UBE2S-mediated deubiquitination of GLUT1 via USP10 regulates glucose metabolic reprogramming and immune microenvironment to promote fibrosis in endometriosis

To explore potential molecular biomarkers for EM, we conducted transcriptomic sequencing on three NESCs and three EESCs. Following our screening strategy (Fig. 1 A), we first removed redundant data and selected the most highly expressed transcripts for genes with multiple isoforms. A total of 20,311 genes were identified (Table S4 ), and by applying a |log2FC| >1 and FDR < 0.001 as selection thresholds, 4,404 differentially expressed genes (DEGs) were identified. We next analyzed the transcriptome data using GO and GSEA enrichment analyses to uncover key biological processes and pathways involved in EM. GO analysis of DEGs revealed significant enrichment of several biological processes, including “positive regulation of proteolysis”, “regulation of protein ubiquitination”, “positive regulation of ubiquitin-dependent protein catabolic process”, and “membrane protein proteolysis” (Fig. S1 A). In parallel, GSEA was performed on a pre-ranked list of all expressed genes, which identified significant enrichment of gene sets such as “Positive Regulation of Protein Metabolic Process”, “Ubiquitin Protein Transferase Activator Activity”, and “Positive Regulation of Proteolysis” (Figs. S1 B–D). These results suggest that ubiquitin-mediated proteolytic pathways are transcriptionally dysregulated in endometriosis and may contribute to disease progression. To further identify genes involved in the ubiquitin-proteasome pathway (UPPGs) that might be closely related to EM, we curated a list of 181 UPPGs [ 13 ] and intersected them with the 4,404 DEGs, revealing 28 UPPGs with differential expression in EESCs (8 upregulated and 20 downregulated genes). Based on these results, we generated a volcano plot (Fig. 1 B), rendering the 28 differentially expressed UPPGs and a heatmap to depict their expression patterns visually (Fig. 1 C). In this study, the immune score refers to the quantitative estimation of immune cell abundance, which was calculated using the xCell algorithm implemented in the R package IOBR [ 14 ]. Specifically, we assessed the immune cell abundance and performed Pearson’s correlation (cor.test) between the expression of the 28 ubiquitin–proteasome pathway-related genes and immune cell infiltration levels. After filtering with p -value 0.9, we identified three genes with the strongest correlations, which were visualized in a Venn diagram (Fig. 1 D). Further analysis of the PPI network of the 28 UPPG-related DEGs using the STRING database revealed that UBE2S exhibited the highest degree of interaction (Fig. 1 E). Although STRING analysis highlights hub genes based on network connectivity, such findings alone cannot establish functional importance in endometriosis. Therefore, we next validated UBE2S expression in EESCs and performed functional experiments to assess its biological role in EM. Fig. 1 Upregulation of UBE2S expression in EM compared to that in the normal endometrium. ( A ) A diagram illustrating the screening strategy for identifying UBE2S with selective dependency in EM sequencing data. ( B ) Volcano plot illustrating the differential expression of UPPGs between the two groups. ( C ) Heatmap displaying the differential expression of UPPGs between the two groups. ( D ) Venn diagram depicting the overlap of RNA sequencing data from EM, UPPGs, and immune signatures. ( E ) STING database analysis of protein-protein interactions for 28 genes. ( F ) Western blot analysis of UBE2S expression in NM, Eu, and EC tissues. ( G ) qRT-PCR analysis of UBE2S mRNA expression levels in NESCs, EuSCs, and EESCs. ( H ) Western blot analysis of UBE2S protein levels in NESCs, EuSCs, and EESCs. * p  < 0.05, ** p  < 0.01, *** p  < 0.001, **** p  < 0.0001 Upregulation of UBE2S expression in EM compared to that in the normal endometrium. ( A ) A diagram illustrating the screening strategy for identifying UBE2S with selective dependency in EM sequencing data. ( B ) Volcano plot illustrating the differential expression of UPPGs between the two groups. ( C ) Heatmap displaying the differential expression of UPPGs between the two groups. ( D ) Venn diagram depicting the overlap of RNA sequencing data from EM, UPPGs, and immune signatures. ( E ) STING database analysis of protein-protein interactions for 28 genes. ( F ) Western blot analysis of UBE2S expression in NM, Eu, and EC tissues. ( G ) qRT-PCR analysis of UBE2S mRNA expression levels in NESCs, EuSCs, and EESCs. ( H ) Western blot analysis of UBE2S protein levels in NESCs, EuSCs, and EESCs. * p  < 0.05, ** p  < 0.01, *** p  < 0.001, **** p  < 0.0001 To confirm UBE2S expression in EM, we performed western blot analysis on 12 paired samples of EC tissue, Eu tissue, and NM tissues. The findings indicated that UBE2S protein levels were significantly elevated in EC tissues compared to normal tissues (Fig.  1 F). Additionally, we measured UBE2S mRNA and protein levels in EESCs, EuSCs, and NESCs, revealing a significant upregulation of UBE2S at both the mRNA (Fig.  1 G) and protein (Fig.  1 H) levels in EESCs. These results suggest that UBE2S is overexpressed in both endometriotic tissues and cells. To investigate the biological function of UBE2S in EESCs, we generated UBE2S overexpression and knockdown cell lines and assessed their effects on cell proliferation and migration. Initially, we evaluated the impact of UBE2S expression on EESCs’ proliferation using CCK-8 and EdU assays. The results indicated that UBE2S overexpression significantly promoted EESCs’ proliferation (Fig.  2 A and C; p  < 0.05), whereas UBE2S knockdown significantly inhibited cell proliferation (Fig.  2 B and D; p  < 0.01). Furthermore, we performed Transwell and scratch assays to examine the effect of UBE2S on EESCs migration. The data demonstrated that overexpression of UBE2S significantly enhanced EESCs migration (Fig.  2 E and G; p  < 0.05), whereas UBE2S knockdown inhibited migration (Fig.  2 F and H; p  < 0.001). These in vitro experiments indicated that UBE2S promotes both the proliferation and migration of EESCs, suggesting its significant role in the pathogenesis of EM. Fig. 2 UBE2S promotes the proliferation and migration of EESCs in vitro. ( A - B ) CCK-8 assay to evaluate EESC proliferation after overexpression and knockdown of UBE2S. ( C - D ) EdU assay to assess the proliferation of EESCs following UBE2S overexpression and knockdown. ( E - F ) Transwell assay to evaluate the migration capacity of EESCs after UBE2S overexpression and knockdown. ( G - H ) Scratch assay to assess the migration ability of EESCs with UBE2S overexpression and knockdown. * p  < 0.05, ** p  < 0.01, *** p  < 0.001, **** p  < 0.0001 UBE2S promotes the proliferation and migration of EESCs in vitro. ( A - B ) CCK-8 assay to evaluate EESC proliferation after overexpression and knockdown of UBE2S. ( C - D ) EdU assay to assess the proliferation of EESCs following UBE2S overexpression and knockdown. ( E - F ) Transwell assay to evaluate the migration capacity of EESCs after UBE2S overexpression and knockdown. ( G - H ) Scratch assay to assess the migration ability of EESCs with UBE2S overexpression and knockdown. * p  < 0.05, ** p  < 0.01, *** p  < 0.001, **** p  < 0.0001 To further explore the molecular mechanisms by which UBE2S contributes to EM, we transfected Flag-UBE2S into EESCs and used a Flag-empty vector as a negative control. IP analysis revealed an additional band in the Flag-UBE2S group (Fig.  3 A). After excising and subjecting the specific bands to LC–MS/MS analysis, we identified GLUT1 as a potential binding partner of UBE2S. To further confirm the interaction between UBE2S and GLUT1, we performed endogenous immunoprecipitation using UBE2S and GLUT1 monoclonal antibodies in lysates from EESCs. The results confirmed