Research on the mechanism of CRMP4-mediated cytoskeletal actin polymerization in promoting invasive migration of stromal cells in endometriosis

Using the XianTao Academic software, we intersected datasets GSE7305 , GSE25628 , and GSE36537 and obtained 19 significant genes that may affect M2-macrophage-related EMs progression (Fig.  1 A). To validate our bioinformatics findings, we conducted whole-transcriptome sequencing analysis on nine endometrial samples (CON n  = 3, EU n  = 3, CU n  = 3). The results revealed that the expression level of CRMP4 markedly higher in CU than in EU (Fig.  1 B-C). Though not directly compared in whole-transcriptome sequencing analysis, CU vs. CON can be inferred as log₂-FC ≈ 3.54 (> 11-fold). RT-qPCR and WB verified elevated expression of CRMP4 at both protein and mRNA levels in CU tissue compared to EU and CON samples (Fig.  1 D-F). We then performed immunohistochemical staining on endometrial tissues from 31 control subjects (CON) and 32 EMs patients, including both in situ (EU) and ectopic (CU) endometrial tissues. In the 31 CON samples, CRMP4 protein was weakly expressed primarily in the cytoplasm of stromal cells, with almost no expression in glandular epithelium. In contrast, CRMP4 expression was significantly elevated in the stromal cells of the 32 EU samples, while glandular epithelium remained low in expression. Furthermore, the CU samples exhibited diffuse strong expression of CRMP4 in stromal cells (Fig.  1 G). Quantitative scoring of CRMP4 expression across all slides using H-score indicated that levels were elevated in both EU and CU groups compared to CON group, with a significant increase observed in CU group(Fig.  1 H). Spearman correlation analysis showed no correlation between CRMP4 H-score in the EU group and r-ASRM (Fig.  1 I), whereas CRMP4 H-score in the CU group correlated positively with r-ASRM (Fig.  1 J). In conclusion, these results confirm that CRMP4 is significantly upregulated in ectopic tissues of endometriosis and correlates with disease severity, suggesting its important role in the progression of the disease. Fig. 1 CRMP4 is significantly upregulated in endometriosis. ( A ): Venn diagram to identify differentially expressed -related genes in M2 macrophage polarization and endometriotic lesions. ( B ): The volcano plot shows no significant difference in CRMP4 expression between the control group (CON, normal endometrium) and the eutopic endometrium (EU) in cases of endometriosis (EMs). ( C ): CRMP4 is significantly upregulated in the ectopic endometrium (CU) compared to the eutopic endometrium (EU) among the differentially expressed genes (|log2FC| = 2.46, P  = 0.022). ( D ): The results of RT-qPCR indicated that the mRNA expression level of CRMP4 in ectopic endometrium (CU) was significantly upregulated compared to the eutopic endometrium (EU), and the expression level in the EU group was higher than that in the normal endometrium (CON). ( E ): Protein immunoblotting analysis confirmed that CRMP4 protein was specifically overexpressed in the CU group, with band intensity significantly greater than that in the EU and CON groups. ( F ): Quantitative analysis of the protein bands in panel B was performed using ImageJ software, and the results showed that CRMP4 protein expression in the CU group was significantly elevated compared to both the EU and CON groups. ( G ): Representative immunohistochemical images of CRMP4 protein in normal endometrial tissue and ectopic endometrial tissue (EMs). ( H ): Statistical analysis of CRMP4 expression in the three groups of endometrial tissue, evaluated using H-Score. ( I ): The correlation between CRMP4 H-score and ASRM score in Eutopic endometrium tissue (EU) of EMs. ( J ): The correlation between CRMP4 H-score and ASRM score in ectopic endometrial tissue (CU) of EMs. (Data are mean ± SEM. Groups analysed by one-way ANOVA followed by Tukey’s post-test. * P  < 0.05, ** P  < 0.01, *** P  < 0.001, **** P  < 0.0001) CRMP4 is significantly upregulated in endometriosis. ( A ): Venn diagram to identify differentially expressed -related genes in M2 macrophage polarization and endometriotic lesions. ( B ): The volcano plot shows no significant difference in CRMP4 expression between the control group (CON, normal endometrium) and the eutopic endometrium (EU) in cases of endometriosis (EMs). ( C ): CRMP4 is significantly upregulated in the ectopic endometrium (CU) compared to the eutopic endometrium (EU) among the differentially expressed genes (|log2FC| = 2.46, P  = 0.022). ( D ): The results of RT-qPCR indicated that the mRNA expression level of CRMP4 in ectopic endometrium (CU) was significantly upregulated compared to the eutopic endometrium (EU), and the expression level in the EU group was higher than that in the normal endometrium (CON). ( E ): Protein immunoblotting analysis confirmed that CRMP4 protein was specifically overexpressed in the CU group, with band intensity significantly greater than that in the EU and CON groups. ( F ): Quantitative analysis of the protein bands in panel B was performed using ImageJ software, and the results showed that CRMP4 protein expression in the CU group was significantly elevated compared to both the EU and CON groups. ( G ): Representative immunohistochemical images of CRMP4 protein in normal endometrial tissue and ectopic endometrial tissue (EMs). ( H ): Statistical analysis of CRMP4 expression in the three groups of endometrial tissue, evaluated using H-Score. ( I ): The correlation between CRMP4 H-score and ASRM score in Eutopic endometrium tissue (EU) of EMs. ( J ): The correlation between CRMP4 H-score and ASRM score in ectopic endometrial tissue (CU) of EMs. (Data are mean ± SEM. Groups analysed by one-way ANOVA followed by Tukey’s post-test. * P  < 0.05, ** P  < 0.01, *** P  < 0.001, **** P  < 0.0001) To investigate the impact of CRMP4 on the biological behavior of endometrial stromal cells (EMs), we first isolated stromal cells from control endometrium (CON-EuESCs), eutopic endometrium of endometriosis patients (EM-EuESCs), and ectopic endometrium of endometriosis patients (EM-EcESCs). Mesenchymal identity was confirmed by Vimentin⁺/CK19⁻/E-cadherin⁻ immunofluorescence (Fig.  2 A). Western blotting and qRT-PCR analyses revealed that CRMP4 protein and mRNA expression were significantly upregulated in EM-EcESCs and EM-EuESCs compared to CON-EuESCs, with the most pronounced increase observed in EM-EcESCs (Fig.  2 B-D). This suggests a potential pathogenic role for CRMP4 in ectopic stromal cells. To validate this, we knocked down CRMP4 expression in EM-EcESCs using siRNA targeting CRMP4 (si-CRMP4), which resulted in a significant reduction in CRMP4 mRNA and protein levels (Fig.  2 E-G). Conversely, we overexpressed CRMP4 in EM-EcESCs using a lentiviral vector, and the transfection efficiency was confirmed by Western blotting and qRT-PCR (Fig.  2 H-J). Scratch wound healing and Transwell assays demonstrated that si-CRMP4 suppressed migration and invasion abilities of EM-EcESCs, while CRMP4 overexpression enhanced these abilities compared to the CON group (Fig.  2 K-N). Thus, CRMP4 promotes migration and invasion of EM-EcESCs and may drive endometriotic progression. Fig. 2 Overexpression CRMP4 