA-sequencing to evaluate the effect of epithelial and stro-\nmal endometriosis cells on neuronal signaling pathways (Supplemen-\ntary Movie 1). Canonical makers for epithelial organoids ( KRT8,\nEPCAM), fibroblasts ( PDGFRA, MME ) and mature neurons ( MAP2,\nRBFOX3) were used to validate the experimental design and support\nthat these co-cultures mimic the major cell types recapitulated in Epi/\nClose Stroma environments (Fig. 6c and Supplementary Fig. 9a). We\nnext performed DEG analysis between the different conditions. PSBOs\ncultured alone differentiated towards TH + sympathetic and PRPH +\nsensory neurons with a distribution of 55: 45 percent. However, when\nthe PSBOs were co-cultured with epithelial organoids (PSBO+Org) or\nwith both organoids and fibroblasts (PSBO + Org + Fib), this propor-\ntion shifted to 40% for TH+s y m p a t h e t i ca n d6 0 %PRPH+s e n s o r y\nneurons (ratio in Fig. 6d, normalized expression Supplementary\nFig. 9b). Fibroblasts alone were not able to drive this neuronal shift in\nPSBOs. These data suggest that the presence of e pithelial organoids,\nendometrial and lesion-derived, drive neurons toward a sensory\nneuron fate.\nWe performed pathway analysis between ovarian and peritoneal\nco-cultures for the PSBO + Org + Fib conditions. We extracted the\npathways involved in neuronal signaling. Genes associated with sen-\nsory neuron and neuroin flammation signaling pathways were\nincreased in both lesion types, which further validates our spatial DEG\nanalysis (Fig. 6e). We also noted that motor neuron and central ner-\nvous system signaling-related genes were downregulated in both\nlesion type-derived co-cultures. These data suggest that epithelial\norganoids in combination with MME+fibroblasts push neurons toward\na sensory fate and increase the neuroinflammation response signaling\npathway alongside pain receptor expression, validating spatial tran-\nscriptomic findings.\nFinally, we next compared the PSBO neuronal gene signatures to\ngene sets used to de fine the neuronal subpopulations in the spatial\ntranscriptomics (Fig.6e). Inflammatory genes IL6R and TNFRSF1A are\nboth increased in the presence of organoids, while RAMP1 and CSF1R\nseem to be driven by the presence of fibroblasts (Supplementary\nArticle https://doi.org/10.1038/s41467-026-78453-5\nNature Communications Article in Press\n\nFig. 9c). We also observe increased gene expression of inflammation-\nrelated genes, such as NINJ1, IL1B, TNF, IL18 in PSBO+Org+Fib co-\ncultures (Supplementary Fig. 9d). Mechanoreceptor genes, such as\nPVALB and NTRK2, are downregulated in the Org+Fib conditions for all\nof the tissue-derived cells. Expression of the nociceptor geneCALCA is\nspecifically upregulated in the peritoneal and ovarian lesion-derived\norganoid and fibroblast co-cultures compared to EMB co-cultures.\nHowever, some background signal is still present in the epithelial\norganoid– only and fibroblast‑only conditions, suggesting additional\nsources of baseline activity. Future in depth work is needed to\ndelineate the speci fic signaling pathways that drive the neuroin-\nflammatory crosstalk between cell types.\nab\neOvarian Lesion Peritoneal Lesiond\nPan-Neuronal Markers\nOvarian Lesion Peritoneal Lesion\nf\nOvarianPeritoneal\nProportion of Predicted Neuron\n Types\nEpi EpiClose CloseFar FarOther Other\nSpatial Neuronal Cell Type Assignment\nCD10Merge βIIITub\nTRKAMerge TRKB\nc\ngh\nNociceptor \nTranscipt Counts\nEpi Close FarOvarian Peritoneal \nSpatial1\nSpatial 2\nSpatial1\nSpatial 2\ni TRPV1\nCALCA\nSCN9A\nNTRK1\nTRPA1\nEPCAM\nFig. 5 | Spatial distribution of neuronal subtypes. a Dot plot of pan-neuronal\ngenes used to identify neuron bins.b Dot plot of the neuron only bins showing gene\nmarkers used to define the subtypes of neurons.c D o tp l o to fp r e v i o u s l yp u b l i s h e d\nsingle-cell dataset10 from endometriosis lesion tissue showing specificity of neuron\nsubtype genes to neurons. d Spatial plots of representative ovarian (sample O2;\nleft) and peritoneal lesions (sample P2; right) showing the distribution of neuronal\nbins across the tissues, and the dot size was increased to visualize distribution\npatterns. Epithelial cells are plotted to localize the lesion area. Scale bar in (d)\nrepresents 1 mm.e Bar plot of average neuronal subtypes in ovarian and peritoneal\nlesions across the spatial clusters.f IHC images of a representative peritoneal lesion\n(sample P10) showing CD10+fibroblasts andβIII-tubulin+ neurons surrounding the\nepithelial gland (top row) and TRKA + and TRKB + neurons close to the epithelial\ngland (bottom row). Scale bar represents 100µm. g Boxplots show the distribution\nof neuron fraction (βIII-tubulin+ cells in region/ total βIII-tubulin+ cells in tissue)\nacross spatial regions (Epi, Close, and Far) within ovarian (n = 4) and peritoneal\nlesions (n = 4) from Imaging mass cytometry (IMC). Each point represents an\nindependent tissue sample, overlaid as jittered dots. Boxes indicate the median and\ninterquartile range, and whiskers extend to 1.5 × the interquartile range. Statistical\ndifferences between regions were assessed separately within each lesion type using\ntwo-sided Wilcoxon rank-sum tests with Benjamini– Hochberg false discovery rate\n(FDR) correction across comparisons (Close vs Epi, Close vs Far, and Epi vs Far) (*\nFDR = 0.00119).h Box-and-whisker plots with overlaid jittered points show noci-\nceptor transcript counts in Epi, Close, and Far stromal regions across independent\ntissues (n = 4) from Xenium. Boxes denote the median and interquartile range, and\nwhiskers extend to 1.5 × IQR; individual points indicate biological replicates.\nComparisons between regions were performed using a two-sided Wilcoxon rank-\nsum test with Benjamini– Hochberg false discovery rate (FDR) correction, resulting\nin no significant comparisons. i Representative image of Xenium nociceptor tran-\nscripts in a peritoneal lesion (sample P11). Source data are provided as a Source\nData file.\nArticle https://doi.org/10.1038/s41467-026-78453-5\nNature Communications Article in Press\n\nFollicle AMH levels correlate with lesion epithelial distance\nThe link between infertility and endometriosis remains unclear, parti-\ncularly how the presence of endometriosis lesions impacts the quality\nof ovarian follicles. We unexpectedly observed the presence of follicles\nclose to the endometriosis lesion in one of the ovarian samples\n( s a m p l eO 3 ) .W ei d e n t ified follicles based on H&E staining and the\nexpression of speci fic follicular markers ( ZP3, GDF9,OOSP2 and\nFIGLA)\n44(Fig.7a). Anti-Mullerian Hormone (AMH), which is specifically\nexpressed by small, growing follicles, is widely used as a marker for\novarian reserve and ovarian stimulation capacity\n45. Patients with higher\nAMH levels have been associated with better fertility outcomes46.A l l\nidentified follicles were located in the Far Stroma (Fig. 7b), with a\nvariable range ofAMH levels per follicle (Fig.7c). We noted a significant\npositive correlation between a follicle’ s AMH gene expression level and\nits distance from the epithelial gland (Fig.7d), suggesting that follicles\nmay be impacted by the lesion. While these data are exploratory and a\nsimilar analysis needs to be replicated in a larger cohort, they do\ndemonstrate the remarkable capacity of continuous high-resolution\nspatial transcriptomics to infer potential mechanisms associated with\nendometriosis infertility.\nab\ne\nCell type\nTissue type\nEPCAM\nKRT8\nPDGFRA\nMME\nc\nNeuronal Pathway comparison in Lesion Org+Fib\nCell type Tissue type\nd\nβIIITub/ KI67 /DAPI TRKA / DAPI\nDay 35Day 56Day 95\nProliferation\nMaturation\nLog2FC(PRPH/TH) \nPSBO PSBO\n+Fib\nPSBO\n+Org\nPSBO\n+Org +Fib\n1.0\n0.5\n0.0\n - 0.5\nTissue type\nPeritoneal\nOvarian\nFig. 6 | In vitro cell model of neuronal-epithelial-fibroblasts interactions in\nendometriosis. aSchematic of experimental design. Primary cell lines for epithelial\norganoids (Org) and fibroblasts (Fib) were derived from endometriosis patient\nendometrium, peritoneal or ovarian lesion tissues. Peripheral sensory brain orga-\nnoids (PSBO) were derived from a female iPSC-line following the growth factor\nprotocol outlined in the Methods section and Supplementary Fig. 10. Cells were co-\ncultured for 10 days in different combinations for each tissue type: PSBO, PSBO +\norg, PSBO +fib, PSBO + org +fib. Bulk RNA-sequencing was performed on all the co-\nculture conditions (n = 2). The resulting transcriptomics signatures were compared\nfor DEGs and pathway analysis. Supplementary Movie 1 of co-cultures. Created in\nBioRender. Haney, C. (2026) https://BioRender.com/as0z5ynis licensed under CC\nBY 4.0. b IHC staining was performed on day 35, 56 and 95 PSBOs, confirming the\nsuccessful differentiation of iPSC to mature neurons (βIII-tubulin), sensory noci-\nceptor (TRKA + ) and a decrease in proliferation (KI67). Scale bar represents 25µm.