that GLUT1 was present in the samples precipitated by UBE2S (Fig.  3 B), and UBE2S was detected in GLUT1-precipitated samples (Fig.  3 C). To validate this interaction, we co-transfected Flag-tagged GLUT1 and HA-tagged UBE2S plasmids into 293T cells, followed by immunoprecipitation with Flag-agarose beads to isolate GLUT1. Western blot analysis revealed the co-precipitation of HA-UBE2S with Flag-GLUT1 and vice versa (Fig. S2 A-B). Further analysis using UBE2S truncation mutants (HA-UBE2S-∆C, HA-UBE2S-∆N, HA-UBE2S-∆Core) (Fig. S2 C) in Co-IP assays revealed that GLUT1 interacted with UBE2S-WT, UBE2S-∆C, and UBE2S-∆Core, but not with UBE2S-∆N (Fig. S2 D). These results indicate that UBE2S and GLUT1 form a complex in cells and that the N-terminal domain of UBE2S is crucial for this interaction. Fig. 3 UBE2S forms a complex with GLUT1 to participate in EESCs’ glycolysis. ( A ) Mass spectrometry screening of UBE2S-interacting protein GLUT1. ( B - C ) Immunoprecipitation (IP) of whole-cell lysates from EESCs using monoclonal antibodies against UBE2S or GLUT1, followed by western blotting to detect the expression levels of GLUT1 and UBE2S. ( D ) Glucose uptake, lactate, and pyruvate levels measured in EM cells after UBE2S overexpression. ( E ) Glucose uptake, lactate, and pyruvate levels measured in EM cells after UBE2S knockdown. ( F ) Flag-UBE2S overexpression in UBE2S-KO EESCs and Western blot analysis of GLUT1 protein levels. ( G ) UBE2S shRNA transfection in EESCs and Western blot analysis of GLUT1 protein levels. ( H - J ) Glucose uptake, pyruvate and lactate measurements in EESCs transfected with the indicated plasmids. ( K - M ) Glucose uptake, pyruvate and lactate measurements in EESCs transfected with the indicated plasmids. ( N - P ) Glucose uptake, pyruvate and lactate measurements in EESCs transfected with the indicated plasmids. * p  < 0.05, ** p  < 0.01, *** p  < 0.001, **** p  < 0.0001 UBE2S forms a complex with GLUT1 to participate in EESCs’ glycolysis. ( A ) Mass spectrometry screening of UBE2S-interacting protein GLUT1. ( B - C ) Immunoprecipitation (IP) of whole-cell lysates from EESCs using monoclonal antibodies against UBE2S or GLUT1, followed by western blotting to detect the expression levels of GLUT1 and UBE2S. ( D ) Glucose uptake, lactate, and pyruvate levels measured in EM cells after UBE2S overexpression. ( E ) Glucose uptake, lactate, and pyruvate levels measured in EM cells after UBE2S knockdown. ( F ) Flag-UBE2S overexpression in UBE2S-KO EESCs and Western blot analysis of GLUT1 protein levels. ( G ) UBE2S shRNA transfection in EESCs and Western blot analysis of GLUT1 protein levels. ( H - J ) Glucose uptake, pyruvate and lactate measurements in EESCs transfected with the indicated plasmids. ( K - M ) Glucose uptake, pyruvate and lactate measurements in EESCs transfected with the indicated plasmids. ( N - P ) Glucose uptake, pyruvate and lactate measurements in EESCs transfected with the indicated plasmids. * p  < 0.05, ** p  < 0.01, *** p  < 0.001, **** p  < 0.0001 Immunohistochemical analysis revealed that both UBE2S and GLUT1 were more highly expressed in EC tissues than in NM tissues (Fig. S3 A). Western blot analysis of EC tissue ( n  = 12) versus eutopic endometrial tissue (Eu, n  = 12) and NM tissue ( n  = 12) confirmed that GLUT1 expression was significantly elevated in both EC (Fig. S3 B) and EESCs (Fig. S3 C) compared to that in NM and Eu ( p  < 0.001). GLUT1, a key transporter of glucose uptake, is crucial in aerobic glycolysis. To investigate whether UBE2S regulates aerobic glycolysis in EESCs, we assessed its effect on glucose uptake, lactate, and pyruvate production in EESCs. The findings indicated that UBE2S overexpression significantly increased glucose uptake and lactate and pyruvate production (Fig.  3 D), whereas UBE2S knockdown decreased these metabolic processes (Fig.  3 E). In UBE2S-KO EESCs with graded UBE2S expression, we observed a significant increase in GLUT1 protein levels (Fig.  3 F). Additionally, in EESCs transfected with UBE2S shRNA and a negative control plasmid, western blot analysis revealed that UBE2S knockdown led to a significant reduction in GLUT1 protein expression (Fig.  3 G; p  < 0.05). To further investigate whether UBE2S promotes glycolysis through GLUT1 regulation, we constructed GLUT1 knockdown plasmids (Fig. S4 ). The findings revealed that silencing GLUT1 and UBE2S significantly reduced glucose uptake and lactate and pyruvate production in EESCs (Fig.  3 H–J), and silencing GLUT1 reversed the increase in glucose uptake and lactate and pyruvate production induced by UBE2S overexpression (Fig.  3 K–M). Moreover, GLUT1 overexpression counteracted the minimized glucose uptake and lactate and pyruvate production after UBE2S knockdown (Fig.  3 N–P). These results demonstrated that UBE2S improves aerobic glycolysis in EESCs and promotes lactate production by regulating GLUT1. To investigate the influence of UBE2S on glucose metabolism under defined nutrient conditions, we used glucose-free culture medium and supplemented it with higher concentrations of glucose (0, 5, and 25 mM). Lactate production was measured in cells with either UBE2S overexpression or knockdown. In control cells, lactate levels increased in a glucose concentration-dependent manner. UBE2S overexpression further improved lactate accumulation across all glucose concentrations (Fig. S5 A), suggesting that UBE2S promotes glycolytic flux. Conversely, UBE2S knockdown significantly minimized lactate production (Fig. S5 B). These findings suggest that UBE2S facilitates glycolysis and lactate production in a glucose-dependent context, underscoring its central role as a metabolic modulator. UBE2S, a member of the E2 conjugating enzyme family, can extend ubiquitin chains on substrate proteins, thereby regulating protein degradation [ 15 , 16 ]. To investigate whether UBE2S regulates GLUT1 through the proteasomal pathway, we treated cells with the proteasome inhibitor MG132 (10 µM, 8 h). The findings indicated that in EESCs UBE2S-KO cells, the expression of GLUT1 was significantly higher in the MG132-treated group than in the untreated group (Fig. 4 A; p 0.05). Fig. 4 UBE2S reduces K48-linked polyubiquitination of GLUT1. ( A ) Proteasome inhibitor MG132 (10 µM, 8 h) treatment of EESCs UBE2S KO and WT cells, followed by western blotting to detect GLUT1 expression levels. ( B ) CHX (50 µg/mL) treatment of EESCs UBE2S-KO and WT cells at different time points, followed by western blotting to detect GLUT1 expression levels. ( C ) shUBE2S transfection in EESCs and detection of GLUT1 ubiquitination. ( D ) Flag-UBE2S transfection in EESCs and detection of GLUT1 ubiquitination. ( E ) 293T cells transfected with HA-ub-WT, Myc-UBE2S, or Flag-GLUT1 plasmids, followed by detection of GLUT1 ubiquitination levels. ( F ) 293T cells transfected with HA-Ub (K11R, K48R, K63R), Flag-GLUT1, and Myc-UBE2S plasmids, followed by Co-IP to detect GLUT1 ubiquitination levels. ( G ) 293T cells transfected with Myc-UBE2S, Flag-GLUT1, and HA-Ub (K11, K48, and K63) plasmids, followed by Co-IP to detect GLUT1 ubiquitination levels. ( H ) UBE2S-KO 293T cells transfected with Myc-UBE2S WT and its mutants (C95S, C118A, and DM), followed by western blotting to detect GLUT1 expression. ( I ) UBE2S-KO 293T cells transfected with Myc-UBE2S WT and its mutants (C95S, C118A, and DM), Flag-GLUT1, and HA-Ub-K48, followed by Co-IP to detect GLUT1 ubiquitination levels. * p  < 0.05, ** p  < 0.01, *** p  < 0.001, **** p  < 0.0001 UBE2S reduces K48-linked polyubiquitination of GLUT1. ( A ) Proteasome inhibitor MG132 (10 µM, 8 h) treatment of EESCs UBE2S KO and WT cells, followed by western blotting to detect GLUT1 expression levels. ( B ) CHX (50 µg/mL) treatment of EESCs UBE2S-KO and WT cells at different time points, followed by western blotting to detect GLUT1 expression levels. ( C ) shUBE2S