{"paper_id":"f56a8cfd-51ca-4133-a22d-b96d6d7c117c","body_text":"1\nVol.:(0123456789)Scientific Reports |         (2022) 12:1590  | https://doi.org/10.1038/s41598-022-05499-y\nwww.nature.com/scientificreports\nEffects of endometriosis \non immunity and mucosal microbial \ncommunity dynamics in female \nolive baboons\nNhung Le1, Melissa Cregger2, Asgerally Fazleabas3 & Andrea Braundmeier‑Fleming1*\nEndometriosis is defined as the growth of endometrial tissue in ectopic locations, and is associated \nwith altered immune and microbial phenotypes. It is unclear if these changes are the result of the \ndisease or may be causative. We induced endometriosis in non‑human primates (Papio Anubis) to \ntest our hypothesis that the growth of endometriotic lesions results in alterations in immune and \nmicrobial dynamics that may advance disease progression. Baboon samples were collected pre‑\ninoculation (prior to disease induction), at 3, 6, 9, and 15 months after disease induction. Tolerant \nregulatory T‑cells (Tregs) and inflammatory T‑helper 17 (Th17) cells were identified in peripheral \nblood and within the eutopic/ectopic endometrial tissues. Microbiome communities were identified \nin fecal/urine samples. The induction of endometriosis decreased peripheral Tregs cells while Th17 \ncells increased at all post‑induction collections, thus reducing the Tregs:Th17 cells ratio, indicating \nsystemic inflammation. Microbiome diversity and abundance were altered at each sample site after \ndisease induction. Thus, induction of endometriosis in baboons caused an immune shift toward an \ninflammatory profile and altered mucosal microbial profiles, which may drive inflammation through \nproduction of inflammatory mediators. Immune and microbial profiling may lead to innovative \ndiagnostic tools and novel therapies for endometriosis treatment.\nEndometriosis is defined as the presence of endometrial glands and stroma outside of the uterine cavity. In the \nUS, endometriosis occurs in 1 out of every 10 women of reproductive age (over 200 million women  worldwide1,2), \nand 45% of women with disease report sub-fertility or  infertility3. Despite surgical and hormonal treatment of \nendometriosis, which can have initial positive response rates as high as 70% 4, most women experience symp-\ntom recurrence requiring additional surgical intervention within 2   years5. Although endometriosis was first \ncharacterized in 1920’s, our understanding of the pathogenesis and pathophysiology of endometriosis remains \nan enigma, due to numerous reasons: (1) diagnosis of disease can take more than 10   years6, (2) variability in \ndisease presentation and (3) lack of suitable animal models to investigate disease pathophysiology. In this study, \nwe utilized the baboon (Papio anubis) model of induced experimental endometriosis to investigate physiological \nchanges which occur at disease onset and throughout disease progression. The baboon model for endometriosis \nhas been developed since the early  2000s7, and has several advantages over other animal models of the disease. \nThese are (1) naturally menstruating primates, (2) demonstrate changes in the eutopic endometrium during \nthe window of uterine receptivity, (3) develop spontaneous endometriosis similar to humans 8 and (4) adequate \nanimal body size which allows for repetitive sample collection and sufficient sample size.\nEndometriosis is a reproductive immune disorder where ectopic endometrial lesions display altered inflam-\nmatory profiles compared to a normal endometrium in humans and animal  models 9,10. Immune tolerant \nregulatory T cells (Tregs), including natural Tregs (nTregs;  CD4+CD25+Foxp3+) and inducible Tregs (iTregs; \n CD4+CD25− Foxp3+), are prominent immune populations within reproductive tissues. Tregs are characterized \nby their expression of the Forkhead transcription factor p3 (Foxp3) 11,12 and through the production of anti-\ninflammatory cytokines, IL-10 and TGFβ, which inhibit activation of T helper cells (Th). Inflammatory Th17 \ncells are derived from the CD4 lineage of T cells that secrete interleukin-17 (IL17s), IL21, IL22 and express the \nRAR-related orphan receptor gamma transcription factor (RORγt), to activate cytotoxic  CD8+ T cells. Plasma and \nperitoneal fluid levels of IL-17 are elevated in endometriosis patients and we, along with others, have reported \nOPEN\n1Department of Obstetrics and Gynecology, Department of Medical Microbiology, Immunology and Cell Biology, \nSouthern Illinois University School of Medicine, Springfield, IL 62702, USA. 2Oak Ridge National Laboratory, Oak \nRidge, TN, USA. 3Michigan State University, Grand Rapids, MI, USA. *email: abraundmeier88@siumed.edu\n\n2\nVol:.(1234567890)Scientific Reports |         (2022) 12:1590  | https://doi.org/10.1038/s41598-022-05499-y\nwww.nature.com/scientificreports/\nthat Tregs and Th17 immune cell populations contribute to the systemic and tissue specific inflammatory profile \nof patients with  endometriosis13,14. Briefly we found low levels of peripheral Tregs, greater Th17 cell populations \nand an increase in the Th17/Treg ratio, indicative of systemic  inflammation14. We also reported that Tregs and \nTh17 localizations were enhanced within the ectopic endometrial implant, which promotes lesion development \nvia induce  angiogenesis14. Earlier studies using the baboon model demonstrated that the induction of endome-\ntriosis resulted in reduction of Tregs in the peripheral circulation and  endometrium12. Together, these findings \nprompted us to characterize both Tregs and Th17 cell populations during the pathogenesis of endometriosis in \nthe induced non-human primate model of disease.\nHumans co-exist with resident microbiome that is composed of bacteria, viruses and fungi. The continuous \nand complex interactions between the mucosal surface and resident microbes shapes the host immunity. The \nmucosal immune homeostasis has not only protected the mucosal barrier surface against pathogens, it has also \nprotects the commensal microbes against external  insults15,16. The gastrointestinal (GI) microbiota plays an \nimportant role in the development of  CD4+ T cells as evident by that germ free mice have smaller spleens and \nmesenteric lymph nodes and a decreased level of serum  immunoglobulin15,17. The GI microbiota regulates the \ndifferentiation of Tregs and/or Th17 to induce a tolerant and/or inflammatory response toward a pathogenic or \ncommensal-derived foreign antigen. For example, polysaccharide A molecules from Bacteroides fragilis can signal \ntoll like receptor 2 (TLR2) on Tregs to subsequently suppress a Th17  response18 and segmented filamentous bacte-\nria are potent inducers of Th17  cells19,20. The presence of specific bacterial species (Lactobacillus, Bifidobacterium \ninfantis, etc.) in the urogenital (UG) tract protects mucosal epithelial cells by creating an unsuitable environment \nfor pathogens to survive via the production of lactic acid which lowers the pH of the vaginal  environment21. A \nshift in the composition of the GI/UG microbiota can cause either a pathological or protective outcome that is \nmediated by the regulation of Tregs and Th17 cells induced by the mucosal  microbiota22. Therefore, it is reason-\nable to hypothesize that inflammation associated with endometriosis can shift the microbial dynamics of the GI/\nUG tracts and that these shifts may correlate with presence of the disease and the disease severity.\nWe hypothesize that induction of endometriosis in baboons (P.  