{"paper_id":"f91a499b-58a6-406c-a069-40580735e982","body_text":"RESEARCH\nInflammation (2025) 48:4428–4442\nhttps://doi.org/10.1007/s10753-025-02335-z\nlining the uterus possesses the remarkable ability to regen -\nerate without scarring [2]. However, when the endometrium \nsustains damage, such as curettage, it can lead to fibrotic \nregeneration, resulting in conditions like intrauterine adhe -\nsions (IUA) [3, 4] and thin endometrium [5, 6]. The preva-\nlence of IUA varies depending on the type of injury, ranging \nfrom 16 to 24% in women undergoing pregnancy-related \ncurettage and 31% to 45% after hysteroscopic myomec -\ntomy, with an average recurrence rate of more than 30% \n[7, 8]. On the other hand, the prevalence of thin endome -\ntrium is 24–85 cases per 1,000 individuals [ 7, 9]. These \nregenerative disorders of the endometrium can contribute to \ninfertility and negatively impact individuals’quality of life. \nCurrently, clinical treatment for fibrotic endometrium pri -\nmarily involves surgical separation of adhesions, followed \nby adjuvant therapy with estrogens. While this approach \nis effective for most mild to moderate cases, it is prone to \nrecurrence in severe cases [4]. Therefore, exploring the cel-\nlular and molecular mechanisms underlying endometrial \nrepair after injury is crucial for the development of new \ntherapeutic strategies.\nThe initial stage of endometrium shedding is regarded as \nan inflammatory event [9, 10]. Analogous to the physiologic \nshedding of the endometrium, inflammation also arises in \nIntroduction\nThe human endometrium undergoes a monthly process of \nremodeling, shedding, and regeneration, commonly referred \nto as menstruation [ 1]. This multicellular, bi-layered tissue \n \r Guangfeng Zhao\nzhaoguangfeng@nju.edu.cn\nSunan Shen\nshensn@nju.edu.cn\nYayi Hou\nyayihou@nju.edu.cn\n1 The State Key Laboratory of Pharmaceutical Biotechnology, \nDivision of Immunology, Medical School, Nanjing \nUniversity, Nanjing 210093, China\n2 Department of Obstetrics and Gynecology, Nanjing Drum \nTower Hospital, Affiliated Hospital of Medical School, \nNanjing University, Nanjing 210093, China\n3 Jiangsu Key Laboratory of Molecular Medicine,  \nNanjing 210093, China\n4 Jiangsu International Laboratory of Immunity and \nMetabolism, The Department of Pathogenic Biology and \nImmunology, Xuzhou Medical University, Xuzhou  \n221004, China\nAbstract\nMacrophages play a significant role in the repair of endometrial injuries. While large peritoneal macrophages (LPMs) \nhave been reported to migrate to injured organs and repair tissues within the peritoneal cavity, their involvement in the \nrepair of injured endometrium remains unclear. In this study, we utilize a mouse model of endometrial injury that does not \ninvolve laparotomy, a procedure that typically results in a substantial loss of LPMs. Strikingly, we find that LPMs reach \nthe endometrium within 6 h post-modeling. By depleting or supplementing LPMs, our results reveal that these cells are \ncapable of engulfing dead cells in the endometrium and resolving inflammation. Additionally, we observe that the migra -\ntion efficiency of LPMs is enhanced with increased levels of 17β-estradiol (E2) in mice. In vitro assays further confirm \nthat E2 accelerates the migration of LPMs towards apoptotic endometrial stromal cells. Overall, our findings demonstrate \nthat LPMs rapidly migrate into injured endometrium in relation to E2 levels and facilitate the process of tissue repair.\nKeywords Large peritoneal macrophages · Endometrium · Endometrial injury · Endometrial repair · 17β-estradiol\nReceived: 21 March 2025 / Revised: 23 May 2025 / Accepted: 17 June 2025 / Published online: 2 August 2025\n© The Author(s) 2025\nLarge Peritoneal Macrophages Promote the Resolution of \nInflammation in Injured Endometrium\nJingman Li1 · Lijie Yin1 · Jiali Wang1 · Yuchen Pan4 · Chen Peng1 · Yue Dong1 · Sunan Shen1,3 · Yayi Hou1,3 · \nGuangfeng Zhao2\n1 3\n\nInflammation (2025) 48:4428–4442\nthe early phases of pathologic endometrial injury [ 11–13], \nleading to the infiltration of neutrophils and monocytes/\nmacrophages [ 14]. These cells carry out vital functions, \nincluding the clearance of cellular debris and the resolution \nof inflammation [ 15, 16]. Among them, macrophages are \ndeemed the most crucial cells contributing to the repair and \nregeneration of the endometrium [17]. Current research sug-\ngests that the macrophages involved in endometrial repair \noriginate from two distinct sources: in situ proliferation of \nresident macrophages and macrophages derived from mono-\ncytes [18]. However, the potential roles of other sources of \nmacrophages in endometrial repair remain unexplored.\nIn recent years, an increasing number of studies have \nconcentrated on the role of a specific group of resident \nmacrophages found in the mammalian peritoneal cavity, \nknown as large peritoneal macrophages (LPMs) [19]. LPMs \nare not a homogeneous population but exhibit heterogene -\nity in terms of their origin, phenotype, and function. The \nontogeny of LPMs involves both embryonic-derived resi -\ndent macrophages and monocyte-derived macrophages that \ncan replenish the LPM pool under certain conditions [ 20, \n21]. This heterogeneity and ontogeny are crucial for under-\nstanding the diverse roles LPMs play in health and disease. \nLPMs constitute the primary population of peritoneal mac -\nrophages, with their main function being the phagocytosis \nof apoptotic cells [ 22]. Research has indicated that LPMs \npossess the ability to rapidly migrate to the site of injury \nand facilitate tissue repair in mouse models of liver injury \n[23] and damage to the intestinal serosal layer [24]. Further-\nmore, when the peritoneal cavity wall is injured, clusters of \nLPMs physically seal the wound, promoting swift recovery \nin mouse model [ 25]. Given that the uterus is also a vital \norgan situated within the peritoneal cavity, it remains to \nbe further explored whether LPMs similarly migrate to the \nendometrium to exert their effects.\nIn this study, we employed our previously established \nmouse model of endometrial injury, which obviates the \nneed for peritoneal cavity incision [ 11]. This model was \ninitially designed to minimize the risk of systemic infec -\ntions associated with open surgical procedures and to more \nclosely mimic the conditions of human endometrial trauma. \nNotably, within this model, we observed the presence of \nLPMs in the damaged uterine tissue. These LPMs appeared \nearlier than monocyte—derived macrophages and nearly \nsimultaneously with neutrophils. We utilized flow cytom -\netry assays to demonstrate the capacity of LPMs to migrate \nfrom the peritoneal cavity to the injured endometrium. Fur-\nthermore, through experiments involving the depletion and \nsupplementation of LPMs, we elucidated their pivotal role \nin clearing dead cells and resolving inflammation at the site \nof injury. Additionally, our findings revealed that estradiol \n(E2) facilitates the migration of LPMs both in vivo and in \nvitro. These results collectively highlight the crucial func -\ntion of LPMs in endometrial repair and provide evidence \nthat administering E2 following uterine cavity surgery can \nexpedite the migration of LPMs, thereby enhancing their \nreparative capabilities.