transfection in EESCs and detection of GLUT1 ubiquitination. ( D ) Flag-UBE2S transfection in EESCs and detection of GLUT1 ubiquitination. ( E ) 293T cells transfected with HA-ub-WT, Myc-UBE2S, or Flag-GLUT1 plasmids, followed by detection of GLUT1 ubiquitination levels. ( F ) 293T cells transfected with HA-Ub (K11R, K48R, K63R), Flag-GLUT1, and Myc-UBE2S plasmids, followed by Co-IP to detect GLUT1 ubiquitination levels. ( G ) 293T cells transfected with Myc-UBE2S, Flag-GLUT1, and HA-Ub (K11, K48, and K63) plasmids, followed by Co-IP to detect GLUT1 ubiquitination levels. ( H ) UBE2S-KO 293T cells transfected with Myc-UBE2S WT and its mutants (C95S, C118A, and DM), followed by western blotting to detect GLUT1 expression. ( I ) UBE2S-KO 293T cells transfected with Myc-UBE2S WT and its mutants (C95S, C118A, and DM), Flag-GLUT1, and HA-Ub-K48, followed by Co-IP to detect GLUT1 ubiquitination levels. * p  < 0.05, ** p  < 0.01, *** p  < 0.001, **** p  < 0.0001 We then applied the protein synthesis inhibitor cycloheximide (CHX, 50 µg/mL) to stimulate EESCs and assessed GLUT1 protein expression at different time points after stimulation. The results revealed that compared to EESCs UBE2S-WT cells, the half-life of GLUT1 protein was significantly reduced in EESCs UBE2S-KO cells following CHX treatment (Fig.  4 B; p  < 0.01). These findings indicate that UBE2S stabilizes GLUT1 protein by inhibiting its proteasomal degradation, thereby increasing GLUT1 expression in endometrial stromal cells in EM. To explore whether UBE2S affects the ubiquitination levels of GLUT1, we assessed the ubiquitination of GLUT1 in EESCs with overexpressed or knocked-down UBE2S. We observed that knocking down UBE2S in EESCs led to an increase in GLUT1 ubiquitination (Fig.  4 C). Conversely, UBE2S overexpression in EESCs decreased GLUT1 ubiquitination (Fig.  4 D). We also transfected Myc-UBE2S, HA-Ub-WT, and Flag-GLUT1 into 293T cells, followed by Co-IP to assess GLUT1 ubiquitination levels using HA-Ub expression. The results indicated that UBE2S overexpression significantly reduced the ubiquitination of GLUT1 (Fig.  4 E). Since ubiquitination can mediate protein degradation or signal transduction through K11-, K48-, and K63-linked ubiquitin chains [ 17 ], we further investigated the specificity of GLUT1 ubiquitination in 293T cells by overexpressing specific ubiquitin mutants (K11R, K48R, and K63R). The results revealed that UBE2S mediates GLUT1 ubiquitination through the K48 linkage pathway (Fig. 4 F). The overexpression of Myc-UBE2S, Flag-GLUT1, and HA-Ub-K11, K48, K63 in 293T cells demonstrated that UBE2S overexpression significantly reduced K48-linked polyubiquitination of GLUT1 (Fig. 4 G). These findings demonstrated that UBE2S specifically regulated K48-linked ubiquitination of GLUT1, rather than exerting broad effects on other ubiquitin chain types. UBE2S exhibits E2 conjugating enzyme activity and has E3 ligase activity [ 18 , 19 ]. UBE2S exerts its ubiquitin E2 and E3 enzyme activities through two cysteine residues (Cys95 and Cys118) in the UBC domain of its molecular structure. To verify whether the regulation of GLUT1 by UBE2S depends on E2 or E3 enzyme activity, we constructed three UBE2S mutants: UBE2S-C95S (E2 activity loss), UBE2S-C118A (E3 activity loss), and UBE2S-DM (loss of both E2 and E3 activities). We transfected Myc-UBE2S WT and these three mutants into 293T UBE2S-KO cells. The findings indicated that both UBE2S WT and the mutants elevated GLUT1 expression (Fig. 4 H). All mutants significantly reduced GLUT1 ubiquitination (Fig. 4 I). These findings suggested that UBE2S-mediated K48 modification of GLUT1 does not depend on E2 or E3 enzymatic activity. Although UBE2S is an E2 ubiquitin-conjugating enzyme, our findings suggest that it may participate in the regulation of GLUT1 deubiquitination through yet unidentified deubiquitinases, and thus warrants further mechanistic investigation. Mass spectrometry analysis identified USP10 as a protein that interacts with UBE2S in EESCs. Based on this, we hypothesized that UBE2S recruits the deubiquitinase USP10 to remove K48-linked polyubiquitination on GLUT1 in EESCs. To determine whether the interaction between UBE2S and USP10 is direct, we performed a GST pull-down assay using recombinant proteins expressed in E. coli . Specifically, GST-tagged UBE2S was expressed and purified from E. coli using the pGEX-4T-1-UBE2S construct, while His-tagged USP10 was obtained from the pET-24a(+)-USP10 plasmid. Following incubation, GST-UBE2S successfully pulled down His-USP10, as confirmed by immunoblotting with anti-His antibody. These results indicate a direct physical interaction between UBE2S and USP10 in vitro (Fig.  5 A). Next, we lysed EESCs and performed immunoprecipitation using a monoclonal antibody against USP10 and Protein A + G agarose beads. The findings revealed that USP10 interacts with both UBE2S and GLUT1 in cells (Fig.  5 B). Immunoprecipitation using a monoclonal antibody against UBE2S also confirmed the interaction between UBE2S and USP10 in EESCs (Fig.  5 C). Fig. 5 UBE2S recruits USP10 to reduce K48-linked polyubiquitination of GLUT1. ( A ) Purified GST or GST-UBE2S proteins were incubated with purified His-USP10. After pull-down using glutathione agarose beads, the eluted proteins were subjected to SDS-PAGE followed by immunoblotting with anti-His antibody. Input and pull down fractions are shown. GST served as a negative control. ( B - C ) Immunoprecipitation of whole-cell lysates from EESCs using monoclonal antibodies against USP10 or UBE2S, followed by western blotting to detect the expression levels of USP10 and UBE2S. ( D - E ) Co-IP analysis of whole-cell lysates from 293T cells transfected with the indicated plasmids. ( F ) Co-IP analysis of whole-cell lysates from 293T cells transfected with the indicated plasmids. ( G ) Schematic representation of UBE2S and its truncation mutants (top), and co-immunoprecipitation (Co-IP) analysis showing the interaction between FLAG-tagged USP10 and HA-tagged full-length or truncated UBE2S mutants (∆N, ∆C, and ∆Core) in HEK293T cells (bottom). ( H ) UBE2S-KO 293T cells transfected with Myc-UBE2S, Flag-GLUT1, Myc-USP10, and HA-ub-K48 plasmids, followed by Co-IP analysis of whole-cell lysates. ( I ) USP10-KO 293T cells transfected with Myc-UBE2S, Flag-GLUT1, Myc-USP10, and HA-ub-K48 plasmids, followed by Co-IP analysis of whole-cell lysates UBE2S recruits USP10 to reduce K48-linked polyubiquitination of GLUT1. ( A ) Purified GST or GST-UBE2S proteins were incubated with purified His-USP10. After pull-down using glutathione agarose beads, the eluted proteins were subjected to SDS-PAGE followed by immunoblotting with anti-His antibody. Input and pull down fractions are shown. GST served as a negative control. ( B - C ) Immunoprecipitation of whole-cell lysates from EESCs using monoclonal antibodies against USP10 or UBE2S, followed by western blotting to detect the expression levels of USP10 and UBE2S. ( D - E ) Co-IP analysis of whole-cell lysates from 293T cells transfected with the indicated plasmids. ( F ) Co-IP analysis of whole-cell lysates from 293T cells transfected with the indicated plasmids. ( G ) Schematic representation of UBE2S and its truncation mutants (top), and co-immunoprecipitation (Co-IP) analysis showing the interaction between FLAG-tagged USP10 and HA-tagged full-length or truncated UBE2S mutants (∆N, ∆C, and ∆Core) in HEK293T cells (bottom). ( H ) UBE2S-KO 293T cells transfected with Myc-UBE2S, Flag-GLUT1, Myc-USP10, and HA-ub-K48 plasmids, followed by Co-IP analysis of whole-cell lysates. ( I ) USP10-KO 293T cells transfected with Myc-UBE2S, Flag-GLUT1, Myc-USP10, and HA-ub-K48 plasmids, followed by Co-IP analysis of whole-cell lysates In 293T cells, co-transfection of Flag-GLUT1 and HA-USP10, followed by Co-IP, confirmed that GLUT1 directly interacts with USP10 (Fig.  