An u b i s) results in chronic systemic and tissue \nspecific inflammation through regulation of Th17 and Treg populations. Further, the induction of endometriosis \naltered gastrointestinal/urogenital microbial communities that are distinct from non-diseased animals. Utilizing \nthis model, the aims of our study were to (1) identify the immune phenotype as well as microbial phenotypes in \nnon-diseased (pre-inoculation) phase and throughout the progression of disease pathogenesis; and (2) to inves-\ntigate the correlation of the immune phenotypes with microbial communities at each surgical collection time \npoint to further understand how these dynamics shift after experimental induction of endometriosis. Overall, \nthese investigations have allowed us to profile immune and microbial profiles at the onset of endometriosis and \nthroughout disease progression.\nResults\nPeripheral Treg and Th17 cell populations were altered by induction of endometriosis. To \ndetermine if induction of endometriosis altered peripheral immune cell populations, we identified nTregs \n (CD4+CD25+Foxp3+), iTregs  (CD4+CD25− Foxp3+) and Th17 cell populations in blood samples collected over \ntime (Fig. 1). We demonstrated that post induction, animals exhibited systemic inflammation through an altera-\ntion of tolerant and inflammatory T cell populations (Fig. 1).\nInitially, the induction of disease significantly reduced peripheral nTregs at 3 months and 9 months post-inoc-\nulation (Fig. 1A). The iTregs cell population was reduced at 3 months post-inoculation and remained decreased \nat each following time point (Fig.  1B). Conversely, the peripheral Th17 cell population was increased at each \npost-inoculation timepoint (Fig. 1C). To determine if the induction of endometriosis altered peripheral immune \nhomeostasis (i.e., balance between inflammation and tolerance), we analyzed the ratio of Th17 to Tregs (induc-\nible + natural) cell populations. The ratio of Th17 to Tregs populations increased at 3 months post-inoculation \nand remained elevated at each following surgical time points, indicative of systemic inflammation (Fig.  1D). \nThese results suggested that the induction of endometriosis altered both tolerant and inflammatory immune \npopulations which disrupted immune homeostasis and this disruption was maintained as long as peritoneal \nendometriotic lesions were present.\nFoxp3 and RORγt in eutopic and ectopic endometrium of non‑human primates. To investigate \nthe effect of disease induction and disease progression on activation of Tregs and Th17 in eutopic endometrial \ntissues, we measured the expression of transcription factor RORγt (Th17) and Foxp3 (Tregs) in eutopic endo-\nmetrial tissues collected at each surgical time point (Fig. 2). We observed an elevation of Foxp3 and RORγt tran-\nscripts levels in the eutopic endometrium at each time point after disease induction (Fig. 2A,B). Overall, within \nthe eutopic endometrium the fold induction of RORγt transcripts were significantly higher than the Foxp3 \ntranscripts following the induction of endometriosis; thus, driving an inflammatory profile in the eutopic endo-\nmetrium, (Fig. 2A,B). At 15 months of the disease, we collected ectopic endometrial tissues which allowed us to \ncompare RORγt and Foxp3 transcript expression in matched eutopic and ectopic endometrial tissues. RORγt \nand Foxp3 transcripts were elevated in ectopic endometrium compared to matched eutopic samples (Fig.  2B). \nThese data indicated that the eutopic endometrium has enhanced expression of RORγt and Foxp3 transcripts \nthroughout disease progression and that these expression patterns are even more augmented in ectopic endome-\ntrial tissues. Altogether, a lower number of Treg cells in the peripheral blood suggested an increase in systemic \ninflammation via removal of the inhibition of Th17 cell function by Tregs; but a higher number of Treg cells in \nthe eutopic and ectopic endometrium allows ectopic endometrial tissues to attain immune tolerance from the \ninnate immune system and may promote disease establishment.\n\n3\nVol.:(0123456789)Scientific Reports |         (2022) 12:1590  | https://doi.org/10.1038/s41598-022-05499-y\nwww.nature.com/scientificreports/\nBacterial community diversity following induction of endometriosis. To compare differences in \nmicrobial diversity over disease progression, among all animals by sample types, we performed beta-diversity \nanalyses that used phylogenetic information with unweighted and weighted UniFrac distance metrics. The \nunweighted Unifrac (qualitative) measured the fraction of branch length in a phylogenetic tree that leads to \ndescendants of one sample or the other while a weighted UniFrac (quantitative) directly accounted for dif-\nferences in relative abundances of each type of organism. GI bacterial communities were significantly differ -\nent between pre-inoculation and throughout the disease progression, except at 9  months post-inoculation \n(Fig. 3A,B). There was no changes in microbial diversity between study time points for the UG tract (vaginal \nswabs) or peritoneal cavity (peritoneal fluid) (Supplementary Fig. S1). Thus, the induction of endometriosis, in \nnon-human primates, altered GI bacterial diversity as disease progressed but this change was not evident in UG \nor peritoneal bacterial communities.\nBacterial composition alterations with the induction of endometriosis. To assess if the induction \nof endometriosis caused differences in bacterial community compositions (species richness and uniformity), we \nperformed alpha-diversity analysis with PERMANOV A for Simpson’s evenness, Simpson’s diversity and Faith’s \nphylogenetic diversity. Simpson’s evenness measures how evenly the abundance was distributed among spe-\ncies while Simpson’s diversity measured the number of species presented in a community. Additionally, Faith’s \nphylogenetic diversity incorporated phylogenetic difference between species via summing branch lengths on \nphylogenetic trees. Overall, GI bacterial evenness was reduced at 3  months post-inoculation, but recovered \nat 6 months to 15 months post-inoculation (Fig.  4A). Faith’s phylogenetic diversity was lower for all animals \nat 6 months post-inoculation compared to pre-inoculation (Fig.  4B). There was no difference in GI Simpson’s \ndiversity (species richness) at each time point. Urinary bacterial alpha-diversity (Simpson’s evenness and Simp-\nson’s diversity) was reduced at 3, 6 and 15 months post-inoculation (Fig. 4C,D). The vaginal tract and the peri-\ntoneal cavity bacterial alpha diversities did not alter during the disease induction and throughout the disease \nprogression (Supplementary Table S1). Thus, the induction of endometriosis, in non-human primates, altered \nGI and urinary bacterial alpha-diversity.\nTaxonomic variation with induction of endometriosis. Induction of endometriosis altered mucosal \nmicrobiota in the GI/UG tracts. Firmicute  species were abundant in the GI tract of all animals, regardless of \ndisease status (Fig.  5A). The most abundant phylum identified from fecal samples was Firmicutes followed \nby Bacteroidetes and Proteobacteria (Fig.  5A, black underline). At the genus level, prior to disease induction, \nPrevotella dominated GI bacterial communities; this was followed by Megasphaera, Lactobaccillus, Oscillospira, \nAnaerovibrio, 02d06, Treponema, Succinivibrio and CF231 (Fig.  5B). Disease induction resulted in decreased \nlevels of Succinivibrio, Prevotella, Megasphaera, Lactobaccillus and CF231 at 3 months post-inoculation, but the \nFigure 1.  