\nc Heatmap showing the expression of epithelial markers (EPCAM, KRT8)a n d\nfibroblast markers (PDGFRA, MME) in analyzed co-culture conditions.d Bar plot of\nthe proprotion of PRPH+ sensory neurons over TH + sympathetic neurons in all co-\nculture conditions, normalized to the PSBO alone condition.n =3P S B O ,2\nPSBO + fib, 2 PSBO + org, 5-8 PSBO + org + fib per tissue type. Error bars repre-\nsent +/− SE. e Dot plot of neuronal pathways significantly upregulated (NES > 0) or\ndown regulated (NES < 0) in the ovarian and peritoneal co-culture PSBO + org +fib\nconditions. NES is calculated by normalizing raw enrichment scores against the\nmean of permutaiton-based enrichement scores for each gene set.\nArticle https://doi.org/10.1038/s41467-026-78453-5\nNature Communications Article in Press\n\nDiscussion\nThis study presents a high-resolution spatial transcriptomic profiling\nof human ovarian and peritoneal endometriosis lesions. Previously, we\nhad established a comprehensive single-cell atlas, based on dis-\nsociative approaches, of similar lesions. Yet this atlas lacked the spatial\ninformation needed to understand dynamic interactions and ecosys-\ntems driven by the spatial distributions of cells. In addition, due to\ntechnical limitations of dissociative approaches, neurons, a known\ncomponent of lesions and a key driver of endometriosis symptoms,\nwere not detected. Now we have established a comprehensive spatial\natlas of endometriosis lesions, at sub-cellular resolution, and provide\nan unbiased survey of the neuronal landscape within these structures.\nAlthough endometriosis is a high ly heterogeneous condition, we\nemployed spatially defined regions to identify shared pathways across\ntwo of the most prominent lesion types (Fig. 8). This approach\nemphasizes the importance of understanding the disease’ sc o m p l e x i t y\nand spatial patterning to identify common features and mechanisms\namong different lesion types. While spatial transcriptomics provides a\npowerful approach for inferring ecosystems that play a critical func-\ntion in lesion pathobiology, it provides only a snapshot view of the\ndynamic cellular interactions underlying a complex microenviron-\nment. Thus, we developed a 3D co-culture model, using patient-\nderived primary lines, which recapitulates some of the critical cell\ntypes and their functional interactions to begin to understand\ndynamics between neuron-epithelial partners.\nFrom our spatial transcriptomic analysis, we defin e ds h a r e ds p a -\ntial niches across lesion types, labeled Epi, Close and Far Stroma, based\non mutual distance from glandular structures and shared expression\nsignatures. Surprisingly, we found compelling similarities between\nspatial clusters of super ficial peritoneal and ovarian lesions, two\nmorphologically distinct lesions, with variable clinical stages and\ntreatments. Our data support findings of a recent endometrioma\nspatial analysis from Liu and colleagues, where spatial niches were also\nobserved, with fibroblasts proximal to epithelial glands implicated in\nwound healing, whereas distal fibroblasts were de fined as pro-\ninflammatory\n12. The authors also determined that low-resolution spa-\ntial transcriptomics, while informative, does not provide suf ficient\nclarity for comprehensive analysis of immune cells and neuronal\norganization. Indeed, we also demonstrated that low-resolution spatial\ntranscriptomics did not reliably depict the positioning of immune and\nneuronal cells and only provided a discontinuous map of lesions.\nOur spatial transcriptomic analysis showed the presence of B cell\naggregates in the Far cluster consistently across all ovarian and peri-\ntoneal lesions. Previous reports have described an increased presence\nab\nc d\nOvarian Lesion AMH Gene Expression\n in Follicules\nFig. 7 | Ovarian follicleAMH level is correlated to distance from epithelial gland.\na Spatial plot of ovarian lesion (sample O3) with Epi, Close Stroma, Far Stroma and\nFollicles labeled.b AMH expression in Follicle bins.c Boxplot of average AMH levels\nper follicle in the Far Stroma. We did not identify follicles in the Close Stroma. Each\npoint represents a follicle. Box indicates the median and interquartile range, and\nwhiskers extend to 1.5 × the interquartile range.d Linear model comparing the\nfollicle distance from the centroid of the epithelial gland to the averageAMH level\nof the follicle. Blue line is the bestfit linear regression model (y =− 0.3 + 5.8 × 10\n−5x).\nShaded gray is 95% confidence interval. The two-sided Pearson’ s correlation was\ncalculated to be R = 0.53 and the p = 0.00099. n = 1 ovarian lesion, 35 follicles.\nArticle https://doi.org/10.1038/s41467-026-78453-5\nNature Communications Article in Press\n\nof peripheral and peritonealfluid B cells in endometriosis patients, and\nshowed that plasma cells produce autoantibodies against the endo-\nmetrium and autoantibodies commonly found in autoimmune\ndiseases\n29,47. However, most of the B-cell biology in female reproduc-\ntive pathologies remain unknown 48. Our work further suggests the\ncontribution of B cell aggregates to the Far Stroma niche likely\ninvolved in promoting a chronic inflammatory environment. Parallels\nbetween endometriosis and autoimmune diseases, or immune-related\ndisorders\n27, have been suggested in previous studies; however, our\nstudy further explores this perspective and identifies these aggregates\nas a common immunological mark across lesion subtypes, targeting of\nwhich could represent a promising therapeutic approach.\nAn unresolved question in the field is whether neurogenesis\nresults from chronic in flammation or is driven by the presence of\nlesion-specific epithelial cells and fibroblasts in human tissues. With\nour data, we show an increase in nociceptor neurons expressing pain\nreceptors close to epithelial glands across both peritoneal and ovarian\nlesions. Our in vitro 3D cellular model utilizing sensory brain organoids\nprovides functional evidence of epithelial cells andfibroblasts in pro-\nmoting sensory pain neurons. Further spatial analysis supports the\nincreased presence of activated macrophages in areas where noci-\nceptor neurons and epithelial glands colocalize. These data suggest a\npotential synergistic effect whereby epithelial cells influence neurons\nto express pain receptors. In turn, these neurons likely increase the\nproduction of inflammatory molecules (such asNINJ1), promoting the\nrecruitment and activation of immune cells, particularly macrophages,\nand contributing to the sustained in flammatory microenvironment.\nIndeed, NINJ1 was shown to mediate tissue remodeling and in flam-\nmation by regulating leukocyte entry, adhesion, activation, and\nmigration\n49, and represents a potential candidate factor involved in\nendometriosis neuroimmune crosstalk. Further functional and\nmechanistic studies with complex 3D systems, recapitulating all these\ncentral cellular components, will help elucidate the role of macro-\nphages or mast cells in this process.\nEndometriosis is an estrogen-dependent disease, and steroid\nhormone cyclic variation influences the immune landscape in repro-\nductive female organs and related-diseases. Technical and cost lim-\nitations of high-resolution spatial transcriptomics hinder the profiling\nof large clinical datasets. In this study, we chose to focus on lesions\nfrom patients under hormonal contraception (HC) to minimize the\nimpact of the menstrual cycle variation on our dataset. HC is also a\nstandard treatment for endometriosis patients not seeking pregnancy.