enhances EM-EcESCs migration, invasion.Knockdown of CRMP4 inhibits EM-EcESCs proliferation, migration, invasion. ( A ): The immunofluorescence identification results of three groups of stromal cells: Vimentin (a mesenchymal marker) shows strong green fluorescence expression, while the epithelial markers E-cadherin and CK19 exhibit weak red fluorescence, confirming that the cells are of a purified mesenchymal cell type. ( B ): WB assays revealed CRMP4 levels in three types of stromal cells. ( C ): Statistical quantification of CRMP4 protein expression levels in stromal cells. ( D ): qRT-PCR assays revealed CRMP4 mRNA levels in three types of stromal cells. ( E - G ): qRT-PCR and WB assays showed the knockdown efficiency of si-CRMP4 in EM-EcESCs. ( H - J ): qRT-PCR and WB assays showed the over-expression efficiency of CRMP4 in EM-EcESCs. ( K - L ): Migration capability of EM-EcESCs treated with or without si-CRMP4 and oe-CRMP4 were assessed by wound-healing assay and the wound closure rate was quantified (white area, wound-healing area; original magnification 40×). ( M - N ): Migration capability of EM-EcESCs treated with or without si-CRMP4 and oe-CRMP4 were assessed by Transwell assay and the migration cells were quantified (original magnification 200×). (Data are mean ± SEM. Groups analysed by one-way ANOVA followed by Tukey’s post-test. * P  < 0.05, ** P  < 0.01, *** P  < 0.001, **** P  < 0.0001) Overexpression CRMP4 enhances EM-EcESCs migration, invasion.Knockdown of CRMP4 inhibits EM-EcESCs proliferation, migration, invasion. ( A ): The immunofluorescence identification results of three groups of stromal cells: Vimentin (a mesenchymal marker) shows strong green fluorescence expression, while the epithelial markers E-cadherin and CK19 exhibit weak red fluorescence, confirming that the cells are of a purified mesenchymal cell type. ( B ): WB assays revealed CRMP4 levels in three types of stromal cells. ( C ): Statistical quantification of CRMP4 protein expression levels in stromal cells. ( D ): qRT-PCR assays revealed CRMP4 mRNA levels in three types of stromal cells. ( E - G ): qRT-PCR and WB assays showed the knockdown efficiency of si-CRMP4 in EM-EcESCs. ( H - J ): qRT-PCR and WB assays showed the over-expression efficiency of CRMP4 in EM-EcESCs. ( K - L ): Migration capability of EM-EcESCs treated with or without si-CRMP4 and oe-CRMP4 were assessed by wound-healing assay and the wound closure rate was quantified (white area, wound-healing area; original magnification 40×). ( M - N ): Migration capability of EM-EcESCs treated with or without si-CRMP4 and oe-CRMP4 were assessed by Transwell assay and the migration cells were quantified (original magnification 200×). (Data are mean ± SEM. Groups analysed by one-way ANOVA followed by Tukey’s post-test. * P  < 0.05, ** P  < 0.01, *** P  < 0.001, **** P  < 0.0001) Due to the limited passage number of primary cells, we generated stable CRMP4-overexpressing and CRMP4-knockdown cell lines in the immortalized human endometrial stromal cell line ihESC [ 26 ] to facilitate subsequent mechanistic studies. The efficiency of CRMP4 overexpression and knockdown in these stable cell lines was confirmed by RT-qPCR and Western blotting, demonstrating significant differences in CRMP4 mRNA and protein levels compared to control cells (Fig. 3 A-F). Scratch assays at 12 h showed reduced migration in CRMP4-knockdown group of ihESC, whereas overexpression yielded no significant increase (although an upward trend was observed), which we speculate may be due to compensatory effects arising from the inherently high proliferative and migratory capacity of the immortalized cell line (Fig. 3 G-H). Furthermore, Lentiviral CRMP4 overexpression in ihESC reproduced the migratory/invasive phenotype observed in primary EM-EcESCs (Fig. 3 I-J). These findings confirm that CRMP4 has a clear pro-migratory function in cells derived from endometriotic lesions, and that the magnitude of this effect is modulated by the biological characteristics of the cell model. Finally, we validated the overexpression efficiency of CRMP4 in CON-EuESCs 72 h after lentiviral transduction (Fig. 3 K-M); Scratch wound healing and Transwell assays showed accelerated migration and invasion (Fig. 3 N–Q), indicating that CRMP4 can drive the transformation of normal endometrial stromal cells towards a pathological phenotype. Fig. 3 Overexpression CRMP4 enhances ihESC and CON-EuESCs migration, invasion. ( A-C ): qRT-PCR and WB assays showed the knockdown efficiency of sh-CRMP4 in ihESC. ( D - F ): qRT-PCR and WB assays showed the over-expression efficiency of CRMP4 in ihESC. ( G - H ): Migration capability of ihESC treated with or without sh-CRMP4 and oe-CRMP4 were assessed by wound-healing assay and the wound closure rate was quantified (white area, wound-healing area; original magnification 40×). ( I - J ): Migration capability of ihESCs treated with or without sh-CRMP4 and oe-CRMP4 were assessed by Transwell assay and the migration cells were quantified (original magnification 200×). ( K - M ): qRT-PCR and WB assays showed the over-expression efficiency of oe-CRMP4 in CON-EuESCs. ( N - O ): Migration capability of CON-EuESCs treated with oe-CRMP4 were assessed by wound-healing assay and the wound closure rate was quantified (white area, wound-healing area; original magnification 40×). ( P - Q ): Migration capability of CON-EuESCs treated with oe-CRMP4 were assessed by Transwell assay and the migration cells were quantified (original magnification 200×). (Data are mean ± SEM. Groups analysed by one-way ANOVA followed by Tukey’s post-test. * P  < 0.05, ** P  < 0.01, *** P  < 0.001, **** P  < 0.0001) Overexpression CRMP4 enhances ihESC and CON-EuESCs migration, invasion. ( A-C ): qRT-PCR and WB assays showed the knockdown efficiency of sh-CRMP4 in ihESC. ( D - F ): qRT-PCR and WB assays showed the over-expression efficiency of CRMP4 in ihESC. ( G - H ): Migration capability of ihESC treated with or without sh-CRMP4 and oe-CRMP4 were assessed by wound-healing assay and the wound closure rate was quantified (white area, wound-healing area; original magnification 40×). ( I - J ): Migration capability of ihESCs treated with or without sh-CRMP4 and oe-CRMP4 were assessed by Transwell assay and the migration cells were quantified (original magnification 200×). ( K - M ): qRT-PCR and WB assays showed the over-expression efficiency of oe-CRMP4 in CON-EuESCs. ( N - O ): Migration capability of CON-EuESCs treated with oe-CRMP4 were assessed by wound-healing assay and the wound closure rate was quantified (white area, wound-healing area; original magnification 40×). ( P - Q ): Migration capability of CON-EuESCs treated with oe-CRMP4 were assessed by Transwell assay and the migration cells were quantified (original magnification 200×). (Data are mean ± SEM. Groups analysed by one-way ANOVA followed by Tukey’s post-test. * P  < 0.05, ** P  < 0.01, *** P  < 0.001, **** P  < 0.0001) Actin plays a fundamental role in driving cell shape changes and migration by forming microfilament networks with diverse structures [ 27 ]. The core driving force behind cell migration originates from the extension of lamellipodia at the leading edge, structures composed of