Immune cell population in peripheral blood samples of non-human primates. (A–C) Peripheral Treg \nand Th17 cell populations were measured for 8 baboons at pre-inoculation and post-inoculations: 3, 6, 9 and \n15 months. (A) Natural Tregs (nTregs); (B) Inducible Tregs (iTregs); (C) Th17 population. (D) Th17/Tregs ratio. \n*Indicates significance between groups. Mann–Whitney U-test, p-value < 0.05.\n\n4\nVol:.(1234567890)Scientific Reports |         (2022) 12:1590  | https://doi.org/10.1038/s41598-022-05499-y\nwww.nature.com/scientificreports/\nlevels of Succinivibrio, Prevotella, and CF231 increased throughout disease progression from 6 to 9 months post-\ninoculation (Fig. 5B, blue box). However, as the disease progressed, levels of other genera such as Megasphaera, \nTreponema and Prevotella were also decreased (Fig. 5B, green box).\nSimilar to fecal samples, disease induction shifted urinary microbial dynamics upon comparison of pre-\ninoculation samples to those collected at 3- to 15-months post-inoculation (Fig. 6A). The Firmicutes phylum was \npredominant followed by Bacteroidetes, Antinobacteria and Proteobacteria (Fig. 6B, black underline). Prior to the \ndisease induction, urinary bacterial communities were dominated by the genera Porphyromonas, Pseudomonas, \nCampylobacter, Corynebacterium, Acitinobaculum, and Streptococcus (Fig. 6B). After the disease induction, the \nlevels of Corynebacterium, Pseudomonas, and Streptococcus increased (Fig. 6B, red box), while other genera were \ndecreased in these animals at 3 months post-inoculation (Fig.  6B). Pseudomonas, Porphyromonas, Garnerella, \nand Helcococcus were increased in the urinary tract at 6 months and 9 months post-inoculation (Fig.  6B, blue \nbox). As the disease further progressed, the levels of multiple genera once again decreased in the urinary tract \nat 15 months post-inoculation (Fig. 6B, right panel).\nIn the vagina, Bacteroidetes was the predominant phylum, followed by Firmicutes, Antinobacteria, and \nFusobacteria (Fig. S2). At the genus level, prior to disease induction, Porphyromonas , Mobiluncus, Treponema, \nCampylobacter, Prevotella, and Streptobacillus dominated vaginal bacterial communities. However, throughout \nthe disease progression these genera within the vagina diminished and were never restored. We also observed a \ndominant increase of the phylum Firmicutes (Peptoniphilus and Dialister) after the disease induction.\nFor the peritoneal cavity, an unclassified group of bacteria was dominant in the peritoneal bacterial communi-\nties and proteobacteria was the next predominant phylum, followed by Firmicutes and Antinobacteria (Fig. S2). \nPrior to disease induction, the dominant genera were Lactobacillus, 02d06, Campylobacter and Succinivibrio but \nthese diminished upon disease induction and did not restore during the disease development. We observed an \nincrease of the phyla Firmicutes (Phascolarctobacterium and Helcococcus) and Proteobacteria (Campylobacter) \nin these animals upon disease induction.\nFinally, we wanted to determine if induction of disease was caused the development of an inflammatory \ndisorder. While typically only analyzed in gut, we investigated the ratio of Firmicute/Bacteroidetes within each \nsample type. The GI tract had the lowest ratio of Firmicute/Bacteroidetes pre-inoculation, but this gradually \nFigure 2.  Foxp3 and RORγt quantitative RT–PCR of non-human primates eutopic and ectopic endometrial \ntissues. (A) Foxp3, and (B) RORγt transcript levels were measured in the eutopic endometrium tissues of 8 \nbaboons at pre-inoculation: 3, 6, 9 and 15 months. The relative fold induction of Foxp3 and RORγt genes was \nnormalized to H3.3 endogenous gene for all experimental conditions. *Indicates significant difference between \ncompared groups. (C) Fold induction for each eutopic to matched ectopic endometrial tissues at 15 months \ncollection. The ectopic endometrial tissues was normalized to matched eutopic endometrial tissues for all 8 \nanimals. Mann–Whitney U-test, p-value < 0.05.\n\n5\nVol.:(0123456789)Scientific Reports |         (2022) 12:1590  | https://doi.org/10.1038/s41598-022-05499-y\nwww.nature.com/scientificreports/\nincreased at each of the collection time point during the disease development (in GI tract: pre-inoculation = 1.34; \npost-inoculation: 3 m = 1.9; 6 m = 1.8; 9 m = 1.8; 15 m = 2.4). However, in urine, we observed a decrease in the \nratio at each study time point after the induction of endometriosis (pre-inoculation = 1.68; post-inoculation: \n3 m = 0.8; 6 m = 0.9; 9 m = 0.9; 15 m = 1.3). In the vagina and peritoneal cavity, the Firmicute/Bacteroidetes \nratio fluctuated throughout the disease establishment and progression (vaginal tract: pre-inoculation = 0.6; post-\ninoculation: 3 m = 1.2; 6 m = 1.1; 9 m = 0.5; 15 m = 0.8; peritoneal cavity: pre-inoculation = 4; post-inoculation: \n3 m = 5; 6 m = 1.8; 9 m = 4).\nTaken all together, these results indicated that the induction of endometriosis altered the mucosal microbiota \nin multiple sites and additionally, the presence of endometriotic lesions altered the quantity and composition of \nmicrobial species. Disease establishment and progression further distinguished the microbiome profiles of these \nanimals from the healthy control time point (pre-inoculation).\nAssociation of Immune populations with microbial dynamics. To expand upon our analysis of how \nmicrobial dynamics impacted immune status, we performed correlative analyses to determine if an alteration \nof microbial diversities via the induction of endometriosis was associated with peripheral immune cells popula-\ntions (iTregs, nTregs and Th17) (Supplementary Table S2). The GI microbial alpha-diversity was positively cor-\nrelated with circulating nTregs cells at 3 months post-inoculation (p = 0.008), while iTregs cell populations was \nassociated with GI alpha-diversity at 9 months post-inoculation (p = 0.03). Urinary microbial alpha-diversity \ncorrelated with peripheral nTregs at 3 months and 15 months post-inoculation (p = 0.002; p = 0.02 respectively) \nand correlated with Th17 cell population at 6 months post-inoculation (p = 0.04). These results showed that the \ninduction of disease altered bacterial diversity at both sites (GI/urine) moreover, changes in microbial diversity \nwere associated with distinct sub-types of T cells at different stages of disease progression. Immune tolerant cells \n(Treg sub-types) were associated with microbial diversity during early and later stages of disease progression \nwhereas inflammatory cells (Th17) was associated with microbial diversity in the middle of our timeline. These \ndata suggest that not only are the microbial and immune profiles transient but there exists a dynamic relation-\nship between microbial and immune parameters during disease progression.\nAssociation between microbial species and peripheral immune cells with induction of endo ‑\nmetriosis. To examine if bacterial community structure was associated with immune phenotypes, we per -\nformed Pearson’s correlation coefficient for each bacterial site with the peripheral iTregs, nTregs and Th17 cell \npopulations. In the GI tract, prior to the induction of endometriosis, the phyla Bacteroidetes, Firmicutes, and \nProteobacteria (Genera: Clostridium, Coprococcus, Defluviitalea, Oscillospira, Prevotella, RFN20) were negatively \ncorrelated with level of peripheral nTregs and iTregs cells; meanwhile Prevotella  and Sutterella were positively \ncorrelated with level of Th17 cell populations (Fig.  7A). At 3 months post-inoculation, there were additional \nphyla (Actinobacteria, Euryarchaeota, Fusobacteria, Lentisphaerae, Spirochaetes, and Synergistetes) in GI com-\nmunities that had a positive correlation with peripheral nTregs and Th17 cell populations; only Porphyromonas, \nPrevotella, and WAL were negatively correlated with the level of iTregs cells (Fig.  7B, left panel). At 6 months \nFigure 3.  