Mechanisms by Which the Cervicovaginal Microbiota Mediates Endometrial Repair Following Uterine Injury: A Review of Recent Advances

The normal vaginal microbiota is dominated by Lactobacillus species. The abundance and stability of common vaginal lactobacilli play a critically important role in maintaining a healthy vaginal microenvironment and suppressing unfavorable microbial factors. 26 , 27 Using hierarchical clustering analysis, Ravel et al classified the vaginal microbiota into five community state types (CST) based on the composition and relative abundance of dominant species. CST I, II, III, and V are characterized by the predominance of Lactobacillus crispatus , L. gasseri , L. iners , and L. jensenii , respectively, and are associated with relatively low microbial diversity. 27 In contrast, CST IV is characterized by a reduced proportion of Lactobacillus and an elevated proportion of strictly anaerobic bacteria, and is closely associated with a dysbiotic state. 28 The protective capacity of different CSTs varies considerably. CST I is regarded as the most protective type, requiring a relative abundance of L. crispatus of at least 90% to effectively exclude non-commensal microorganisms. 29 CST III exhibits comparatively poor stability and represents a transitional state prone to shift toward dysbiosis. 30 CST IV is strongly associated with bacterial vaginosis, characterized by the enrichment of anaerobic bacteria, most notably Gardnerella vaginalis, and is associated with a markedly elevated risk of reproductive tract infection. 31 The composition of the vaginal microbiota undergoes dynamic changes across the female life cycle. During early puberty, when estrogen levels are low, vaginal microbial diversity is relatively high and Lactobacillus species have not yet attained dominance. As gonadal maturation proceeds and estrogen levels rise, the vaginal epithelium thickens and accumulates glycogen, providing a nutritional substrate for lactobacillary proliferation, and the microbiota gradually shifts toward Lactobacillus -dominant compositions. Following menopause, declining estrogen levels weaken Lactobacillus dominance, microbial diversity increases once again, and the risk of reproductive tract infection rises accordingly. 32 , 33 The traditional view held that the closed cervix and its mucus barrier maintained the uterine cavity in a sterile state. With the development of culture-independent molecular techniques, particularly 16S rRNA gene sequencing, microbial signals have been detected in uterine samples, challenging the concept of a completely sterile uterine cavity. 34 , 35 Studies based on samples obtained directly from the uterine cavity have reported heterogeneous microbial profiles. 36 , 37 Using 16S rRNA gene sequencing, Moreno et al 38 found that during the receptive phase of the endometrium, the uterine microbiota is highly stable, with Lactobacillus accounting for more than 90% of the microbial community in some women. In contrast, Chen et al 39 reported a progressive increase in microbial diversity and a marked decrease in Lactobacillus abundance along the ascending reproductive tract. Although Lactobacillus constituted more than 99.7% of the vaginal microbiota, it accounted for only 30.6% of the uterine microbiota, where Pseudomonas and Acinetobacter were more abundant. These findings highlight the substantial variability in reported uterine microbial profiles across studies. These discrepant findings should be interpreted in light of several methodological considerations. The uterus is a low-biomass environment, making microbial detection particularly susceptible to contamination from cervicovaginal carryover during sampling, as well as from reagents/kits and laboratory handling and processing. 40 The origin of microorganisms detected in the uterine cavity remains uncertain. Proposed sources include ascending migration from the lower reproductive tract, hematogenous or lymphatic dissemination from distant sites such as the gut or oral cavity, and introduction during intrauterine procedures or embryo transfer. 41 , 42 Therefore, whether the detected microbial signals represent a distinct and stable resident uterine microbiome remains controversial, and current findings regarding uterine microbial composition should be interpreted cautiously. The endometrial immune microenvironment changes dynamically across the menstrual cycle. Neutrophils and macrophages participate in menstrual tissue breakdown and repair, whereas uterine natural killer cells, dendritic cells, and T-helper subsets contribute to immune regulation and vascular remodeling. 43 Innate mucosal defenses, including the epithelial barrier, antimicrobial peptides, and pattern-recognition receptors, maintain local homeostasis. Disruption of these defenses may trigger TLR4-mediated inflammatory signaling. 44 Microbiota-derived metabolites, particularly Lactobacillus -derived lactic acid, also mediate host–microbe crosstalk by modulating the local immune environment. CVM dysbiosis may compromise these defenses and perpetuate inflammatory signaling, thereby potentially impairing endometrial regeneration and repair. In chronic endometritis (CE), studies based on uterine or endometrial samples have reported reduced Lactobacillus abundance together with increased detection of potential pathogenic bacteria, including Enterococcus , Escherichia coli , Gardnerella , and Klebsiella . 45 , 46 In intrauterine adhesions (IUA), Qiu et al 47 demonstrated that the endometrial microbiota of IUA patients differs significantly from that of infertile women without intrauterine lesions. At the phylum level, the dominant taxa in IUA patients included Proteobacteria, Firmicutes, and Bacteroidetes; at the genus level, the proportions of Klebsiella , Shewanella , and Lactobacillus were all higher in IUA patients than in non-IUA women.