5 D). Similarly, the co-transfection of Flag-UBE2S and HA-USP10 revealed an interaction between UBE2S and USP10 (Fig.  5 E). Furthermore, co-transfection of recombinant Flag-GLUT1, Myc-UBE2S, and HA-USP10 into 293T cells demonstrated that both Myc-UBE2S and HA-USP10 were present in the immunoprecipitated Flag-GLUT1 samples, further supporting the interaction between GLUT1, USP10, and UBE2S (Fig.  5 F). To identify the structural domains of UBE2S required for its interaction with USP10, we co-transfected three truncated mutants of UBE2S (HA-UBE2S-∆C, HA-UBE2S-∆N, and HA-UBE2S-∆Core) and Flag-USP10 into 293T cells and performed Co-IP. The results revealed that USP10 interacted with UBE2S-WT, UBE2S-∆C, and UBE2S-∆Core, but not with UBE2S-∆N (Fig.  5 G). These findings suggested that the N-terminal domain of UBE2S is essential for its interaction with USP10. To explore whether UBE2S and USP10 cooperatively reduce the K48-linked ubiquitination of GLUT1, we first assessed whether USP10 directly interacts with GLUT1. Co-transfection of Flag-GLUT1 with Myc-UBE2S and Myc-USP10 in 293T UBE2S-KO cells demonstrated that in the absence of UBE2S, USP10 did not directly associate with GLUT1. However, in the presence of UBE2S, USP10 interacted with GLUT1 (Fig.  5 H). We further transfected Flag-GLUT1, Myc-UBE2S, and Myc-USP10 into 293T USP10-KO cells. The results revealed that overexpression of UBE2S alone did not reduce K48-linked ubiquitination of GLUT1. However, the co-expression of UBE2S and USP10 significantly decreased the K48-linked polyubiquitination of GLUT1 (Fig.  5 I). These data demonstrate that neither UBE2S nor USP10 alone can regulate the K48 deubiquitination of GLUT1. Only the simultaneous presence of UBE2S and USP10 can effectively exert deubiquitination activity. UBE2S uses its N-terminal domain as a scaffold to form a trimeric complex with GLUT1 and USP10, thereby reducing K48-linked polyubiquitination of GLUT1 and maintaining GLUT1 protein stability. UBE2S exhibits a high immune score in breast cancer, glioma, bladder cancer, and liver cancer, signifying its strong correlation with the immune microenvironment [ 20 , 21 ]. Abnormal infiltration of M2 macrophages is a hallmark of the chronic inflammatory changes observed in EM. To investigate whether UBE2S influences M2 macrophage polarization in EM, THP-1 cells were induced to differentiate into M0 macrophages using PMA, followed by co-culture with EESCs from different treatment groups. Macrophages were collected to assess M2 polarization markers (Fig. 6 A). The findings indicated that compared to M0 macrophages co-cultured with control EESCs, macrophages co-cultured with UBE2S-overexpressing EESCs exhibited significantly increased mRNA expression of CD163 (Fig. 6 B), Arg-1 (Fig. 6 C), and CD206 (Fig. 6 D). Moreover, knocking down GLUT1 in EESCs reversed the elevated M2 macrophage polarization induced by UBE2S overexpression (Fig. 6 B–D). Western blot analysis confirmed these findings at the protein level (Fig. 6 E). Fig. 6 UBE2S overexpression in EESCs enhances glycolysis and promotes M2 macrophage polarization. ( A ) Diagram of co-culture model. ( B - D ) qRT-PCR analysis of CD163, CD206, and Arg-1 mRNA levels in macrophages co-cultured with EESCs transfected with different plasmids for 48 h. ( E ) Western blot analysis of CD163, CD206, and Arg-1 protein levels in macrophages co-cultured with EESCs transfected with different plasmids. ( F - H ) EESCs overexpressing UBE2S treated with 2-DG or CPM for 24 h, followed by the collection of conditioned media (CM) to treat M0 macrophages. qRT-PCR analysis of CD163, CD206, and Arg-1 mRNA levels in macrophages. ( I ) EESCs overexpressing UBE2S treated with 2-DG or CPM for 24 h, followed by the collection of CM to treat M0 macrophages. Western blot analysis of CD163, CD206, and Arg-1 protein levels in macrophages. * p  < 0.05, ** p  < 0.01, *** p  < 0.001, **** p  < 0.0001 UBE2S overexpression in EESCs enhances glycolysis and promotes M2 macrophage polarization. ( A ) Diagram of co-culture model. ( B - D ) qRT-PCR analysis of CD163, CD206, and Arg-1 mRNA levels in macrophages co-cultured with EESCs transfected with different plasmids for 48 h. ( E ) Western blot analysis of CD163, CD206, and Arg-1 protein levels in macrophages co-cultured with EESCs transfected with different plasmids. ( F - H ) EESCs overexpressing UBE2S treated with 2-DG or CPM for 24 h, followed by the collection of conditioned media (CM) to treat M0 macrophages. qRT-PCR analysis of CD163, CD206, and Arg-1 mRNA levels in macrophages. ( I ) EESCs overexpressing UBE2S treated with 2-DG or CPM for 24 h, followed by the collection of CM to treat M0 macrophages. Western blot analysis of CD163, CD206, and Arg-1 protein levels in macrophages. * p  < 0.05, ** p  < 0.01, *** p  < 0.001, **** p  < 0.0001 The small-molecule inhibitor cephalomannine (CPM) inhibits UBE2S expression [ 22 ]. After transfecting Flag-UBE2S into EESCs (designated as OE-UBE2S-EESCs), the cells were cultured in a serum-free medium for 24 h. They were then pretreated with DMSO, 1 mM of the glycolysis inhibitor 2-DG, or 100 µM CPM for 24 h, followed by medium replacement and continued culture. After 24 h, the supernatant was collected and mixed with fresh medium at a 1:1 ratio to treat M0 macrophages for 48 h. Macrophages were then collected to assess M2 polarization. The results revealed that with the addition of 2-DG or CPM to the OE-UBE2S-EESCs culture medium and co-cultured with macrophages, M2 polarization of macrophages was significantly reduced at both the mRNA and protein levels (Fig. 6 F–I), with statistically significant differences. To directly assess the role of lactate as a downstream effector in the UBE2S/GLUT1 axis-mediated M2 macrophage polarization, we conducted rescue experiments using exogenous sodium lactate. Macrophages were co-cultured with UBE2S-shRNA or GLUT1-shRNA-transfected EESCs in the presence or absence of sodium lactate (10 mM). Western blot analysis revealed that the expression levels of M2 markers, including CD206, Arg1, and CD163, were significantly reduced in co-cultures with UBE2S- or GLUT1- silencing. However, supplementation with exogenous lactate partially restored the expression of these M2 polarization markers in both models (Figs. S6 A-B), suggesting that lactate is a functional mediator of UBE2S/GLUT1-induced M2 polarization. These findings indicated that UBE2S promotes GLUT1 expression, enhances glycolysis in EESCs, and facilitates M2 macrophage polarization in endometriotic lesions. Inhibition of UBE2S expression or glycolysis in EESCs reduced the polarization of M2 macrophages in endometriotic lesions. EM is a progressive fibrotic process characterized by tissue hardening, particularly excessive deposition of extracellular matrix during the fibrotic tissue remodeling (FMT) process [ 23 ]. Studies have indicated that during EM development, endogenous macrophage polarization influences tissue remodeling, with M2 macrophages exhibiting a positive correlation with fibrosis in EM. M2 macrophages secrete TGF-β1, which induces fibrosis in endometrial cancer cells [ 24 ]. Excessive migration and invasion of endometrial stromal cells are key indicators of fibrosis. In this context, we measured the expression of TGF-β1 and discovered that after co-culture with UBE2S-overexpressing EESCs (OE-UBE2S-EESCs), the TGF-β1 expression in macrophages was significantly increased (Fig. S7 A). Conversely, treatment with 2-DG or CPM markedly decreased TGF-β1 expression (Fig. S7 B). After investigating the effects of ectopic stromal cells on macrophage polarization, we examined the impact of macrophages on the biological functions of endometriotic stromal cells. We co-cultured UBE2S-overexpressing EESCs (M0-UBE2S) and control EESCs transfected with an empty vector (M0-NC) with macrophages for 48 h, followed by the