β diversity in gastrointestinal tract (GI) from pre- and post-inoculation of endometriosis using \nthe ANOSIM algorithm. (A) Unweighted Unifrac, (B) Weighted Unifrac. The horizontal lines inside the \nboxes indicate the median, whereas the lower lines and upper lines of the boxes indicate the 25th and the \n75th percentiles, respectively. The dots outside the boxes indicate the outliers. *Indicates significance between \ngroups, p-value < 0.05. (Pre-inoc and 3 m post-inoc: unweighted p = 0.38, weighted p = 0.021; pre-inoc and 6 m \npost-inoc: unweighted p = 0.04, weighted p = 0.029; pre-inoc and 9 m post-inoc: unweighted p = 0.11, weighted \np = 0.38; pre-inoc and 15 m post-inoc: unweighted p = 0.043, weighted p = 0.04).\n\n6\nVol:.(1234567890)Scientific Reports |         (2022) 12:1590  | https://doi.org/10.1038/s41598-022-05499-y\nwww.nature.com/scientificreports/\npost-inoculation, GI bacteria positively correlated with the level of iTregs cell population, whereas a lower \namount of GI species was positively correlated with peripheral nTregs cells (Fig. 7B, middle panel). At 9 months \nand 15 months post-inoculation, we detected a higher number of GI bacteria that negatively correlated with \nthe level of iTregs cells compared to 3 months and 6 months post-inoculation (Fig.  7B, right panel). Overall, \nthe induction of endometriosis resulted in a higher number of GI bacteria that correlated with immune cell \npopulations (nTregs, Th17) at 3 months and 6 months post-inoculation; meanwhile at 9 months and 15 months \npost-inoculation, there was a reduction in the number of GI species that correlated with these immune cell \npopulations (Fig. 7B).\nIn urine, at pre-inoculation, the phyla Actinobacteria, Bacteroidetes, Firmicutes, and Proteobacteria were \nnegatively correlated with the levels of iTregs and Th17 cell populations (Fig.  8, left panel). At 3 months post-\ninoculation, urinary bacteria were negatively correlated with the peripheral iTregs cell population but were \npositively correlated with the level of Th17 cells (Fig. 8, middle panel). No correlation was noted between urinary \nspecies and circulating nTregs cells at pre-inoculation and at 3 months post-inoculation (Fig.  8). At 6 months \npost-inoculation, there were positive correlations between the urinary bacteria and levels of peripheral iTregs \nand nTregs cells, while only Straptobacillus was negatively correlated with circulating Th17 cell population (Fig. 8, \nmiddle panel). A negative correlation was identified for urinary species and the peripheral iTregs cell population \nat 9 months post-inoculation. No correlation was detected between urinary species and the circulating nTregs \nFigure 4.  α diversity from pre- and post-inoculation of endometriosis in GI tract (A, B) and urine (C, D). (A): \nSimpson’s evenness in GI tract (Simpson’s evenness [species uniformity]: pre-inoc and 3 m post-inoc: p = 0.09; \n3 m and 6 m post-inoc: p = 0.02; 3 m and 9 m post-inoc: p = 0.034; 3 m and 15 m post-inoc: p = 0.04), (B) Faith’s \nphylogenetic diversity in GI tract (p = 0.01), (C) Simpson’s evenness in urine (Simpson’s evenness: pre-inoc and \n15 m post-inoc: p = 0.03; 3 m and 15 m post-inoc: p = 0.04), (D) Simpson’s diversity in urine (Simpson’s diversity: \npre-inoc and 6 m post-inoc: p = 0.03). The horizontal lines inside the boxes indicate the median, whereas the \nlower lines and upper lines of the boxes indicate the 25th and the 75th percentiles, respectively. The dots outside \nthe boxes indicate the outliers. *Indicates significance between groups, p-value < 0.05.\n\n7\nVol.:(0123456789)Scientific Reports |         (2022) 12:1590  | https://doi.org/10.1038/s41598-022-05499-y\nwww.nature.com/scientificreports/\nFigure 5.  Taxonomical analysis for fecal samples of 8 non-human primates from pre- and post-inoculation of \nendometriosis. (A) Level 2 (phyla) taxonomical summary plots. (B) The top 50 abundant bacterial genera in GI \ntract. Samples were collected from animal at the pre-inoculation (pre-Inoc) (left panel) and following disease \ninduction (right panel).\nFigure 6.  Taxonomical analysis for urine samples of 8 non-human primates from pre- and post-inoculation \nof endometriosis. (A) Level 2 (phyla) taxonomical summary plots. (B) The top 50 abundant bacterial genera in \nurinary tract. Samples were collected from animal at the pre-inoculation (pre-inoc) (left panel) and following \ndisease induction (right panel).\n\n8\nVol:.(1234567890)Scientific Reports |         (2022) 12:1590  | https://doi.org/10.1038/s41598-022-05499-y\nwww.nature.com/scientificreports/\ncells at 9 months and at 15 months post-inoculation (Fig.  8). Finally, at 15 months post-inoculation, there was \na positive correlation for urinary bacteria and the level of Th17 cell population (Fig.  8). We did not detect any \ncorrelation between immune populations and vaginal or peritoneal samples (Supplementary Table S3).\nIn summary, the induction of peritoneal endometriosis altered bacterial communities within the GI and UG \ntracts of animals with endometriosis. Specifically, the dysbiosis of GI/UG microbial communities was associated \nwith aberrant levels of peripheral nTregs, iTregs (immune tolerant), and Th17 (inflammatory) cell populations.\nDiscussion\nThe immune system and the commensal bacterial species in the gut and the female reproductive tract play a major \nrole in preserving overall homeostasis for the host’s health. These microbiota support immunological regulation \nof reproductive functions, and are influenced by factors such as immunological responses, metabolic changes and \nthe  environment23. Therefore, a shift in the commensal microbial communities are indicative of potential shifts \nin immune signaling and function. Utilizing a non-human primate animal model of induced endometriosis in \nolive baboons, this study investigated the alteration of immune populations and microbial dynamics in response \nto establishment of disease.\nOur first goal was to define the endometriosis-associated inflammation throughout the progression of endome-\ntriosis by characterizing Tregs and Th17 profiles using the non-human primate induced model of endometriosis.  \nIncreased Th17 cells and their cytokine profiles have been observed in the peritoneal fluid of women with \n endometriosis24, and excessive IL-17 from Th17 cells is associated with the severity of  disease25. However, there \nare limited publications regarding the levels of Th17 cells in animal models of endometriosis. Consistent with \nprevious reports from human studies, Th17 populations in baboons were expressed most abundantly in the \nperipheral circulation throughout the disease  pathogenesis12,26,27. Similar to previous reports from Braundmeier \net al.,  the induction of endometriosis resulted in a rapid decrease in both nTregs and iTregs in the peripheral \ncirculation. Thus, our data demonstrates a systemic immune imbalance (enhanced Th17/Treg ratio) after the \ninduction of endometriosis; furthermore, this immune phenotype persists throughout the progression of the \ndisease (15 months). In addition, we observed an upregulation of both Foxp3 and RORγt transcripts in the \neutopic and matched ectopic endometrial tissues, which supports similar reports in human  studies28. The immune \nsystem’s failure to down regulate Foxp3 expression in the eutopic endometrial tissues during the disease may \nFigure 7.  