parallel actin bundles that continuously extend towards the cell membrane, propelling cell movement [ 28 ]. In neurons, CRMP4 overexpression induces filopodia formation and neurite branching [ 29 ]. Initially, we observed ultrastructural differences in the morphology of three groups of stromal cells using TEM. Statistical analysis revealed that EMs-EcESCs bore more and longer filopodia than EM-EuESCs or CON-EuESCs, whereas EM-EuESCs displayed increased filopodia number only (Fig. 4 A-C). To further investigate the role of CRMP4 in filopodia formation in EMs-EcESCs, we used siRNA to knock down CRMP4 expression and lentivirus to overexpress CRMP4. The results showed that CRMP4 knock-down shortened and reduced filopodia, while overexpression lengthened and increased them (Fig. 4 D-F), confirming CRMP4 controls filopodial growth. Fig. 4 CRMP4 regulates skeletal remodeling of endometriosis stromal cells (EM EcESCs). ( A ): Representative transmission electron microscopy fields of cytoskeleton ultrastructure of three groups of stromal cells. Arrow indicates filopodia. The scale bar = 10 μm, 500 nm. ( B-C ): Quantification of average filopodia length, filopodia number in per cell and per perimeter. Bar chart of filopodia was determined from at least 16 cells, and data was obtained over 4 independent experiments. ( D ): Representative transmission electron microscopy fields of cytoskeleton ultrastructure of EM-EcESCs treated with or without sh-CRMP4 and oe-CRMP4.Arrow indicates filopodia. The scale bar = 10 μm, 500 nm. ( E-F ): Quantification of average filopodia length, filopodia number in per cell and per perimeter. Bar chart of filopodia was determined from at least 16 cells, and data was obtained over 4 independent experiments. ( G ): The fluorescence intensity of Phalloidin and Vinculin in EM-EcESCs with either CRMP4 overexpression or silencing was assessed, with green fluorescence indicating the distribution of Phalloidin. ( H ): Quantitative results of Phalloidin green fluorescence intensity across different groups analyzed using ImageJ. ( I ): Quantitative results of Vinculin green fluorescence intensity across different groups analyzed using ImageJ. (J-M ): Western blot analysis was performed to evaluate the expression of precipitated P (F-actin) and supernatant S (G-actin) in CRMP4-silenced and CRMP4-overexpressing EM-EcESCs, processed using a G-actin/F-actin separation kit after ultrahigh-speed treatment. The ratio of the grayscale values for S (G-actin) to P (F-actin) was statistically analyzed. (Data are mean ± SEM. Groups analysed by unpaired two-tailed Student’s t-test and by one-way ANOVA followed by Tukey’s post-test. * P  < 0.05, ** P  < 0.01, *** P  < 0.001, **** P   0.05) CRMP4 regulates skeletal remodeling of endometriosis stromal cells (EM EcESCs). ( A ): Representative transmission electron microscopy fields of cytoskeleton ultrastructure of three groups of stromal cells. Arrow indicates filopodia. The scale bar = 10 μm, 500 nm. ( B-C ): Quantification of average filopodia length, filopodia number in per cell and per perimeter. Bar chart of filopodia was determined from at least 16 cells, and data was obtained over 4 independent experiments. ( D ): Representative transmission electron microscopy fields of cytoskeleton ultrastructure of EM-EcESCs treated with or without sh-CRMP4 and oe-CRMP4.Arrow indicates filopodia. The scale bar = 10 μm, 500 nm. ( E-F ): Quantification of average filopodia length, filopodia number in per cell and per perimeter. Bar chart of filopodia was determined from at least 16 cells, and data was obtained over 4 independent experiments. ( G ): The fluorescence intensity of Phalloidin and Vinculin in EM-EcESCs with either CRMP4 overexpression or silencing was assessed, with green fluorescence indicating the distribution of Phalloidin. ( H ): Quantitative results of Phalloidin green fluorescence intensity across different groups analyzed using ImageJ. ( I ): Quantitative results of Vinculin green fluorescence intensity across different groups analyzed using ImageJ. (J-M ): Western blot analysis was performed to evaluate the expression of precipitated P (F-actin) and supernatant S (G-actin) in CRMP4-silenced and CRMP4-overexpressing EM-EcESCs, processed using a G-actin/F-actin separation kit after ultrahigh-speed treatment. The ratio of the grayscale values for S (G-actin) to P (F-actin) was statistically analyzed. (Data are mean ± SEM. Groups analysed by unpaired two-tailed Student’s t-test and by one-way ANOVA followed by Tukey’s post-test. * P  < 0.05, ** P  < 0.01, *** P  < 0.001, **** P   0.05) F-actin, the core component of microfilaments, regulates key cellular activities such as cell migration, adhesion, division, and apoptosis through the dynamic equilibrium between its polymerized (F-actin) and monomeric (G-actin) states. Adhesion plaques, serving as bridges between transmembrane integrin clusters and the intracellular cytoskeleton, anchor the extracellular matrix (e.g., fibronectin) to the F-actin network via linker proteins such as Talin, Vinculin, and Paxillin, mediating mechanical signal transduction and regulating cell adhesion, migration, and proliferation [ 30 , 31 ]. Vinculin, a core scaffolding protein of adhesion plaques, directly participates in the mechanical coupling of actin filaments and adhesion plaques [ 32 ]. In this study, we used phalloidin, which specifically binds to F-actin, in immunofluorescence assays to assess the regulatory effects of CRMP4 overexpression and knockdown on polymerized actin filaments (F-actin) and the adhesion plaque protein Vinculin in EMs-EcESCs and ihESCs. Phalloidin staining revealed that CRMP4 overexpression raised F-actin and vinculin in EM-EcESCs, whereas knock-down lowered both (Figs. 4 G-I). In CRMP4-overexpressing ihESC, vinculin rose but phalloidin signal did not, which we speculate is related to a compensatory effect due to the inherent high proliferation and migration characteristics of immortalized cell lines leading to enhanced cytoskeleton polymerization; silencing CRMP4 reduced both markers (Fig. S1 A-C). To verify that CRMP4 regulates F-actin expression by promoting the polymerization of monomeric G-actin into F-actin, we used ultracentrifugation to separate G-actin (supernatant S) and F-actin (pellet P) in EM-EcESCs [ 33 ], and then detected the expression levels of G-actin in the supernatant S and F-actin in the pellet P by Western blotting. The results showed silencing CRMP4 decreased F-actin and increased G-actin, raising the G/F ratio, whereas overexpression had the opposite effect, lowering the G/F ratio, confirming that CRMP4 regulates the cytoskeleton by promoting the polymerization of G-actin into F-actin (Fig. 4 J-M). Overall, these results indicate that CRMP4 participates in the regulation of cytoskeleton remodeling in EMs-EcESCs, likely through its effect on the polymerization of G-actin into F-actin. G-actin/F-actin ratio governs the transcriptional activity of myocardin-related transcription factor (MRTF) and serum response factor (SRF) [ 34 , 35 ]; MRTF links actin dynamics to SRF-driven migration [ 36 ], and target genes FOSB, JUNB, ATF3, ZEP3 promote invasion [ 37 ]. Initially, we performed qRT-PCR, WB, and