removal of endometriotic cells. Fresh serum-free medium was then added to continue co-culture with macrophages for an additional 24 h. The supernatant was collected and mixed with a fresh serum-free medium at a 1:1 ratio to prepare a conditioned medium (CM) for EESC treatment (Fig.  7 A). The results revealed that treatment with M0-UBE2S-CM significantly enhanced the proliferation (Fig.  7 B–C) and migration (Fig.  7 D) abilities of endometriotic cells. The increased proliferation and migration abilities were significantly reduced after 2-DG treatment in the M0-UBE2S group (Fig.  7 B–D), with statistically significant differences. Similarly, after co-culturing UBE2S knockdown EESCs with macrophages for 48 h (designated M0-sh-UBE2S), a conditioned medium was prepared and used to treat EESCs. The results demonstrated that this treatment group reduced the proliferation and migration abilities of EESCs, whereas adding 10 mmol/L lactate to the macrophage-conditioned medium stimulated their proliferation and migration (Fig.  7 E–G; p  < 0.05). Fig. 7 M2-polarized macrophages promote proliferation, migration, and fibrosis in EM stromal cells. ( A ) Diagram of the model of EESCs treated with conditioned media (CM). ( B ) EdU assay to assess EESC proliferation. ( C ) CCK-8 assay to assess EESC proliferation. ( D ) Scratch assay to evaluate the migration ability of EESCs after treatment with conditioned media. ( E ) EdU assay to assess EESC proliferation. ( F ) CCK-8 assay to assess EESC proliferation. ( G ) Scratch assay used to evaluate the migration ability of EESCs after treatment with the conditioned media. ( H ) Western blot analysis of FMT markers (α-SMA, FN, Col-1) in EESCs after 48 h of conditioned medium treatment. ( I ) Western blot analysis of FMT markers in EESCs after 48 h of conditioned media treatment. * p  < 0.05, ** p  < 0.01, *** p  < 0.001, **** p  < 0.0001 M2-polarized macrophages promote proliferation, migration, and fibrosis in EM stromal cells. ( A ) Diagram of the model of EESCs treated with conditioned media (CM). ( B ) EdU assay to assess EESC proliferation. ( C ) CCK-8 assay to assess EESC proliferation. ( D ) Scratch assay to evaluate the migration ability of EESCs after treatment with conditioned media. ( E ) EdU assay to assess EESC proliferation. ( F ) CCK-8 assay to assess EESC proliferation. ( G ) Scratch assay used to evaluate the migration ability of EESCs after treatment with the conditioned media. ( H ) Western blot analysis of FMT markers (α-SMA, FN, Col-1) in EESCs after 48 h of conditioned medium treatment. ( I ) Western blot analysis of FMT markers in EESCs after 48 h of conditioned media treatment. * p  < 0.05, ** p  < 0.01, *** p  < 0.001, **** p  < 0.0001 We then evaluated the expression of FMT-related markers in the various EESC treatment groups. The results indicated that M0-UBE2S-CM treatment significantly increased the expression of α-SMA, Col-1, and FN in EESCs (Fig.  7 H). Moreover, 2-DG treatment reduced the FMT process in EESCs (Fig.  7 H). Conversely, the M0-sh-UBE2S-CM group exhibited decreased expression of α-SMA, Col-1, and FN. However, the addition of lactate to the macrophage-conditioned medium reversed the reduction in FMT markers (Fig.  7 I), with statistically significant differences. These findings suggested that co-culturing M2 macrophages with UBE2S-overexpressing EESCs promotes stromal cell proliferation, migration, and FMT in EM. To evaluate the specificity of CPM as a UBE2S inhibitor, we first performed structure-based molecular docking using a panel of enzymes associated with ubiquitination, glycolysis, and fibrosis [ 25 , 26 ]. CPM indicated the robust binding affinity regarding UBE2S (Fig. S8 A and Table S5 ), with markedly increased binding free energies (ΔG) observed for other ubiquitin-related enzymes (UBE2C, UBE2D1, USP10, USP7, and OTUB1), glycolytic enzymes (HK2, PFKFB3, PKM2, and LDHA), and fibrosis-correlated kinases (TGF-βR1, SMAD3, MAPK1, JNK1, and mTOR), suggesting low likelihood of off-target interactions. To determine whether the inhibitory effects of CPM are specifically mediated through UBE2S, we performed rescue experiments using UBE2S KO EESCs. As expected, UBE2S deletion significantly reduced lactate production in EESCs. Notably, CPM treatment did not further suppress lactate levels in the UBE2S-KO background, indicating that CPM’s glycolysis-inhibiting effects require the presence of UBE2S and are unlikely to result from off-target mechanisms (Fig. S8 B). To evaluate whether the antifibrotic effects of CPM are specifically mediated through UBE2S, we used a macrophage-EESCs co-culture model (Fig.  7 A). Conditioned medium (CM) was collected from co-cultures of macrophages with either UBE2S-KO (M0-KO) or WT (M0-WT) EESCs, and applied to naïve EESCs. Western blot analysis revealed that M0-KO-CM markedly reduced the expression of fibrosis-related markers, including α-SMA, Col-1, and FN, compared with M0-WT-CM (Fig. S8 C). Furthermore, CM derived from the co-culture of CPM-treated UBE2S-deficient (M0-KO) EESCs and macrophages did not further change the expression of fibrotic markers, indicating that CPM has no additional effects in the absence of UBE2S. In contrast, CM generated from CPM-treated WT EESCs co-cultured with M0 macrophages significantly downregulated fibrotic markers in EESCs compared to CM from untreated M0-WT co-cultures, with statistically significant differences (Fig. S8 C). These results suggest that CPM exerts its antifibrotic effects through a UBE2S-dependent mechanism, underscoring its specific role in modulating glycolysis and fibrosis pathways through UBE2S. We developed a UBE2S knockout (UBE2S −/− ) transgenic mouse model to further investigate the impact of UBE2S on uterine endometrial activity in vivo. During reproduction, we observed a significant decrease in fertility in UBE2S −/− homozygous female mice; therefore, we used UBE2S −/− male mice and UBE2S +/− female mice for breeding. An EM model was established using endometrial tissues from UBE2S −/− female mice to study the effect of UBE2S on the occurrence of EM in vivo (Fig.  8 A). The results revealed that the size of endometriotic lesions was significantly reduced in UBE2S −/− mice (Fig.  8 B). By calculating the volume and weight of all endometriotic lesions in both the experimental and control groups (Fig.  8 C-D), we determined that the lesions in the experimental group were inhibited. Therefore, in vivo UBE2S knockdown can suppress EM development. Fig. 8 UBE2S promotes EM development in vivo. ( A ) Establishment of UBE2S knockout mouse model of EM. ( B - D ) Body weight and lesion volume in experimental ( n  = 6) and control ( n  = 6) groups after EM lesion removal. ( E ) H&E (10 x) staining and IHC (20 x) for UBE2S, GLUT1, CD163, and α-SMA (scale bar, 50 μm), Quantification of staining intensity was performed using ImageJ software. ( F ) CPM treatment in a C57BL/6J mouse model of EM. ( G ) Representative images of the largest ectopic tissues from each mouse. ( H - I ) Volume ( H ) and weight ( I ) of ectopic lesions. ( J ) Schematic representation of UBE2S-mediated metabolic reprogramming and immune modulation in EM. * p  < 0.05, ** p  < 0.01, *** p  < 0.001 UBE2S promotes EM development in vivo. ( A ) Establishment of UBE2S knockout mouse model of EM. ( B - D ) Body weight and lesion volume in experimental ( n  = 6) and control ( n  = 6) groups after EM lesion removal. ( E ) H&E (10 x) staining and IHC (20 x) for UBE2S, GLUT1, CD163, and α-SMA (scale bar, 50 μm), Quantification of staining intensity was performed using ImageJ software. ( F ) CPM treatment in a C57BL/6J mouse model of EM. ( G ) Representative images of the largest ectopic tissues from each mouse. ( H - I ) Volume ( H ) and weight ( I ) of ectopic lesions. ( J ) Schematic representation of UBE2S-mediated metabolic reprogramming and immune modulation