Pearson’s correlation coefficient of GI bacterial communities with level of peripheral immune cell \npopulations (iTregs, nTregs, Th17) in 8 non-human primates from pre- and post-inoculation of endometriosis. \n(A) Pre-inoculation. (B) Post-inoculation.\nFigure 8.  Pearson’s correlation coefficient of urinary bacterial communities with level of peripheral \nimmune cell populations (iTregs, nTregs, Th17) in 8 non-human primates from pre- and post-inoculation of \nendometriosis.\n\n9\nVol.:(0123456789)Scientific Reports |         (2022) 12:1590  | https://doi.org/10.1038/s41598-022-05499-y\nwww.nature.com/scientificreports/\nenhance ectopic endometrial growth through inhibition of immune  clearance29. Alternatively, the higher expres-\nsion of Th17 and lower Treg transcript within endometriotic lesions may also stimulate the continuous RORγt \ncell proliferation in endometriosis, which leads to the infiltration of inflammatory cells and mediators into the \nlesion(s), thus promoting tissue remodeling.\nOur second goal was to define microbial composition after the induction of endometriosis and identify the \npotential variations associated with disease. We started by assessing the bacterial species richness and diversity \nat pre-inoculation and throughout the progression of endometriosis. The GI microbiome diversity is crucial in \nhealth maintenance as microbiota and their metabolites (short chain fatty acids [SCFAs] and microbially trans-\nformed bile acids) have been proven to play a fundamental role in immune cell regulation and  signaling30,31. \nIn this regard, the gut microbial communities (e.g. Clostridia  spp.) serve as a source of SCFAs such as butyrate \nand propionic acid, which help to maintain Tregs cell expansion (immunosuppressive function) and promote \nan intestinal  homeostasis30,31. The microbiota within the UG tract is dominated by Lactobacillus species. These \ncommensal bacteria in the UG tract create an unsuitable environment for pathogens and keep them from colo-\nnizing and causing infection by producing lactic acid as a fermentation byproduct that lowers the pH of the UG \ntract environment (pH < 4.5). A reduction in microbial diversity due to dysbiosis and inflammation reduces \ntheir metabolic activity, which might alter immune homeostasis. We observed that the GI bacterial communities \nwere altered in their diversity and richness after the induction of the disease but recovered later as the disease \nprogressed (6 months to 15 months post-inoculation). Urinary bacterial diversity was reduced after the induction \nof endometriosis and remained altered throughout disease progression. These results showed that induction of \nthe disease altered bacterial diversity at both sites (GI/urogenital tracts), but unlike the urogenital tract the GI \nmicrobial dynamics were transient in the response to disease induction.\nConsistent with previous studies, more than 80% of all study animals’ microbiota were composed of Actino-\nbacteria, Bacteroidetes, Firmicutes, and Proteobacteria16,32,33. An elevated ratio of Firmicutes/Bacteroidetes has \nbeen associated with  obesity34,35, colorectal  cancer36, and rheumatoid  arthritis37. A recent review from Magne \net al., suggested that using Firmicutes/Bacteroidetes  ratio to determine health status would be a challenge due \nto multiple discrepancies such as lifestyle associated factors and the sampling process. In this study, the ratio of \nFirmicute/Bacteroidetes in the GI tract gradually increased at each of the collection timepoints during disease \ndevelopment. However, in the urinary tract, we observed a decrease in the ratio of Firmicute/Bacteroidetes  at \neach study timepoint after the induction of endometriosis. The reduction in the levels of Succinivibrio, Megas -\nphaera and Prevotella spp. (phylum of Proteobacteria, Firmicutes, and Bacteroidetes respectively) observed after \nthe induction of endometriosis may play a role in endometriosis-associated inflammation at 3 months post-\ninoculation. Indeed, Prevotella spp. are known to stimulate the production of anti-inflammatory cytokines, such \nas IL-10, using  Foxp3+ regulatory T cells through the production of propionate, succinate and acetate during the \ncarbohydrate fermentation of organic  acids38. The level of propionic acid increases during infection to reduce \ninflammation and protects tissues during the immune response to an  infection38–40. Similarly, high levels of \nClostridium spp. (Firmicutes phylum) induce colonic  Foxp3+ regulatory T cells and activate T cell dependent \nimmunoglobulin A  production38,41,42. Based on a comparison between pre-inoculation samples to those collected \nat 3- to 15-months post-inoculation, disease induction increased the pathogenic bacteria (e.g. Corynebacterium, \nPseudomonas, and Streptococcus ) in urine. Additionally, the presence of endometriotic lesions altered both \nquantity and microbial species composition.\nAdditionally, we investigated the association between the GI and urinary microbiome and the peripheral \ncirculating immune cells in non-human primates with endometriosis. Overall, disease induction resulted in \nmore GI bacteria that were positively correlated with immune cell populations (nTregs, Th17) at 3 months and \n6 months post-inoculation; however, a reduction in GI species that correlated with these immune cells popu-\nlations was observed at 9 months. The association between urinary bacteria and the peripheral immune cells \nchanged throughout the disease pathogenesis as urinary bacteria were negatively correlated with the peripheral \niTregs cell population but were positively correlated with the level of Th17 cells. Our results indicate a dysbiosis \nin the GI and urinary microbiomes of animals with endometriosis that is concomitant with alterations in the \nlevels of peripheral nTregs, iTregs (immune tolerant), and Th17 (inflammatory) cell populations. However, the \nmechanism(s) of action between specific GI/UG species and host immunity during endometriosis still needs \nfurther investigation.\nIn summary, our findings provide evidence that there may be a unique microbiome “signature” in the GI and \nUG tracts, as well as a distinct immune profile that is associated with induction of endometriosis. The major \nfindings of this study are the following: (1) the mucosal microbiomes (GI, UG) exhibited a unique profile at pre-\ninoculation vs. post-inoculation; (2) a systemic inflammatory phenotype via an increase in the ratio of Th17:Treg \ncells upon the induction of endometriosis; (3) a correlation between inflammation and alteration of microbial \ncommunities throughout disease progression. Our results support interaction between the immune system and \nmucosal microbial dynamics in patients with endometriosis and warrant further investigations to elucidate how \nthese physiological systems impact the pathogenesis of endometriosis.\nLimitation of the study: using non-human primate offers a tremendous advantage such as physiologic similar-\nity to humans and reproducibility of experimental results. But the disadvantage of using these animals include \nthe difficulty of availability and relatively high cost. We acknowledged that there were no control animals in \nthe study over a period of 15 months. However, environmental factors such as diet, infections, antibiotics, and \ngenetic background were controlled in the study. Additionally, the pre-inoculation stage was used as the control \nto be compared to the disease progression over a period of 15 months after inoculation. Thus, the study design \nestablishes an internal baseline for all biological measurement, and therefore reduces animal variability. This \ndesign results in a reduction of error and an increase in power with a limited sample size.