immunofluorescence analyses on three groups of endometrial tissues to assess MRTF and SRF expression. Results show that CU group showed higher MRTF mRNA levels than EU group, yet WB revealed lower total MRTF protein in CU (Fig. 5 A-C). We attribute this mismatch to established MRTF biology: miRNAs and RNA-binding proteins dock on the MRTF 3′UTR, blocking translation and accelerating nascent-peptide degradation [ 38 ], while nuclear G-actin or phosphorylation exposes the nuclear-export signal(NES), enabling exportin Crm1-mediated re-export to the cytoplasm for proteasomal destruction [ 39 ]. Thus, post-transcriptional control keeps steady-state MRTF protein in CU below that in EU despite elevated transcription. SRF protein expression was significantly elevated in both EU and CU tissues (Fig. 5 A, B), whereas SRF mRNA expression was only significantly increased in CU (Fig. 5 C). Immunofluorescence staining showed stronger MRTF (green) and SRF (red) signals in CU versus CON, with SRF also elevated in EU (Fig. 5 D-F). RT-qPCR confirmed FOSB, JUNB, ATF3 and ZEP3 up-regulation in CU, and JUNB/ZEP3 in EU (Fig. 5 G). These findings suggest that aberrant activation of the MRTF/SRF signaling axis and its downstream target genes may be an important molecular mechanism underlying CRMP4-mediated migration and invasion of EMs. Subsequently, we examined MRTF and SRF protein and gene expression levels in stromal cells from the three groups using WB and RT-qPCR. WB revealed unchanged MRTF protein, whereas SRF protein and both MRTF/SRF transcripts were elevated in EM-EcESCs (Fig. 5 H-J), suggesting pathological engagement of the pathway. Building upon these findings and our previous demonstration that CRMP4 promotes G-actin polymerization into F-actin, thereby reducing free G-actin levels, we further investigated the subcellular localization and expression changes of MRTF (green fluorescence) and SRF (red fluorescence) in CRMP4-overexpressing and CRMP4-silenced EM-EcESCs models using immunofluorescence. The results showed that, in EM-EcESCs, CRMP4 overexpression boosted MRTF nuclear accumulation; silencing reduced nuclear MRTF and SRF levels (Fig. 5 K-M). This confirms that CRMP4 influences SRF expression activity by regulating MRTF nuclear translocation. To further elucidate the regulatory mechanism of CRMP4 on the MRTF-SRF signaling axis, we established stable CRMP4-overexpressing and CRMP4-silenced ihESC cell models using stable transfection techniques to address the limitations of primary cell passage and the insufficient efficiency of siRNA transient transfection. WB and qPCR analysis showed that CRMP4 manipulation did not alter MRTF protein but decreased MRTF mRNA when silenced; SRF mRNA and protein rose with CRMP4 overexpression and fell with knockdown (Fig. 5 N-P). Thus, CRMP4 does not control MRTF protein abundance but governs its nuclear translocation, thereby indirectly dictating SRF-dependent transcription. Fig. 5 MRTF/SRF pathway was activated in endometriosis and is regulated by the expression of CRMP4. ( A-C ): The expression levels of MRTF and SRF in the endometrial tissues of the three groups were assessed using Western blotting (WB) and RT-qPCR. ( D-F ): Immunofluorescence revealed the fluorescence intensity of MRTF (green) and SRF (red) in the endometrial tissues across the three groups, which was quantified using ImageJ analysis. ( G) : RT-qPCR analysis demonstrated the expression levels of MRTF/SRF downstream target genes FOSB, JUNB, ATF3, and ZEP3 in the endometrial tissues of the three groups. ( H-J ): The expression levels of MRTF and SRF in three groups of stromal cells were assessed using Western blotting (WB) and RT-qPCR. ( K ): The localization characteristics of MRTF and SRF within the nucleus under different CRMP4 expression levels (overexpression/silencing) was used by immunofluorescence examined. ( L-M ): Quantitative analysis of the nuclear-to-cytoplasmic ratio of MRTF (green fluorescence signal) and the expression levels SRF (red fluorescence signal) were performed using ImageJ software. ( N-P ): The expression levels of CRMP4,MRTF and SRF were measured in stable transfected cell lines with overexpression and silencing of CRMP4 using WB and RT-qPCR techniques.(Data are mean ± SEM. Groups analysed by one-way ANOVA followed by Tukey’s post-test. * P  < 0.05, ** P  < 0.01, *** P  < 0.001, **** P   0.05) MRTF/SRF pathway was activated in endometriosis and is regulated by the expression of CRMP4. ( A-C ): The expression levels of MRTF and SRF in the endometrial tissues of the three groups were assessed using Western blotting (WB) and RT-qPCR. ( D-F ): Immunofluorescence revealed the fluorescence intensity of MRTF (green) and SRF (red) in the endometrial tissues across the three groups, which was quantified using ImageJ analysis. ( G) : RT-qPCR analysis demonstrated the expression levels of MRTF/SRF downstream target genes FOSB, JUNB, ATF3, and ZEP3 in the endometrial tissues of the three groups. ( H-J ): The expression levels of MRTF and SRF in three groups of stromal cells were assessed using Western blotting (WB) and RT-qPCR. ( K ): The localization characteristics of MRTF and SRF within the nucleus under different CRMP4 expression levels (overexpression/silencing) was used by immunofluorescence examined. ( L-M ): Quantitative analysis of the nuclear-to-cytoplasmic ratio of MRTF (green fluorescence signal) and the expression levels SRF (red fluorescence signal) were performed using ImageJ software. ( N-P ): The expression levels of CRMP4,MRTF and SRF were measured in stable transfected cell lines with overexpression and silencing of CRMP4 using WB and RT-qPCR techniques.(Data are mean ± SEM. Groups analysed by one-way ANOVA followed by Tukey’s post-test. * P  < 0.05, ** P  < 0.01, *** P  < 0.001, **** P   0.05) Cytoplasmic MRTF levels mirror the G-actin/F-actin balance; nuclear MRTF triggers SRF-dependent transcription [ 40 ]. To clarify whether the regulatory effect of CRMP4 on MRTF nuclear translocation is mediated through the actin polymerization pathway, this study employed actin dynamics-interfering drugs in a stable transfection model of ihESC cells. The drugs used include Cytochalasin D (which binds to free G-actin with high affinity to inhibit actin polymerization) and Jasplakinolide (which enhances the stability of F-actin structures). We treated CRMP4-silenced or -overexpressing ihESC with Cytochalasin D (40 nM, IC₅₀) or Jasplakinolide (7 µM, IC₅₀) (Fig. 6 A-B). The expression of MRTF in the nucleus and cytoplasm was detected using protein immunoblotting (Fig. 6 .C-F) and by assessing the fluorescence intensity of MRTF (green) through cellular immunofluorescence (Fig. 6 .G-I). The results indicate that in sh-CRMP4 cells, nuclear MRTF fell and cytoplasmic MRTF rose relative to control; Jasplakinolide restored nuclear accumulation(Fig. 6 C-D, G-H). In contrast, OE-CRMP4 increased nuclear MRTF; Cytochalasin D reversed this shift(Fig. 6 E-F, G-I). These results clearly demonstrate that the nuclear translocation of MRTF is dependent on the actin polymerization activity mediated by CRMP4. Fig. 6 CRMP4 regulates actin polymerization in endometrial stromal cells by facilitating the nuclear translocation of MRTF, which