\nMaterials and Methods\nAnimals and Experimental Protocol\nFemale BALB/c mice (8–10 weeks old) and female \nC57BL/6 mice (8–10 weeks old) were brought from Jiangsu \nHuachuang Xinnuo Pharmaceutical Technology Co., Ltd. \n(Taizhou, China) and were housed in a pathogen-free condi-\ntion in a 12-h light and dark cycle. All procedures involved \nin mice were approved by the institutional guidelines for \nanimal care and used based the Animal Care Committee at \nNanjing University and followed all the ARRIVE guide -\nlines [26].\nEndometrial Injury Time-Course Experiment\nBalb/c female mice (n = 4–6) were anesthetized with isoflu-\nrane and fixed in the supine position. The probe of electric \nscratching tool was inserted into the uterus of mice through \nthe vaginal opening. The switch was pressed, vibrated for \n8 s, and then paused for 8 s. This process was repeated \ntwice more. The probe was slowly pulled out and the mouse \nwas placed under a warm lamp for recovery from anesthe -\nsia. Samples were harvested at 0, 3, 6, 12, 24 or 48 h after \nmodeling.\nCLL Eliminate LPMs Experiment\nBalb/c female mice ( n = 4–6) were injected intraperitone -\nally with 100 μL PBS/con-CLL (7 mg/mL, FormuMax, \nF70101-A)/CLL (7 mg/mL, FormuMax, F70101C-A) 24 h \nin advance. Next, endometrium of mice was injured using \nthe electric scratching tool for 12 h.\nIntrauterine Supplementation of LPMs\nVaginal secretions from female BALB/c mice (8–10 weeks \nold) were scraped using a sterile disposable 200 μL suction \ntip, suspended in PBS, and placed under a light microscope \nto observe cell morphology. The mouse model was created \nat estrus based on the cell morphology in vaginal secretion \nsmear of mice. Balb/c female mice (n = 3–5) were anesthe-\ntized with isoflurane and fixed in the supine position. They \nwere then disinfected and hair from their abdominal sur -\nfaces were removed. The abdominal skin and muscles were \n1 3\n4429\n\nInflammation (2025) 48:4428–4442\nthen cut. One side of the uterine horn was found. A syringe \nneedle (30 G) was inserted and the lining of the uterus \nwas scratched 50 times. 10 µL of saline solution contained \nLPMs (8 × 104) or not was then injected into the uterine cav-\nity. Both ends of the uterine horn were held with forceps for \n5 min to ensure that liquid is fully absorbed by the tissue. \nThe muscle layer and skin were sutured. The mouse was put \nunder a warm lamp for recovery from anesthesia. Samples \nwere harvested after 12 h.\nIsolation of Primary Mouse Cells\nLarge Peritoneal Macrophage (LPMs)\nLPMs were obtained from female Balb/c mice (8–10 weeks \nold). The mice need to be fasted for 8 h and sacrificed by cer-\nvical vertebra dislocation. Then inject cold DMEM medium \n5 mL into peritoneal cavity of mice and gently massage \nthem for about 2 min. Subsequently, collect the peritoneal \nfluid and 300 × g centrifugate for 5 min to obtain cells. Lyse \nerythrocytes if necessary. Inoculate the peritoneal fluid cells \ninto cell culture plates with warm DMEM medium supple -\nmented with 10% FBS for 4 h. Then abandon the superna -\ntant and wash cells attached to the bottom with warm PBS \nbuffer solution to removal suspension cells. The remaining \ncells are LPMS for subsequent experiments.\nEndometrial Stromal Cells (ESCs)\nUterine tissues from mice were cut into pieces and digested \nwith trypsin with 0.1% EDTA for 10 min at 37 °C in a 5% \nCO2-humidified atmosphere. After termination of trypsin \ndigestion, the cells were digested in HBSS medium contain-\ning 0.8 mg/mL collagenase type I for 2 h at 37 °C in a 5% \nCO2-humidified atmosphere. Then digestive fluid contained \ntissue residue were fractionated with 40-μm cell strainers. \n300 × g centrifugate the filtrate for 5 min and the cell precip-\nitates were cultured in cell culture dishes with DMEM/F12 \nmedium containing 10% FBS, 100 U/mL penicillin, and 100 \nU/mL streptomycin at 37 °C in a 5% CO2-humidified atmo-\nsphere. After 2–3 days of culture, the cells that could con -\ntinuously proliferate were ESCs. Primary ESCs from P2/P3 \nwere used for subsequent experiments.\nCells Treatment and Experiments in vitro\nESCs Apoptosis Inducing Method\nESCs cultured to passage 2 or 3 were seeded in 24-well \nplates (1 × 104 cells/well) and treated with 20% ethanol for \n1, 5 and 25 min or not. 20% ethanol system was absolute \nethanol and DMEM/F12 medium containing 10% FBS \n(1:4). The subsequent experiments used 5 min as treatment \ncondition.\nPhagocytosis Assay of LPMs on ESCs\nESCs cultured to passage 2 or 3 were seeded in 24-well \nplates (1 × 104 cells/well) and treated with carboxyfluores -\ncein succinimidyl amino ester (CFSE, Invitrogen, catalog # \n65–0850-85) for 10 min at 37 °C in a 5% CO 2-humidified \natmosphere. Then the cells labeled by CFSE were induced \napoptosis and cocultured with LPMs (1:3) isolated from \nmice for 12 h.\nE2 and Fulvestrant Treatment of LPMs\nThe LPMs isolated from mice were seeded in 24-well plates \n(5 × 104 cells/well) and treated by fulvestrant (100 nM) or \nnot for 1 h. Next, the LPMs treated by E2 (1, 10 or 100 \nnM) or not for 24 h. The fulvestrant and E2 was purchased \nfrom MedChemExpress (MCE) (catalog # HY-13636 and \nHY-B0141).\nLPMs Migration Model and Crystal Violet Staining\nLPM was inoculated into the transwells (1 × 104 cells/well) \nand pre-treated by fulvestrant or E2 or not for 24 h. The tran-\nswells within LPMs were cocultured with apoptotic ESCs in \n24 well plates for 6 h. Then the transwells were fixed with \n4% paraformaldehyde in phosphate buffer for 15 min at \nroom temperature. After washed twice with phosphate buf -\nfer, the transwells were fixed with 1% crystal violet solution \n(biosharp, catalog # BL802A) in phosphate buffer for 15 \nmin. They were washed with phosphate buffer until there \nwas no residual crystal violet solution. Cells on the side of \nthe transwells were gently wiped off with a cotton swab, \nleaving only the bottom cells. Finally, they were viewed \nand photographed under a Nikon Eclipse Ti-U microscope \nequipped with a digital camera (DS-Ri1, Nikon).