\nInterestingly, this study also suggests that HC may not be effective\nagainst the neurogenic and pain-modulating processes driven by the\npresence of epithelial glands and close stroma. Additional studies have\nshown that estrogens play an important role in the mast and\nmacrophage-neuron communication in endometriosis\n33,50. Therefore,\nfurther research is necessary to evaluate the impact of HC on lesions in\nthe context of neuroimmune signaling and to investigate how this may\nbe altered in non-hormonally treated patients, and across menstrual\ncycle phases.\nFinally, one limitation of spatial transcriptomics is that it encom-\npasses only a single spatial modality. Future integration of spatial\nproteomics, ECM organization and other modalities such as lipidomics\nwill provide additional insights, further enhancing our understanding\nof the complex microenvironment associated with endometriosis\nlesions. Spatial transcriptomics infers co-localization of cells but can-\nnot prove direct cell-cell contact. Further in vitro cellular modeling is\nneeded to truly address cell-cell contact and functional impact. Our\ninitial spatial profiling focused on the two most common subtypes --\nsuperficial peritoneal and ovarian lesions. However, considering the\nsignificant heterogeneity of endometriosis, it is essential to investigate\nall lesion subtypes, including deep in filtrating endometriosis and\nextra-pelvic lesions, in order to identify shared pathways that could\ncontribute to the development of comprehensive treatment strategies\nfor all lesion variants.\nBy integrating whole transcriptome data with tissue morphology,\nspatial transcriptomics has revolutionized our understanding of dis-\neases such as cancer and irritable bowel disease\n51,52. Here, we fill a\nFig. 8 | Representation model of common cellular features across lesions.\nSchematic of proposed model for spatial cluster cell type distributions in ovarian\nand peritoneal lesions, highlighting common features across lesions. Epi Stroma\ncluster showed increased expression in Nociceptor neurons. The Close Stroma\ncluster showed increased expression of CD10+fibroblasts, Nociceptor neurons,\nactivated macrophages in both lesion types and NK cells in ovarian lesions only\n(indicated with *). The Far Stroma cluster showed increased vascular structures, B\ncells and lymphatic vessels in both lesion types. Globally, sympathetic neurons\nwere sparse, and there was a large increase in sensory neurons in the lesion\nmicroenvironment. Lastly, the data showed mast cells are enriched close to neu-\nrons in both the Close and Far Stroma clusters. Created in BioRender. Courtois, E.\n(2026) https://BioRender.com/husfc4nis licensed under CC BY 4.0.\nArticle https://doi.org/10.1038/s41467-026-78453-5\nNature Communications Article in Press\n\ncritical gap in endometriosis research by revealing precise spatial\ncellular interactions within lesions and establishing parallels between\ntwo of the most common endometriosis lesions. We found that ovarian\nand peritoneal lesions share spatial features. Our findings highlight a\nprominent role for immune cells, uncover fibroblast compartments\nsurrounding glands, and define the spatial distribution of neuronal and\nmacrophage subsets. Understanding the cellular and molecular drivers\nof pain— typically among the earliest symptoms— may offer a path to\npreventing or mitigating the disease, particularly its most debilitating\nmanifestations, and informing the development of therapeutic\nstrategies.\nMethods\nHuman endometriosis tissue collection\nThis study complies with all relevant ethical regulations and was\napproved by The Jackson Laboratory and the University of Connecticut\nHealth Center Institutional Review Board (IRB). Patients ’ enrollment\nwas conducted under either The Jackson Laboratory IRB 19-145J-2 and\nthe Endometriosis Research Innovation Support & Education Bior-\nepository (EndoRISE) CT biorepository IRB 24-111JSF-1, with all parti-\ncipants providing written informed consent, including agreement to\nshare demographic information. Individuals assigned female at birth,\nages 18 and above, were consented and enrolled in the CT EndoRISE\nbiorepository for longitudinal endometriosis research, agreeing to\nshare biospecimens (tissues, whole blood, urine, peritonealfluid) and\nphenotypic data. Informed consent included secondary usage of tis-\nsues for endometriosis research, such as organoids and cell line gen-\neration, and subsequent research applications. Study participants\ndonating tissue samples for this study did not receive compensation.\nTissues and phenotypic data (including quality of life, clinical and\nmedical history data) were collected by endometriosis specialists at\nUCHC following the World Endometriosis Research Foundation\n(WERF) Endometriosis Phenome and Biobanking Harmonization Pro-\nject (EPHect) guidelines and standard operating procedures. Speci-\nmens were immediately transported on ice to the biorepository at The\nJackson Laboratory for processing and long-term storage (within 1 h of\ncollection). The EndoRISE Sample Access Review Committee (com-\nprised of staff from UCHC and The Jackson Laboratory) reviewed the\nrequest for the data and biospecimen distribution for this project and\napproved the request unanimously. Patient information and use of\nspecimens in this project can be found in Supplementary Data 1.\nTissues were collected at the University of Connecticut Health\nCenter and transported to The Jackson Laboratory on ice in MACS\ntissue storage solution (Miltenyi, Cat:130-100-008). Tissues werefixed\nin 10% Formalin in Phosphate Buffered Saline (PBS) for 24 h at 4\noC.\nFollowing fixation, tissues were washed three times in PBS and stored\nin 70% Ethanol at 4 oC until sent to embedded in paraf fin blocks.\nFormalin-fixed Paraffin-embedded (FFPE) tissue blocks were sectioned\nat 5 µm and stained with Hematoxylin and Eosin (H&E) to visualize\nhistology or Picrosirius Red (PS) to identify areas of dense collagen\nstaining in lesions. For tissues where epithelial glands surrounded by\nfibroblasts were not obvious with the histological staining, we per-\nformed immunohistochemistry using anti-CD10 (R&D Systems,\nCat:AF1182) to identifyfibroblasts and anti-Pan-cytokeratin (Biolegend,\nCat:914204) to identify epithelial glands as previously described\n10 (see\nbelow IMC, IHC and confocal imaging section).\nTissue processing for spatial transcriptomics\nOnce a lesion area was identi fied with epithelial glands and CD10+\nfibroblasts, a 6.5 × 6.5 mm region of interest (ROI) was selected for\nspatial transcriptomics. RNA quality control was performed on the\nselected FFPE blocks with the Qiagen RNAeasy FFPE kit (Cat: 72504),\nfollowed by Agilent tapestation analysis of RNA. Only FFPE blocks with\na DV200 above 30% were used for all spatial transcriptomics\nplatforms.\nFFPE blocks that passed all quality control metrics were sectioned\nat 5 µm and put on glass slides. The sectioned tissues were either\nprocessed following the 10X Visium V2 protocol, Visium HD V1 pro-\ntocol, or Xenium Human 5 K panel by the Single Cell Biology Core at\nThe Jackson Laboratory. Visium V2 was run on three samples, eutopic\nendometrium (EMB), peritoneal lesion and ovarian lesion, from two\npatients. Visium HD was run on eight samples, four peritoneal lesions\nand four ovarian lesions, from seven patients. Supplementary Data 2\nincludes tissue information and spatial metrics. Xenium was run on\nthree peritoneal lesions and one ovarian lesion.\nSpatial transcriptomics pre-processing and cluster\nidentification\nThe Illumina base call libraries were converted to FASTQfiles. The 10x\nGenomics Space Ranger pipeline (v3.0.0) was used to align the\nsequencing reads to the human reference genome GRCh38-2020-A.\nFor tissues processed with the Visium V2, we kept the spots that\nexpress more than 100 genes and have less than 10% of mitochondrial\ngene counts. QuPath was used to segment the nuclei in the H&E image.\nThe transcriptional data was used to manually label the spots con-\ntaining EPCAM + epithelial glands, MME + closefibroblasts and OGN +\nfar fibroblasts.