Resveratrol alleviates endometriosis via restoration of hypoxic milieu and inhibition of autoantibody deposition Authors: Yi Luo, Mio Kobayashi, Di Zhang, Yosuke Ono, Osamu Yoshino, Masahiko Kanehira Published: 2026 ABSTRACT Endometriosis, a chronic disease that affects about 10% of women of reproductive age, is characterized by severe pain and growth of endometrium-like tissues outside the uteri. Although several theories have been advanced for the cause of endometriosis, no theory fully explains why endometriosis occurs. For this study, we strove to elucidate a new mechanism of endometriosis leading to establishment of a new therapeutic strategy for its treatment. In our murine endometriosis model, a marked hypoxic state was induced in the peritoneal cavity, as indicated by increased expression of HIF-1α in peritoneal exudate cells. Moreover, innate immune response, such as TLR4 and IL-1β expression, was upregulated. To restore the hypoxic milieu in endometriosis, we chose resveratrol (RSV), a natural polyphenolic compound, as a hypoxia-reducing agent. As expected, RSV administration decreased the number of HIF-1α–positive cells. Furthermore, RSV administration reduced the size and weight of endometriotic lesions significantly. However, RSV administration did not change TLR4 and IL-1β expression levels in peritoneal exudate cells. In the process of seeking mechanisms of endometriosis other than TLR4-dependent and IL-1β-dependent pathways, we detected autoantibody deposition in endometriotic lesions. In endometriotic lesions from RSV-administered mice, accumulation of autoantibodies was inhibited compared to a control group. As described herein, we proposed a new concept by which autoantibody deposition can cause endometriosis, raising the possibility of RSV as a promising therapeutic option for endometriosis. Introduction Endometriosis, a prevalent gynecological disorder, is characterized by the implantation of endometrial tissue outside the uterus, such as the ovaries, fallopian tubes, pelvic peritoneum, bladder, or the large or small intestine. These ectopic endometrium-like tissues grow in response to estrogen during the menstrual cycle, similarly to the normal endometrium. The growth can cause recurring pelvic pain, painful periods, and dyspareunia. Moreover, it is associated with infertility [1]. Earlier studies have suggested that endometriosis affects approximately 10% of women of reproductive age, and that 25–50% of women with endometriosis are infertile [2]. Although the endometriosis pathogenesis remains unclear, the most widely accepted model is that menstrual blood flows back into the abdominal cavity during menstruation and that endometrial cells are subsequently implanted to the surface of adjacent organs. Other mechanisms have been suggested as well: e.g., coelomic metaplasia, immune dysregulation, genetic causes, and environmental and lifestyle factors [3]. Numerous studies have demonstrated estrogen as a main driver of the progression, inflammation, and pain of endometriosis. Therefore, progesterone administration is used currently as the most common regimen for endometriosis because of its beneficial therapeutic effects and limited side effects [4]. However, in some cases, patients with endometriosis display progesterone resistance, which might result from the unresponsiveness of the endometrium to progesterone [5]. For this study, we strove to elucidate hidden mechanism(s) of endometriosis and to identify new therapeutic target(s) in addition to an estrogen-dependent pathway. In the peritoneal milieu, resident cells of several types including macrophages, lymphocytes, and mesothelial cells help to regulate the fine-tuned balance between anti-inflammatory and pro-inflammatory immune responses and homeostasis. Especially, peritoneal resident macrophages are crucially important for clearance of pathogens and cellular debris originating from necrotic or apoptotic cells. Endometrial tissues disseminated to the peritoneal cavity are known to be eliminated by peritoneal macrophages and the cooperation of immune cells such as natural killer cells [6, 7]. Because the most widely accepted etiology of endometriosis is that endometrial tissues fail to be eliminated because of the dysfunction of immune systems [8], we have inferred that rebalancing the dysregulated immune system is a promising therapeutic approach for endometriosis. We have emphasized resveratrol (RSV), a plant-derived polyphenol that is found originally in berries, peanuts, wine, and the seeds of grapes. The plant-derived polyphenol RSV has been reported widely to have many biological effects such as anti-inflammatory, anti-oxidant, anti-angiogenic, and anti-cancer activity effects, and regulation of lipid metabolism [9, 10]. With our murine model of endometriosis, we show that systemic administration of RSV improves hypoxic state in peritoneal cavity, and that RSV inhibits autoantibody deposition against endometrial tissues. Our findings highlight RSV as a promising option for women with endometriosis. Materials and Methods Animals Seven-week-old to ten-week-old female BALB/c mice and C57BL/6J