activates SRF. ( A ): CCK-8 assay to determine the optimal concentration of the actin polymerization inducer Jasplakinolide in induced human embryonic stem cells (ihESCs). ( B ): CCK8 assay to identify the optimal concentration of the actin depolymerizing agent Cytochalasin D in ihESCs. ( C - D ): Western blot analysis to detect the expression levels of MRTF in the nucleus and cytoplasm across different groups: Control group, Control group with Jasplakinolide, sh-CRMP4 group, and sh-CRMP4 group with Jasplakinolide. ( E - F ): Western blot analysis to assess the expression levels of MRTF in the nucleus and cytoplasm across different groups: Control group, Control group with Cytochalasin, oe-CRMP4 group, and oe-CRMP4 group with Cytochalasin. ( G - I ): Immunofluorescence analysis to evaluate MRTF expression (green fluorescence) across different groups and to quantify the ratio of MRTF expression in the nucleus. ( J - K ): RT-qPCR to measure the expression levels of mRNA for MRTF, SRF, FOSB, JUNB, ATF3, and ZFP3 across different groups. (Data are mean ± SEM. Groups analysed by one-way ANOVA followed by Tukey’s post-test. *P < 0.05, **P < 0.01, ***P < 0.001, ****P 0.05) CRMP4 regulates actin polymerization in endometrial stromal cells by facilitating the nuclear translocation of MRTF, which activates SRF. ( A ): CCK-8 assay to determine the optimal concentration of the actin polymerization inducer Jasplakinolide in induced human embryonic stem cells (ihESCs). ( B ): CCK8 assay to identify the optimal concentration of the actin depolymerizing agent Cytochalasin D in ihESCs. ( C - D ): Western blot analysis to detect the expression levels of MRTF in the nucleus and cytoplasm across different groups: Control group, Control group with Jasplakinolide, sh-CRMP4 group, and sh-CRMP4 group with Jasplakinolide. ( E - F ): Western blot analysis to assess the expression levels of MRTF in the nucleus and cytoplasm across different groups: Control group, Control group with Cytochalasin, oe-CRMP4 group, and oe-CRMP4 group with Cytochalasin. ( G - I ): Immunofluorescence analysis to evaluate MRTF expression (green fluorescence) across different groups and to quantify the ratio of MRTF expression in the nucleus. ( J - K ): RT-qPCR to measure the expression levels of mRNA for MRTF, SRF, FOSB, JUNB, ATF3, and ZFP3 across different groups. (Data are mean ± SEM. Groups analysed by one-way ANOVA followed by Tukey’s post-test. *P < 0.05, **P < 0.01, ***P < 0.001, ****P 0.05) We assessed the mRNA levels of FOSB, JUNB, ATF3, and ZFP3 across different experimental groups using RT-qPCR. The results indicated that the expression of FOSB, JUNB, ATF3, and ZFP3 was significantly downregulated in the sh-CRMP4 group.; Jasplakinolide partially rescued FOSB, JUNB and ZFP3 (Fig.  6 J). In the oe-CRMP4 group, the expression of FOSB, JUNB, ATF3, and ZFP3 was significantly upregulated, whereas Cytochalasin D attenuated their induction (Fig.  6 K). FOSB, JUNB, ATF3, and ZFP3 are associated with cellular differentiation, proliferation, apoptosis, and migration/invasion. This study confirms that CRMP4 promotes MRTF nuclear translocation through actin polymerization, thereby activating the SRF signaling pathway. This activation specifically upregulates the expression of target genes, ultimately driving the invasive and migratory capabilities of ectopic mesenchymal cells. In the field of cytoskeletal research related to endometriosis (EMs), the Rho signaling pathway has been confirmed to play a critical regulatory role. Wu et al. found that the expression intensity of key molecules in the Rho signaling pathway within endometriotic lesions is significantly positively correlated with disease staging, suggesting that the Rho pathway may be linked to the pathogenesis and severity of EMs [ 41 ]. Notably, RhoA, a core member of this pathway, has been shown to be abnormally activated and closely associated with the invasive characteristics of EMs cells [ 42 ]. Based on this, to further investigate the therapeutic potential of CRMP4 in vivo, we administered lentivirus containing CRMP4 knockdown and overexpression sequences via intraperitoneal injection in a previously established mouse model of endometriosis (Fig. S2 ). This study established a dual control system: in addition to a conventional PBS control group, a RhoA inhibitor was used as a positive control for treatment, aiming to compare the effects of CRMP4 gene intervention with those of RhoA and to deeply analyze the molecular regulatory mechanisms of CRMP4 in the pathological process of EMs. After 28 days, typical endometriotic lesions adhering to surrounding tissues were observed in the peritoneal cavities of the modeled mice, with lesions infiltrating surrounding tissues and forming abnormal blood vessels (Fig. 7 A). After isolating the lesions and performing HE staining, typical glandular and stromal structures of the formed endometriotic lesions were evident, Immunofluorescence staining of the lesions revealed E-cadherin-positive red fluorescence in the endometrial glandular epithelium and Vimentin-positive green fluorescence in the endometrial stromal structures (Fig. 7 B). Compared to the control group, CRMP4 knock-down markedly reduced lesion volume (Fig. 7 C-E); efficacy was verified by WB, qPCR and IHC (Fig. 7 F-K). Fig. 7 CRMP4 regulating endometriosis progression in vivo. ( A ): The EMs lesions formed in the peritoneal cavity of mice from different groups. ( B ): The HE staining of mouse EMs lesions reveals typical glandular structures (highlighted by the red box) and stromal components; Immunofluorescence staining of mouse EMs lesions shows that Vimentin (green fluorescence) marks the stromal areas, while E-Cadherin (red fluorescence) highlights the glandular regions. ( C ): Typical ectopic lesions in the abdominal cavity of EMs mice in different groups. ( D-E ): The number and volume of EM lesions formed in different groups. ( F-H ): The expression levels of CRMP4 and RhoA in endometrial tissues from mice in different groups were analyzed. ( I-K ): The expression levels of CRMP4, RhoA, MRTF and SRF in EMs lesions of different groups were assessed using immunohistochemistry. ( L-N ) The expression levels of MRTF and SRF in EMs lesions of different groups were assessed using qPCR, Western blotting. ( O-P ): The expression levels of MRTF and SRF in EMs lesions of different groups were assessed using immunofluorescence. ( Q ): The expression levels of FOSB, JUNB, ATF3, and ZFP36 in the lesions of endometriosis (EMs) in mice. (Data are mean ± SEM. Groups analysed by one-way ANOVA followed by Tukey’s post-test. * P  < 0.05, ** P  < 0.01, *** P  < 0.001, **** P   0.05) CRMP4 regulating endometriosis progression in vivo. ( A ): The EMs lesions formed in the peritoneal cavity of mice from different groups. ( B ): The HE staining of mouse EMs lesions reveals typical glandular structures (highlighted by the red box) and stromal components; Immunofluorescence staining of mouse EMs lesions shows that Vimentin (green fluorescence) marks the stromal areas, while E-Cadherin (red fluorescence) highlights the glandular regions. ( C ): Typical ectopic lesions in the abdominal cavity of EMs mice in different groups. ( D-E ): The