\nFor tissues processed with Visium HD V1, count matrices of\naggregated 8 × 8µm bins resulting from SpaceRanger (10x Genomics)\nwere processed with custom R and Python scripts. Bins with low UMI\ncounts were excluded using sample-specific UMI cutoffs. Each cutoff\nwas established by first de fining a tissue mask of contiguous bins\nintersecting at least one nucleus. The UMI cutoff was then taken as that\nUMI count that best separated bins within the tissue mask from those\noutside the tissue mask, as quantified by F1 score [i.e., 2 x precision x\nrecall / (precision + recall)]. Nuclei were counted in Python by applying\nthe ‘ 2D_versatile_he’ model of stardist to an image normalized to the\n5th and 95th percentiles. Nuclei overlapping each bin were counted by\nperforming spatial joins between the nuclei coordinates and bin\ncoordinates using geopandas in Python. The tissue mask was then\ndefined by applying morphological smoothing operations to bins\nintersecting at least one nucleus using OpenCV in Python. Specifically,\nislands on the periphery of the initial mask of non-empty bins were\nconnected using morphological closing (via morphologyEx and a\nsquare kernel, each of whose sides were 80 µms — i.e., ten times the\nsize of an aggregated bin). Small, isolated groups of bins were then\nremoved via morphological opening (via the same function and ker-\nnel). This resulted in a refined tissue mask. For each tissue individually,\n8 µm bins were removed outside of the refined tissue mask, and 8 µm\nbins were removed with a UMI count less than the threshold that\nseparates the foreground and the background of the re fined tissue\nmask. The H&E was used to identify the epithelial glands, but unsu-\npervised clustering was used to identify thefibroblast compartments.\nThe clusters closest to the glands and highest inMME expression were\nlabeled as ‘ Close Stroma’ . The clusters outside of the MME+ areas and\nhigher in OGN were labeled as ‘ Far Stroma’ . The left-over clusters that\ndid not fitt h i sd efinition were labeled as ‘ Other’ .\nFor tissues processed with Xenium 5 K human panel, each tissue\nwas run through the Points2Region pipeline\n26,53 using a subsample of\n750k transcripts. This created a spatial unsupervised clustering of the\ntranscripts was projected back to the entire tissue. The clusters were\nlabeled as Epi, Close or Far based on the expression of the top DEGs\nfrom the VisiumHD data (Fig. 2h and Supplementary Fig. 3).\nVisium HD sample integration and clustering\nDue to large sample size (Supplementary Data 2) sample integration\nwas performed on a representative set of cells from each sample. This\nsampling approach is based on a geometric sketching method, and it\nensures that rare populations are also included in the subset\n54.T h e\nintegration of eight samples was done with Seurat using the Harmony\nArticle https://doi.org/10.1038/s41467-026-78453-5\nNature Communications Article in Press\n\nmethod on Principal Component space. Then, the integration was\nexpanded to the full dataset.\nAn aggregated pseudo-bulk-based workflow was used, where each\ncell is an aggregated sample of one cluster in each sample. Then\nclustering was done using the complete method and Euclidean dis-\ntance (Fig. 3e). After cell aggregation, cell type- speci fic differential\nexpression was performed using the DESeq2 method and the Wilcoxon\nrank sum test. Pathway analysis for the biological process gene\nontology terms (GO:BP) was run.\nVisium HD cell type assignment and distance analysis\nTo predict the major cell type within each bin, Seurat ’ s AddModule-\nScore function was used with known markers for the following cell\ntypes: Epithelial Cells, Fibroblast Cells, Vascular Cells, Neurons, Mac-\nrophages, NK cells, T Cells, B Cells, DCs and Mast Cells (Supplementary\nData 3). The resulting scores were scaled and filtered for a z-score\nabove 1 to be labeled as a cell type. The spatial location of all the cell\ntypes was plotted using SpatialDimPlot, and increased the point size to\nvisualize the patterns. The proportion of each cell type per sample in\neach defined fibroblast compartment (Epi, Close or Far) was calculated\nby taking the number of bins for that cell type in the fibroblast com-\npartment divided by the total number of bins in that fibroblast com-\npartment and graphed it in a proportion plot. To assess which cell type\nproportions were signi ficantly different between the stroma com-\npartments across lesion types, post-hoc analysis was performed of\nlinear mixed-effect models fit independently for each cell type and\nlesion type, with compartment as the fix e de f f e c ta n ds a m p l ea st h e\nrandom effect (Supplementary Fig. 3b)\nTo identify subpopulations of neurons, all the bins labeled as\nneurons from the previous global cell typing were subsetted and run\nthrough AddModuleScore with another set of genes speci fict ot h e\nsubpopulations (Supplementary Data 3). Again, the scores were scaled,\nbut a z-score above 0.01 was used to identify the subpopulations of\nneurons, as these genes are often low in abundance. The same process\nwas performed to identify macrophage subpopulations using the gene\nmarkers from a published single-cell dataset of endometriosis tissue\n10\n(Supplementary Data 3). The same linear mixed-effect model was used\nto calculate the differences between the subtypes of neurons and\nmacrophages in the fibroblast compartments (Supplementary Fig. 5b\nand Fig. 5g).\nTo understand the spatial connectivity of cells, Squidpy neighbors\nenrichment analysis was performed on each VisiumHD sample indivi-\ndually across the global cell types. This resulted in a heatmap for each\nsample showing which cell types are spatially clustered close to each\nother (Fig. 3c, d and Supplementary Fig. 5a– f).\nEnrichment analysis between neurons and immune cells\nTo check whether immune cells are enriched around neuron bins, the\nproportion of each immune cell type in the bins directly surrounding\nthe neuron bin, de fined by two layers of 8 ×8 µmb i n s( n e u r o np e r -\niphery in Fig. 4a), was calculated. The proportion of each immune cell\ntype around all the non-neuron bins was calculated (non-neuron per-\niphery in Fig. 4a). These values were normalized to show their fold\nchange.\nTo evaluate the enrichment of macrophages in the vicinity of\nneurons within each compartment (Epi, Close Stroma, and Far\nStroma), the distance was calculated between each neuron bin in that\ncompartment and its closest macrophage, noting that the neuron and\nmacrophage bins may not fall in the same compartment. Similarly, the\ndistance between each non-neuron, non-macrophage bin in each\ncompartment was calculated relative to its closest macrophage bin.\nThe adjusted p-value was graphed from the result of a Wilcoxon test\ncorrected using the Benjamini-Hochberg procedure. Within each\nsample, all distances were log normalized to the median of the non-\nneuron/non-macrophage to macrophage distances within that sample\n(Fig. 4b). As such, a log-normalized distance less than zero indicates\nthat a macrophage is closer to a neuron than the median distance\nbetween a macrophage and a non-neuron, non-macrophage.\nLigand-receptor analysis\nWe inferred spatially resolved ligand – receptor– mediated cell – cell\ncommunication using COMMOT55, a framework that reconstructs sig-\nnaling interactions in spatial transcriptomics data through a collective\noptimal transport model. This method incorporates spatial distances\nbetween cells and competition among ligand and receptor species,\nenabling biologically realistic estimation of communication strength\nand directionality.\nSpatial transcriptomics data were preprocessed following stan-\ndard COMMOT guidelines, including log‑transformation and normal-\nization of total transcript counts. Ligand– receptor pairs were obtained\nfrom curated databases of CellChat\n56 and CellPhoneDB57 available\nthrough COMMOT, and were formatted into a structured data frame\nfor downstream analysis. COMMOT then computed pairwise ligand-\nreceptor communication matrices using spatial constraints (distance\nthreshold of 50µm) and support for heteromeric receptor complexes.