mice were purchased from Japan SLC, Inc. (Shizuoka, Japan). Green fluorescent protein (GFP) transgenic mice (C57BL/6 background) were kindly provided by Dr. M. Okabe (Osaka University, Osaka, Japan). All mice were housed in SPF facilities under controlled conditions of temperature (22−26°C), humidity (approximately 60%), and light (12 h dark/light cycle), with free access to food and water. All procedures involving animals, including housing and care, methods of euthanasia, and experiment protocols, were approved by the Animal Care Committee of the University of Yamanashi (Approval no.: A6-2). In vivo study using a murine model of endometriosis Donor female BALB/c mice were injected subcutaneously with β-estradiol (Merck KGaA, Darmstadt, Germany) suspended in sesame oil (non-roasted) (Fujifilm Wako Pure Chemical Corporation, Osaka, Japan) at a dose of 0.2 µg/mouse on days 14 and 7 before the day of collection of uteri (day 0). On day 0, the mice were deeply anesthetized using isoflurane (Fujifilm Wako Pure Chemical Corp.). After whole blood was collected from the femoral artery, it was transferred to a 1.5 ml Eppendorf tube added with 1.0 µl of 0.5 M EDTA. Then, uteri were collected aseptically and were transferred to plastic dish added with PBS (0.4 ml/uterus) and placed on ice. The uteri were cut into small pieces (approximately 1–2 mm diameter) with scissors and a scalpel blade. Concurrently, recipient female BALB/c mice were anesthetized with a mixture of medetomidine, midazolam, and butorphanol at respective doses of 0.3 mg/kg, 4.0 mg/kg, and 5.0 mg/kg. The minced uteri were mixed with blood at a ratio of 1:2. Then 0.6 ml of slurry was injected into the peritoneal cavity of the anesthetized recipient mice with a 1.0 ml syringe and 18 G needle. The control group received an equal volume of PBS instead. Six days after injection, all mice were euthanized by bleeding whole blood under deep anesthesia with isoflurane. The abdominal skin was incised and inspected for endometriotic lesions, which were collected and weighed. Thereafter, they were cut into two pieces. Half of the tissue was fixed in 4% paraformaldehyde (Fujifilm Wako Pure Chemical Corp.) for 24 h at 4°C for histopathological assay. The other half was stored at −80°C for additional RNA extraction. Next, to examine the endometriotic tissue origin, endometriosis was induced using C57BL/6 background GFP+-mice and syngeneic GFP−-mice (C57BL/6) respectively as donors and recipients (and vice versa). Peripheral blood for preparing the slurry of uteri was derived from GFP−-mice. Six days after transplantation, all mice were euthanized by bleeding whole blood under deep anesthesia. Abdominal skin was incised. Fluorescence emitted from endometriotic tissue was detected using an IVIS Lumina X5 System and Living Image software (PerkinElmer, Inc., Shelton, CT, USA). Histopathological assay After the fixed endometrial tissues were embedded in paraffin blocks, they were sectioned at 5 µm thickness using a microtome to prepare tissue slides. After deparaffinization and subsequent rehydration, tissue slides were stained with hematoxylin and eosin (H&E) or Elastica-Masson. For immunostaining, the slides were deparaffinized, rehydrated and heated for 10 min at 95–100°C in 10 mM sodium citrate buffer (pH 6.0) to recover antigenicity. After recovering antigenicity, the slides were blocked with a blocking solution consisting of 0.05% (v/v) Tween 20 (Fujifilm Wako Pure Chemical Corp.) in Tris-buffered saline (TBS-T) supplemented with 2% (v/v) goat serum (Jackson ImmunoResearch Laboratories, Inc., West Grove, PA, USA) for 1 h at room temperature with gentle rocking. Then the sections were reacted with a Goat anti-mouse IgG+IgM (H+L) antibody conjugated with fluorescein isothiocyanate (FITC) (LGC Clinical Diagnostics, Inc., Milford, MA. USA) diluted to 1:200 in blocking solution for 24 h at 4°C. After washing with TBS-T, the sections were covered with mounting medium containing 4’,6-diamidino-2-phenylindole (DAPI; Vector Laboratories, Inc., Newark, CA, USA) and were observed using a fluorescence microscope (BZ-X800L) and viewer software (BZ-X800; Keyence Corp., Osaka, Japan). To detect expression of estrogen receptor α, fixed endometrial tissues were embedded in OCT compound (Sakura Finetek Japan Co., Ltd., Tokyo, Japan), and were frozen at −80°C. Using a cryostat, cryosections were prepared from the frozen blocks at 20 µm thickness. Next, the slides were blocked with blocking solution consisting of TBS-T supplemented with 2% (v/v) goat serum (Jackson ImmunoResearch Laboratories, Inc.) for 1 h at room temperature with gentle rocking. Thereafter, the sections were incubated with anti-Estrogen Receptor alpha antibody (Bioss Inc. Woburn, MA, USA) diluted to 1:200 in blocking solution at 4°C overnight. After washing with TBS-T, the slides were reacted with a goat anti-rabbit IgG horseradish peroxidase (HRP) conjugated antibody (Bio-techne Corp., Minneapolis, MN, USA) for 1 h. The slides were stained with 3,3′-diamino benzidine (DAB) (ScyTek Laboratories, Inc., Logan, UT, USA). Treatment with RSV First, RSV (Sigma-Aldrich