\n\n10\nVol:.(1234567890)Scientific Reports |         (2022) 12:1590  | https://doi.org/10.1038/s41598-022-05499-y\nwww.nature.com/scientificreports/\nMaterials and methods\nAnimal housing and health screening. The study was reported in accordance with ARRIVE guidelines \nfrom PLOS ONE editorial team. In the study, all 8 non-human primates lived under the same housing condi-\ntions, within the same room throughout the duration of the study. All animals received standardized environ-\nmental enrichment such as Kong toys, visual stimulation, and auditory enrichment. They all received the same \nstandardized diet from the same provider, but the food was not irradiated. Animals had been on the same diet \nfor at least 60 days prior to the initiation of the study so the dietary influence on microbial dynamics should be \nwell established.\nTuberculosis (TB) testing and fecal float/smear for parasites were screened for all animals. Animals were \npurchased from a conventional colony that is known to be positive for Papiine herpesvirus 2 and simian T-cell \nleukemia virus. Since all animals had these infectious reagents when we sampled them as controls, infectious \nagents should not affect the changes observed following the induction of endometriosis. Routine health screening \nconsists of the semi-annual TB testing and annual physical exam with CBC/chem panel/fecal flotation. Animals \nthat had a change in health status or require medical intervention were then removed from data analysis. Because \nour animal care facility has a closed colony, we were able to minimize the exposure and transmission of several \ninfectious disease agents.\nInduction of endometriosis and samples process. Endometriosis was experimentally induced in 8 \nolive baboons as described  previously12. Briefly, all animals were of reproductive age and confirmed disease free \nby a laparoscopic viewing of the abdominal cavity prior to inoculation. At the time of inoculation, autologous \nmenstrual endometrium was deposited in the pouch of Douglas, the uterine fundus, the cul de sac, and the \novaries via a pipelle during laparoscopic surgery. A secondary inoculation was performed at the subsequent \nmenses. Disease progression was monitored by laparoscopic visualization at 4 different timepoints over a period \nof 15 months after inoculation.\nAll experiments were performed in accordance with relevant guidelines and regulations. All biological sam-\nples were collected in the animal care facility at the University of Illinois, Chicago, IL, USA under protocol #17-\n037. All procedures were approved by the University of Illinois Institutional Animal Care and Use Committee \n(IACUC) and Michigan State University. Urine and peritoneal fluid samples, fecal and vaginal swab samples, \nheparinated peripheral blood and eutopic endometrium were collected at pre-inoculation (pre-inoc) and at four \npost-inoculation timepoints: 3 months, 6 months, 9 months and 15 months. At 15 months post-inoculation, \nectopic endometrial tissues were collected prior to the animal being euthanized.\nUrine and peritoneal fluid samples (10–50 ml), without preservative, were centrifuged to collect the cellular \ndebris and stored at − 80 °C until DNA extraction was performed. Fecal and vaginal swabs were immediately \nplaced into separate 1 ml sterile  Ca2+/Mg2+ free phosphate-buffered saline (1X PBS) and stored at −  80 °C until \nDNA extraction was performed. Peripheral blood mononuclear cells (PBMCs) were extracted from heparinized \nblood collection vials and stored in 1 ml of freezing medium (90% Fetal bovine serum, 10% Dimethyl sulfoxide) \nin liquid nitrogen until use. All tissues were further processed for RNA extraction for quantitative RT–PCR.\nAnalysis of lymphocytes using fluorescence‑activated cell sorter. Flow cytometric analysis \nwas performed on all PBMCs to detect nTregs  (CD4+CD25+Foxp3+), iTregs  (CD4+CD25− Foxp3+) and Th17 \n (CD4+CD25− RORγt) cells by utilizing the protocol from Braundmeier et al. 2012. Briefly, approximately  105 to \n 106 mononuclear cells were stained directly with anti-human FITC-CD4 (L200; 550628, BD Pharmigen), anti-\nhuman APC-CD25 (BC96; 17-0259, eBioscience), anti-human PE-Foxp3 (PCH101; 12-4776; eBioscience) and \nanti-human PE-RORγt (AFKJS-9; 12-6988; eBioscience) antibodies. Lymphocyte cell populations were sorted \nusing a BD Accuri C6 flow cytometer and its respective software (BD Biosciences). Populations were gated \non CD4 fluorescent intensity;  CD4+ subpopulations were identified by  CD25+,  Foxp3+ and RORγt+ fluorescent \nintensity. Treg and Th17 cells were compared within each animal.\nQuantitative RT–PCR. Real-time PCR analyses were performed using the following primer/probe sets \nfrom Applied Biosystems: Histone 3.3 primers (Forward: GGC GCT CCG TGA AAT TAG AC; Reverse: CGC TGG \nAAG GGA AGT TTG C; Probe: CGC TGG AAG GGA AGT TTG C), Foxp3 (Hs01085834_m1) and RORγt primers \n(Forward: TGG ACC ACC CCC TGC TGA GAAGG; Reverse: CTT CAA TTT GTG TTC TCA TGACT; Probe: GGG \nAGC CAA GGC CGG).\nReal-time PCR amplification and detection were performed in MicroAmp optical 96-well reaction plates \nusing the QuantStudio™ 3 real-time PCR detection system. Relative fold induction of Foxp3 and RORγt were \ncalculated by the ∆Ct method: ∆Ct  =  Cttarget gene –  CtH3.3 gene (presented with  2−∆Ct ) in eutopic endometrium at all \ncollection time points and ectopic endometrium at 15 months post-inoculation. The difference between Foxp3, \nRORγt and H3.3 was normalized to controls (pre-inoculation) for each animal. H3.3 was used as an endogenous \ncontrol gene.\nMicrobial community analysis. DNA extraction was performed on fecal specimens, urine pellets, vaginal \nsamples and peritoneal fluid pellets using a MoBio PowerSoil DNA Isolation kit (Qiagen, Carlsbad, CA). After \nextraction, the DNA stock concentration was measured using a Qubit™ dsDNA BR (Broad-Range) Assay Kit \n(Q32850; Invitrogen).\n16S rRNA gene amplification and sequencing. Bacterial sequencing targeted the V4 region of the 16S rRNA \ngene (archaeal/bacterial) with a two-step polymerase chain reaction (PCR) approach using the Illumina Nextera \n\n11\nVol.:(0123456789)Scientific Reports |         (2022) 12:1590  | https://doi.org/10.1038/s41598-022-05499-y\nwww.nature.com/scientificreports/\nXT sequencing protocol. The forward and reverse primer mixture was modified and amplified as previously \ndescribed, with four variants of 515F and one 806R primer modified for the Illumina MiSeq  platform43. The \nthermal cycler conditions for the primary PCR were: 3 min at 95 °C followed by 35 cycles of 95 °C for 30 s, 55 °C \nfor 30 s and 72 °C for 30 s with a 5 min final extension at 72 °C. The PCR products were purified with Agencourt \nAmpure XP beads (Beckman Coulter, Indianapolis, IN) and each sample was then individually labeled with a \nunique set of forward and reverse indexes through a second PCR. The secondary index PCR cycle was the same \nas above but with only 8 cycles, and the resulting product was again purified with Agencourt Ampure XP beads. \nThese DNA amplicons were normalized, pooled to a final loading concentration of 4 pM with 20% PhiX spike-in \nand sequenced bi-directionally 250 bases using v2 reagents on the MiSeq platform (Illumina, San Diego, CA) at \nthe University of Tennessee Genomics Core.