\nDownstream analyses included quantifying sender and receiver\nactivities for interactions between macrophages and neurons within\nClose Stroma and Far Stroma.\nIMC, IHC and confocal imaging\nWe used previously published Imaging Mass Cytometry (IMC) data\nreanalyzed for markers specific to this project (Fig. 2 & 4)8.I m m u n o -\nhistochemistry (IHC) was adapted from a previously described\nprotocol\n8 for the localization of TRKA and TRKB neurons (Fig. 5).\nBriefly, FFPE tissue sections were melted for 10 minutes at 55 oC,\ndeparaffinized in 2 washes of Histoclear (National Diagnostics, Cat: HS-\n200). The tissues were rehydrated in a gradient of ethanol to water\nwashes. Antigen retrieval was performed in EDTA pH 9 (Vector Labs,\nCat: H-3301-250) for 15 min at 95\noCin a pressure humidity chamber\n(BioSB TintoRetriever). Sections were permeabilized in 0.1% triton for\n10 min, then blocked in 10% donkey serum and 3% BSA solution for\n1 hour at room temperature. Primary antibodies were incubated in 3%\nBSA in PBS overnight at 4\noC. Supplementary Table 1 includes primary\nantibody information. Unbound primary antibody was washed away,\nthen sections were incubated in secondary antibodies for 2 h at room\ntemperature. Secondary antibodies were used at 1:1000 dilution,\nDonkey Anti-Goat IgG-Alexa fluor 647 (Invitrogen, Cat: A32849) and\nDonkey anti-Mouse IgG- Alexa Fluor 555 (Jackson Immunological, Cat:\nAB_3095480). A Dapi counter stain was used. Slides were coverslipped\nwith Flouramount-G (Life Tech, Cat:00-4958-02) (Supplementary\nTable 1).\nIHC images were acquired using a Leica TCS SP8 laser scanning\nconfocal microscope (Leica Microsystems, Wetzlar, Germany) equip-\nped with a DM6000 upright microscope stand. The system was con-\ntrolled by Leica Application Suite X (LAS X) software (version\n3.5.7.23225). Images were obtained using the HC PL APO CS2 40x/1.30\nOil objective (Cat #: 11506358) and HC PL APO CS2 63x/1.30 Glyc\nobjective (Cat #: 11506353). Leica Microsystems’ Type F immersion oil\n(Cat #: 11513859) was used with the 40x objective, while Leica Micro-\nsystems’ Type G Immersion Liquid (Cat #: 11513910) was used with the\n63x objective. Excitation was provided by a combination of a 405 nm\ndiode laser and a White Light Laser (WLL). Laser power was controlled\nvia the software and was typically set between 1 – 10% to minimize\nphotobleaching. Fluorescence emission was detected by a combina-\ntion of photomultiplier tubes (PMTs) and high-sensitivity hybrid\ndetectors (HyDs). The spectral detection bands were tuned to opti-\nmally capture the emission of each fluorophore while minimizing\ncrosstalk. The spectral detection bands were tuned as follows:\n• DAPI: Excitation at 405 nm, emission collected at 410– 494 nm.\nArticle https://doi.org/10.1038/s41467-026-78453-5\nNature Communications Article in Press\n\n• Alexa Fluor 488: Excitation at 499 nm, emission collected at\n509– 572 nm.\n• Alexa Fluor 568: Excitation at 577 nm, emission collected at\n587– 648 nm.\n• Alexa Fluor 647: Excitation at 653 nm, emission collected at\n658– 779 nm.\nImages were acquired as 8-bit files with a frame size of 512 × 512\npixels. A conventional galvanometric scanner was used. For imaging, a\nscan speed of 400 Hz was used with a line averaging of 1. The pinhole\nwas set to 1 Airy unit (AU) to ensure confocality. For Z-stacks, images\nwere acquired using system-optimized Nyquist values for the step size,\nresulting in a step size of 0.333 µm for the 63x objective (total axial\ndistance of 7.66 µm) and 0.346 µm for the 40 x objective (total axial\ndistance of 13.85 µm). Sequential scanning was employed for multi-\nchannel imaging to prevent bleed-through between channels. The scan\nsequence was set to “between frames”. Raw image files (.lif) were\nprocessed and analyzed using LAS X software and Fiji/ImageJ. The\ncaptured signals were max projected.\nIMC quantification\nPer‑cell marker intensities and centroid coordinates were extracted\nfrom an AnnData object and processed independently for each ima-\nging field‑of‑view (FOV). Raw marker values (PanCK, CD10, OGN) were\narcsinh‑transformed (cofactor = 5) and robustly scaled to [0,1] using\nthe 1st– 99th percentiles per FOV. PanCK‑positive cells were defined as\nthose at or above the 90th percentile of normalized PanCK within each\nFOV. Distances from every cell to the nearest PanCK+ cell in the same\nFOV were computed using a cKDTree nearest ‑neighbor search, con-\nverted to microns, and constrained to outside ‑only values by setting\nnegative distances to zero. To quantify spatial gradients, normalized\nmarker intensities were binned by distance (8 µm bins), averaged per\nFOV, and then aggregated across images, followed by light Gaussian\nsmoothing ( σ =1 . 5 b i n s ) . A k e r n e l‑smoothed variant was also com-\nputed by evaluating Gaussian ‑weighted means on a 0 – 200 µmg r i d\n(8 µms p a c i n g ;σ = 1.5 × bin width). The pipeline yielded per‑cell anno-\ntations (*_norm, PanCK + mask, distance values) and cohort‑level dis-\ntance‑dependent profiles for CD10 and OGN.\nAMH-follicle distance analysis\nWe received one lesion that came from an ovarian cortex and con-\ntained follicles near an endometriosis lesion (Fig.7). The Seurat func-\ntion SpatialFeaturePlot was used to plot the AMH level for the follicle\ncluster on the tissue. K-nearest-neighbor was used to spatially localize\nthe follicles into the closest stroma cluster. To analyze the quality of\nthose follicles, the centroid of the epithelial gland and each follicle\nwere calculated. DBscan was used to detect the bins for each follicle,\nresulting in 35 follicles and the average AMH for each follicle was cal-\nculated. The Euclidean distance from the epithelial gland centroid to\nthe follicle centroid was compared to the mean AMH level for each\nfollicle using a linear regression model.\nEndometrial epithelial organoid andfibroblast culture\nTissues were dissociated into single-cell suspension as previously\ndescribed10. The single-cell suspension was resuspended in cold\nMatrigel (Corning, Cat:356231) in 50 µL domes in a 24-well plate. Epi-\nthelial organoid media, previously described58, was added to cover the\ndomes. For thefirst passage, to separate the epithelial organoids from\nthe fibroblasts, the fibroblasts were allowed to grow to the bottom of\nthe dish, then the Matrigel dome was carefully removed to leave the\nfibroblasts in place. The fibroblasts were washed with DPBS, and\nfibroblast media was added. Fibroblast media contained DMEM/F12\n(Gibco, Cat:11320082) supplemented with 10% dextran-coated char-\ncoal stripped FBS (Sigma, Cat:C6241, Gibco, Cat:A5256801), 1% peni-\ncillin-streptomycin, 2 mM L-glutamine, 1 nM β-Estrodiol (Sigma,\nCat:E4389), 2 ug/mL insulin (Sigma, Cat:91077 C). Epithelial organoids\nwere passaged every 7 days according to previously described\nprotocol\n58. Fibroblasts were passaged 1:4 every 7 – 10 days once 80%\nconfluent with a five-minute digestion of TrypLE (Gibco, Cat:1260510)\nto lift the cells from the plate.\nPeripheral sensory brain organoid culture\nHuman induced pluripotent stem cells (hiPSCs) were cultured and\ninduced to brain organoids, adapted from the following protocols43,59.\nAll material and reagents can be found in the protocols.io (dx.doi.org/\n10.17504/protocols.io.kxygxw1ddv8j/v1). The hiPSCs were derived\nfrom two healthy females of African and Puerto Rican ancestry at the\nJackson Laboratory (GeDiX697.3 J and GeDiX653.1 J). This hiPSC lines\nwere fully validated for trilineage differentiation potential and whole\ngenome sequencing. Brie fly, to induce neuronal progenitor cells in\norganoid formation, the dual-SMAD method was adapted as pre-\nviously described\n60 on a shaking platform at 95 rpm inside a 37 oC\nincubator. Briefly, cells were cultured in 100% KSR medium supple-\nmented with 10 μMS B 4 3 1 5 4 2a n d1μML D N 1 9 3 1 8 9f o rd a y s4– 8p o s t -\nseeding. To promote the differentiation to sensory neurons, a protocol\nwas adapted\n61,2 . 5μMS U 5 4 0 2 ,1μMD A P Ta n d1 . 