\nFlow Cytometry Assay\nAll cells were filtered through a 70 μm cell strainer and \nthen washed with PBS to generate single-cell suspensions. \nAn Fc-receptor blocker (CD16/32, eBioscience, catalog \n# 14–0161-82) was used to reduce non-specific antibody \nbinding. Single cell suspensions were labeled with flow \ncytometry antibodies (manufacturer and catalog number \nare showed in the Supplementary material Table S1) and \ndetected by BD FACS Calibur or Beckman Coulter Cyto -\nflex S. Data analysis was performed by FlowJo software. \n1 3\n4430\n\nInflammation (2025) 48:4428–4442\nStatistical Analysis\nAll values with a normal distribution presented on the graphs \nare shown as means ± S.E.M. Unpaired Student’s t-tests was \nused to analyze statistical significance, and P-values < 0.05 \nwere considered statistically significant. All statistics were \nperformed with GraphPad Prism 8.\nResults\nLPMs Migrate Into Injured Endometrium\nTo examine the alterations in inflammation-associated \nimmune cells within the injured endometrium, we induced \nendometrial damage in mice using an electric scratching \ntool, as previously described [ 11], at various time points: \n3, 6, 12, 24, and 48 h prior to sample collection (Fig. 1A). \nFlow cytometry results showed that, after endometrium was \ninjured, the proportion of neutrophils ( CD11bintLy6G+) in \nuterus increased to the peak at 6 h and gradually disappeared \n(Fig. 1B). The increase in monocytes ( CD11b+CCR2+) and \nuterine resident macrophages ( CD11b+CD102−F4/80int) \nwere slightly later, reach the peak at 12 h (Fig. 1C and D). \nRemarkably, there were a group of cells characterized as \nCD11bhigh, CD102+ and F4/80high. Their phenotype dif -\nfered from that of neutrophils but increased at a similar rate \n(Fig. 1E). CD102 is the specific marker of cavity resident \nmacrophages [ 27]. Given that the uterus is located in the \nperitoneal cavity, we supposed that these cells might be \nperitoneal macrophages.\nNext, we measured the proportions of LPMs and SPMs \nin peritoneal cavity at each moment of endometrial injury by \nflow cytometry. The CD11b + myeloid cells in the peritoneal \nAfter removing dead cells and adherent cells, we refer the \nCD11b+CD102+ cells as LPMs, CD11b+CD102−MHCII+ \ncells as SPMs, CD11b+CD102−MHCII− cells as new cells, \nCD11bintLy6G+ cells as neutrophils, CD11b+CCR2+ cells \nas monocytes, and CD11b+F4/80intCD102− cells as uterine \nresident macrophages.\nImmunofluorescence Staining\nMouse uterus tissues or primary mouse cells were fixed with \n4% paraformaldehyde in phosphate buffer. Paraffin-embed-\nded samples were sectioned at 2 μm. Tissue slices or cells on \nglass slides were incubated with primary antibodies (manu-\nfacturer and catalog number are showed in the Supplemen -\ntary material Table S2) overnight at 4 °C. After rinsing three \ntimes in PBS, the samples were incubated with secondary \nantibody (manufacturer and catalog number are showed in \nthe Supplementary material Table S2) for 1.5 h at room tem-\nperature in the dark, and then the nuclei were stained with \nDAPI (Bioword, China). The slides were visualized using a \nFV3000 Laser Scanning Confocal Microscope (Olympus).\nEnzyme-Linked Immunosorbent Assay (ELISA) and \nBiochemical Detection\nThe levels of IL-1β, IL-6 and TNF-α in mouse serum were \ndetected using the corresponding mouse ELISA kit accord-\ning to the manufacturer’s instructions (Biolegend, catalog # \n432601, 431301 and 430901). The level of E2 was detected \nusing the corresponding mouse ELISA kit according to the \nmanufacturer’s instructions (Cloud-Clone Corp., catalog # \nCEA461Ge). Serum E2 levels were determined at Service -\nbio Biotechnology Co. LTD (Wuhan, Hubei, China).\nQuantitative real-Time PCR\nQuantitative real-time PCR was carried out as previously \ndescribed [14]. The primer sequences we used are showed \nin the Supplementary material Table S3.\nRNA Sequencing Assay\nMouse LPMs samples were analyzed via RNA sequencing \nin Hangzhou Lianchuan Biotechnology Co., Ltd. (Hang -\nzhou, China). Preliminary progression was analyzed after \nobtaining raw data. Genes differential expression analysis \nwas performed by DESeq2 software between E2 and con -\ntrol groups. The fold value represented the degree of differ-\nential expression between E2 and control groups. The genes \nwith the parameter of false discovery rate (FDR) below 0.05 \nand absolute fold change ≥ 2 was considered differentially \nexpressed genes.\nFig. 1 LPMs migrate into injured endometrium. A The endometrium \nof Balb/c female mice (8–10 weeks old) was injured using the elec -\ntric scratching tool and samples were collected at 3-, 6-, 12-, 24- and \n48-h post-injury. B-E The statistical figures of percentages of neutro -\nphils (CD11bint Ly6G+), monocytes (CD11b+CCR2+), uterine resident \nmacrophages ( CD11b+F4/80intCD102−) and LPMs ( CD11b+F4/80\n+CD102+) in uterus determined by flow cytometry assay. F The statis-\ntical figures of percentages of LPMs (CD11b+CD102+F4/80+) in peri-\ntoneal cavity determined by flow cytometry assay. G Flow cytometry \nassay data of the F4/80 and GATA6 expressing levels on cells from \nabove gates. Among them, Q2 represented LPMs. H Immunofluores-\ncence staining images (cross section) of CD102 and GATA6 expres-\nsions in uterus at 0 h (control) and 6 h after injured. CD102 (red) was \nlocated on the cell membrane and GATA6 (green) was located in the \nnucleus. Bar = 50 μm. I The statistical figures of percentages of LPMs \n(CD11b+CD102+F4/80+GATA6+) in uterus determined by flow cytom-\netry assay. J Immunofluorescence staining images (cross section) of \nTUNEL in uterus at 0 h (control) and 6 h after injured. Bar = 100 μm. K \nThe level of IL-1β, IL-6 and TNF-α in serum detected by ELISA assay. \nBar graphs show the mean ± SEM; unpaired Student’s t test was used \nto compare the experimental groups. Bar = 50 μm. n = 4–6; *P < 0.05, \n**P < 0.01, ***P < 0.001, ****P < 0.0001\n1 3\n4431\n\nInflammation (2025) 48:4428–4442\n \n1 3\n4432\n\nInflammation (2025) 48:4428–4442\nliver, or bone marrow, except for a marked reduction in \nmacrophages in the spleen (Supplementary Fig. 3B). More-\nover, level of IL-6 in serum was slightly increased, whereas \nthe levels of IL-1β and TNF-α did not change significantly \n(Supplementary Fig. 3 C). Taken together, these results sug-\ngested that intraperitoneal injection of CLL could deplete \nLPMs from the peritoneal cavity of mice without causing \ntoo many additional effects.