Normal endometrial function is a critical determinant of female reproductive capacity. Once the endometrium is damaged, its structural integrity and physiological function are compromised, interfering with embryo implantation and reducing pregnancy rates. 1 , 2 Endometrial regeneration and repair has therefore emerged as a major research focus in obstetrics and gynecology. Current clinical strategies for promoting endometrial repair include pharmacological treatment, bioactive materials and physical barrier therapies, and stem cell-based approaches. Although these modalities have demonstrated some efficacy, they are hampered by considerable inter-individual variability in response, inconsistent therapeutic outcomes, and incompletely understood mechanisms. 3 , 4 The repair mechanisms underlying endometrial injury remain insufficiently elucidated, highlighting an urgent need to explore new perspectives and potential breakthroughs. With the advancing progress of the Human Microbiome Project (HMP), research into the female reproductive tract microecosystem has attracted widespread attention. The HMP revealed that the female reproductive tract microbiota accounts for approximately 9% of the total human microbiome, and its composition is closely associated with female reproductive health. 5 Dysbiosis of the reproductive tract microbiota can trigger local inflammatory responses, impair endometrial receptivity, and has been closely linked to recurrent implantation failure and pregnancy loss. 6 , 7 Recent studies have further highlighted the clinical relevance of interactions among vaginal, urinary, and intestinal microbial ecosystems, suggesting that microbial dysbiosis at different anatomical sites may contribute to local immune disturbances and disease susceptibility. 8 , 9 The traditional view held that the cervix serves as a barrier preventing the ascending migration of vaginal microorganisms, and that the uterine cavity exists in a state of relative sterility. 10 However, the application of high-throughput sequencing technologies has demonstrated that the cervix does not constitute an absolute barrier to microbial ascent. Cervicovaginal microbiota (CVM) refers to the microbial communities of the vagina and cervix. Some studies have proposed that microorganisms detected in the uterine cavity may partly originate from the lower reproductive tract, suggesting a potential role in endometrial injury and repair. 11 , 12 Nevertheless, the specific mechanisms by which the CVM mediates endometrial repair remain poorly understood, and reviews on this topic are scarce. This review examines the mechanistic role of the CVM in post-injury endometrial repair, discusses relevant intervention strategies and future research directions, and aims to provide a theoretical basis for the management of endometrial injury.

A narrative literature search was conducted in PubMed, Embase, and Web of Science for relevant English-language publications available up to April 30, 2026. Search terms included combinations of “cervicovaginal microbiota,” “vaginal microbiota,” “endometrial microbiota,” “endometrial injury,” “endometrial repair,” “intrauterine adhesions,” “chronic endometritis,” “TGF-β1/Smad,” “TLR4/NF-κB,” and “Wnt/β-catenin.” Relevant original studies, experimental studies, clinical studies, and reviews were considered according to their relevance to the mechanisms linking the cervicovaginal microbiota with endometrial injury and repair.