5μM CHIR99021 was\nadded on days 5– 13 of differentiation. The cells were transitioned from\n100% KSR media to 100% Neural Induction Media (NIM) over a 7 day\ngradient from day 8– 14. Once in 100% NIM on day 14, the supplements\nwere switched to 10 ng/mL FGF2 and 10 ng/mL EGF2 for 6 days,\nwashing every other day. The cells were finally put into Terminal Dif-\nferentiation (TD) media supplemented with 20 ng/mL BDNF, 20 ng/mL\nGDNF, 25 ng/mL NGF, 10 ng/mL NT-3, 200 µMA s c o r b i cA c i da n d\n0.5 mM dbcAMP on day 20 post feeding. The cells were washed every\n3 days with fresh media and supplements (Supplementary Fig. 10).\nTo verify the neural induction to peripheral sensory organoids,\nthe brain organoids were were fixed in 4% PFA for 4 hours at r4\noC,\nwashed with PBS 3 times for 5 min, then put in 30% sucrose for 48 h at\n4\noC. Once the organoids sunk in the sucrose solution, they were\nembedded in OCT and sectioned at 14 µm. The sections were then\nstained for anti-β-III-tubulin (Promega, Cat: G7121), anti-KI67 (Invitro-\ngen, Cat:14-5698-82), TRKA (R&D Systems, Cat:AF175) to detect the\ndecrease in proliferation and increase in maturation. The immune\nfluorescence protocol was as follows: cells were washed 3 times with\nPBS, permeabilized with 0.25% Triton-X100 in PBS for 15 minutes,\nwashed 3 time with PBS, blocked with 10% donkey serum, 1% BSA in\nPBS for 1 h at room temperature, and incubated in primary antibodies\nat 1:100 dilution in 0.1% Triton-X100, 1% BSA in PBS overnight at 4oC.\nOn the second day, cells were washed 3 times for 5 min in PBS, incu-\nbated in secondary antibodies 1:1000 for 2 h room temperature,\nwashed 3 times for 5 min with PBS, incubated with DAPI for 5 min,\nwashed 3 times for 5 min with PBS. PSBO IHCs were imaged on Leica\nTCS SP8 laser scanning confocal microscope using a 63x objective as\ndescribed above. Images were optimized for brightness and contrast\nacross the entire field of view with FIJI.\nCo-culture and bulk RNA-seq of brain organoids, epithelial\norganoids and fibroblasts\nTo assess the molecular changes that occur when endometriosis epi-\nthelial cells and fibroblasts are in contact with sensory neurons, co-\nculture experiments were performed. Eutopic endometrial, peritoneal\nlesion and ovarian lesion epithelial organoids and fibroblasts were\ncultured in different combinations (denoted in Fig. 6a, patient line\ninformation in Supplementary Data 1) with 86 day sensory peripheral\nbrain organoids. There were 6 conditions: PSBO alone (n =4 ) ,e p i t h e l i a l\norganoids alone ( n =4 ) , fibroblasts alone ( n =3 ) , P S B O+fib( n =2 ) ,\nPSBO + epi (n =2 ) ,P S B O+fib+e p i( n = 20, 4 EMB, 8 ovarian, 8 perito-\nneal). After 10 days of co-culturing, the Matrigel domes were melted on\nice for 30 min with Cell Recovery Solution (Corning, Cat: 354253), the\ncells were spun down 520 g for 5 min at 4\noC, cells were washed with\nArticle https://doi.org/10.1038/s41467-026-78453-5\nNature Communications Article in Press\n\nPBS once and spun down again. The cells went right into RNA extrac-\ntion following the Qiagen RNeasy Mini Kit protocol (Cat: 74106). RNA\nquality was assessed with Agilent Tapestation, and only samples with a\nRIN above 6 were used for bulk RNA sequencing. Samples were\nprepped for whole-genome library prep. Then sequenced on Novaseq\nXP l u s1 0 Bw i t h3 0 0c y c l e sw i t ha na v e r a g eo f2 . 9m i l l i o nr e a d sp e r\nsample.\nAnalysis of bulk RNA-seq\nReads were aligned with the human genome GRCh38-2020-A,filtered\nwith nf-core/rnaseq (v.3.19) utilizing the STAR aligner 62,a n dc o u n t s\nwere normalized to counts per million. DeSeq2 was used to subtract\nthe brain organoid alone conditions from all the rest of the conditions\nto get DEGs. The VST scaled matrix was used to assess gene expression\ndifferences between groups. Then, the different co-culture conditions\nwere compared using the FGSEA pathway analysis. The neuronal-\nrelated pathways were selected to compare between groups.\nStatistics and reproducibility\nAll statistical tests were run as described in the above methods sec-\ntions. Adjustedp-values or FDR were used when possible. No statistical\nmethod was used to predetermine sample size due to limited avail-\nability to precious human samples. No data were excluded from the\nanalyses. The experiments were not randomized. The Investigators\nwere not blinded to allocation during experiments and outcome\nassessment.\nReporting summary\nFurther information on research design is available in the Nature\nPortfolio Reporting Summary linked to this article.\nData availability\nThe raw FASTQ data for Visium, VisiumHD, and bulk-RNA sequencing\nis deposited on the dbGAP public data repository (Accession #\nphs004891.v1.p1 [ http://www.ncbi.nlm.nih.gov/projects/gap/cgi-bin/\nstudy.cgi?study_id=phs004891.v1.p1]). Access requests are submitted\nthrough dbGAP’ ss t a n d a r dD a t aA c c e s sR e q u e s tp r o c e s sa n dr e v i e w e d\nby the relevant Data Access Committee, with a typical response time of\n2– 4 weeks. Xenium data and Visium Space Ranger outputs have been\ndeposited on EGA (EGAS50000002140). Access requests are sub-\nmitted through EGA and reviewed by the Data Access Committee\n(EGAC50000001139). Requests will be required to complete a Data\nAccess Agreement through JAX and the requesters institution. Source\ndata are provided for Xenium and IMC data quanti fications in the\nSource Data file.\nRestrictions on use for all data include: (1) research must be for\nthe sole purpose of advancing endometriosis research, diagnostic\ntools, therapeutic development and advancing patient care. (2) users\nare prohibited from posting or making available data through open-\naccess public websites and/or databases (controlled access databases\nare acceptable). (3) Users shall appropriately acknowledge and refer-\nence the EndoRISE Biorepository in all publications, presentations, and\nother scholarly or scienti fic communications that include data or\nfindings derived, in whole or in part, from Materials obtained through\nthe EndoRISE Biorepository. Source data are provided in this paper.\nCode availability\nAll code developed and used for this study is available on GitHub:\nhttps://github.com/TheJacksonLaboratory/endometriosis-visium-hd.\n(https://doi.org/10.5281/zenodo.21826389)\n63.\nReferences\n1. Chapron, C., Marcellin, L., Borghese, B. & Santulli, P. Rethinking\nmechanisms, diagnosis and management of endometriosis.Nat.\nRev. Endocrinol.15,6 6 6–682 https://doi.org/10.1038/s41574-019-\n0245-z (2019).\n2. Missmer, S. A. et al. Impact of Endometriosis on Life-Course\nPotential: A Narrative Review.Int. J. Gen. Med. 14,9 –25 https://doi.\norg/10.2147/IJGM.S261139(2021).\n3. Harder, C., Velho, R. V., Brandes, I., Sehouli, J. & Mechsner, S.\nAssessing the true prevalence of endometriosis: A narrative review\nof literature data.Int. J. Gynecol. Obstet.167, 883–900 https://doi.\norg/10.1002/ijgo.15756(2024).\n4 . G o d i n ,S .K . ,W a g n e r ,J . ,H u a n g ,P .&B r e e ,D .T h er o l eo fp e r i p h e r a l\nnerve signaling in endometriosis.FASEB BioAdvances3,8 0 2–813\nhttps://doi.org/10.1096/fba.2021-00063(2021).\n5. Morotti, M., Vincent, K., Brawn, J., Zondervan, K. T. & Becker, C. M.\nPeripheral changes in endometriosis-associated pain.Hum.\nReprod. Update 20,7 1 7–736 https://doi.org/10.1093/humupd/\ndmu021 (2014).\n6. Velho, R. V., Sehouli, J. & Mechsner, S. Mechanisms of peripheral\nsensitization in endometriosis patients with peritoneal lesions and\nacyclical pain. Arch. Gynecol. Obstet.308 1327–1340, https://doi.\norg/10.1007/s00404-023-07110-9(2023).\n7 . A s - S a n i e ,S .e ta l .E n d o m e t r i o s i s :Ar e v i e w .Jama 334,6 4–78 https://\ndoi.org/10.1001/jama.2025.2975(2025).\n8. Sasamoto, N. et al. Trends in pelvic pain symptoms over 2 years of\nfollow-up among adolescents and young adults with and without\nendometriosis.Pain 164,6 1 3–624 https://doi.org/10.1097/j.pain.\n0000000000002747(2023).\n9. Ortega-Gutiérrez, M., Muñoz-Gamez, A. & Girón-Prieto, M. S. Pri-\nmary care approach to endometriosis: diagnostic challenges and\nmanagement strategies—a narrative review. J. Clin. Med. 14,\n4757 (2025).