\nWe subsequently depleted LPMs in the peritoneal cav -\nity using CLL 24 h prior to experimentally inducing endo -\nmetrial damage in mice with our electric scratching tool, \nwith samples collected 12 h later (Fig. 2A). Flow cytometry \nresults indicated that CLL effectively hindered the migration \nof LPMs from the peritoneal cavity to the uterus, whereas \ncon-CLL did not exert such an effect (Fig. 2B and C). \nThe absence of LPMs prevented their accumulation in the \ninjured uterus. Furthermore, the lack of LPMs exacerbated \nboth the elevation of inflammatory factors in the serum \nresulting from uterine injury (Fig. 2D) and the accumula -\ntion of dead cells in the endometrium (Fig. 2E). Based on \nthese observations, we hypothesized that LPMs may play a \ncrucial role in phagocytosing dead cells, mitigating inflam -\nmation, and facilitating endometrial repair.\nLPMs Relieve Inflammation and Phagocytose Dead \nCells in Injured Endometrium\nTo validate our hypothesis regarding the function of LPMs \nin the injured endometrium, we conducted a LPMs trans -\nplantation experiment. In this study, we accessed the peri -\ntoneal cavity of the mice and induced endometrial damage \nusing a syringe needle. This procedure mirrored the estab -\nlishment of the mechanical damage IUA mouse model \ndescribed in our earlier publications [11, 28] (Fig. 3A). Our \nfindings revealed that the introduction of LPMs was capa -\nble of abrogating the upregulation of serum inflammatory \nfactors ( IL-1β, IL-6, and TNF-α) triggered by endometrial \ninjury (Fig. 3B). We also noted that LPMs were colocalized \nwith dead cells within the endometrium (Fig. 3C), indicat-\ning their capacity to phagocytose these cells in that location. \nTo further substantiate this observation, we conducted an \nin vitro phagocytosis assay. Given that endometrial stromal \ncells (ESCs) constitute the primary cell population in the \nendometrium, we utilized ESCs derived from mouse uteri as \nthe target for phagocytosis. ESCs were seeded in well plates \nand subsequently treated with ethanol for 5 min to induce \napoptosis (Supplementary Fig. 4 A and B). The immuno -\nfluorescence images obtained in vitro demonstrated that \nLPMs engaged in trogocytosis, shredding, and endocytos -\ning apoptotic ESCs, confirming their ability to phagocytose \nthese cells (Fig. 3D). Collectively, these results established \nthat LPMs can alleviate inflammation and phagocytose dead \nCD102 (red) was located on the cell membrane and GATA6 \n(green) was located in the nucleus. They were divided into \nthree groups based on the levels of CD102 and MHCII, \nincluding LPMs ( CD102+), SPMs ( CD102−MHCII+) and \nan undefined subgroup called new cells ( CD102−MHCII−) \n(Supplementary Fig. 1B). Among them, LPMs had the \nhighest proportion (> 90%). Moreover, LPMs exhibited \nF4/80high and GATA6+ specifically, which was different \nfrom the other two populations (Supplementary Fig. 1 C and \nSupplementary Fig. 1D). Flow cytometry results showed \nthat, after damage to the endometrium, the proportion of \nLPMs in peritoneal cavity was decreased at 3 h and came \nto the valley bottom at 6 h (Fig. 1F). Then the proportion \ngradually returned to normal. This pattern of change was \njust corresponded to that of CD11b+CD102+F4/80+ cells in \ninjured endometrium (Fig. 1E). In addition, the proportion \nof SPMs was increased at 24 h after damage (Supplemen -\ntary Fig. 1E) and new cells were increased to the peak at 6 h \n(Supplementary Fig. 1 F). These data indicated that LPMs \nmay be the cell population which migrated from peritoneal \ncavity into the damaged endometrium.\nTo validate this hypothesis, we conducted a detailed \ncharacterization of CD11b+, CD102+, and F4/80+ cells \nwithin the injured endometrium and discovered that these \ncells expressed GATA6, a specific marker for LPMs (Sup -\nplementary Fig. 1 A, Fig. 1G and H). Our flow cytometry \nanalysis revealed that, subsequent to endometrial injury, \nthe proportion of LPMs in the uterus increased notably at \n3 h, reaching a peak at 6 h (Fig. 1I), followed by a gradual \ndecline. Correspondingly, this treatment caused the increase \nof dead cells number in the endometrium (Fig. 1J) and the \nupregulation of inflammatory factors ( IL-1β、IL-6 and \nTNF-α) in serum (Fig. 1K). All the above experiments were \ncarried out on Balb/c mice. A similar phenomenon was also \nobserved in C57BL/6 mice (Supplementary Fig. 2 A, B, C). \nOverall, these findings indicate that LPMs could transfer \nfrom peritoneal cavity to injured endometrium. This process \nwas as rapid as the infiltration of neutrophils, and could be \nobserved in different mouse strains.\nLPMs Absence Enhances Inflammation and \nAccumulation of Dead Endometrial Cells\nOur findings underscored the capacity of LPMs to swiftly \nmigrate to the site of uterine injury. To delve deeper into the \nroles of LPMs in the injured uterus, we employed a strat -\negy to deplete LPMs by administering clodronate liposomes \n(CLL) intraperitoneally to the mice, with con-CLL served \nas the control. Notably, a substantial majority of LPMs were \neffectively eliminated within 48 h post-injection (Supple -\nmentary Fig. 3 A). Additionally, we observed no significant \nalterations in the proportions of macrophages in the uterus, \n1 3\n4433\n\nInflammation (2025) 48:4428–4442\nwere numerous large scaly cells, i.e. the keratinized endo -\nmetrial epithelial cells (Supplementary Fig. 4 A). Indeed, the \nlevel of E2 was higher in the serum of proestrus mice than \nthat in the serum of estrus mice (Supplementary Fig. 4B). \nWe then divided mice into estrus and proestrus groups by the \nabove characteristics, and damaged their endometrium with \nour electric scratching tool at 6 h before sample harvesting \nrespectively (Fig. 4A). The results showed that the migra -\ntion efficiency of LPMs in proestrus mice was significantly \nhigher than that in estrus mice, although their migration to \nthe uterus occurred in both stages of mice (Fig. 4B and C). \ncells in the injured endometrium, which is crucial for uter -\nine repair.\nEstradiol can Accelerate the Migration of LPMs to \nthe Damaged Endometrium\nWe identified the estrous periods according to the cell mor-\nphology in mouse vaginal secretion smear of mice. In the \nvaginal secretions of proestrus mice, there were mainly \nnucleated, round, large cells, i.e. the normal endometrial epi-\nthelial cells. In the vaginal secretions of estrus mice, there \nFig. 2 LPMs absence enhances \ninflammation and accumulation \nof dead endometrial cells. A The \nschematic diagram of modelling \nprocess. Balb/c female mice (8–10 \nweeks old) were injected intra-\nperitoneally with 100 μL PBS/con-\nCLL/CLL 24 h in advance. Next, \nendometrium of mice was injured \nusing the electric scratching tool \nfor 12 h. B The statistical figures \nof percentages of LPMs (CD11b+C\nD102+F4/80+) in peritoneal cavity \ndetermined by flow cytometry \nassay. C The statistical figures of \npercentages of LPMs (CD11b+CD\n102+F4/80.