The TGF-β1/Smad pathway is a well-established regulator of endometrial fibrosis. Upon binding to its receptor, TGF-β1 induces phosphorylation of Smad2 and Smad3, which then associate with Smad4 to form a trimeric complex that translocates to the nucleus and drives transcription of fibrosis-related genes such as COL1A1 . Smad3 can directly bind to collagen gene promoters to promote collagen expression and simultaneously suppresses matrix metalloproteinase activity in fibroblasts, thereby impeding extracellular matrix degradation. Smad7, an inhibitory regulator of this pathway, exerts negative control by competitively binding to the TGF-β type I receptor to block Smad2/3 phosphorylation and by promoting receptor degradation. 48 In both human IUA tissues and experimental models, increased TGF-β1 and Smad3 expression together with reduced Smad7 expression has been observed, supporting the involvement of this pathway in uterine scar formation. 49 , 50 Altered miRNA expression, particularly reduced miR-29, has also been associated with enhanced TGF-β1/Smad signaling and fibrotic changes in experimental and human endometrial stromal cell studies. 51 , 52 Available clinical evidence and review-level data suggest that reproductive tract dysbiosis may be associated with altered inflammatory and fibrotic responses. 53 , 54 Reduced Lactobacillus abundance and lactic acid levels have been proposed to favor a pro-inflammatory environment characterized by increased IL-6, IL-1, IL-8, and TNF-α. IL-6 may enhance Th1 cell survival and IFN-γ production, whereas IL-33 and TSLP released from injured endometrial epithelial cells may promote Th2 responses and fibrotic remodeling. 55 Mechanistic support is stronger in animal models. Wu et al 56 showed that vaginal administration of L. crispatus in IUA mice reduced uterine TGF-β1, p-Smad2, and p-Smad3 expression, increased MMP-9 activity, and attenuated extracellular matrix deposition and fibrosis ( Figure 1 ). Similarly, vaginal delivery of engineered L. crispatus suppressed TGF-β1/Smad signaling while restoring vaginal microbiota homeostasis in mice. 57 These findings support a potential interaction between the microbiota and the TGF-β1/Smad pathway, although confirmation in human studies is still needed. Figure 1 Proposed mechanism of L. crispatus in alleviating IUA via the TGF-β1/Smad pathway. The diagram illustrates the proposed mechanism of L. crispatus in modulating the TGF-beta1/Smad pathway. In the extracellular space, L. crispatus is shown to inhibit the TGF-beta1 receptor, reducing phosphorylation of Smad2 and Smad3. These phosphorylated Smad proteins form a complex with Smad4 in the cytoplasm and translocate to the nucleus. The pathway includes CTGF and MMP-9, which are involved in ECM aberrant deposition. Smad7 inhibits the phosphorylation of Smad2 and Smad3. Activation and inhibition interactions are indicated by arrows, with proposed activation and inhibition shown by dashed lines. The diagram spans the extracellular, cytoplasm and nucleus compartments. L. crispatus impacts TGF-beta1/Smad pathway, showing interactions in extracellular, cytoplasm and nucleus. Abbreviations : TGF-β1, transforming growth factor-β1; Smad, small mothers against decapentaplegic; CTGF, connective tissue growth factor; MMP, matrix metalloproteinase; ECM, extracellular matrix; ( P ) phosphorylation. Proposed mechanism of L. crispatus in alleviating IUA via the TGF-β1/Smad pathway. Toll-like receptor 4 (TLR4) is a central receptor of the innate immune system responsible for recognizing pathogen-associated molecular patterns. Upon binding of pathogen-derived lipopolysaccharide (LPS) to TLR4, the downstream adaptor protein MyD88 is sequentially activated, leading to phosphorylation of the NF-κB inhibitory protein (IκB), release of NF-κB, and subsequent transcription of large quantities of pro-inflammatory cytokines including IL-1β, IL-6, and TNF-α. 58 , 59 In human endometrial tissue, aberrant activation of the TGF-β1/NF-κB pathway further dysregulates the activity of collagen metabolism-related proteases: mRNA expression of ADAM15, ADAM17, and PAI-1 is markedly increased, while expression of uPA, which promotes fibrinolysis, is significantly reduced. Together, these changes promote aberrant collagen deposition in the extracellular matrix, suppress fibrinolysis, and ultimately drive tissue fibrosis and IUA formation. 