Inc., St. Louis, MO, USA) was dissolved in 100% ethanol at 22.8 mg/ml (ca. 100 mM) concentration to create a stock solution, which was stored at −80°C. Before use, the stock solution was diluted to 4 mg/ml with PBS. Two days after induction of endometriosis, all mice were divided randomly into RSV-treated (n=5) and control (n=5) groups. Mice for RSV-treated group were administered 20 mg/kg of RSV intraperitoneally (i.p.). For control group mice, an equivalent amount of PBS was administered instead of RSV. Six days later, all mice were euthanized and were inspected for endometriotic lesions as described above. Collection of peritoneal exudate cells Peritoneal exudate cells were collected from healthy mice, and from mice in RSV-treated and control groups. Briefly, after euthanization, 1.0 ml of ice-cold PBS was injected into the peritoneal cavity using a 1.0 ml syringe and an 18 G needle. The lavage fluid was collected. The procedure was conducted again to obtain 2 ml of lavage fluid. After the total cell number was ascertained, 20 µl of the lavage fluid was placed on a slide glass, air-dried, and stored at −80°C for additional analysis. The lavage fluid was centrifuged at 3,000 × g for 5 min. The cell precipitates were used for RNA analyses. RT-PCR Total RNA from uterus and endometriotic tissues was purified using Sepasol RNA I Super G (Nacalai Tesque, Inc., Kyoto, Japan) according to the manufacturer’s instructions. Then 2 µg of total RNA was reverse-transcribed to cDNA using ReverTra Ace (Toyobo Co., Ltd., Osaka, Japan) combined with primers annealed to the downstream mRNAs for carbohydrate antigen 125 (CA125) (Muc16) and estrogen receptor alpha (Esr1) according to the manufacturer’s instructions. The following primers were used for this study: Muc16, 5′-GACCCAAATGTTACATCAGGAG-3′; Esr1, 5′-GGCTTGCTGTTGTCCACGTATA -3′. The generated cDNA was amplified by PCR using thermal cycler (GeneAmp PCR System 9700; Thermo Fisher Scientific Inc.,Waltham, MA, USA) and Takara ExTaq DNA polymerase (Takara Bio Inc., Japan) for 35 cycles at 95°C for 30 s, 56°C for 30 s, and 72°C for 30 s. The products were resolved on a 2% agarose gel and were detected using ethidium bromide staining. The following primer pairs were used for this study: Muc16 forward, 5′-TAAAAGCAAAGCACAGCACAGC-3′, reverse, 5′-GTATTGGTCTGAGCACTGAAGT-3′; and Esr1 forward, 5′-CCATTGTCTAGCTTAAGAAGGC-3′, reverse, 5′-AAGGGTCATGGT CATGGTAAGT-3′. Quantitative RT-PCR (qRT-PCR) Total RNA from peritoneal exudate cells was purified using RNeasy Plus Mini kit (Qiagen Inc., Germany) according to the manufacturer’s instruction. Then 350 ng of total RNA was reverse-transcribed to complementary DNA (cDNA) using ReverTra Ace (Toyobo Co., Ltd.) combined with primers annealed to the downstream mRNAs for IL-6 (Il6), Tumor necrosis factor-α (Tnfa), IL-1β (Il1b), Toll-like receptor 4 (Tlr4), Vascular endothelial growth factor A (Vegfa), and beta-2-microglobulin (B2m) according to the manufacturers’ instructions. The following primers were used for this study: Il6, 5′-CCGAGTAGATCTCAAAGTGACT-3′; Tnfa, 5′-GAGGTTGACTTTCTCCTGGTAT-3′; Il1b, 5′-TGCCTGAAGCTCTTGTTGATGT-3′; Tlr4, 5′-GGATAAATCCAGCCACTGAAGT-3′; Vegfa, 5′-GAAGATGTACTCTATCTCGTCG-3′; and B2m, 5′-TGAATTCAGTGTGAGCCAGGAT-3′. Quantitative RT-PCR was performed using a StepOne Plus Real Time PCR System (Thermo Fisher Scientific Inc.) using Thunderbird SYBR qPCR mix (Toyobo Co., Ltd.) according to the manufacturers’ instructions. The primer sequences used for this study were the following: Il6 forward, 5′-ACAAAGCCAGAGTCCTTCAGAGA-3′, reverse, 5′-TCCTCTCGTAACCTTTAACCCCA-3′; Tnfa forward, 5′-TAGCCCACGTCGTAGCAAAC-3′, reverse, 5′-TGTTCCATGTTGGGTAGCCG-3′; Il1b forward, 5′-TTCCTTGTGCAAGTGTCTGAAG-3′, reverse, 5′-ATCTTTTGGGGTCCGTCAACTT-3′; Tlr4 forward, 5′-AGGAAGCTTGAATCCCTGCATA-3′, reverse, 5′-TGAAGAAGGAATGTCATGAGGG-3′; Vegfa forward, 5′-CATGAACTTTCTGCTCTCTTGG-3′, reverse, 5′-ATCACTTCATGGGACTTCTGCT-3′; and B2m forward, 5′-TCTTTCTGGTGCTTGTCTCACT-3′, reverse, 5′-AGCAGTTCAGTATGTTCGGCTT-3′. Ct was calculated automatically using StepOne software (Thermo Fisher Scientific Inc.). All data were normalized against Ct for beta-2-microglobulin as a housekeeping gene and were calculated using the 2−ΔΔCt method. Collagen assay After endometriotic tissues of mice from RSV-treated and control groups were collected and weighed, the total collagen content of each tissue was found using Sirius Red Total Collagen Detection Assay Plate Kit (Chondrex Inc., Woodinville, WA, USA) according to the manufacturer’s instructions. The contents of total collagen in samples were normalized with the weights of the respective tissues. Absorbance at 530 nm was found using a microplate reader (SpectraMax 340; Molecular Devices, LLC, San Jose, CA, USA). Giemsa staining and immunofluorescence staining Peritoneal exudate cells placed on a slide glass were stained with staining of blood smears (Hemacolor Rapid; Merck KGaA, Darmstadt, Germany) according to the manufacturer’s instructions. The percentages of cell types were found. Also, the slide glasses were stained for HIF-1α, a marker of hypoxia. Briefly, the cells on slide glasses were fixed and