in EM. * p  < 0.05, ** p  < 0.01, *** p  < 0.001 Subsequently, we performed immunohistochemical staining to evaluate UBE2S and GLUT1 expression. The results indicated a significant reduction in UBE2S and GLUT1 levels in endometriotic lesions from the experimental group (Fig.  8 E; p  < 0.001). Additionally, we assessed the expression of α-SMA and CD163-related markers, which further confirmed that UBE2S promotes macrophage M2 polarization and fibrosis during EM development in vivo (Fig.  8 E; p  < 0.01). To further investigate the in vivo role of UBE2S in regulating GLUT1-related macrophage polarization and fibrosis, we performed immunofluorescence staining on ectopic lesion tissues from UBE2S WT and UBE2S⁻ / ⁻ mice. In WT lesions, GLUT1⁺ endometriotic cells were frequently surrounded by CD163⁺ M2 macrophages. This spatial correlation was markedly reduced in UBE2S⁻ / ⁻ lesions, which exhibited substantially lower GLUT1 expression and decreased infiltration of CD163⁺ cells (Fig. S9 A). Quantitative analysis confirmed a significant reduction in CD163⁺ cell density around lesions in UBE2S-deficient mice (Fig. S9 A, right panel). Similarly, α-SMA⁺ fibroblasts were enriched around GLUT1⁺ cells in WT lesions, aligning with active fibrotic remodeling. In contrast, UBE2S⁻ / ⁻ lesions indicated reduced GLUT1 expression and significantly reduced α-SMA signal intensity (Fig. S9 B). Quantification supported these findings (Fig. S9 B, right panel). Collectively, these results provide direct in vivo evidence that UBE2S promotes macrophage polarization and fibrotic remodeling through GLUT1-dependent mechanisms. To investigate the effect of the UBE2S inhibitor CPM on EM in vivo, we established an EM mouse model using C57BL/6J mice (Fig.  8 F). The experimental group received intraperitoneal injections of CPM (10 mg/kg) once weekly. After four weeks, endometriotic tissues were excised for measurement and analysis (Fig.  8 G). Compared to the control group, the volume and weight of the endometriotic lesions were significantly reduced in the treatment group (Fig.  8 H-I). These findings suggested that the UBE2S inhibitor CPM demonstrates a significant therapeutic effect on EM in vivo.

Endometrial tissue samples were collected from 20 patients with ovarian ectopic cysts and other non-endometriotic diseases (including uterine fibroids and cervical intraepithelial neoplasia) who underwent total hysterectomy at the First Affiliated Hospital of Harbin Medical University. The samples included 20 cases of ectopic endometrial (EC) tissue, 20 cases of eutopic endometrial (Eu) tissue, and 20 cases of normal endometrial (NM) tissue from non-endometriotic patients. Among the 20 clinical specimens collected per group, 5 were allocated for immunohistochemical analysis, the remaining 15 underwent primary endometrial stromal cell isolation. Following rigorous dual immunophenotyping (Vimentin⁺/Cytokeratin⁻), only cell cultures maintaining > 95% purity through passages 2–3 were utilized for subsequent experiments, including transcriptomic sequencing (RNA-seq), protein/RNA extraction, proliferation/migration assays, and metabolic assays. The inclusion criteria were as follows: Patients aged ≥ 18 and ≤ 50 years, premenopausal women with a normal menstrual cycle (28 ± 7 days), no use of oral contraceptives, contraceptive injections, implants, or intrauterine devices, and no hormone replacement therapy within at least three months prior to sampling. The exclusion criteria were as follows: Age  50 years, postmenopausal status, patients in the experimental group who were not diagnosed with EM by postoperative pathology, and patients in the control group with pathological changes in endometrial tissue. All samples were approved by the Ethics Committee of the First Affiliated Hospital of Harbin Medical University, and informed consent was obtained from all participants. Ectopic endometrial stromal cells (EESCs), eutopic endometrial stromal cells (EuSCs), and normal endometrial stromal cells (NESCs) were isolated by enzymatic digestion with collagenase type IV (1 mg/mL, Biosharp) at 37 °C. After filtration and centrifugation, the cells were resuspended in DMEM/F12 medium (GIBCO, NY, USA) containing 10% FBS and cultured at 37 °C in a 5% CO₂ incubator. Total RNA was extracted using Trizol reagent (Invitrogen, CA, USA). RNA purity and integrity were evaluated by NanoDrop 2000 (Thermo Fisher, USA) and Bioanalyzer 2100 (Agilent, USA), and degradation was checked by 1.5% agarose gel. Poly-T magnetic beads were used to isolate mRNA, and libraries were prepared with the VAHTS Universal V6 RNA-seq Library Kit for MGI (Vazyme, China). Library quality was assessed by Qubit 3.0 (Thermo Fisher, USA) and Bioanalyzer 2100, followed by sequencing on the MGI-SEQ 2000 platform (Frasergen, Wuhan, China). Raw data were filtered with SOAPnuke (v2.1.0) to remove adaptor, low-quality, and high-N reads. Clean reads were aligned to the reference genome using HISAT2 (v2.1.0) and Bowtie2 (v2.3.5). Transcript abundance was quantified with RSEM (v1.3.1) and normalized to FPKM. Differential expression was analyzed using DESeq2 (v1.22.2), with |log2FC| >1 and false discovery rate (FDR) < 0.05 as the significance thresholds. Differentially expressed genes (DEGs) were subjected to GO enrichment analysis using the enrichGO function implemented in the R package clusterProfiler (v3.6.3). Pathways with p value ≤ 0.05 were considered significantly enriched. Visualization of the GO results was performed using the R package ggplot2 (v3.5.1). For GSEA, the pre-ranked list of all expressed genes was used as input, and enrichment was assessed against the Molecular Signatures Database (MSigDB, Hallmark gene sets). The GSEA was performed using the GSEA function in clusterProfiler (v3.6.3), with the following parameters: minimum gene set size = 0, maximum gene set size = 20,000, p value cutoff = 0, multiple testing correction using the Benjamini–Hochberg (BH) method, and 1,000 permutations. Gene sets with a normalized enrichment score (NES) > 1 (or < − 1), p value < 0.05, and FDR < 0.25 were regarded as significantly enriched. Visualization of GSEA results was performed using the gseaNb function from the R package GseaVis (v0.0.5). Protein–protein interaction (PPI) analysis of DEGs was performed using the STRING database ( https://string-db.org/ ). The resulting PPI network was visualized and further analyzed using Cytoscape software (v3.9.1) and analyzed to identify potential hub genes and key regulatory modules. THP-1 cells (ScienCell) were differentiated into M0 macrophages by treatment with 100 ng/ml phorbol 12-myristate 13-acetate (PMA; Sigma-Aldrich, St. Louis, MO, USA) for 24 h. When cells were observed under the microscope to adhere to the plate and exhibit spreading morphology, the medium was replaced with complete medium without PMA to terminate the stimulation. A transwell chamber with a pore size of 0.4 μm (Corning, NY, USA) was used. EESCs subjected to different treatments were seeded in the upper chamber, while M0 macrophages were plated in the lower chamber. After 48 h of co-culture, cells in the lower chamber were harvested for analysis of M2 polarization markers (CD163, CD206, ARG-1). Proteins were excised from Coomassie Brilliant Blue–stained gels, reduced with DTT, alkylated with iodoacetamide, and digested overnight with trypsin. The resulting peptides were extracted with acetonitrile, desalted using ZipTip C18, and analyzed by LC–MS/MS on a nanoLC-Q Exactive mass spectrometer (Thermo Scientific, USA). Mass spectrometric data were searched against the UniProt human proteome database (2022 release) using the SEQUEST HT algorithm in Proteome Discoverer (v1.4). Search parameters were set as follows: trypsin specificity with up to two missed cleavages allowed, precursor ion mass tolerance of 10 ppm, fragment ion mass tolerance of 0.02 