60 , 61 Experimental evidence suggests that pathogenic bacterial overgrowth associated with CVM dysbiosis may promote persistent activation of the TLR4/NF-κB pathway and local inflammatory responses, whereas Lactobacillus may exert anti-inflammatory effects through modulation of this pathway. In an IUA mouse model, vaginal administration of engineered L.crispatus carrying CXCL12 reduced serum and uterine IL-1β and TNF-α levels, suppressed TLR4/NF-κB signaling, increased Lactobacillus abundance, and reduced Klebsiella abundance. 57 A schematic summary of this proposed mechanism is shown in Figure 2 . Figure 2 Proposed mechanism of CVM-mediated regulation of endometrial repair via the TLR4/NF-κB pathway. The diagram illustrates the TLR4/NF-kappa B pathway. LPS activates TLR4 on the cell membrane, leading to MyD88 activation in the cytoplasm. This triggers phosphorylation of I kappa B, releasing NF-kappa B, which translocates to the nucleus. In the nucleus, NF-kappa B promotes transcription of pro-inflammatory factors IL1B, IL6 and TNF. This results in increased expression of ADAM15, ADAM17, PAI-1 and decreased uPA, contributing to ECM aberrant deposition. CXCL12-engineered L. crispatus is shown to inhibit this pathway, reducing inflammation. Activation is indicated by solid arrows, while proposed inhibition is shown with dashed lines. TLR4/NF-kB pathway: LPS activates, CXCL12-L. crispatus inhibits, impacting inflammation and ECM deposition. Notes : ↑ indicates increased expression; ↓ indicates decreased expression. Abbreviations : LPS, lipopolysaccharide; TLR4, Toll-like receptor 4; CXCL12, C-X-C motif chemokine ligand 12; MyD88, myeloid differentiation primary response 88; IκB, inhibitor of nuclear factor kappa B; NF-κB, nuclear factor kappa B; ADAM, a disintegrin and metalloproteinase; PAI-1, plasminogen activator inhibitor-1; uPA, urokinase-type plasminogen activator; ECM, extracellular matrix; P, phosphorylation. Proposed mechanism of CVM-mediated regulation of endometrial repair via the TLR4/NF-κB pathway. The Wnt/β-catenin signaling pathway is a core regulatory system governing cell proliferation, differentiation, and tissue regeneration, and plays an indispensable role in both cyclic endometrial regeneration and post-injury repair. In the canonical Wnt signaling pathway, binding of Wnt ligands to the cell surface receptor Frizzled and co-receptor LRP5/6 inhibits the activity of the β-catenin destruction complex comprising APC, Axin, and GSK-3β. This prevents β-catenin from undergoing phosphorylation-dependent degradation, allowing it to accumulate in the cytoplasm and subsequently translocate to the nucleus, where it associates with TCF/LEF transcription factors to activate downstream target gene transcription, promoting endometrial epithelial cell proliferation and differentiation and driving endometrial regeneration and repair. 62 In human adhesion endometrial tissue from patients with IUA, increased expression of non-phosphorylated β-catenin, Wnt5a, and Wnt7a has been observed, together with the presence of CD140b⁺CD146⁺ endometrial stem cell-like cells. 63 Animal studies further suggest that the effects of Wnt signaling are context- and duration-dependent, as transient activation may facilitate endometrial repair, whereas sustained activation may promote fibrosis. 64 These findings support an important role for Wnt/β-catenin signaling in endometrial repair. Evidence linking microbiota alterations to Wnt/β-catenin signaling is still emerging. In an IUA rat model, Zeng et al 65 reported reduced endometrial microbial diversity and β-catenin expression, while combined growth hormone and estrogen treatment was associated with restoration of microbial diversity, increased β-catenin expression, and reduced fibrosis. Molecular docking suggested a potential interaction between the microbial metabolite citraconic acid and β-catenin, and antibiotic intervention reduced the therapeutic effect. These findings