\n1 0 . T a n ,Y .e ta l .S i n g l e - c e l la n a l y s is of endometriosis reveals a coor-\ndinated transcriptional programme driving immunotolerance and\nangiogenesis across eutopic and ectopic tissues.Nat. Cell Biol. 24,\n1306–1318 https://doi.org/10.1038/s41556-022-00961-5(2022).\n11. Fonseca, M. A. S. et al. Single-cell transcriptomic analysis of\nendometriosis.Nat. Genet. 55,2 5 5–267 https://doi.org/10.1038/\ns41588-022-01254-1(2023).\n12. Liu, S. et al. Single-cell and spatial transcriptomic profiling revealed\nniche interactions sustaining growth of endometriotic lesions.Cell\nGenomics 5,1 0 0 7 3 7https://doi.org/10.1016/j.xgen.2024.\n100737 (2025).\n13. Clower, L., Fleshman, T., Geldenhuys, W. J. & Santanam, N. Tar-\ngeting oxidative stress involved in endometriosis and its pain.Bio-\nmolecules 12, https://doi.org/10.3390/biom12081055 (2022).\n14. Garcia Garcia, J. M. et al. Endometriosis: cellular and molecular\nmechanisms leading tofibrosis.Reprod. Sci.30,1 4 5 3–1461 https://\ndoi.org/10.1007/s43032-022-01083-x(2023).\n15. Bacci, M. et al. Macrophages are alternatively activated in patients\nwith endometriosis and requiredfor growth and vascularization of\nlesions in a mouse model of disease. Am. J. Pathol. 175,5 4 7–556\nhttps://doi.org/10.2353/ajpath.2009.081011(2009).\n16. Capobianco, A. & Rovere-Querini, P. Endometriosis, a disease of the\nmacrophage.Front Immunol. 4,9 https://doi.org/10.3389/fimmu.\n2013.00009(2013).\n1 7 . F a t t o r i ,V .e ta l .N o c i c e p t o r-to-macrophage communication\nthrough CGRP/RAMP1 signaling drives endometriosis-associated\np a i na n dl e s i o ng r o w t hi nm i c e .Sci. Transl. Med 16,e a d k 8 2 3 0\nhttps://doi.org/10.1126/scitranslmed.adk8230(2024).\n18. Forster, R. et al. Macrophage-derived insulin-like growth factor-1 is a\nkey neurotrophic and nerve-sensitizing factor in pain associated\nwith endometriosis.Faseb J. 33, 11210–11222 https://doi.org/10.\n1096/fj.201900797R(2019).\n19. Greaves, E. et al. Estradiol is a critical mediator of macrophage-\nnerve cross talk in peritoneal endometriosis.Am. J. Pathol. 185,\n2286–2297 https://doi.org/10.1016/j.ajpath.2015.04.012(2015).\nArticle https://doi.org/10.1038/s41467-026-78453-5\nNature Communications Article in Press\n\n20. Scheerer, C., Frangini, S., Chi antera, V. & Mechsner, S. Reduced\nSympathetic Innervation in Endometriosis is Associated to Sema-\nphorin 3C and 3F Expression.Mol. Neurobiol.54,5 1 3 1–5141 https://\ndoi.org/10.1007/s12035-016-0058-1(2017).\n21. Zhu, H., Wang, Y., He, Y. & Yu, W. In flammation-mediated macro-\nphage polarization induces TRPV1/TRPA1 heteromers in endome-\ntriosis. Am. J. Transl. Res. 14,3 0 6 6–3078 (2022).\n22. Barcena, M. L. et al. In fluence of Nerve Growth Factor in\nEndometriosis-Associated Symptoms.Reprod. Sci. 18,1 2 0 2–1210\nhttps://doi.org/10.1177/1933719111410711(2011).\n23. Yan, D., Liu, X. & Guo, S. W. Neuropeptides Substance P and Cal-\ncitonin Gene Related Peptide Accelerate the Development and\nFibrogenesis of Endometriosis.Sci. Rep.9, 2698 https://doi.org/10.\n1038/s41598-019-39170-w(2019).\n24. Vercellini, P. et al. Association between endometriosis stage, lesion\ntype, patient characteristics andseverity of pelvic pain symptoms: a\nmultivariate analysis of over 1000 patients.Hum. Reprod. 22,\n266–271 https://doi.org/10.1093/humrep/del339(2006).\n25. Imperiale, L., Nisolle, M., Noël, J. C. & Fastrez, M. Three types of\nendometriosis: pathogenesis, diagnosis and treatment. State of the\nArt. J. Clin. Med. 12, https://doi.org/10.3390/jcm12030994 (2023).\n26. Andersson, A. et al. Points2Region s: Fast, interactive clustering of\nimaging-based spatial transcriptomics data.Cytom. A105,6 7 7–687\nhttps://doi.org/10.1002/cyto.a.24884(2024).\n27. Rahmioglu, N. et al. The genetic basis of endometriosis and\ncomorbidity with other pain and inflammatory conditions.Nat.\nGenet. 55,4 2 3–436 https://doi.org/10.1038/s41588-023-01323-\nz (2023).\n28. Chishima, F. et al. Peritoneal and p eripheral B-1-cell populations in\npatients with endometriosis.J. Obstet. Gynaecol. Res. 26,1 4 1–149\nhttps://doi.org/10.1111/j.1447-0756.2000.tb01298.x(2000).\n29. Riccio, L. G. C., Baracat, E. C., Chapron, C., Batteux, F. & Abrão, M. S.\nThe role of the B lymphocytes in endometriosis: A systematic\nreview. J. Reprod. Immunol.123,2 9–34 https://doi.org/10.1016/j.jri.\n2017.09.001\n(2017).\n30. Riccio, L. G. C. et al. B lymphocyt es inactivation by Ibrutinib limits\nendometriosis progression in mice.Hum. Reprod. 34, 1225–1234\nhttps://doi.org/10.1093/humrep/dez071(2019).\n31. Gagné, D. et al. Blood leukocyte s ubsets are modulated in patients\nwith endometriosis.Fertil. Steril.80,4 3–53 https://doi.org/10.1016/\ns0015-0282(03)00552-1(2003).\n32. Anaf, V. et al. Pain, mast cells, and nerves in peritoneal, ovarian, and\ndeep infiltrating endometriosis.Fertil. Steril.86,1 3 3 6–1343 https://\ndoi.org/10.1016/j.fertnstert.2006.03.057(2006).\n33. McCallion, A. et al. Estrogen mediates in flammatory role of mast\ncells in endometriosis pathophysiology.Front. Immunol.13, 961599\nhttps://doi.org/10.3389/fimmu.2022.961599(2022).\n34. Henlon, Y. et al. Single-cell analysis identi fies distinct macrophage\nphenotypes associated with prodisease and proresolving functions\nin the endometriotic niche.Proc. Natl. Acad. Sci. USA 121,\ne2405474121https://doi.org/10.1073/pnas.2405474121(2024).\n35. Bohonyi, N. et al. Local upregulation of transient receptor potential\nankyrin 1 and transient receptor potential vanilloid 1 ion channels in\nrectosigmoid deep infiltrating endometriosis.Mol. Pain. 13,\n1744806917705564https://doi.org/10.1177/\n1744806917705564(2017).\n36. Tokushige, N., Markham, R., Russell, P. & Fraser, I. S. Nerve fibres in\nperitoneal endometriosis.Hum. Reprod.21,3 0 0 1–3007 https://doi.\norg/10.1093/humrep/del260(2006).\n37. Tokushige, N. et al. Nerve fibers in ovarian endometriomas.Fertil.\nSteril. 94,1 9 4 4–1947 https://doi.org/10.1016/j.fertnstert.2009.12.\n074 (2010).\n38. Yan, D., Liu, X. & Guo, S.-W. Nerve fibers and endometriotic lesions:\npartners in crime in inflicting pains in women with endometriosis.\nEur. J. Obstet. Gynecol. Reprod. Biol.209,1 4–\n24 https://doi.org/10.\n1016/j.ejogrb.2016.06.017(2017).\n39. Moon, K. M. et al. Proprioception, the regulator of motor function.\nBMB Rep. 54,3 9 3–402 https://doi.org/10.5483/BMBRep.2021.54.\n8.052 (2021).\n40. Morotti, M., Vincent, K. & Becker, C. M. Mechanisms of pain in\nendometriosis.E u r .J .O b s t e t .G y n e c o l .R e p r o d .B i o l .209,8 –13\nhttps://doi.org/10.1016/j.ejogrb.2016.07.497(2017).\n41. Arnold, J. et al. Imbalance between sympathetic and sensory\ninnervation in peritoneal endometriosis.Brain, Behav., Immun.26,\n132–141 https://doi.org/10.1016/j.bbi.2011.08.004(2012).\n42. Lu, T. et al. Decoding transcrip tional identity in developing human\nsensory neurons and organoid modeling.Cell 187,7 3 7 4–7393\nhttps://doi.org/10.1016/j.cell.2024.10.023(2024).\n43. Rigamonti, A. et al. Large-Scale Production of Mature Neurons from\nHuman Pluripotent Stem Cells in a Three-Dimensional Suspension\nCulture System. Stem Cell Rep. 6,9 9 3–1008 https://doi.org/10.\n1016/j.stemcr.2016.05.010(2016).\n44. Wagner, M. et al. Single-cell analysis of human ovarian cortex\nidentifies distinct cell populations but no oogonial stem cells.Nat.\nCommun. 11, 1147 https://doi.org/10.1038/s41467-020-14936-\n3 (2020).\n45. Moolhuijsen, L. M. E. & Visser, J . A. Anti-Müllerian Hormone and\nOvarian Reserve: Update on Assessing Ovarian Function.J. Clin.\nEndocrinol. Metab.105, 3361–3373 https://doi.org/10.1210/clinem/\ndgaa513 (2020).\n4 6 . T r e m e l l e n ,K .P . ,K o l o ,M . ,G i l m o r e ,A .&L e k a m g e ,D .N .A n t i -\nmullerian hormone as a marker of ovarian reserve.A u s t .N .Z .J .\nObstet. Gynaecol.45,2 0–24 https://doi.org/10.1111/j.1479-828X.\n2005.00332.x(2005).\n47. Osuga, Y. et al. Lymphocytes in Endometriosis. Am. J. Reprod.\nImmunol.65,1 –10 https://doi.org/10.1111/j.1600-0897.2010.00887.\nx (2011).\n48. Shen, M. et al. The role of endometrial B cells in normal endome-\ntrium and benign female reproductive pathologies: a systematic\nreview. Hum. Reprod. Open 2022,h o a b 0 4 3https://doi.org/10.\n1093/hropen/hoab043(2022).\n49. Lee, H.-J., Ahn, B. J., Shin, M. W., Choi, J.-H. & Kim, K.