permeabilized for 5 min at room temperature in methanol (Fujifilm Wako Pure Chemical Corp.). Then slides were washed with TBS-T and blocked with blocking solution consisting of TBS-T added with 2% (w/v) bovine serum albumin (Sigma-Aldrich Inc.) for 1 h at room temperature. Next, slides were reacted with a mAb anti-HIF-1 alpha antibody [DyLight488] (Bio-techne Corp.) diluted to 1:200 in blocking solution for 24 h at 4°C. After washing with TBS-T, the slides were covered with mounting media containing 4’,6-diamidino-2-phenylindole (DAPI; Vector Laboratories, Inc.). They were observed using fluorescence microscopy (BZ-X800L) and analyzer software (BZ-X800; Keyence Corp.). The positivity of HIF-1α was calculated as the ratio of HIF-1α-positive nuclei to total nuclei. Statistical analysis All data are presented as the mean and standard deviation (SD) of the indicated numbers of samples and replicates. Statistical analyses were conducted (Microsoft Excel; Microsoft Corp., Redmond, WA, USA) or Statcel 4 software (OMS Publishing Inc., Saitama, Japan). Significance was inferred based on Student t-test results or one-way ANOVA followed by Tukey-Kramer test. P values of <0.05 were inferred as significant. Results A murine endometriosis model established for this study faithfully reproduces the clinical phenotypes of human endometriosis First, we intended to establish a murine endometriosis model that is comparable to human endometriosis in terms of etiology and pathology. Because differences in female reproductive cycles exist between human and rodents, such as the estrogen and progesterone concentrations in a non-pregnant state, we incorporated them into consideration. As presented in Fig. 1AFig. 1.Establishment of a mouse model of endometriosis. A. Scheme of endometriosis induction. B. Representative image of endometriotic lesion (left, red square) and macroscopic finding of endometriotic tissue (inset). Scale bar in inset: 5 mm. Representative image of H&E staining of endometriotic tissue (right). The area to the right of the dashed line is ectopic endometrial tissue consisting of stromal cells and endometrial glands (denoted as asterisks). Scale bar: 200 µm. C. Representative image of immunohistochemical staining of endometriotic tissue for estrogen receptor α (ERα). Scale bar: 100 µm. D. RT-PCR analysis for detection of mRNA for MUC16 (Muc16) and ERα (Esr1) in endometriotic tissues (EM). Normal uterus tissues (UT) were used as controls. E. Identification of endometriotic tissue sources. Uterus tissue from GFP transgenic mice (C57BL/6 background) was implanted into isogenic C57BL/6 recipient (GFP to B6), and vice versa (B6 to GFP). Red squares represent macroscopic endometriotic lesions. White squares represent the area corresponding to red square under fluorescent imaging., donor mice were injected with β-estradiol once a week for 2 weeks to induce uterine endometrium growth. Thereafter, all donor mice were euthanized. Whole blood and uteri were collected aseptically. Then recipient syngeneic mice were anesthetized. Slurry consisting of minced uteri and blood was transplanted intraperitoneally. All recipient mice were euthanized and were inspected on day 6 after transplantation. As depicted in Fig. 1B, tissue mass adhered to the abdominal wall was observed. In addition, H&E staining of the tissue specimen revealed the presence of typical stromal tissues and endometrial glands. Furthermore, immunohistochemical analysis showed the expression of estrogen receptor α (ERα) in the specimen (Fig. 1C). Consistently, RT-PCR analysis showed expression of mRNAs for Mucin 16 (Muc16) and estrogen receptor α (Esr1), which are also known as potential biomarkers for diagnosis of endometriosis, in the specimen (Fig. 1D). Based on these observations, the tissue specimen formed for this study was diagnosed pathologically as endometriosis. Furthermore, we tried to identify the source of endometriotic lesions. As shown in Fig. 1E, the endometriotic lesion did not emit fluorescent light when C57BL/6 mouse and GFP-transgenic mice were used, respectively, as donor and recipient. By contrast, the endometriotic lesion did emit fluorescence light when GFP-transgenic mice and C57BL/6 mice were used respectively as the donor and recipient. This result suggests that endometriotic lesions formed in our mouse model consist mainly of donor-derived uterus tissues. Establishment of a mouse model of endometriosis. A. Scheme of endometriosis induction. B. Representative image of endometriotic lesion (left, red square) and macroscopic finding of endometriotic tissue (inset). Scale bar in inset: 5 mm. Representative image of H&E staining of endometriotic tissue (right). The area to the right of the dashed line is ectopic endometrial tissue consisting of stromal cells and endometrial glands (denoted as asterisks). Scale bar: 200 µm. C. Representative image of immunohistochemical staining of endometriotic tissue for estrogen receptor α (ERα). Scale bar: 100 µm. D. RT-PCR analysis for detection of mRNA for MUC16 (Muc16) and ERα (Esr1) in endometriotic tissues (EM). Normal uterus tissues (UT) were used as