raise the possibility that microbiota-derived metabolites may influence Wnt/β-catenin signaling, but they do not establish a fixed causal pathway. In an in vitro study, Fazeli et al 66 found that L.crispatus extract increased AXIN2 expression in human endometrial stem cells, providing preliminary cell-level evidence of a possible interaction between Lactobacillus-derived components and Wnt signaling. A schematic summary of these proposed interactions is shown in Figure 3 . Figure 3 Proposed mechanism of CVM-mediated regulation of endometrial repair via the Wnt/β-catenin pathway. The diagram illustrates the Wnt/beta-catenin signaling pathway and the proposed microbiota-related interactions. Extracellular components include L. crispatus and citraconic acid. L. crispatus is linked to the Wnt/beta-catenin pathway via a proposed modulation arrow. Citraconic acid is shown with a predicted binding arrow to beta-catenin. The Wnt ligand binds to Frizzled and LRP5/6, inhibiting the beta-catenin destruction complex composed of APC, AXIN and GSK-3 beta. This inhibition leads to the accumulation of beta-catenin in the cytoplasm, which then translocates to the nucleus. In the nucleus, beta-catenin binds to TCF/LEF, activating target gene transcription. The process promotes cell proliferation and differentiation. Activation, inhibition, proposed modulation and predicted binding are indicated by different arrow styles. L. crispatus and citraconic acid regulate Wnt/beta-catenin pathway, impacting cell growth and gene transcription. Note : ↑ indicates increased expression. Abbreviations : LRP5/6, low-density lipoprotein receptor-related protein 5/6; AXIN, axis inhibition protein; APC, adenomatous polyposis coli; GSK-3β, glycogen synthase kinase-3 beta; TCF/LEF, T-cell factor/lymphoid enhancer-binding factor; P, phosphorylation. Proposed mechanism of CVM-mediated regulation of endometrial repair via the Wnt/β-catenin pathway.

Microbiota-targeted interventions are being explored for endometrial injury-related conditions, but the strength of evidence varies substantially across different approaches. Key preclinical and clinical studies on microbiota-mediated endometrial repair and microbiota-targeted interventions are summarized in Table 1 . Table 1 Summary of Key Preclinical and Clinical Studies on Microbiota-Mediated Endometrial Repair Study Study Design/Model Condition and Intervention/Exposure Main Findings Salma et al (2016) 49 Human case-control + rabbit IUA model IUA; Smad3 inhibitor SIS3 in rabbits TGF-β1/Smad3 increased and Smad7 decreased in IUA; SIS3 attenuated adhesions in rabbits. Zhao et al (2020) 53 Prospective observational study IUA; vaginal/cervical microbiota and cytokine profiles IUA was associated with higher bacterial burden, lower diversity, and a pro-inflammatory/pro-fibrotic cytokine profile. Wu et al (2023) 56 IUA mouse model + randomized clinical comparison after TCRA IUA; vaginal L. crispatus ; estrogen comparator In mice, L. crispatus suppressed TLR4/NF-κB and TGF-β1/Smad signaling and reduced fibrosis; clinically, re-adhesion was reduced versus estrogen. Kong et al (2022) 57 Preclinical IUA mouse model IUA; CXCL12-engineered L. crispatus Reduced IL-1β and TNF-α, inhibited TLR4/NF-κB and TGF-β1/Smad signaling, and promoted repair. Wang et al (2017) 58 Human endometrial tissue + rat model Asherman syndrome/IUA; NF-κB activation NF-κB p65 expression and nuclear localization increased in adhesion endometrium; rat findings supported sustained activation. Xue et al (2022) 63 Human tissue/cell + rat model Asherman syndrome/IUA; temporal Wnt/β-catenin activation Transient Wnt activation supported repair, whereas prolonged activation was associated with fibrosis. Zeng et al (2025) 65 IUA rat model + cell/multi-omics analyses IUA; growth hormone + estrogen; citraconic acid Treatment restored microbial diversity and β-catenin and reduced fibrosis; citraconic acid showed a potential β-catenin interaction. Fazeli et al (2025) 66 In vitro human endometrial stem-cell model L. crispatus extract Increased AXIN2 expression, suggesting possible Wnt-pathway modulation. He et al (2025) 67 CE mouse model + clinical study