-W. Ninjurin1: a\npotential adhesion molecule and its role in inflammation and tissue\nremodeling.Mol. Cells 29, 223–228 https://doi.org/10.1007/\ns10059-010-0043-x(2010).\n50. Greaves, E. et al. Estrogen rece ptor (ER) agonists differentially\nregulate neuroangiogenesis in peritoneal endometriosis via the\nrepellent factor SLIT3.Endocrinology155,4 0 1 5–4026 https://doi.\norg/10.1210/en.2014-1086(2014).\n51. Garrido-Trigo, A. et al. Macrophage and neutrophil heterogeneity at\nsingle-cell spatial resolution in human inflammatory bowel disease.\nNat. Commun. 14,4 5 0 6https://doi.org/10.1038/s41467-023-\n40156-6 (2023).\n52. Khaliq, A. M. et al. Spatial transcriptomic analysis of primary and\nmetastatic pancreatic cancers highlights tumor microenviron-\nmental heterogeneity.Nat. Genet 56,2 4 5 5–2465 https://doi.org/\n10.1038/s41588-024-01914-4(2024).\n53. Marco Salas, S. et al. Optimizing Xenium In Situ data utility by\nquality assessment and best-practice analysis workflows. Nat.\nMethods 22,8 1 3–823 https://doi.org/10.1038/s41592-025-02617-\n2 (2025).\n5 4 . H i e ,B . ,C h o ,H . ,D e M e o ,B . ,B r y s o n ,B .&B e r g e r ,B .G e o m e t r i c\nSketching Compactly Summarizes the Single-Cell Transcriptomic\nLandscape.Cell Syst. 8,4 8 3–493 https://doi.org/10.1016/j.cels.\n2019.05.003(2019).\n55. Cang, Z. et al. Screening cell –cell communication in spatial tran-\nscriptomics via collective optimal transport.Nat. Methods 20,\n218–228 https://doi.org/10.1038/s41592-022-01728-4(2023).\nArticle https://doi.org/10.1038/s41467-026-78453-5\nNature Communications Article in Press\n\n56. Jin, S. et al. Inference and analysis of cell-cell communication using\nCellChat. Nat. Commun. 12,1 0 8 8https://doi.org/10.1038/s41467-\n021-21246-9(2021).\n57. Efremova, M., Vento-Tormo, M., Teichmann, S. A. & Vento-Tormo, R.\nCellPhoneDB: inferring cell–cell communication from combined\nexpression of multi-subunit ligand–receptor complexes.Nat. Pro-\ntoc. 15,1 4 8 4–1506 https://doi.org/10.1038/s41596-020-0292-\nx (2020).\n58. Boretto, M. et al. Patient-derived organoids from endometrial dis-\nease capture clinical heterogeneity and are amenable to drug\nscreening.Nat. Cell Biol. 21,1 0 4 1–1051 https://doi.org/10.1038/\ns41556-019-0360-z(2019).\n59. Bozal, S. B. et al. Development of an automated 3D high content cell\nscreening platform for organoid phenotyping.SLAS Discov. 29,\nhttps://doi.org/10.1016/j.slasd.2024.100182 (2024).\n60. Chambers, S. M. et al. Combined small-molecule inhibition accel-\nerates developmental timing and converts human pluripotent stem\ncells into nociceptors.Nat. Biotechnol.30,7 1 5–720 https://doi.org/\n10.1038/nbt.2249(2012).\n61. Hiranuma, M., Okuda, Y., Fujii, Y., Richard, J.-P. & Watanabe, T.\nCharacterization of human iPSC-derived sensory neurons and their\nfunctional assessment using multi electrode array.Sci. Rep. 14,\n6011 https://doi.org/10.1038/s41598-024-55602-8(2024).\n62. Ewels, P. A. et al. The nf-core framework for community-curated\nbioinformatics pipelines.Nat. Biotechnol.38,2 7 6–278 https://doi.\norg/10.1038/s41587-020-0439-x(2020).\n63. Haney, C., Alizadeh, E., White, B.& Courtois, E. High-Resolution\nSpatial Transcriptomics Reveals Fibroblast and Neuroimmune\nMicroenvironments in Endometriosis Lesions Code.Zenodohttps://\ndoi.org/10.5281/zenodo.21826389 (2026).\nAcknowledgements\nThe authors would first like to thank all of the study participants that\ngenerously donated their tissues tohelp us better understand the biol-\nogy of endometriosis. Without their participation, none of this would be\npossible. We heartfully thank the EndoRISE Biorepository for enabling\npart of this research through access to biospecimens, and in particular\nKayceety Mullaj and Makayla Murphy for their support. Next, we would\nlike to thank all of The Jackson Laboratory core facilities that supported\nthis work and in particular: Genome Technologies, Histology, Micro-\nscopy and Cyberinfrastructure High-performance computing resources.\nOf particular importance, we gratefully acknowledge the advanced\nexpertise of the Single Cell Biology Lab and their continuous support.\nThese shared services are supported in part by the JAX Cancer Center\n(P30 CA034196). We thank the OB/Gyn department and UConn Health\nSurgery Center personnel for their assistance in consenting patients and\ncollecting specimens. A special thank you to the Robson and Skarnes\nLabs at the Jackson Laboratory for the use of iPSC cells, especially\nJuliana Alcoforado Diniz, Enricca Pellegrino and Cole Loring for their\nguidance and Adrian Melo Carrilo for his help with the bulk-RNA-seq\nanalysis.\nThis study makes use of data and/or resources generated by The\nJackson Laboratory’s Genetic Diversity (GeDi) hiPSC initiative and made\navailable through internal data access and sharing mechanisms. Data\nfrom the following GeDi samples were used: GeDiX697.3 J and\nGeDiX653.1 J.\nAuthor contributions\nE.T.C conceived and designed the study. C.M.H. and E.T.C performed\nand supervised experiments. D.E.L collected consent from participants\nand clinical samples. C.M.H., J.L., J.K, and E.T.C. processed and\nbiobanked samples for downstream experiments. J.K. provided EndoR-\nISE biospecimen and de-identified patient information. C.M.H., J.L., and\nM.S. performed IHC experiments and imaging. C.M.H. conceived and\nperformed in vitro experiments and bulk-RNA seq. C.M.H and W.F.F\nperformed bulk-RNA-seq analysis. P.R. provided knowledge and\nresources for iPSC brain organoid cultures. C.M.H, E.A., and B.S.W.\nperformed computational data analysis for visium HD. B.S.W, W.F.F., and\nE.T.C. provided crucial data interpretation. C.M.H. and E.T.C. wrote the\nmanuscript and generatedfigures and schematics. All authors reviewed\na n de d i t e dt h em a n u s c r i p t .\nFunding\nThis study was supported by The Mayday Fund and the State of Con-\nnecticut, through HB6672, which funds the CT Endometriosis Tissue and\nData Biorepository, which is part of the EndoRISE program. The iPSC line\nused in this study was derived with support from The Jackson Labora-\ntory, Director’s Innovation Fund, award number 19000-22-12.\nCompeting interests\nThe authors declare no competing interests.\nAdditional information\nSupplementary informationThe online version contains\nsupplementary material available at\nhttps://doi.org/10.1038/s41467-026-78453-5\n.\nCorrespondenceand requests for materials should be addressed to\nElise T. Courtois.\nPeer review informationNature Communicationsthanks Carlos Simon\nand the other anonymous reviewer(s) for their contribution to the peer\nreview of this work. A peer review file is available.\nReprints and permissions informationis available at\nhttp://www.nature.com/reprints\nPublisher’s note Springer Nature remains neutral with regard to jur-\nisdictional claims in published maps and institutional affiliations.\nOpen Access This article is licensed under a Creative Commons\nAttribution-NonCommercial-NoDerivatives 4.0 International License,\nwhich permits any non-commercial use, sharing, distribution and\nreproduction in any medium or format, as long as you give appropriate\ncredit to the original author(s) and the source, provide a link to the\nCreative Commons licence, and indicate if you modified the licensed\nmaterial. You do not have permission under this licence to share adapted\nmaterial derived from this article or parts of it. The images or other third\nparty material in this article are included in the article's Creative\nCommons licence, unless indicatedotherwise in a credit line to the\nmaterial. If material is not included in the article's Creative Commons\nlicence and your intended use is not permitted by statutory regulation or\nexceeds the permitted use, you will need to obtain permission directly\nfrom the copyright holder. To view a copy of this licence, visit http://\ncreativecommons.org/licenses/by-nc-nd/4.0/\n.\n© The Author(s) 2026\nThis Article in Press is shared early to give you faster access to new\nresearch. It is citable and carries a permanent DOI. The final edited\nversion will replace it automatically.\nArticle https://doi.org/10.1038/s41467-026-78453-5\nNature Communications Article in Press","source_license":"CC0","license_restricted":false}