controls. E. Identification of endometriotic tissue sources. Uterus tissue from GFP transgenic mice (C57BL/6 background) was implanted into isogenic C57BL/6 recipient (GFP to B6), and vice versa (B6 to GFP). Red squares represent macroscopic endometriotic lesions. White squares represent the area corresponding to red square under fluorescent imaging. In endometriosis, the peritoneal milieux under hypoxic and innate immune responses are evoked via upregulation of HIF-1a Next, we strove to elucidate the unknown etiologies of endometriosis. We collected peritoneal exudate cells (PECs) by lavaging peritoneal cavities of healthy mice and mice into which endometriosis had been induced. As Fig. 2AFig. 2.Phenotypic characteristics of peritoneal exudate cells in endometriosis. A. Representative images of Wright–Giemsa staining of peritoneal exudate cells collected from healthy and endometriosis mice (upper). Scale bar: 100 µm. The percentage (lower left) and the total number (lower right) of peritoneal exudate cells collected from healthy BALB/c mice and BALB/c mice induced endometriosis. The bar graph shows mean ± SD. B. Representative images of immunofluorescence staining of peritoneal exudate cells collected from healthy and endometriosis mice for HIF-1α (upper). Scale bar: 200 µm. The ratio of the number of HIF-1α -positive nucleus to the total number of nuclei in the section (lower) is shown. The bar graph shows mean ± SD. C. Quantitative RT-PCR analysis for TLR4 (Tlr4) and VEGF-A (Vegfa) mRNAs in peritoneal exudate cells collected from healthy and endometriosis mice. N.D. denotes not detected. The bar graph presents mean ± SD. D. Quantitative RT-PCR analysis for IL-1β (Il1b), IL-6 (Il6), and TNF-α (Tnfa) mRNAs in peritoneal exudate cells collected from healthy and endometriosis mice. N.D. denotes not detected. The bar graph shows mean ± SD. Significance was inferred from results of Student’s two-tailed t-tests. Abbreviations: Lym, lymphocyte; Mph, macrophage; Neu, neutrophil. shows, both microscopic findings and population of PECs of endometriosis mice were almost identical to those of healthy mice. The number of PECs of endometriosis mice was about 70% less than that of healthy mice. Because peripheral blood was also transplanted with uterine tissue in our endometriosis model, we hypothesized that a hypoxic microenvironment would be induced in the peritoneal cavity by the oxidation of blood contents, especially iron in hemoglobin. To test the hypothesis, we stained PECs for HIF-1α, a well-known marker of hypoxia. As presented in Fig. 2B, almost 100% of PECs collected from endometriosis mice expressed HIF-1α in their nucleus compared to less than 5% of PECs from healthy mice. Next, we attempted to ascertain the factors linking hypoxia and endometriosis. Among many of the genes which are upregulated by HIF-1α, we monitored the expression levels of Toll-like receptor 4 (TLR4) and vascular endothelial growth factor-A (VEGF-A) genes in PECs. As Fig. 2C shows, Tlr4 gene was upregulated about 2.5 times in PECs from endometriosis mice compared to that from healthy mice, although Vegfa genes were below detectable levels in PECs collected from both mice. Furthermore, we monitored the expression levels of Il1b, Il6, and Tnfa genes, known as proinflammatory cytokines. Their expressions were increased by TLR4 activation. As Fig. 2D shows, Il1b gene was upregulated about six times in PECs from endometriosis mice compared to that from healthy mice, although Il6 and Tnfa genes were below detectable levels in PECs collected from both mice. Based on these results, one can infer that upregulation of HIF-1α followed by innate immune response via TLR4-IL-1β axis can be a potential inducer of endometriosis. Phenotypic characteristics of peritoneal exudate cells in endometriosis. A. Representative images of Wright–Giemsa staining of peritoneal exudate cells collected from healthy and endometriosis mice (upper). Scale bar: 100 µm. The percentage (lower left) and the total number (lower right) of peritoneal exudate cells collected from healthy BALB/c mice and BALB/c mice induced endometriosis. The bar graph shows mean ± SD. B. Representative images of immunofluorescence staining of peritoneal exudate cells collected from healthy and endometriosis mice for HIF-1α (upper). Scale bar: 200 µm. The ratio of the number of HIF-1α -positive nucleus to the total number of nuclei in the section (lower) is shown. The bar graph shows mean ± SD. C. Quantitative RT-PCR analysis for TLR4 (Tlr4) and VEGF-A (Vegfa) mRNAs in peritoneal exudate cells collected from healthy and endometriosis mice. N.D. denotes not detected. The bar graph presents mean ± SD. D. Quantitative RT-PCR analysis for IL-1β (Il1b), IL-6 (Il6), and TNF-α (Tnfa) mRNAs in peritoneal exudate cells collected from healthy and endometriosis mice. N.D. denotes not detected. The bar graph shows mean ± SD. Significance was inferred from results of Student’s two-tailed t-tests. Abbreviations: Lym, lymphocyte; Mph, macrophage; Neu, neutrophil. RSV, a hypoxia-reducing agent, alleviates endometriosis symptoms Our findings raised the possibility that restoration