Da, carbamidomethylation of cysteine as a fixed modification, and oxidation of methionine as a variable modification. Protein identifications were filtered with Percolator at a 1% FDR to ensure high peptide confidence. USP10⁻ / ⁻ and UBE2S⁻ / ⁻ cell lines were generated using the CRISPR/Cas9 system. To generate USP10⁻ / ⁻ cells, sgRNAs targeting USP10 were cloned into the PX459 vector and transfected into HEK293T cells using Lipofectamine 3000. After 24 h, cells were selected with puromycin (2 µg/mL) for 48 h. Single-cell clones were isolated, expanded, and validated by PCR and Western blot. Similarly, UBE2S⁻ / ⁻ knockout was conducted in HEK293T cells and EESCs using a similar approach. The sgRNAs targeting UBE2S were inserted into PX459, followed by transfection and puromycin selection. PCR and Western blot validated clones to confirm successful knockout. The sgRNA sequences used for USP10 and UBE2S are provided in Table S1 . Recombinant plasmids used in this study included pCAGGS-UBE2S (Flag, HA, and Myc), pCMV-Flag-UBE2S ∆N, pCMV-Flag-UBE2S ∆C, and pCMV-Flag-UBE2S ∆Core, obtained from Professor Changjiang Weng at Harbin Veterinary Research Institute. We synthesized specific shRNAs targeting UBE2S and GLUT1, along with scrambled shRNA oligonucleotides as controls (shRNA sequences are listed in Table S1 ). The synthesized shRNA oligonucleotides were annealed, digested with BamH I and EcoR I, and ligated into the pLVX-shRNA1 vector (Clontech, Palo Alto, USA). Transfections were done using Lipofectamine 3000. For lentivirus packaging, HEK293T cells were co-transfected with target plasmids and packaging vectors psPAX2 and pMD2.G. Cells were seeded at 2000 cells/well in 96-well plates and cultured for 24, 48, and 72 h. Absorbance at 450 nm was measured after adding 10 µL of CCK-8 solution (Meilunbio, Dalian, China). Cells were seeded at 1 × 10⁵ cells/well in 12-well plates, incubated with EdU (Beyotime Biotechnology, Shanghai, China) for 4 h, fixed, and stained. Fluorescence was imaged using a fluorescence microscope. The treated cells were reseeded in six-well plates. When cell confluence reached 90%, a vertical scratch was made using a pipette tip. After washing with PBS to remove suspended cells and debris, images were captured and recorded at 0 h. The cells were cultured for 24 h, and images were captured again. Cells (2.5 × 10 5 cells/mL) were seeded in the upper chamber of a 24-well transwell plate (Corning, NY, USA) with an 8 μm pore filter and cultured in a serum-free medium. The lower chamber was filled with DMEM supplemented with 20% FBS. After 24 h of incubation at 37 °C, the cells were fixed with 4% paraformaldehyde for 15 min, washed with PBS, stained with crystal violet for 20 min, washed with PBS, and imaged for statistical analysis. After transfection, the cells were reseeded in six-well plates and cultured for 12–16 h. Cells were collected, and glucose, lactate, and pyruvate concentrations were measured using glucose, lactate, and pyruvate assay kits (mlbio, Shanghai, China). For H&E staining, tissue  sections (5 μm) were deparaffinized with xylene, stained with hematoxylin and eosin, and examined under a light microscope. For IHC, tissue  sections (4 μm) were deparaffinized, rehydrated, and treated with 3% H 2 O 2 for 15 min to inactivate endogenous peroxidases. After antigen retrieval in citrate buffer (pH 9.0) and blocking with goat serum for 30 min, the sections were incubated overnight at 4 °C with primary antibodies UBE2S (1:200, Proteintech) or GLUT1 (1:200, Proteintech). The next day, the sections were incubated with HRP-conjugated secondary antibody for 1 h, followed by DAB staining and microscopic analysis. Quantification of staining intensity and the percentage of positive area was performed using ImageJ software (NIH, Bethesda, MD, USA). For each sample, five randomly selected, non-overlapping fields (20×) were analyzed to calculate the average staining intensity. All image analyses were conducted in a blinded manner. Cells were collected and lysed in lysis buffer (50 mM Tris, pH 7.6, 0.5 mM EDTA, 0.1% NP40, and 0.5 mM PMSF). After centrifugation, the protein concentration was measured using a BCA assay (Beyotime, Shanghai, China). One-fifth of the total protein was used for input experiments, while the remaining was incubated with an antibody against the target protein and mixed with agarose overnight at 4 °C. After washing with buffer, the complexes were boiled for 10 min and subjected to western blotting or mass spectrometry analysis. The pGEX-4T-1-UBE2S plasmid was expressed in E. coli BL21 (DE3) cells induced with 1 mM IPTG. Bacterial pellets were collected, lysed in buffer with protease inhibitors, and sonicated. Following centrifugation (12,000 × g, 20 min, 4 °C), the supernatant containing GST-UBE2S was incubated with glutathione agarose beads overnight at 4 °C. Purified His-tagged USP10 protein, expressed employing the pET-24a(+)-USP10 plasmid, was added to the GST-UBE2S-bound beads and incubated for 2 h at 4 °C. After washing the beads 4–5 times, bound proteins were eluted by boiling in sodium dodecyl sulfate (SDS) loading buffer and analyzed by SDS-polyacrylamide gel electrophoresis followed by immunoblotting with an anti-His antibody. Total RNA was extracted from cell samples using Trizol (Invitrogen, CA, USA), and the RNA concentration was measured using NanoDrop (Thermo Fisher Scientific, MA, USA). Reverse transcription was performed with a reverse transcription kit (Seven, Beijing, China) to synthesize cDNA. RT-qPCR was conducted using S6 Universal SYBR qPCR Mix (Enzy Artisan, Shanghai, China). The primers used for qPCR are shown in Table S2 . To evaluate the binding specificity of CPM toward UBE2S and assess potential off-target interactions, structure-based molecular docking was performed using AutoDock Vina software (v1.2.5). Protein structures were downloaded from the database ( www.uniprot.org ) and preprocessed using ADFRsuite-1.0. Receptor PDBQT files were prepared while preserving the original protonation and charge states. Additionally, ligand structures of CPM were processed using ADFRsuite-1.0 to generate the corresponding PDBQT files. Docking was conducted on 32 parallel CPU threads with a grid spacing of 0.375; other parameters were kept at default. Binding affinities were assessed based on Vina docking scores (ΔG), where lower (more negative) values indicate stronger predicted binding. To evaluate the role of UBE2S in endometriosis in vivo, we employed a syngeneic transplantation model. The UBE2S-KO (UBE2S –/– ) mice on a C57BL/6J background were generated by homologous recombination (Biocytogen, Beijing, China). Genotyping primers are listed in Table S3 . Female C57BL/6J and UBE2S –/– mice (6–8 weeks old) received intraperitoneal estrogen injections (1 µg/mL) on days 1, 4, and 7. On day eight, donor mice were euthanized under anesthesia, and endometrial tissue from the uterine horns was collected and suspended in PBS. Each 0.6 mL tissue suspension was injected intraperitoneally into two immunocompetent C57BL/6J recipient mice. Estrogen was administered weekly thereafter. Following four weeks, ectopic lesions were harvested. For drug intervention, mice were randomly assigned to receive CPM (10 mg/kg, intraperitoneally, once weekly) or saline starting one week post-surgery ( n  = 6/group). After four weeks of treatment, lesions were collected for analysis. All animal procedures were approved by the Animal Ethics Committee of the First Affiliated Hospital of Harbin Medical University and conducted in accordance with institutional and national guidelines. Data were analyzed using SPSS (version 20.0) and GraphPad Prism (version 8.0) software. A one-way analysis of variance, followed by Tukey’s post-hoc test, was employed to compare the differences between the groups. A p  < 0.05 was considered statistically significant.