number and volume of EM lesions formed in different groups. ( F-H ): The expression levels of CRMP4 and RhoA in endometrial tissues from mice in different groups were analyzed. ( I-K ): The expression levels of CRMP4, RhoA, MRTF and SRF in EMs lesions of different groups were assessed using immunohistochemistry. ( L-N ) The expression levels of MRTF and SRF in EMs lesions of different groups were assessed using qPCR, Western blotting. ( O-P ): The expression levels of MRTF and SRF in EMs lesions of different groups were assessed using immunofluorescence. ( Q ): The expression levels of FOSB, JUNB, ATF3, and ZFP36 in the lesions of endometriosis (EMs) in mice. (Data are mean ± SEM. Groups analysed by one-way ANOVA followed by Tukey’s post-test. * P  < 0.05, ** P  < 0.01, *** P  < 0.001, **** P   0.05) Subsequently, we assessed the expression levels of MRTF and SRF in the lesions of mice across different groups using WB, qPCR, and IHC. Our results indicated that sh-CRMP4 lowered MRTF mRNA without affecting protein (Fig.  7 I-N), consistent with cellular data. OE-CRMP4 up-regulated both MRTF and SRF at mRNA and protein levels, more prominently at the RNA level (Fig.  7 L-N). The immunohistochemical staining results of MRTF and SRF in the lesions among different groups (Fig.  7 O-P), where the SRF positive expression rate in sh-CRMP4 group was significantly downregulated compared to the control group. Meanwhile, in the RhoA inhibitor group, SRF positive expression was decreased but not statistically significant (Fig.  7 P). Immunofluorescence confirmed higher MRTF/SRF signals in oe-CRMP4 lesions and lower SRF intensity in sh-CRMP4 lesions; Rhosin produced downward trends without significance (Fig.  7 O-P). Additionally, we used RT-qPCR to assess the expression levels of FOSB, JUNB, ATF3, and ZFP36 in the lesions of different groups (Fig.  7 Q). The results revealed oe-CRMP4 significantly elevated JUNB, ZFP36 and ATF3, leaving FOSB unchanged; sh-CRMP4 down-regulated JUNB, ZFP36 and FOSB, with ATF3 unaltered; Rhosin reduced JUNB and FOSB only. Collectively, CRMP4 regulates the core components of the MRTF/SRF signaling axis, selectively modulating the expression of downstream effector molecules such as JUNB, ZFP36, FOSB, and ATF3. This mechanism plays a key regulatory role in the pathological progression of endometriosis.

This study followed the ethical guidelines of The General Hospital of Ningxia Medical University(Approval No.KYLL-2024-0685). This study we recruited 32 women diagnosed with endometriosis by laparoscopy and histology at the General Hospital of Ningxia Medical University, and 31 infertile patients without endometriosis or adenomyosis served as controls. All subjects underwent combined laparoscopic and hysteroscopic exploration; postoperative endometrial biopsy confirmed normal endometrial. Clinical characteristics—age, BMI, height, pregnancies, age at menarche, menstrual cycle and duration, and cesarean section rate—were recorded. Endometriosis cases were ASRM-staged (American Society for Reproductive Medicine), had no comorbidities, and had not used hormones or related drugs within 3 months. Within 10 min after surgery, tissues were collected under sterile conditions: one portion was transported on ice for primary stromal-cell isolation, and another was snap-frozen in liquid nitrogen. Detailed patient data are in Table  1 ; informed consent was obtained from all participants. Table 1 Clinical characteristics of women with and without (control) endometriosis Age (years) Control ( n  = 31) Endometriosis group ( n  = 32) P -value 36.65 ± 7.718 38.38 ± 5.558 0.313 BMI (kg/m2) 23.01 ± 7.718 22.8 ± 3.2867 0.757 Gravidity 1.94 ± 1.569 1.72 ± 1.17 0.538 Parity 1.19 ± 1.138 1.03 ± 0.647 0.492 Menarche 13.45 ± 1.207 13.5 ± 1.459 0.887 Menstrual average cycle (days) 28.1 ± 2.508 28.88 ± 2.97 0.265 Menstrual duration (days) 5.68 ± 1.536 5.44 ± 1.625 0.55 C/S (%) 8(31) 11(32) 0.459 ASRM stage  Stage 1 3(32) (9.38%)  Stage 2 8(32) (25.00%)  Stage 3 10(32) (31.25%)  Stage 4 11(32) (34.37%) Statistical analysis was performed using Student’s t-test. Data are mean ± SEM. ASRM: American Society for Reproductive Medicine Clinical characteristics of women with and without (control) endometriosis Statistical analysis was performed using Student’s t-test. Data are mean ± SEM. ASRM: American Society for Reproductive Medicine The ihESCs (human endometrial ectopic stromal cells lines) were purchased from IMMOCELL (Xiamen, Fujian China). Extraction of Primary EM-EcESCs: velvety, pale-yellow areas of ectriotic cyst wall were minced in ice-cold PBS, digested 30–40 min at 37 °C with 0.25% collagenase II + 0.15% collagenase IV, pipetted every 10 min. Digest was quenched with 10 vol PBS–1% FBS, sequentially filtered (70 μm, 40 μm), centrifuged 300 g 5 min, and plated in Advanced DMEM/F12-10% FBS-1% P/S. Extraction process of Control-Endometrial stromal cells (CON-EuESCs) and Endometriosis-Endometrial stromal cells (EM-EuESCs) were similarly; excess erythrocytes were removed by triple PBS rinse and red blood cell lysis buffer. Mesenchymal clumps attached within 2 h; medium was changed next day. Cells were passaged 1:3 every 3–4 d and used at passage 3. Purity: Cells on coverslips were fixed (4% paraformaldehyde, 30 min), permeabilised (0.5% Triton, 30 min), blocked (2% fetal bovine serum, 1 h) and stained: vimentin⁺/E-cadherin⁻/CK19⁻ defined stromal purity. Validated ESCs were used for subsequent experiments. Each patient isolate was validated once and assayed in triplicate. Normal (CON, n  = 31), eutopic (EU, n  = 32) and ectopic (CU, n  = 32) endometrium were paraffin-embedded, sectioned, and deparaffinized, blocked with 3% H₂O₂, antigen-retrieved in citrate buffer, and incubated with anti-CRMP4 (13661-1-AP; Proteintech 1:300) followed by HRP-conjugated secondary antibody(Proteintech, China). Mouse ectopic lesions ( n  = 6) were stained identically using anti-MRTF (TA5302S, Abmart,1∶400) and anti-SRF (66742-1-Ig Abmart,1∶200). Expression was quantified with H-scores from three random 400× fields by an investigator blinded to clinical status; mouse sections were analysed with ImageJ. The tissue or cultured cells RNA extraction process based on the Axyprep™ total RNA kit (Axygen, Bioscience, Hangzhou, China), followed by reverse transcription to produce cDNA. Quantitative real-time PCR (qRT–PCR) analysis was performed using SYBR Green PCR Master Mix (Seven, Beijing, China), with ACTB, GAPDH, serving as the reference genes. The qRT-PCR primers are listed in Table  2 . The mRNA expression of the samples was normalized using the 2−∆∆CT method. Table 2 The sequence of the primers for mRNAs. human(h), mouse(m) Gene Name Forward and Reverse Primer CRMP4 (h) F:5’TGTGGTGCCTGAGCCTGAGTC3’ (h) R:5’GCTGTCATTCCAGTGGGTGATGTC3’ (m) F: 5’GCCCACTGTACCTTTAGCACTGC3’ (m) R: 5’CATACGCTCCTCCACGCCATTG3’ SRF (h) F: 5’GACCTCACGCAGACCTCCTC3’ MRTF (h) R: 5’CAGTTGTGGGCACGGATGAC3’ (m) F: 5’CACCGTGCTCAATGCCTTCTC3’ (m) R: 5’CACCTGGCTCCTGGACCTG3’ (h) F: 5’CTGGACGCCTGGAGGACTTC3’ (h) R: 5’TCTGGCTATGGAGGTCGTCAATG3’ (m) F: 5’TGCTGCGTCCTGCTGTCTAAG3’ (m) R: 5’TCCTCAATCTGCTTGTCCTTCTCC3’ FOSB (h) F: 5’GCTGGCGGAGGTGAGAGATTTG3’ (h) R: 5’GCGTCTTGGCTGGTCTGGAAG3’ (m) F: 5’GACCCTTATGACATGCCAGGAACC3’ (m) R: 5’ACTGGTGGTTGTGCTGGTTGAAG3’ JUNB (h) F: 5’TACCCGACGACCACCATCAGC3’ (h) R: 5’ACGGTCTGCGGTTCCTCCTTG3’ (m) F: 5’TGGCAGCGGTGGAGGTACAG3’ (m) R: 5’ACGTGGTTCATCTTGTGCAGGTC3’ ATF3 (h) F: 5’CCTCGGGGTGTCCATCACAAAAG3’ (h) R: 5’AGGCACTCCGTCTTCTCCTTCTTC3’ (m) F: 5’GGAAAAGGAGGCGGCGAGAAAG3’ (m) R: 5’CAGCTCAGCATTCACACTCTCCAG3’ ZFP36 (h) F: 5’ATGTCGGACCTTCTCAGAGAGTGG3’ GAPDH ACTB (h) R: 5’TCTTGTATTTGGGGTGGCGATTGG3’ (m) F: 5’AGCCCATCTGCCCACTCTCTG3’ (m) R: 5’CCCAAACACCCCTGCCTCAAAG3’ (h) F: 5’CAGGAGGCATTGCTGATGAT3’ (h) R: 5’GAAGGCTGGGGCTCATTT3’ (m) F: 5’GGTTGTCTCCTGCGACTTCA3’ (m) R: 5’TGGTCCAGGGTTTCTTACTCC3’ (h) F: 5’CCTGGCACCCAGCACAAT3’ (h) R: 5’GGGCCGACTCGTCATAC3’ (m) F: 5’GTGCTATGTTGCTCTAGACTTCG3’ (m) R: 5’ATGCCACAGGATTCCATACC3’ The sequence of the primers for mRNAs. human(h), mouse(m) (h) F:5’TGTGGTGCCTGAGCCTGAGTC3’ (h) R:5’GCTGTCATTCCAGTGGGTGATGTC3’ (m) F: 5’GCCCACTGTACCTTTAGCACTGC3’ (m) R: 5’CATACGCTCCTCCACGCCATTG3’ (h) R: 5’CAGTTGTGGGCACGGATGAC3’ (m) F: 5’CACCGTGCTCAATGCCTTCTC3’ (m) R: 5’CACCTGGCTCCTGGACCTG3’ (h) F: 5’CTGGACGCCTGGAGGACTTC3’ (h) R: 5’TCTGGCTATGGAGGTCGTCAATG3’ (m) F: 5’TGCTGCGTCCTGCTGTCTAAG3’ (m) R: 5’TCCTCAATCTGCTTGTCCTTCTCC3’ (h) R: 5’GCGTCTTGGCTGGTCTGGAAG3’ (m) F: 5’GACCCTTATGACATGCCAGGAACC3’ (m) R: 5’ACTGGTGGTTGTGCTGGTTGAAG3’ (h) R: 5’ACGGTCTGCGGTTCCTCCTTG3’ (m) F: 5’TGGCAGCGGTGGAGGTACAG3’ (m) R: 5’ACGTGGTTCATCTTGTGCAGGTC3’ (h) R: 5’AGGCACTCCGTCTTCTCCTTCTTC3’ (m) F: 5’GGAAAAGGAGGCGGCGAGAAAG3’ (m) R: 5’CAGCTCAGCATTCACACTCTCCAG3’ GAPDH ACTB (h) R: 5’TCTTGTATTTGGGGTGGCGATTGG3’ (m) F: 5’AGCCCATCTGCCCACTCTCTG3’ (m) R: 5’CCCAAACACCCCTGCCTCAAAG3’ (h) F: 5’CAGGAGGCATTGCTGATGAT3’ (h) R: 5’GAAGGCTGGGGCTCATTT3’ (m) F: 5’GGTTGTCTCCTGCGACTTCA3’ (m) R: 5’TGGTCCAGGGTTTCTTACTCC3’ (h) F: 5’CCTGGCACCCAGCACAAT3’ (h) R: 5’GGGCCGACTCGTCATAC3’ (m) F: 5’GTGCTATGTTGCTCTAGACTTCG3’ (m) R: 5’ATGCCACAGGATTCCATACC3’ Whole-cell and nuclear proteins were extracted, and the proteins were transferred to polyvinylidene fluoride membrane and incubated with primary antibody [CRMP4 (13661-1-AP; Proteintech1:3000), MRTF (TA5302S; Abmart,1:2000), SRF (66742-1-Ig, Abmart, 1:1000), RhoA (66733-Ig; Proteintech,1:1000), LaminA/C (10298-1-AP; Proteintech 1∶2000), GAPDH (10494-1-Ig; Proteintech 1:10000), ACTB (66009-1-Ig; Proteintech 1∶20000)] and the secondary antibodies. The membrane was subjected to an enhanced chemiluminescence solution, and the results were evaluated with ImageJ. Membrane and cytosolic proteins were extracted using a membrane and cytosol protein extraction kit (BB-3102, Bestbio). Transfection of primary endometrial ectopic stromal cells with siRNA to silence the expression of CRMP4: Three CRMP4-targeting siRNAs (synthesised by General Biol) were first tested individually; the most effective sequence, siRNA1, was selected for all downstream experiments. The siRNAs were synthesized by General Biol. Their target sequences of the siRNAs were as follows: forward, 5’CCUUUUAUGCUGAUAUUUA3’ and reverse, 5’UAAAUAUCAGCAUAAAAGG3’ for siRNA1; forward, 5’GGCUUAUAAGGAUUUGUAU3’ and reverse, 5’AUACAAAUCCUUAUAAGCC3’ for siRNA2; forward, 5’CAGGAAAAAGGAAAUGUA3’ and reverse, 5’UACAUUUCCUUUUUUCCUG3’ for siRNA3. Cells were transfected with 50 nM siRNA using LipofectamiTM 3000(L3000015, ThermoFisher, China). The ectopic endometrial stromal cells overexpressing CRMP4 were constructed using viral solution containing Polybrene (4 µg/mL) at an MOI of 30. Stable overexpression and knockout lines were constructed by He Yuan Biotechnology (Shanghai) Co., Ltd. Transfection efficiency was evaluated after 72 h using qRT-PCR and Western blotting (WB). Only EcESCs from the third to sixth generations were utilized for cell transfection experiments. Migration of EM-EcESCs and CON-EuESCs were evaluated using wound-healing and Transwell assays. For wound-healing, EM-EcESCs or CON-EuESCs were inoculated at 5 × 10 5 cells per well. Once confluence exceeds 95%, a 200 µL pipette tip created vertical scratches. After aspirating the medium, wells were washed three times with PBS and replace with serum-free medium. Images were captured at time 0 h using a microscope, the plate was then incubated and migration images re-captured at 6 h and 12 h. Scratch area was quantified with ImageJ for statistical analysis. For Transwell assays, thawed matrix gel was mixed with serum-free medium at 1:8 and carefully layered in the upper chamber, avoiding bubbles. After 3 h at 37℃ for gel formation, residual liquid was aspirated and serum-free medium was added for an additional hour of hydration. Cells at 70% confluence were digested and resuspended to 2.5 × 10 5 cells/mL; 200 µL of this suspension was seeded into each upper chamber (the lower chamber contained 10% FBS medium). After 24 h incubation, the cells were fixed and stained (4% paraformaldehyde for 25 min, 0.4% crystal violet for 5 minutes). Following a rinse with distilled water, five fields were examined under a microscope, and averages were calculated for statistical analysis. We digested 5 × 10 6 treated EM-EcESCs with trypsin and centrifuged at 1000 rpm for 5 min to discard the supernatant. Fix the samples in 0.25% glutaraldehyde at 4 °C for 4 h, then in 1% osmium tetroxide at room temperature for 1 h. Dehydrate the samples through a graded ethanol (50%, 70%, 90%, 100%) and 100% acetone, 15 min each. After dehydration, embed the samples in epoxy resin for 4 h, and prepare 70 nm ultrathin sections. Retrieve the sections onto copper grids and perform double staining with uranyl acetate and lead citrate. Examine the samples using transmission electron microscopy at 50,000× magnification to observe cell membrane pseudopodia morphology, and at 500,000× to analyze actin filament structure within the pseudopodia. Randomly select 4 fields of view, measuring pseudopodia length and number in 16 cells total, and analysis the data statistically. 2 × 10 5 EM-EcESCs or ihESCs were seeded on slides overnight, then subjected to overexpression or silencing. After twice 37 °C PBS washes, cells were fixed (4% paraformaldehyde, 30 min), permeabilised (0.1% Triton X-100, 3 min), and washed (PBS, 3 × 10 min).Under dark conditions, slides were incubated 1.5 h with Penitrem A green-fluorescent or Vinculin primary antibody; EM-EcESCs were further labelled 1 h at room temperature with rhodamine–anti-SRF (1:200, Abmart, 66742-1-Ig) and fluorescein isothiocyanate (FITC) labeled anti-MRTF (1:400, Abmart, TA5302S). Nuclei were counterstained with DAPI (Abcam) and images acquired on a fluorescence microscope; fluorescence area was quantified in ImageJ. Following the G-F-Actin In Vivo Assay Kit (Cytoskeleton, BK037): aspirate medium completely, add pre-warm LAS2, scrape cells at a 30°angle, incubate lysate 10 min at 37 °C, then transfer 100 µL to a new EP tube. Centrifuge 2000 rpm, 5 min, room temperature, and move supernatant to a new EP tube. Ultracentrifuge 100,000 g, 1 h, 37 °C; collect supernatant (G-actin) gently, leaving F-actin pellet. Add 100 µL depolymerization buffer to pellet, ice 1 h with gentle mixing every 15 min(label F-actin). Add 25 µL 5× SDS buffer to both fractions, mix well for analysis. We used CCK-8 to Screen the Optimal Drug Concentration of Cytochalasin D and Jasplakinolide. Based on the literature, a concentration gradient for Cytochalasin