of a hypoxic microenvironment in the peritoneal cavity can alleviate endometriosis symptoms. RSV, a phytochemical derived from fruits and vegetables, attenuates hypoxic events as well as angiogenesis, inflammation, and cellular senescence. For endometriosis in mice, RSV was chosen as a candidate treatment option. As Fig. 3AFig. 3.Evaluation of RSV effects for the treatment of endometriosis. A. Representative images of endometriotic lesion treated with PBS (Control) or RSV (upper; red square) and macroscopic finding of endometriotic tissue (lower). Scale bar: 10 mm. Weight of collected endometriotic tissues (right). The bar graph shows mean ± SD. B. Representative images of Elastica–Masson staining of endometriotic tissues treated with PBS (Control) or RSV (upper). Scale bar: 100 µm. Quantification of total collagen in endometriotic tissues treated with PBS (Control) or RSV (lower). The bar graph shows mean ± SD. Significance was inferred from results of Student’s two-tailed t-tests. shows, intraperitoneal administration of RSV inhibited growth of endometriotic lesions compared to endometriosis mice treated with vehicle (Control), by almost half in terms of the lesions’ size and weight. As shown in Fig. 3B, endometriotic lesions from RSV-treated endometriosis mice contained fewer inflammatory cells and about three times the collagen compared to those of control endometriosis mice. From these results, it can be inferred that RSV administration halts progression of endometriosis by inhibition of inflammatory responses, and that it promotes fibrotic tissue formation as a result. Evaluation of RSV effects for the treatment of endometriosis. A. Representative images of endometriotic lesion treated with PBS (Control) or RSV (upper; red square) and macroscopic finding of endometriotic tissue (lower). Scale bar: 10 mm. Weight of collected endometriotic tissues (right). The bar graph shows mean ± SD. B. Representative images of Elastica–Masson staining of endometriotic tissues treated with PBS (Control) or RSV (upper). Scale bar: 100 µm. Quantification of total collagen in endometriotic tissues treated with PBS (Control) or RSV (lower). The bar graph shows mean ± SD. Significance was inferred from results of Student’s two-tailed t-tests. RSV is insufficient to decrease innate immune responses despite its anti-HIF-1a effect Next, we attempted to ascertain whether the RSV-induced alleviation of endometriosis can be attributed to restoration of hypoxic microenvironment in the peritoneal cavity. Although RSV administration changed neither microscopic findings nor populations of PECs in endometriosis mice, it recovered the number of PECs (Fig. 4AFig. 4.Phenotypic characteristics of peritoneal exudate cells in endometriosis treated with RSV. A. Representative images of Wright–Giemsa staining of peritoneal exudate cells collected from BALB/c mice induced endometriosis and treated with PBS (Control) or RSV (upper). Scale bar: 100 µm. The percentage (lower left) and the total number (lower right) of peritoneal exudate cells collected from endometriosis mice treated with PBS (Control) or RSV are shown. The bar graph shows mean ± SD. B. Representative images of immunofluorescence staining of peritoneal exudate cells collected from endometriosis mice treated with PBS (Control) or RSV for HIF-1α (upper). Scale bar: 200 µm. The ratio of the number of HIF-1α -positive nucleus to the total number of nuclei in the section (lower). The bar graph shows mean ± SD. C. Quantitative RT-PCR analysis for TLR4 (Tlr4) and IL-1β (Il1b) mRNAs in peritoneal exudate cells collected from endometriosis mice treated with PBS (Control) or RSV. The bar graph shows mean ± SD. Significance was inferred from results of Student’s two-tailed t-tests.). As shown in Supplementary Fig. 1, the population of PECs was consisting of CD19-positive B cells as the main population, and CD3ε-positive T cells, F4/80-positive macrophages, and Gr-1-positive granulocytes as the minor population in both healthy and endometriosis mice. As expected, RSV administration improved hypoxic state and decreased HIF-1α expression in the nuclei of PECs in endometriosis mice (Fig. 4B, and Supplementary Figs. 2A and B ) almost half of control. However, as Fig. 4C shows, RSV administration did not change the expression of either Tlr4 or Il1b genes in PECs in endometriosis mice. From these results, we inferred that RSV administration is insufficient to decrease innate immune response in endometriosis. Moreover, we inferred that there exist pathways other than the TLR4-IL-1β axis for deterioration of endometriosis. Phenotypic characteristics of peritoneal exudate cells in endometriosis treated with RSV. A. Representative images of Wright–Giemsa staining of peritoneal exudate cells collected from BALB/c mice induced endometriosis and treated with PBS (Control) or RSV (upper). Scale bar: 100 µm. The percentage (lower left) and the total number (lower right) of peritoneal exudate cells collected from endometriosis mice treated with PBS (Control) or RSV are shown. The bar graph shows mean ± SD. B. Representative images of immunofluorescence staining of