Abnormal glycolytic metabolism in endometriotic tissue has been extensively reported [ 27 , 28 ]; however, the molecular mechanisms underlying these changes remain unclear. Based on transcriptomic data, we found that UBE2S is significantly upregulated in EESCs. Previous studies have shown that UBE2S has a key regulatory function in tumor metabolic reprogramming. For instance, in hepatocellular carcinoma (HCC), UBE2S stabilizes VHL protein, indirectly enhancing HIF-1α stability, which promotes the expression of glycolysis-related genes, thereby enhancing glucose uptake and lactate production to adapt to the tumor hypoxic microenvironment [ 10 ]. In contrast to the aforementioned HIF pathway-dependent mechanism, we discovered that UBE2S recruits USP10, specifically mediating K48-linked deubiquitination of GLUT1, inhibiting its protein degradation and enhancing its stability. This regulatory mechanism may stem from the fact that the endometriotic lesions are not fully hypoxic. Further functional experiments showed that UBE2S-mediated stabilization of GLUT1 significantly promotes glucose uptake and glycolytic flux, driving EESCs into a high-glycolysis metabolic state. This finding provides the first evidence of the regulatory role of UBE2S in glycolytic reprogramming, expanding our understanding of the molecular mechanisms underlying glycolytic abnormalities in EM. In most studies, UBE2S exerts pathogenic effects by mediating ubiquitination of substrate proteins through its E2 or E3 enzymatic activity. For example, UBE2S promotes the activity of APC/C through K11-linked ubiquitination, accelerating the cell cycle [ 29 ]; it also enhances Wnt/β-catenin signaling by regulating the stability of β-catenin, facilitating invasion and metastasis in colorectal cancer [ 30 ]. However, studies have shown that UBE2S can recruit USP15 and exert deubiquitination functions, suggesting that UBE2S may have dual regulatory capabilities [ 19 ]. Our study further supports this hypothesis and, for the first time, identifies that UBE2S recruits USP10 through its N-terminal domain to co-regulate K48 deubiquitination of GLUT1. We confirm that USP10 alone cannot independently accomplish this modification process; it must rely on UBE2S to form a complex. This finding reveals a novel function of UBE2S in regulating GLUT1 stability through a deubiquitination mechanism, expanding the complex biological roles of UBE2S. EM is a chronic inflammatory disease characterized by a significantly disrupted immune microenvironment [ 31 ]. M2 macrophages dominate the lesion microenvironment, promoting inflammation and fibrosis. UBE2S-mediated GLUT1 stability was observed to enhance glycolytic levels in EESCs and promote M2 macrophage polarization through lactate accumulation. Lactate modulates macrophage phenotypic switching through HIF-1α and STAT3 signaling pathways [ 5 , 32 ]. Further investigations revealed that UBE2S-mediated GLUT1 overexpression significantly increased lactate levels and activated M2-related markers (CD206, Arg-1, and IL-10). Moreover, the downregulation of UBE2S or treatment with glycolytic inhibitors reversed this effect, indicating that this pathway is essential in regulating macrophage polarization. Chronic inflammation and fibrosis in EM lesions are key factors contributing to disease progression and recurrence [ 33 – 35 ]. M2 macrophages secrete TGF-β1 and PDGF, which activate fibroblasts and promote myofibroblast differentiation through the TGF-β1/Smad signaling pathway, accelerating the fibrotic process [ 4 , 36 , 37 ]. In vitro and animal model experiments demonstrated that UBE2S-mediated M2 macrophage polarization promoted the conversion of EM stromal fibroblasts into myofibroblasts, increasing the expression levels of α-SMA and Col1. Treatment with the UBE2S inhibitor CPM significantly inhibited this effect, suggesting that UBE2S may be a therapeutic target for EM-related fibrosis. Notably, while UBE2S deficiency reduces fertility in female mice, this phenotype is primarily due to early embryonic lethality caused by APC/C inactivation and meiotic arrest, rather than hormonal imbalance or gonadal dysfunction [ 38 ]. Our observations did not indicate changes in sexual maturation, estrous cyclicity, or ovarian reserve in UBE2S –/– females. Furthermore, to reduce hormonal variability during model induction, all recipient and donor mice were subjected to standardized estrogen priming and estrous cycle synchronization. These measures ensure that hormonal or developmental biases do not confound the fibrotic and metabolic phenotypes observed, but rather reflect the intrinsic role of UBE2S in adult endometrial physiology. Despite these findings, this study has some limitations. This research primarily utilized cell and mouse models, and clinical validation with patient samples remains necessary to enhance the translational potential of these findings. GLUT1-mediated metabolic reprogramming may also involve broader downstream signaling pathways, including the PI3K/Akt/mTOR axis [ 39 , 40 ]. Future studies must investigate these molecular mechanisms. Although our in silico docking and UBE2S-KO rescue experiments support the specificity of CPM, we acknowledge the absence of full-scale DUB or kinase profiling. Computational prediction cannot fully capture in-cell binding dynamics. Future studies using chemical proteomics may help confirm CPM’s direct targets under physiological conditions. In conclusion, this study uncovers a novel mechanism by which UBE2S-mediated deubiquitination of GLUT1 regulates glucose metabolism, the immune microenvironment, and fibrosis in EM, proposing UBE2S as a potential therapeutic target (Fig.  8 J). These findings offer novel insights into EM pathophysiology and lay a theoretical foundation for future clinical interventions.

Endometriosis (EM) is a chronic gynecological disorder characterized by the growth of endometrial-like tissue outside the uterine cavity, leading to persistent inflammation, fibrosis, and symptoms such as pelvic pain and infertility [ 1 , 2 ]. Although its pathogenesis has been widely investigated, the underlying molecular mechanisms remain unclear. Increasing evidence suggests that metabolic reprogramming and immune dysregulation play key roles in EM progression, particularly via altered glucose metabolism and immune cell dysfunction [ 3 , 4 ]. Endometriosis (EM) lesions display a glycolysis-driven metabolic signature, with lactate, produced by dysregulated glucose metabolism, serving as a crucial mediator between ectopic lesions and immune cells. Accumulated lactate promotes macrophage M2 polarization, reshaping the immune microenvironment and enhancing immune evasion [ 5 ]. M2 macrophages further contribute to fibrosis by inducing the differentiation of fibroblasts into myofibroblasts [ 4 , 5 ]. Thus, lactate-driven immune modulation may link glucose metabolism with fibrotic progression in EM. Targeting metabolic enzymes could offer new diagnostic and therapeutic strategies, yet the specific regulatory mechanisms remain to be elucidated. Ubiquitination is a crucial post-translational modification that regulates protein stability and function, playing a crucial role in cellular metabolism and immune homeostasis across different diseases [ 6 , 7 ]. UBE2S, a unique E2 ubiquitin-conjugating enzyme with both E3-dependent and E3-independent ligase activities [ 8 ], has been implicated in protein degradation and tumor progression [ 9 , 10 ]. However, its role in EMs has not been elucidated. Using IP/MS, we identified a specific correlation between UBE2S and the glucose transporter GLUT1. This key glycolytic regulator facilitates glucose uptake and promotes lactate accumulation [ 11 ]. Lactate, beyond its metabolic role, is a critical immunoregulatory metabolite known to drive M2 macrophage polarization [ 12 ]. Although UBE2S and GLUT1 have been individually associated with ubiquitination and metabolic regulation, their functional interplay in EMs remains unknown. This study aims to explore whether UBE2S modulates GLUT1 ubiquitination and stability, thereby contributing to the metabolic and immune changes related to EMs. In this study, we investigated the role of UBE2S in EM and proposed that UBE2S enhances GLUT1 stability by promoting its deubiquitination through Ubiquitin-Specific Peptidase 10 (USP10). This process facilitates metabolic reprogramming, macrophage polarization, and lesion fibrosis. By elucidating this pathway in vitro and in vivo, we aimed to uncover new insights into the metabolic-immune interplay in EM and identify potential therapeutic targets.

Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 1 Supplementary Material 2 Supplementary Material 2 Supplementary Material 3 Supplementary Material 3 Supplementary Material 4 Supplementary Material 4