+) in uterus determined \nby flow cytometry assay. D The \nlevel of IL-1β, IL-6 and TNF-α in \nserum detected by ELISA assay. \nE Immunofluorescence staining \nimages (cross section) of TUNEL \nin uterus. Bar graphs show the \nmean ± SEM; unpaired Student’s \nt test was used to compare the \nexperimental groups. Bar = 100 \nμm. n = 4–6; *P < 0.05, **P \n< 0.01, ***P < 0.001, ****P \n< 0.0001\n \n1 3\n4434\n\nInflammation (2025) 48:4428–4442\nFig. 3  LPMs relieve inflammation and phagocytose dead cells in \ninjured endometrium. A The schematic diagram of modelling pro -\ncess. Balb/c female mice (8–10 weeks old) were carried out surgery \nand intrauterine injected LPMs (8 × 104). Samples were harvested \nafter 12 h. B The level of IL-1β, IL-6 and TNF-α in serum detected by \nELISA assay. C Immunofluorescence staining images (cross section) \nof CD102 (pink), GATA6 (green) and TUNEL (yellow) expressions in \nuterus from ‘injury + LPMs’ group mice. D Immunofluorescence stain-\ning images of LPMs (F4/80 +, red) and ESCs (CFSE labeled, green) \nisolated in vitro. Bar graphs show the mean ± SEM; unpaired Student’s \nt test was used to compare the experimental groups. Bar = 50 μm. n \n= 3–5; *P < 0.05, **P < 0.01\n \n1 3\n4435\n\nInflammation (2025) 48:4428–4442\nfactors in proestrus mice were significantly lower than those \nin estrus mice (Fig. 4F). These data suggested that the differ-\nent E2 levels may influence the migration of LPMs into the \nimpaired uterus to suppress inflammation.\nMoreover, at 6 h after endometrium injury, the proportion \nof LPMs in peritoneal cavity of mice was inversely pro -\nportional to the level of E2 (Fig. 4D), but directly propor -\ntional in uterus (Fig. 4E). The serum levels of inflammation \nFig. 4 The migration efficiency of LPMs into injured uterus is influ -\nenced by E2 level. A The schematic diagram of modelling process. \nBalb/c female mice (8–10 weeks old) were divided into estrus group \nand proestrus group. The endometrium was injured using the elec -\ntric scratching tool for 6 h. B The statistical figures of percentages of \nLPMs (CD11b + CD102 + F4/80 +) in peritoneal cavity determined by \nflow cytometry assay. C The statistical figures of percentages of LPMs \n(CD11b + CD102 + F4/80 +) in uterus determined by flow cytometry \nassay. D Correlation analysis graph of percentage of LPMs in perito -\nneal cavity and E2 of serum. E Correlation analysis graph of percent-\nage of LPMs in uterus and E2 of serum. F The level of IL-1β, IL-6 and \nTNF-α in serum detected by ELISA assay. Bar graphs show the mean \n± SEM; unpaired Student’s t test was used to compare the experimental \ngroups. n = 4–8; *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001\n \n1 3\n4436\n\nInflammation (2025) 48:4428–4442\nnon-vascular route, subsequently contributing to tissue \nrepair [23]. Subsequent research has further elucidated the \nmigratory behavior and tissue repair capabilities of LPMs \nin both mouse intestinal [ 24] and peritoneal cavity wall \n[25] injuries. Furthermore, pericardial macrophages, which, \nakin to LPMs, belong to the category of cavity—resident \nmacrophages, have been demonstrated to infiltrate damaged \ncardiac tissue and exert a reparative influence in a mouse \nmodel [21]. Notably, GATA6⁺ macrophages have been iden-\ntified within human pericardial fluid. This finding provides \nsubstantial support for the hypothesis that the reparative \nfunction observed in the mouse model holds clinical rele -\nvance in human cardiac diseases [21]. Given that the uterus \nis also situated within the peritoneal cavity, we are intrigued \nby the possibility that LPMs may similarly respond to endo-\nmetrial injury.\nIt is known that injury to the peritoneal cavity wall causes \nLPMs to form adherent aggregates at the wound site, which \nleads to a significant reduction of LPMs in peritoneal fluid \n[25]. In addition, peritoneal cavity infection caused by lapa-\nrotomy can also trigger a large loss of LPMs [19]. Therefore, \nthe laparotomy during the establishment of animal models \nis bound to interfere with the study of LPMs. Our modeling \ntool skillfully avoided disruption of the normal peritoneal \nE2/ER-β Axis is Necessary for the Migration of LPMs \nto Apoptotic ESCs\nAbove results hinted E2 may drive the migration of LPMs to \ndamaged uterus. To further explore roles of E2 in migration \nof LPMs, we intraperitoneally injected E2 to estrus mice, and \nthen damaged their endometrium with the electric scratch -\ning tool at 6 h before sample harvesting (Fig. 5A). Dimethyl \nsulfoxide (DMSO) was used as control. As showed in flow \ncytometry assay results, E2 promoted the accumulation of \nLPMs in impaired uterus, while there was no difference in \nthe percentage of LPMs in peritoneal cavity between ‘Injury \n+ DMSO’ group and ‘Injury + E2’ group (Fig. 5B and C). \nIn addition, the transcriptome RNA sequencing (RNA-seq) \nresults showed that E2 increased the expression of many \ngenes related to cytoskeleton, adhesion and migration in \nLPMs (Fig. 5D). Indeed, we found that E2 could elongate \nthe cytoskeleton and increase pseudopodia of LPMs in vitro \n(Fig. 5E), which were both necessary for cells to migrate. \nTo better simulate the migration of LPMs to the damaged \nuterus, an in vitro migration assay was designed. ESCs are \nconsidered as the main cells in endometrium. ESCs were \nthus seeded in well plates, and then treated with ethanol for \n5 min to induce apoptosis. Correspondingly, LPMs were \nseeded in the upper transwells for co-culture with apop -\ntotic ESCs. The results showed that E2 indeed promote the \nmigration of LPMs to apoptotic ESCs in a dose-dependent \nmanner (Fig. 5F).\nAccumulated evidence has revealed that there are three \ntypes of estrogen receptors (ERα, ERβ, and GPR30). Impor-\ntantly, LPMs were detected to express only ER-β (gene \nnamed Esr-2) specifically, but not ER-α (gene named Esr-1) \nand GPR30 (gene named Gpr30) (Fig. 5G). Moreover, E2 \n(10 nM) significantly up-regulated the expression of ER-β \nin LPMs (Fig. 5H and I), indicating ER-β in LPMs specifi -\ncally responded to the stimulation of E2. Estrogen receptor \nantagonist fulvestrant (Ful) is showed to completely inhibit \nestrogen-mediated changes in gene transcription. In in vitro \nmigration assay, Ful eliminated the promotion of E2 on the \nmigration of LPMs to apoptotic ESCs (Fig. 5J). Combined \nwith all above results, we concluded that E2/ ER-β axis was \nnecessary for the migration of LPMs to apoptotic ESCs.