peritoneal exudate cells collected from endometriosis mice treated with PBS (Control) or RSV for HIF-1α (upper). Scale bar: 200 µm. The ratio of the number of HIF-1α -positive nucleus to the total number of nuclei in the section (lower). The bar graph shows mean ± SD. C. Quantitative RT-PCR analysis for TLR4 (Tlr4) and IL-1β (Il1b) mRNAs in peritoneal exudate cells collected from endometriosis mice treated with PBS (Control) or RSV. The bar graph shows mean ± SD. Significance was inferred from results of Student’s two-tailed t-tests. RSV inhibits autoantibody deposition in endometriotic lesions Finally, we considered other pathways that deteriorate endometriosis in mice. It has been known that peritoneal cavity of endometriosis patients is in a proinflammatory state. Beside TLR4-IL-1b axis, an aberrant balance between proinflammatory and anti-inflammatory immune cell subset, such as M1/M2 macrophage and Th17/Treg, is thought to play a key role in endometriosis. In addition, some evidence have been accumulating that production of autoantibodies correlates to severity of endometriosis and, of note, circulating antibodies that react to endometrium or ovary were detected from endometriosis patients. B-1 cells are known as a population that resides in peritoneal and pleural cavities and which contributes to early defense against bacterial and viral infection. In addition, B-1 cells spontaneously produce natural antibodies that are reactive to multiple antigens with low affinities. Mainly, the class of these antibodies is IgM. Reportedly, some B-1 cells are implicated in the development of autoimmune diseases by producing autoantibodies. According to these reports, we strove to ascertain whether B-1 cells, and especially autoantibodies produced from them, are associated with deterioration of endometriosis in mice. It is particularly interesting that endometriotic lesions from endometriosis mice showed deposition of autoantibodies in the tissues (see Fig. 5Fig. 5.Decreased deposition of autoantibody in endometriotic tissues treated with RSV. A. Representative images of immunofluorescence staining of endometriotic tissues treated with PBS (Control) or RSV for anti-mouse IgM/IgG. Scale bar: 200 µm., left). However, administration of RSV drastically decreased the deposition of autoantibodies in endometriotic lesions (see Fig. 5, right). Consistently, although abundant CD5+/CD19+ B-1 cells were detected in endometriotic lesion from endometriosis mice (Supplementary Fig. 3, left), administration of RSV efficiently decreased the accumulation of B-1 cells in endometriotic lesion (Supplementary Fig. 3, right). We also strove to determine the level of total IgM in the peritoneal lavage from healthy mice, and from mice in RSV-treated and control groups. The administration of RSV successfully suppressed the level of total IgM in peritoneal lavage in endometriosis, though a little higher than that in healthy mice (Supplementary Fig. 4). These results suggest that autoantibody deposition is a novel etiology of endometriosis. The effect of RSV in alleviation of endometriosis can be attributed to inhibition of autoantibody production from B-1 cells. Decreased deposition of autoantibody in endometriotic tissues treated with RSV. A. Representative images of immunofluorescence staining of endometriotic tissues treated with PBS (Control) or RSV for anti-mouse IgM/IgG. Scale bar: 200 µm. Discussion The most widely accepted etiology of endometriosis is that reflux of endometrial cells and tissue fragments into peritoneal cavity occurs, with growth outside of the uterus with fibrosis, pain, and long-lasting inflammation [3, 11]. The disease onset and progression are attributed mainly to an aseptic inflammation and hormonal imbalance between estrogen and progesterone [12, 13]. For this study, we established a murine model of endometriosis that is comparable to human endometriosis. Using the model, we demonstrated that peritoneal cavities are hypoxic in endometriosis and that autoantibody deposition occurs in endometriotic lesions. First, in our model, HIF-1α, one marker of hypoxia, is expressed markedly in PECs of endometriosis mice. That expression is attributable to oxidation of blood contents, mainly iron in red blood cells, transplanted with endometrium, which reproduces menstrual blood [14]. Furthermore, the innate immune pathway (e.g., TLR4 and IL-1β) was activated. In this study, we could not identify a ligand for TLR4. There has been evidence that dysbiosis of intra-uterine flora could lead to endometriosis, as well as that menstrual blood of women with endometriosis is highly contaminated with Escherichia coli, a source of lipopolysaccharide (LPS) that can activate TLR4, compared with that with healthy women [15, 16]. In the future study, we would like to explore the relationship between dysregulation of intra-uterine flora and innate immune pathway in endometriosis. In our results, mRNAs for IL-6, TNF-α or Vegfa were not detected in PECs. Reportedly, not only IL-1β but also IL6 and Tnfa are upregulated in endometriotic tissues. A high concentration of VEGF-A is also detected in peritoneal