CE; L. crispatus Chen-01 Reduced inflammatory signaling in mice; clinical improvements in pathology and pregnancy outcomes were reported. Hu et al (2024) 68 Retrospective cohort CE/infertility; doxycycline + vaginal Lactobacillus vs doxycycline No significant improvement in biochemical or clinical pregnancy outcomes; premature rupture of membranes was less frequent. Naghi Jafarabadi et al (2024) 69 Randomized controlled trial Recurrent implantation failure; intravaginal probiotics before FET No significant improvement in chemical or clinical pregnancy rates. Liu et al (2024) 70 Preclinical thin-endometrium mouse model Thin endometrium; G-CSF-engineered L. crispatus Increased endometrial thickness, glands, and vascular density and reduced fibrosis. Lu et al (2022) 71 Human vaginal profiling + mouse endometriosis model Endometriosis; vaginal antibiotics or VMT In mice, VMT and antibiotics reduced lesion progression and NF-κB-associated inflammation. Abbreviations : CE, chronic endometritis; FET, frozen embryo transfer; G-CSF, granulocyte colony-stimulating factor; IUA, intrauterine adhesion; RCT, randomized controlled trial; TCRA, transcervical resection of adhesions; VMT, vaginal microbiota transplantation. Summary of Key Preclinical and Clinical Studies on Microbiota-Mediated Endometrial Repair Abbreviations : CE, chronic endometritis; FET, frozen embryo transfer; G-CSF, granulocyte colony-stimulating factor; IUA, intrauterine adhesion; RCT, randomized controlled trial; TCRA, transcervical resection of adhesions; VMT, vaginal microbiota transplantation. Probiotic preparations currently represent the most established approach for targeting the reproductive tract microbiota, with L. crispatus being the most extensively studied candidate strain. 72 In CE, L. crispatus isolated from healthy women of reproductive age suppressed uterine inflammation and promoted embryo implantation in a mouse model; a subsequent clinical study reported improvements in progesterone levels, pathological findings, and pregnancy outcomes in patients with CE. 67 In IUA, Wu et al 56 reported in a randomized controlled trial that vaginal administration of L. crispatus following transcervical resection of adhesions reduced postoperative intrauterine re-adhesion compared with estrogen therapy. However, the benefits of microbiota-targeted interventions have not been consistent across studies. Hu et al 68 found that the addition of vaginal Lactobacillus to antibiotic treatment in patients with CE did not significantly improve clinical pregnancy or live birth rates compared with antibiotics alone.66 Similarly, randomized trials of intravaginal probiotics before embryo transfer reported no significant improvement in biochemical or clinical pregnancy rates, although potential benefits were observed for selected outcomes or subgroups. 69 These findings suggest that the efficacy of probiotic interventions may depend on patient characteristics, baseline microbiota composition, and treatment protocols. Engineered probiotics represent a more experimental approach. In a mouse model of thin endometrium, intrauterine administration of engineered L. crispatus expressing granulocyte colony-stimulating factor increased endometrial thickness, gland number, and vascular density, while reducing fibrosis and intrauterine Pseudomonas abundance. 70 Although these findings suggest potential advantages of combining microbiota modulation with targeted delivery, evidence remains limited to preclinical studies. Vaginal microbiota transplantation (VMT) is an emerging intervention strategy that has attracted growing interest in recent years. Animal experiments have shown that VMT can effectively reduce local inflammatory levels and promote Lactobacillus colonization, demonstrating efficacy comparable to antibiotic treatment in reproductive tract inflammation-related disease models. 71 However, evidence directly supporting VMT for endometrial repair remains very limited, and its clinical application in this context has not been established. Standardization of donor screening, transplantation protocols, and long-term safety assessment will be required before further clinical translation.