\nDiscussion\nOver the past decade, large peritoneal macrophages (LPMs) \nhave garnered increasing attention. As the primary subpop-\nulation of macrophages within the mammalian peritoneal \ncavity, they fulfill a pivotal immune function in maintain -\ning homeostatic balance. A seminal study illustrated that \nLPMs can swiftly migrate to the site of liver injury via a \nFig. 5 E2/ER-β axis is necessary for the migration of LPMs to apop -\ntotic ESCs. A The schematic diagram of modelling process. The endo-\nmetrium of Balb/c female mice (8–10 weeks old) which were in estrus \nperiod was injured using the electric scratching tool for 6 h. B The sta-\ntistical figures of percentages of LPMs (CD11b + CD102 + F4/80 +) in \nperitoneal cavity determined by flow cytometry assay. C The statistical \nfigures of percentages of LPMs ( CD11b + CD102 + F4/80 +) in uterus \ndetermined by flow cytometry assay. D RNA-seq analysis of LPMs \nisolated from mouse peritoneal cavity treated with or without E2 (10 \nnM) for 24 h. The heat map represented the differential expression of \ngenes, sorted from top to bottom according to fold change (log2). E \nImmunofluorescence staining images of LPMs isolated from mouse \nperitoneal cavity in vitro. LPMs were treated by E2 (10 nM) for 24 \nh. Phalloidin (green) stained the cytoskeleton. Bar = 30 μm. F Crys-\ntal violet staining images of LPMs which migrated from inside of the \ntranswell to the bottom of that. LPMs in transwell were pre-treated \nby E2 (1, 10 or 100 nM) or not for 24 h. The transwell within LPMs \nwas cocultured with Ap-ESC for 6 h. Ap-ESC represents the apoptotic \nESCs. Bar = 100 μm. The column graph was the statistic of the purple \ncell number. G The expression level (2∆Ct) of Esr-1, Esr-2 and gpr30 \nin LPMs (isolated from estrus or proestrus mice) or 3T3-L1 cells \ndetected by qRT-PCR assay. H Relative expression of mRNAs of Esr-\n2 in LPMs from those mice detected by qRT-PCR assay. The LPMs \nisolated from mice were treated by E2 (1, 10 or 100 nM) or not for \n24 h. I Immunofluorescence staining images of LPMs isolated from \nmouse peritoneal cavity in vitro. The LPMs were treated by E2 (10 \nnM) or not for 24 h. J Crystal violet staining images of LPMs which \nmigrated from inside of the transwell to the bottom of that. LPMs in \ntranswell were pre-treated by fulvestrant (Ful, 100 nM) or not for 1 h \nand treated by E2 (10 nM) or not for 24 h. The transwell within LPMs \nwas cocultured with Ap-ESC for 6 h. Bar = 100 μm. The column graph \nwas the statistic of the purple cell number. Bar graphs show the mean \n± SEM; unpaired Student’s t test was used to compare the experimental \ngroups. n = 3; *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001\n1 3\n4437\n\nInflammation (2025) 48:4428–4442\n \n1 3\n4438\n\nInflammation (2025) 48:4428–4442\naccumulation of dead uterine cells. These findings imply \nthat LPMs may contribute to the phagocytosis of dead cells, \nsuppression of inflammation, and ultimately, the promo -\ntion of endometrial repair. To further scrutinize the impact \nof LPMs on damaged endometrium, we administered addi -\ntional LPMs directly into the uterus. Given that intrauterine \ninjection necessitated opening the peritoneal cavity of mice, \nwe adopted the endometrial injury method utilized in the \nconventional IUA mouse model [ 28]. This method offered \nthe advantage of more severely disrupting the endometrium, \nthereby accentuating the effects of LPMs. Our results indi -\ncated that LPMs significantly reduced inflammation levels \nfollowing uterine injury in mice, and we observed colocal -\nization of LPMs with TUNEL-positive cells in the endo -\nmetrium. However, due to technical constraints, it was not \nfeasible to determine whether the TUNEL-positive cells \nwere dead uterine cells or LPMs that gradually disappeared \nafter entering the uterus, potentially due to death and subse-\nquent removal by other phagocytes. Therefore, we investi -\ngated the phagocytosis of LPMs on apoptotic ESCs in vitro, \nhoping to sequentially mimic their possible behavior in the \ndamaged endometrium. In addition, the limitation of this \npart of the study is that the histological changes of the endo-\nmetrium in the early stage (within 12 h) after injury are not \nobvious, so we only indirectly reflected the effect of LPMs \non endometrium repair through the level of inflammation \nand the number of dead cells. More indicators of early stage \nafter endometrium injury need to be further explored.\nAfter surgical separation of adhesions in patients with \nIUA, E2 is usually administered as an adjunctive therapy. \nAlthough there are several evidences showed E2 promotes \nthe regeneration of endometrium [ 29–31], the influence of \nE2 on LPMs is unknown. Since the E2 levels in mice are \nhighest during proestrus and lowest during estrus [ 32], we \ninvestigated whether there was any difference in migration \nefficiency of LPMs into impaired uterus between proestrus \nand estrus in mice. The amount of LPMs migrating in estrus \nmice tended to be significantly lower than in proestrus mice. \nThis finding suggested that female sex hormones may be \nable to influence the migration of LPMs. The main hormones \nthat change during the female cycle are E2 and progesterone \nP4, and their levels alternately rise and fall [ 33]. In human \nhormone cycle, the early stage of endometrium regeneration \nis mainly regulated by E2 [ 18]. Mice undergo an estrous \ncycle similarly to the hormonal cycle of humans, and that \ncan be divided into four stages. Among them, proestrus and \nestrus are dominated by E2, while metestrus and diestrus \nare P4-dominant stages [34]. Our experimental results con-\nfirmed that the E2 level in proestrus mice is higher than that \nin estrus. Considering the high migration efficiency of LPMs \nin proestrus mice, and there was some evidence that E2 or \nE2 analogs could promote cell migration [ 35–37]. We only \nenvironment, allowing us to study the behavior of LPMs in \nresponse to endometrial injury with maximum accuracy. It \nis worth mentioning that during the experiments, we found \nthat intraperitoneal anesthetics also caused the depletion of \nLPMs, so all our animal experiments were performed with \ninhaled anesthetics. In addition, we strictly followed sterile \noperation when injecting drugs or fluorescent dye intraper -\nitoneally. All these efforts have been made to ensure that \nthe behavioral patterns of LPMs can be simulated precisely \nunder normal conditions in endometrial damage.