\nSequence bioinformatics analysis. Data were quality filtered and processed using  QIIME244. First, paired end \nreads were merged with a Phred quality threshold of Q30; then a quality assessment was performed by specific \nfiltering conditions in accordance with QIIME2 quality control process (Trim and truncate primers: trim-left-\nforward and reverse = 10, trunc-len-forward and reverse = 250). Exact sequence variants (ESVs) were clustered \nusing the DADA2  algorithm44 and aligned to the Greengenes-reference v. 13.8 database for archaea/bacteria. \nFinally, artifact sequences or host contamination (i.e. mitochondria, chloroplast or eukaryote) were filtered out.\nSequencing statistic. A total of 932,143 sequences were obtained after quality filtering and sequence process-\ning. The average number of sequences per sample was 58,957 for fecal samples, 12,099 for urine samples, 43,549 \nfor vaginal samples and 2457 for the peritoneal cavity. Rarefraction curves were set to account for variation in \nsequencing depth and according to each sample type: fecal sample with 2500; urine samples with 600; vaginal \nsamples 2500; peritoneal samples with 200. At these cut point of sequences per sample, rarefaction curves pla-\nteaued indicating sufficient sequencing for the discovery and investigation of the GI/UG and peritoneal cavity \nmicrobial communities.\nStatistical analysis. All results in figures and tables are expressed as mean ± SEM, n values in figure legends \nindicate the number of independent experiments, unless otherwise indicated. Alpha-diversity and evenness \nwere estimated for each sample using Simpson’s evenness measure E, Simpson’s index diversity, and Faith’s phy-\nlogenetic diversity metrics (Faith’s PD) calculated in QIIME2. Microbiome alpha-diversity comparisons between \nthe pre-inoculation and all post-inoculation, and the effect of peripheral immune cells with microbiota were \nassessed by ANOV A (Qimme2R and phyloseq packages). Beta-diversity (diversity between samples) on both \nweighted and unweighted UniFrac was conducted to compare the dissimilarity between samples via QIIME2. \nA constrained analysis of principal coordinates ([CAP], capscale function in vegan package) was calculated for \nbacteria in GI/UG, peritoneal cavity samples with the level of peripheral circulating immune cells included as \npredictor variables. Variation in community composition among samples was visualized via a non-metric multi-\ndimensional scaling plot (NMDS) based on weighted and unweighted Unifrac with phyloseq package. Statistical \ndifferences in community composition were assessed using PERMANOV A in QIIME2 with 999 permutations to \nmeasure factors driving bacterial community  composition45,46. For taxon abundance, raw counts were retained \nand normalized by clr transformation; one-way ANOV A was used to study how presence of disease influenced \ntaxon abundances. Pearson correlations were performed using QIIME to assess the relationships between the \nGI/UG diversity and level of immune cells in peripheral blood samples.\nNon-parametric tests were used to determine differences between study time points when the data set was \nnot normally distributed. Mann–Whitney U test was used to determine differences in immune populations for \nboth peripheral blood and endometrial tissue analyses. A value of P < 0.05 was considered statistically significant. \nData analysis was conducted using GraphPad Prism 7.\nReceived: 1 June 2021; Accepted: 11 January 2022\nReferences\n 1. Bulun, S. E. Endometriosis. N. Engl. J. Med. 360, 268–279. https:// doi. org/ 10. 1056/ NEJMr a0804 690 (2009).\n 2. As-Sanie, S. et al. Assessing research gaps and unmet needs in endometriosis. Am. J. Obstet. Gynecol. 221, 86–94. https:// doi. org/ \n10. 1016/j. ajog. 2019. 02. 033 (2019).\n 3. Macer, M. L. & Taylor, H. S. Endometriosis and infertility: A review of the pathogenesis and treatment of endometriosis-associated \ninfertility. Obstet. Gynecol. Clin. N. Am. 39, 535–549. https:// doi. org/ 10. 1016/j. ogc. 2012. 10. 002 (2012).\n 4. Becker, C. M., Gattrell, W . T., Gude, K. & Singh, S. S. Reevaluating response and failure of medical treatment of endometriosis: A \nsystematic review. Fertil. Steril. 108, 125–136. https:// doi. org/ 10. 1016/j. fertn stert. 2017. 05. 004 (2017).\n 5. Bulletti, C., Coccia, M. E., Battistoni, S. & Borini, A. Endometriosis and infertility. J. Assist. Reprod. Gen. 27, 441–447. https:// doi. \norg/ 10. 1007/ s10815- 010- 9436-1 (2010).\n 6. Arruda, M. S., Petta, C. A., Abrão, M. S. & Benetti-Pinto, C. L. Time elapsed from onset of symptoms to diagnosis of endometriosis \nin a cohort study of Brazilian women. Hum. Reprod. 18, 756–759. https:// doi. org/ 10. 1093/ humrep/ deg136 (2003).\n 7. Braundmeier, A. G. & Fazleabas, A. T. The non-human primate model of endometriosis: Research and implications for fecundity. \nMol. Hum. Reprod. 15, 577–586. https:// doi. org/ 10. 1093/ molehr/ gap057 (2009).\n 8. Merrill, J. A. Spontaneous endometriosis in the Kenya baboon (Papio doguera). Am. J. Obstet. Gynecol. 101, 569–570. https:// doi. \norg/ 10. 1016/ 0002- 9378(68) 90572-3 (1968).\n 9. Ahn, S. H. et al. Immune-inflammation gene signatures in endometriosis patients. Fertil. Steril. 106, 1420-1431.e1427. https:// doi. \norg/ 10. 1016/j. fertn stert. 2016. 07. 005 (2016).\n\n12\nVol:.(1234567890)Scientific Reports |         (2022) 12:1590  | https://doi.org/10.1038/s41598-022-05499-y\nwww.nature.com/scientificreports/\n 10. Symons, L. K. et al. The immunopathophysiology of endometriosis. Trends Mol. Med. 24, 748–762. https:// doi. org/ 10. 1016/j. \nmolmed. 2018. 07. 004 (2018).\n 11. Guo, S.-W . Recurrence of endometriosis and its control. Hum. Reprod. Update  15, 441–461. https:// doi. org/ 10. 1093/ humupd/ \ndmp007 (2009).\n 12. Braundmeier, A. et al. Induction of endometriosis alters the peripheral and endometrial regulatory T cell population in the non-\nhuman primate. Hum. Reprod. 27, 1712–1722. https:// doi. org/ 10. 1093/ humrep/ des083 (2012).\n 13. Osuga, Y . et al. Lymphocytes in endometriosis. Am. J. Reprod. Immunol. 65, 1–10. https:// doi. org/ 10. 1111/j. 1600- 0897. 2010. 00887.x \n(2011).\n 14. Le, N. X. H. et al. Alteration of systemic and uterine endometrial immune populations in patients with endometriosis. Am. J. \nReprod. Immunol. 85, e13362. https:// doi. org/ 10. 1111/ aji. 13362 (2021).\n 15. Macpherson, A. J., Martinic, M. M. & Harris, N. The functions of mucosal T cells in containing the indigenous commensal flora \nof the intestine. Cell. Mol. Life Sci. 59, 2088–2096. https:// doi. org/ 10. 1007/ s0001 80200 009 (2002).\n 16. Cregger, M. et al. Reproductive microbiomes: Using the microbiome as a novel diagnostic tool for endometriosis. Reprod. Immunol. \nOpen Access https:// doi. org/ 10. 21767/ 2476- 1974. 100036 (2017).\n 17. Mazmanian, S. K., Liu, C. H., Tzianabos, A. O. & Kasper, D. L. An immunomodulatory molecule of symbiotic bacteria directs \nmaturation of the host immune system. Cell 122, 107–118. https:// doi. org/ 10. 1016/j. cell. 2005. 05. 007 (2005).\n 18. Round, J. L. et al. The Toll-like receptor 2 pathway establishes colonization by a commensal of the human microbiota. Science 332, \n974–977. https:// doi. org/ 10. 1126/ scien ce. 12060 95 (2011).\n 19. Ivanov, I. I. et al. Induction of intestinal Th17 cells by segmented filamentous bacteria. Cell 139, 485–498. https:// doi. org/ 10. 1016/j. \ncell. 2009. 09. 033 (2009).\n 20. Gaboriau-Routhiau, V . et al. The key role of segmented filamentous bacteria in the coordinated maturation of gut helper T cell \nresponses. Immunity 31, 677–689. https:// doi. org/ 10. 