Post-injury endometrial repair is a complex biological process involving coordinated immune responses, fibrosis, and tissue regeneration. Current evidence suggests a biologically plausible interaction among the CVM, local immune regulation, and signaling pathways including TLR4/NF-κB, TGF-β1/Smad, and Wnt/β-catenin. Microbiota-derived metabolites may also contribute to this crosstalk, as suggested by emerging findings such as the potential interaction between citraconic acid and β-catenin. However, these mechanisms are supported mainly by experimental and preclinical studies, while direct causal evidence in women remains limited. Microbiota-targeted approaches, including conventional probiotics, engineered probiotics, and VMT, may represent potential strategies for modulating the reproductive tract microenvironment. Among these, conventional probiotics have been evaluated clinically, whereas engineered probiotics and VMT remain largely experimental in the context of endometrial repair. Therefore, their clinical efficacy, safety, and applicability require further validation. Important methodological challenges also remain, including uncertainty regarding the existence and origin of a stable resident uterine microbiome and the risk of contamination in low-biomass samples. Future studies should prioritize anatomically site-specific sampling, rigorous contamination control, standardized sequencing and bioinformatic methods, and longitudinal follow-up to clarify temporal relationships between microbiota changes and endometrial repair. Mechanistic findings should be further validated in human tissues and well-designed experimental models, followed by large multicenter randomized controlled trials before microbiota-targeted strategies can be considered for routine clinical management.

The endometrium, as a core component of the female reproductive system, possesses remarkable capacity for cyclic regeneration. Anatomically, the endometrium is divided into two layers: the functional layer and the basal layer. The functional layer comprises approximately two-thirds of the total endometrial thickness and consists of the compacta and spongiosa sublayers. Its physiological activity is cyclically regulated by hormonal signals governed by the hypothalamic–pituitary–ovarian axis, manifesting as rhythmic phases of proliferation, secretion, and shedding. This layer must maintain sufficient thickness and adequate blood supply to support successful embryo implantation; consequently, its structural integrity is critical for the implantation of the fertilized ovum. The basal layer constitutes the remaining one-third of endometrial thickness and is composed of glandular bases, dense stromal cells, and a vascular network. Although it does not participate in cyclic shedding, it provides the structural foundation and cellular source for post-injury endometrial repair, and its integrity directly determines the efficiency of endometrial regeneration. 13 , 14 Following physiological shedding or tissue injury, endometrial repair proceeds through a highly coordinated sequence involving inflammatory responses and their resolution, re-epithelialization, cellular proliferation, angiogenesis, and extracellular matrix remodeling. Neutrophils and macrophages participate in the clearance of cellular debris and damaged tissue during the early phase, followed by proliferation of epithelial and stromal cells and reconstruction of the vascular network. A tightly regulated transition from inflammation to regeneration is essential for restoring tissue integrity while preventing excessive extracellular matrix deposition and fibrosis. 15 Endometrial injury is induced by a variety of factors, among which intrauterine procedures, uterine ischemia, and infection represent the principal causes. Data indicate that approximately 90% of intrauterine adhesions are associated with pregnancy-related uterine interventions, and the greater the number of curettage procedures performed, the more severe the resulting endometrial damage. 16 These injurious factors impair endometrial function primarily through the following mechanisms: disruption of the basal layer architecture, leading to a reduction in endometrial stem cell number and functional impairment; damage to the endometrial vascular network, resulting in diminished local blood supply; and induction of chronic inflammatory responses that suppress endometrial regeneration. 17 , 18 In addition, individual differences in genetic background and reproductive tract microecological imbalance may further compromise endometrial repair potential and increase the risk of injury-related endometrial disorders. 19 , 20 The pathological mechanisms underlying endometrial injury involve dysregulation at multiple levels, including basal layer fibrosis, impaired cell differentiation, disrupted angiogenesis, and scar tissue formation. 21–23 At the molecular level, the TGF-β/Smad signaling axis promotes fibroblast activation and excessive extracellular matrix deposition, serving as a central driver of endometrial fibrosis; 24 the Wnt/β-catenin pathway participates in regulating the proliferation and differentiation of endometrial epithelial cells, and its aberrant activation is closely associated with impaired endometrial regeneration. 25 Current clinical interventions targeting these injury mechanisms remain limited by inconsistent therapeutic efficacy and high recurrence rates, suggesting that the regulatory network governing endometrial injury has yet to be fully elucidated. In recent years, the reproductive tract microecosystem has attracted growing attention for its potential role in modulating the endometrial microenvironment, offering a new perspective for understanding the mechanisms of post-injury endometrial repair.