\nIt is recently reported that there are two distinct popu -\nlations of macrophages in the peritoneal cavity—the small \nperitoneal macrophages (SPMs) and LPMs. SPMs are \nbone marrow derived with the F4/80 low, CD11b low, \nCD102 negative and MHCII positive phenotype. LPMs are \nembryo derived and F4/80 high, CD11b high, and CD102 \npositive. They selectively express the zinc finger tran -\nscription factor GATA-binding protein 6 ( GATA6) and are \nmaintained in the peritoneal cavity through self-renewal \n[27]. Using electric scratching tool to injure the endome -\ntrium of mice for different durations, we found that LPMs \ncould transfer from peritoneal cavity to injured endo -\nmetrium rapidly, with a peak at 6 h. In this process, the \nmigration of LPMs was accompanied by a rapid increase \nof CD11b+CD102−MHCII−F4/80−GATA6− cells in the peri -\ntoneal cavity. We called them the ‘new cells’, and consid -\nered them may be a population of undifferentiated myeloid \ncells derived from bone marrow. Immediately thereafter, \nthe proportion of SPMs began to rise. Many studies have \nsuggested that SPMs are the differentiation precursors of \nLPMs, and that new cells may also be the differentiation \nprecursors of SPMs [ 22]. Their amount was increased in \norder to supplement the amount of LPMs. This was essen -\ntial to maintain the homeostasis of peritoneal cavity envi -\nronment. Our researches suggested the ability of LPMs to \nquickly migrate into injured mouse uterus. However, due \nto the limit of experiment condition, we did not verify this \nphenomenon on humans. More efforts should be made to \nexplore the changes of LPMs in human endometrial repair.\nWe observed that there was a large accumulation of \ndead cells in the uterus and an inflammatory response in \nthe body after endometrial injury. These occur at approxi -\nmately the same time as the migration of LPMs into uterus. \nTherefore, we hypothesized that LPMs may play a role in \nclearing apoptotic cells and resolving inflammation in the \nendometrium. We initially employed CLL to deplete LPMs. \nThis approach proved highly effective in eliminating LPMs, \nand by restricting the intraperitoneal injection of CLL to a \n48-h window, we ensured that tissue-resident macrophages \nin the uterus, liver, and bone marrow remained unaffected. \nThe absence of LPMs notably reduced their presence in the \ninjured uterus, resulting in heightened inflammation and an \n1 3\n4439\n\nInflammation (2025) 48:4428–4442\nSupplementary Information  The online version contains \nsupplementary material available at  h t t p  s : /  / d o i  . o  r g /  1 0 . 1  0 0 7  / s 1  0 7 5 3 - 0 \n2 5 - 0 2 3 3 5 - z.\nAuthor Contributions Y .H., S.S., J.L., G.Z. acquired and interpreted \ndata, and drafted the manuscript; Y .H., S.S., J.W. acquired data and \nrevised the manuscript; Y .H., G.Z., S.S., J.L., J.W., L.Y ., Y .P., C.P., \nY .D. contributed to data interpretation and revision of the manuscript; \nY .H., G.Z., J.L. conceived and supervised the work, acquired, and in-\nterpreted data, and drafted the manuscript. All authors read and ap -\nproved the final manuscript.\nFunding This work was supported by the National Key R&D Program \nof China (2023YFC2308200), National Natural Science Foundation \nof China (82471663,82071600, 82271653), and the Key Research and \nDevelopment Program of Jiangsu Province (BE2019706).\nData Availability No datasets were generated or analysed during the \ncurrent study.\nDeclarations\nEthics Approval and Consent to Participate The animal study was ap -\nproved by Nanjing University Animal Care and Use Committee. The \nstudy was conducted inaccordance with the local legislation and insti-\ntutional requirements.\nPatient Consent for Publication Not applicable.\nCompeting interests The authors declare no competing interests.\nOpen Access   This article is licensed under a Creative Commons \nAttribution-NonCommercial-NoDerivatives 4.0 International License, \nwhich permits any non-commercial use, sharing, distribution and \nexplored the effect of E2 on LPMs in this study, but not P4. \nWe found that the amount of LPMs migrating to the dam -\naged uterus in mice was proportional to E2 levels. Further \nin vitro experiments also confirmed that E2 could promote \nLPMs migration by regulating the expression of a series of \nmigration-related genes and promoting cytoskeletal elon -\ngation. Our results also showed that ER-β was specifically \nexpressed in LPMs and its expression level was upregulated \nby E2 stimulation. There may be a positive feedback mecha-\nnism to drive the migration of LPMs to the uterus, that is, \nthe higher the E2 level is, the more ER-β is expressed, and \nthe migration of LPMs to the uterus is promoted. The closer \nto the uterus, the E2 level is further increased, which fur -\nther promotes the recruitment of LPMs there. E2 has been \nwidely applied in the treatment of diseases related to endo -\nmetrium injury [ 29–31], our study may provide some new \nexplanations for its therapeutic effects.\nIn conclusion, our study has demonstrated the capacity of \nLPMs to swiftly migrate into the damaged endometrium of \nthe uterus (Fig. 6). This response occurs rapidly, requiring \nonly a few hours. Once in the uterus, LPMs contribute to \ntissue repair through phagocytosis of dead cells and sup -\npression of inflammation. Additionally, we observed that \nthe migration efficiency of LPMs was greater in mice with \nhigher levels of E2. In vitro experiments further confirmed \nthat E2 promotes the migration of LPMs. These findings \nunderscore the crucial role of LPMs in the rapid repair of \nthe endometrium.\nFig. 6 Diagram of the mechanism \nof LPMs promoting the repair of \ndamaged endothelium\n \n1 3\n4440\n\nInflammation (2025) 48:4428–4442\n16. Zhou, J. Z., S. S. Way, and K. Chen. 2018. Immunology of the \nuterine and vaginal mucosae. Trends in Immunology 39:302–314.\n17. Lv, H., H. Sun, L. Wang, et al. 2023. Targeting CD301+ mac -\nrophages inhibits endometrial fibrosis and improves pregnancy \noutcome. EMBO Molecular Medicine 15:e17601.\n18. Critchley, H. O. D., J. A. Maybin, G. M. Armstrong, A. R. W. Wil-\nliams, et al. 2020. Physiology of the endometrium and regulation \nof menstruation. Physiological Reviews 100:1149–1179.\n19. Salm, L., R. Shim, N. Noskovicova, et al. 2023. Gata6+ large \nperitoneal macrophages: An evolutionarily conserved sentinel \nand effector system for infection and injury. Trends in Immunol-\nogy 44:129–145.\n20. Liu, Z., Y . Gu, S. Chakarov, et al. 2019. Fate mapping via Ms4a3-\nexpression history traces monocyte-derived cells. Cell 178 (6): \n1509-1525.e19.\n21. Deniset, J. F., D. Belke, W. Y . Lee, et al. 2019. Gata6+ pericardial \ncavity macrophages relocate to the injured heart and prevent car-\ndiac fibrosis. Immunity 51:131-140.e5.\n22. Louwe, P. A., L. Badiola Gomez, H. Webster, et al. 2021. \nRecruited macrophages that colonize the post-inflammatory peri-\ntoneal niche convert into functionally divergent resident cells. \nNature Communications 12:1770.\n23. Wang, J., P. A. Kubes, et al. 2016. Reservoir of mature cavity \nmacrophages that can rapidly invade visceral organs to affect tis-\nsue repair. Cell 165:668–78.