1016/j. immuni. 2009. 08. 020 (2009).\n 21. Smith, S. B. & Ravel, J. The vaginal microbiota, host defence and reproductive physiology. J. Physiol. 595, 451–463. https:// doi. org/ \n10. 1113/ JP271 694 (2017).\n 22. Round, J. L. & Mazmanian, S. K. The gut microbiota shapes intestinal immune responses during health and disease. Nat. Rev. \nImmunol. 9, 313–323. https:// doi. org/ 10. 1038/ nri25 15 (2009).\n 23. Palm, N. W ., de Zoete, M. R. & Flavell, R. A. Immune-microbiota interactions in health and disease. Clin. Immunol. 159, 122–127. \nhttps:// doi. org/ 10. 1016/j. clim. 2015. 05. 014 (2015).\n 24. Tarokh, M. et al. Serum and peritoneal fluid cytokine profiles in infertile women with endometriosis. Iran J. Immunol. 16, 151–162. \nhttps:// doi. org/ 10. 22034/ IJI. 2019. 80258 (2019).\n 25. Gogacz, M. et al. Increased percentage of Th17 cells in peritoneal fluid is associated with severity of endometriosis. J. Reprod. \nImmunol. 117, 39–44. https:// doi. org/ 10. 1016/j. jri. 2016. 04. 289 (2016).\n 26. Fazleabas, A. T., Braundmeier, A. & Parkin, K. Endometriosis-induced changes in regulatory T cells: Insights towards developing \npermanent contraception. Contraception 92, 116–119. https:// doi. org/ 10. 1016/j. contr acept ion. 2015. 06. 006 (2015).\n 27. Le, N. X. H. et al. Alteration of systemic and uterine endometrial immune populations in patients with endometriosis. Am. J. \nReprod. Immunol. 85, e13362. https:// doi. org/ 10. 1111/ aji. 13362 (2020).\n 28. Berbic, M. et al. The role of Foxp3+ regulatory T-cells in endometriosis: A potential controlling mechanism for a complex, chronic \nimmunological condition. Hum. Reprod. 25, 900–907. https:// doi. org/ 10. 1093/ humrep/ deq020 (2010).\n 29. Liang, Y ., Wu, J., Wang, W ., Xie, H. & Y ao, S. Pro-endometriotic niche in endometriosis. Reprod. Biomed. Online 38, 549–559. \nhttps:// doi. org/ 10. 1016/j. rbmo. 2018. 12. 025 (2019).\n 30. Blander, J. M., Longman, R. S., Iliev, I. D., Sonnenberg, G. F . & Artis, D. Regulation of inflammation by microbiota interactions \nwith the host. Nat. Immunol. 18, 851–860. https:// doi. org/ 10. 1038/ ni. 3780 (2017).\n 31. Baker, J. M., Al-Nakkash, L. & Herbst-Kralovetz, M. M. Estrogen–gut microbiome axis: Physiological and clinical implications. \nMaturitas 103, 45–53. https:// doi. org/ 10. 1016/j. matur itas. 2017. 06. 025 (2017).\n 32. Leonardi, M., Hicks, C., El-Assaad, F ., El-Omar, E. & Condous, G. Endometriosis and the microbiome: A systematic review. BJOG \nInt. J. Obstet. Gynaecol. 127, 239–249. https:// doi. org/ 10. 1111/ 1471- 0528. 15916 (2020).\n 33. Smith, B. C. et al. The cervical microbiome over 7 years and a comparison of methodologies for its characterization. PLoS ONE 7, \ne40425. https:// doi. org/ 10. 1371/ journ al. pone. 00404 25 (2012).\n 34. Cox, A. J., West, N. P . & Cripps, A. W . Obesity, inflammation, and the gut microbiota. Lancet Diabetes Endocrinol. 3, 207–215. \nhttps:// doi. org/ 10. 1016/ S2213- 8587(14) 70134-2 (2015).\n 35. Muscogiuri, G. et al. Gut microbiota: A new path to treat obesity. Int. J. Obes. Suppl. 9, 1. https:// doi. org/ 10. 1038/ s41367- 019- 0011-7 \n(2019).\n 36. Ahn, J. et al. Human gut microbiome and risk for colorectal cancer. J. Natl. Cancer Inst. 105, 1907–1911. https:// doi. org/ 10. 1093/ \njnci/ djt300 (2013).\n 37. Scher, J. U. & Abramson, S. B. The microbiome and rheumatoid arthritis. Nat. Rev. Rheumatol. 7, 569–578. https:// doi. org/ 10. 1038/ \nnrrhe um. 2011. 121 (2011).\n 38. Picard, C. et al. Review article: Bifidobacteria as probiotic agents—physiological effects and clinical benefits. Aliment. Pharmacol. \nTher. 22, 495–512. https:// doi. org/ 10. 1111/j. 1365- 2036. 2005. 02615.x (2005).\n 39. Baba, N., Samson, S., Bourdet-Sicard, R., Rubio, M. & Sarfati, M. Commensal bacteria trigger a full dendritic cell maturation \nprogram that promotes the expansion of non-Tr1 suppressor T cells. J. Leukoc. Biol. 84, 468–476. https:// doi. org/ 10. 1189/ jlb. 01080 \n17 (2008).\n 40. Donaldson, G. P ., Lee, S. M. & Mazmanian, S. K. Gut biogeography of the bacterial microbiota. Nat. Rev. Microbiol.  14, 20–32. \nhttps:// doi. org/ 10. 1038/ nrmic ro3552 (2016).\n 41. Corrêa-Oliveira, R., Fachi, J. L., Vieira, A., Sato, F . T. & Vinolo, M. A. Regulation of immune cell function by short-chain fatty \nacids. Clin. Transl. Immunol. 5, e73. https:// doi. org/ 10. 1038/ cti. 2016. 17 (2016).\n 42. Kaiko, G. E. et al. The colonic crypt protects stem cells from microbiota-derived metabolites. Cell 165, 1708–1720. https:// doi. org/ \n10. 1016/j. cell. 2016. 05. 018 (2016).\n 43. Cregger, M. A. et al. The Populus holobiont: Dissecting the effects of plant niches and genotype on the microbiome. Microbiome 6, \n31. https:// doi. org/ 10. 1186/ s40168- 018- 0413-8 (2018).\n 44. Callahan, B. J. et al. DADA2: High-resolution sample inference from Illumina amplicon data. Nat. Methods 13, 581–583. https:// \ndoi. org/ 10. 1038/ nmeth. 3869 (2016).\n 45. Kuczynski, J. et al. Using QIIME to analyze 16S rRNA gene sequences from microbial communities. Curr. Protoc. Bioinform. https:// \ndoi. org/ 10. 1002/ 04712 50953. bi100 7s36 (2011).\n 46. Caporaso, J. G. et al. QIIME allows analysis of high-throughput community sequencing data. Nat. Methods  7, 335–336. https://  \ndoi. org/ 10. 1038/ nmeth.f. 303 (2010).\nAcknowledgements\nThe authors thank Ms. Samantha Hrbek for her assistance with samples collection; Dr. Carrel for her assistance \nwith the microbiome analysis. M. Cregger was funded by the Laboratory Director Research and Development \n\n13\nVol.:(0123456789)Scientific Reports |         (2022) 12:1590  | https://doi.org/10.1038/s41598-022-05499-y\nwww.nature.com/scientificreports/\nprogram at Oak Ridge National Laboratory. Funding was provided by National Institutes of Health (Grant No. \nNIH-NICHD) and School of Medicine, Southern Illinois University, (Grant No. ADR-SIU-RSG).\nAuthor contributions\nN.X.H.L. was involved in experimental design, and undertook the laboratory work, analysis and manuscript \npreparation. M.C. assisted with the microbiome analysis. A.F . was involved in experimental design and execution \nof the manuscript. A.B. was involved in funding, experimental design, execution and manuscript preparation. \nAll authors revised, edited and approved the manuscript.\nCompeting interests \nThe authors declare no competing interests.\nAdditional information\nSupplementary Information The online version contains supplementary material available at https:// doi. org/ \n10. 1038/ s41598- 022- 05499-y.\nCorrespondence and requests for materials should be addressed to A.B.-F .\nReprints and permissions information is available at www.nature.com/reprints.\nPublisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and \ninstitutional affiliations.\nOpen Access  This article is licensed under a Creative Commons Attribution 4.0 International \nLicense, which permits use, sharing, adaptation, distribution and reproduction in any medium or \nformat, as long as you give appropriate credit to the original author(s) and the source, provide a link to the \nCreative Commons licence, and indicate if changes were made. The images or other third party material in this \narticle are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the \nmaterial. If material is not included in the article’s Creative Commons licence and your intended use is not \npermitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from \nthe copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.\n© The Author(s) 2022","source_license":"CC0","license_restricted":false}