lavage in endometriosis patients [17]. These discrepancies might derive from differences of the specimens which were used for analyses. Although PECs analyzed for our study consist mainly of lymphocytes and macrophages, endometrial tissues are composed from various cells such as fibroblasts, stromal cells, macrophages, and endothelial cells [18, 19]. Because few reports describe studies that specifically examine PECs in endometriosis, our results might provide new perspectives elucidating the etiology of endometriosis. Next, we evaluated the therapeutic effect on endometriosis of RSV, which is known as a phytochemical that exhibits anti-inflammatory, antioxidant, and anti-fibrotic properties [9, 10]. As expected, after application of RSV as a hypoxia-reducing agent, the treatment reduced the size and weight of endometriotic lesions in endometriosis mice. During the acute phase of endometriosis (e.g., 6 days), RSV treatment halted inflammation and promoted fibrosis with accompanying collagen deposition in endometriotic lesions. Collagen deposition in endometriotic lesions is known to cause adhesion with pelvic and abdominal cavities, which induce pain and infertility [20]. Whether fibrosis by RSV treatment could be a process of wound healing or a result of scar formation is still obscure. More precise analysis is needed. Moreover, evaluation of RSV effects during the chronic phase of endometriosis is necessary. Although RSV treatment can reverse HIF-1α upregulation in PECs, expression levels of mRNAs for Tlr4 and Il1b were unchanged. It remains uncertain whether upregulation of TLR4 and IL-1β in PECs is insufficient but necessary for endometriosis, but RSV alleviated endometriosis by mechanisms other than TLR4 and IL-1β. Reportedly, antibodies to endometrial antigens such as α-enolase, a glycolytic enzyme, and hormones have been detected in sera from endometriosis patients [21]. Autoantibodies which recognize p53, a tumor suppressor protein were detected most commonly in endometriosis patients [22]. Consistent with these reports, autoantibodies were deposited in endometriosis lesions in our model. Identification of the endogenous ligands recognized by autoantibodies persists as a challenge for future study. Among the cells which make up the population of PECs, B-1 cells are a unique subset of B lymphocytes, a distinct population to B-2 cells. Whereas B-2 cells produce antibodies through the cooperation of helper T-cells and antigen-presenting cells after several days, B-1 cells can produce antibodies within a day in the absence of prior exposure to foreign antigens [23, 24]. Surprisingly, some evidence has accumulated to demonstrate that autoantibodies generated from B-1 cells contribute to some autoimmune diseases (e.g., lupus nephritis and autoimmune arthritis) [25, 26]. In our model, because large amounts of autoantibodies were detected in endometriotic lesions in the early phase, these autoantibodies were thought to arise from B-1 cells. We showed that RSV administration decreased autoantibody deposition in endometriotic lesions. Although we did not show how RSV can inhibit generation of autoantibodies, a reasonable explanation is that RSV can inhibit the hyperresponsiveness of immune systems as reported earlier [27, 28]. Besides, it has been reported that RSV ameliorates disease severity of lupus-prone mice through reducing B cell numbers and serum autoantibodies via upregulation of FcγRIIB, an inhibitory Fc receptor [29]. In addition, RSV inhibits signaling from FcεRI, a Fc receptor for IgE, through suppression of IgE binding to FcεRI, and phosphorylation of Syk, a key tyrosine kinase for FcεRI signaling, in human basophils [30]. It has still been speculative that RSV exerts such effects toward B-1 cells, these points seem worthy of further investigation in future studies. Moreover, there are hurdles to overcome when RSV is used for clinical applications, such as dosage, route of administration and duration. There are some reports that the bioavailability of RSV is drastically decreased when administered orally, due to its low solubility, low pH in stomach, and rapid metabolism in liver. Meanwhile, the blood concentration of RSV increases rapidly when administered intravenously and may cause unexpected side effects. It has known that metabolized RSV is mainly turned into three states such like glucuronide, sulfate, and free. Among them, free RSV could be stabilized by binding with blood proteins like albumin and lipoproteins, and readily delivered into target tissues [31]. There is still room for consideration for clinical application of RSV in endometriosis. In summary, we demonstrated RSV as a novel candidate for the treatment of endometriosis. Thousands of patients have endometriosis that is unresponsive to pharmacological and hormonal therapies. Our findings are expected to open the possibility of RSV as an effective dietary supplement that improves the quality of life of many women with endometriosis-related pain and infertility. Conflict of Interest Author M.K. has received research support from Japan Tobacco Inc. Supplementary Material Supplement Files