\n24. Honda, M., M. Kadohisa, D. Yoshii, et al. 2021. Directly recruited \nGATA6 + peritoneal cavity macrophages contribute to the repair \nof intestinal serosal injury. Nature Communications 12:7294.\n25. Zindel, J., M. Peiseler, M. Hossain, et al. 2021. Primordial \nGATA6 macrophages function as extravascular platelets in sterile \ninjury. Science 371:eabe0595.\n26. Percie du Sert, N., V . Hurst, A. Ahluwalia, et al. 2020. The arrive \nguidelines 2.0: Updated guidelines for reporting animal research. \nBritish Journal of Pharmacology 177:3617–3624.\n27. Ghosn, E. E., A. A. Cassado, G. R. Govoni, T. Fukuhara, Y . Yang, \nD. M. Monack, et al. 2010. Two physically, functionally, and \ndevelopmentally distinct peritoneal macrophage subsets. Pro-\nceedings of the National Academy of Sciences USA 107:2568–73.\n28. Wang, J., D. Li, Y . Pan, et al. 2021. Interleukin-34 accelerates \nintrauterine adhesions progress related to CX3CR1+ monocytes/\nmacrophages. European Journal of Immunology 51:2501–2512.\n29. Chen, Y ., W. Fei, Y . Zhao, et al. 2020. Sustained delivery of \n17β-estradiol by human amniotic extracellular matrix (HAECM) \nscaffold integrated with PLGA microspheres for endometrium \nregeneration. Drug Delivery 27:1165–1175.\n30. Li, B., L. Zhang, Y . Xie, et al. 2022. Evaluation of pharmacoki -\nnetics and safety of a long-term estradiol-releasing stent in rat \nuterine. Regenerative Therapy 21:494–501.\n31. Zhang, S. S., X. X. Xu, W. W. Xiang, et al. 2020. Using \n17β-estradiol heparin-poloxamer thermosensitive hydrogel to \nenhance the endometrial regeneration and functional recovery \nof intrauterine adhesions in a rat model. The F ASEB Journal \n34:446–457.\n32. Krause, W. C., R. Rodriguez, B. Gegenhuber, et al. 2021. Oestro-\ngen engages brain MC4R signalling to drive physical activity in \nfemale mice. Nature 599:131–135.\n33. Jain, V ., R. R. Chodankar, J. A. Maybin, et al. 2022. Uterine \nbleeding: how understanding endometrial physiology underpins \nmenstrual health. Nature Reviews Endocrinology 18:290–308.\n34. Ang, C. J., T. D. Skokan, K. L. McKinley, et al. 2023. Mecha -\nnisms of regeneration and fibrosis in the endometrium. Annual \nReview of Cell and Developmental Biology 39:197–221.\n35. Li, S., K. Jiang, J. Li, et al. 2020. Estrogen enhances the prolif -\neration and migration of ovarian cancer cells by activating tran -\nsient receptor potential channel C3. Journal of Ovarian Research \n13:20.\nreproduction in any medium or format, as long as you give appropri -\nate credit to the original author(s) and the source, provide a link to the \nCreative Commons licence, and indicate if you modified the licensed \nmaterial. You do not have permission under this licence to share \nadapted material derived from this article or parts of it. The images or \nother third party material in this article are included in the article’s Cre-\native Commons licence, unless indicated otherwise in a credit line to \nthe material. If material is not included in the article’s Creative Com -\nmons licence and your intended use is not permitted by statutory regu-\nlation or exceeds the permitted use, you will need to obtain permission \ndirectly from the copyright holder. To view a copy of this licence, visit  \nh t t p  : / /  c r e a  t i  v e c  o m m o  n s .  o r g  / l i  c e n  s e s /  b y  - n c - n d / 4 . 0 /.\nReferences\n1. Wang, W., F. Vilella, P. Alama, et al. 2020. Single-cell transcrip-\ntomic atlas of the human endometrium during the menstrual \ncycle. Nature Medicine 26:1644–1653.\n2. Giudice, L. C. 2020. Multidimensional transcriptomic mapping \nof human endometrium at single-cell resolution. Nature Medicine \n26:1513–1514.\n3. Lee, W. L., C. H. Liu, M. Cheng, et al. 2021. Focus on the primary \nprevention of intrauterine adhesions: Current concept and vision. \nInternational Journal of Molecular Sciences 22:5175.\n4. Zhou, Z., H. Wang, X. Zhang, et al. 2022. Defective autophagy \ncontributes to endometrial epithelial-mesenchymal transition in \nintrauterine adhesions. Autophagy 18:2427–2442.\n5. Mouhayar, Y ., J. M. Franasiak, and F. I. Sharara. 2019. Obstetrical \ncomplications of thin endometrium in assisted reproductive tech-\nnologies: A systematic review. Journal of Assisted Reproduction \nand Genetics 36:607–611.\n6. Lv, H., G. Zhao, P. Jiang, et al. 2022. Deciphering the endometrial \nniche of human thin endometrium at single-cell resolution. Proc \nNatl Acad Sci U S A 119:e2115912119.\n7. Zhao, G., J. Dai, and Y . Hu. 2025. Development of regenerative \ntherapies targeting fibrotic endometrium in intrauterine adhesion \nor thin endometrium to restore uterine function. Science China \nLife Sciences.  h t t p s :   /  / d o  i . o  r  g  /  1 0  . 1 0   0 7 /  s 1 1  4 2 7 -  0 2 4 -  2 8 4 2 - 6.\n8. Zhao, G., and Y . Hu. 2024. Mechanistic insights into intrauterine \nadhesions. Seminars in Immunopathology 47 (1): 3.\n9. Zhang, X., H. Lv, Q. Weng, et al. 2025. “Thin endometrium” at \nsingle-cell resolution. American Journal of Obstetrics and Gyne-\ncology 232 (4S): S135–S148.\n10. Evans, J., L. A. Salamonsen, A. Winship, et al. 2016. Fertile \nground: human endometrial programming and lessons in health \nand disease. Nature Reviews Endocrinology 12:654–667.\n11. Li, J., Y . Pan, J. Yang, et al. 2022. Tumor necrosis factor-α-\nprimed mesenchymal stem cell-derived exosomes promote M2 \nmacrophage polarization via Galectin-1 and modify intrauterine \nadhesion on a novel murine model. Frontiers in Immunology  \n13:945234.\n12. Yang, J., J. Li, J. Wang, et al. 2023. Oroxylin A relieves intrauter-\nine adhesion in mice through inhibiting macrophage pyroptosis \nvia SIRT3-SOD2-ROS pathway. International Immunopharma-\ncology 118:110023.\n13. Jiang, Q., J. Li, Y . Pan, et al. 2022. Melatonin-primed MSCs alle-\nviate intrauterine adhesions by affecting MSC-expressed galec -\ntin-3 on macrophage polarization. Stem Cells 40:919–931.\n14. Zindel, J., and P. Kubes. 2020. DAMPs, PAMPs, and LAMPs in \nimmunity and sterile inflammation. Annual Review of Pathology: \nMechanisms of Disease 15:493–518.\n15. Silvestre-Roig, C., Q. Braster, A. Ortega-Gomez, et al. 2020. \nNeutrophils as regulators of cardiovascular inflammation. Nature \nReviews Cardiology 17:327–340.\n1 3\n4441\n\nInflammation (2025) 48:4428–4442\nPublisher’s Note Springer Nature remains neutral with regard to juris-\ndictional claims in published maps and institutional affiliations.\n36. Ariyani, W., W. Miyazaki, I. Amano, et al. 2020. Soy isoflavones \naccelerate glial cell migration via GPER-mediated signal trans -\nduction pathway. Frontiers in Endocrinology 11:554941.\n37. Zhao, X., X. Li, P. Liu, et al. 2022. 17β-estradiol promotes angio-\ngenesis through non-genomic activation of Smad1 signaling in \nendometriosis. Vascular Pharmacology 142:106932.\n1 3\n4442","source_license":"CC0","license_restricted":false}