D (0–8 µM) and Jasplakinolide (10–45 nM) was established. In a 96-well plate, 100 µL of PBS was added to the outer wells to prevent evaporation. The second column served as a no-cell medium control, while the remaining wells were seeded with 3 × 10 3 ihESC cells per well (including 100 µL of complete medium). After culturing at 37 °C with 5% CO 2 for 24 h, gradient drug treatments were applied. Following another 24-hour incubation, 10 µL of CCK-8 solution was added to each well (to avoid bubble interference). After incubating for 1 h, the wells were checked, and any residual bubbles were pierced with a heated needle. The absorbance value at 450 nm was measured using a microplate reader to calculate cell viability. All animal experiments were conducted in accordance with the Guide for the Care and Use of Laboratory Animals published by the US National Institutes of Health. The Institution Animal Care and Use Committee Ningxia Medical University reviewed and approved this study(Approval No. IACUC-NYLAC-2024-170). A mouse EMs model was built by ectopic uterine fragment transplantation. Donors ( n  = 12) and recipients ( n  = 24) were synchronized with 3 µg benzyl estradiol s.c. 72 h before surgery; controls ( n  = 6) received vehicle only. Under tribromoethanol anesthesia, donor uteri were excised, opened, and minced in PBS; fragments were injected 1:2 into recipient peritoneal cavities. Three days later, recipients were randomly assigned ( n  = 6 each): (i) the negative control group(PBS), (ii) the OE-CRMP4 group, (iii) the sh-CRMP4 group, or (iv) the RhoA inhibitor group (Rhosin). Lentiviral stocks (1 × 10⁸ PFU/mL, 200 µL per mouse) or Rhosin(40 mg/kg) were given i.p. on days 7, 14 and 21. Mice were continuously monitored daily and killed on day 28; lesions were measured (length, width, volume) and validated by IHC, qRT-PCR and WB. The flowchart of mouse modeling is shown in Figure S2 . The results are shown as the mean ± SEM deviation from a minimum of three independent experiments. All statistical analyses were conducted using GraphPad Prism 8 software (San Diego, CA, USA). Statistical significance was assessed using Student’s t-test, one-way ANOVA, and two-way ANOVA. In all figures, a P -value of less than 0.05 was deemed significant. NS indicates no significance ( P  ≥ 0.05), * P  < 0.05, ** P  < 0.01, *** P  < 0.001, **** P  < 0.0001).

Endometriosis (EMs), while classified as a benign disease, exhibits malignant biological characteristics that can affect multiple organs in the pelvic and extra-pelvic regions, including the brain [ 5 ]. Ectopic endometrial cells mimic tumour behaviour—aberrant adhesion, colonization, migration, and invasion —with migration now recognised as central to intraperitoneal implantation and differentiation [ 9 – 11 ]. These processes require a dynamic cytoskeleton, and the cytoskeleton is closely linked to various human diseases. Actin filaments, microtubules and intermediate filaments continuously reorganise to generate shape and propulsion [ 12 ]; such remodelling is exaggerated in EMs [ 13 ]. Here we delineate how CRMP4 drives this programme in endometrial stromal cells: by enhancing actin polymerization it triggers MRTF nuclear import, SRF activation and transcription of pro-migratory genes (Fig. 8 ). We used infertile women without endometriosis as controls. This approach is common in endometriosis research, because ethical rules forbid biopsy in healthy fertile volunteers. The control group matched surgery conditions and cycle phase, but our findings may not fully apply to women with normal fertility. In summary, targeting CRMP4 presents a promising strategy for improving the treatment outcomes of endometriosis by modulating the polymerization state of actin in the cytoskeleton of endometrial stromal cells. Fig. 8 Hypothesis mechanism diagram of this study: CRMP4 provides the morphological basis for cell movement by promoting the polymerization of free G-actin into F-actin, which facilitates the growth of cellular protrusions in mesenchymal cells. Additionally, it enhances the nuclear translocation of MRTF, further activating SRF and upregulating the expression of its target genes, including FOSB, JUNB, ATF3, and ZEP3. This process ultimately strengthens the migratory and invasive capabilities of mesenchymal cells Hypothesis mechanism diagram of this study: CRMP4 provides the morphological basis for cell movement by promoting the polymerization of free G-actin into F-actin, which facilitates the growth of cellular protrusions in mesenchymal cells. Additionally, it enhances the nuclear translocation of MRTF, further activating SRF and upregulating the expression of its target genes, including FOSB, JUNB, ATF3, and ZEP3. This process ultimately strengthens the migratory and invasive capabilities of mesenchymal cells Currently, experimental data on cytoskeleton involvement in EMs pathogenesis remain relatively limited. One transcriptomic study in EMs patients found most differentially expressed genes encode focal adhesion, actin cytoskeleton and MAPK signaling proteins [ 43 ]. This research found that, CRMP4 mRNA and protein were highest in ectopic endometrial tissue, intermediate in eutopic tissue, and lowest in normal tissue, forming a stepwise gradient. Immunohistochemical H-Score confirmed CRMP4 levels significantly elevated in both eutopic and ectopic versus normal endometrial tissue, with stronger upregulation in ectopic tissues; these differences were statistically significant, suggesting an important role for CRMP4 in the development of EMs lesions. The formation of ectopic lesions in EMs requires migration, implantation, and adhesion of eutopic endometrial cells [ 44 ]. This cell migration is accompanied by morphological restructuring, with dynamic assembly of actin playing a crucial role. Actin monomers continuously polymerize into filaments at the plasma membrane; when the barbed ends of actin filaments open a gap with the membrane, new monomers are added to the barbed end, promoting filament elongation. The mechanical force generated drives cell membrane deformation and migration [ 27 ]. Our research group compared lamellipodia number and length in three stromal cell groups by transmission electron microscopy. EM-EcESCs group showed significantly more and longer lamellipodia than the other two groups, indicating richer lamellipodial structure. EM-EuESCs also had more lamellipodia than CON-EuESCs, indicating that the stromal cells in the eutopic endometrium of EMs have undergone morphological changes, thereby supporting the eutopic endometrium hypothesis. In EM-EcESCs, CRMP4 silencing by siRNA reduced lamellipodia length and number, while overexpression increased them. This suggests CRMP4 plays a promoting role in the growth of lamellipodia in EM-EcESCs. Immunofluorescence of F-actin bundles in lamellipodia showed CRMP4 silencing decreased fluorescence intensity, while overexpression increased it, confirming CRMP4 promotes lamellipodia growth by facilitating F-actin polymerization. Furthermore, during cell migration, the assembly and disassembly of focal adhesions (FAs) determine the direction of cell movement. This study found that CRMP4 could enhance the formation of FAs in EM-EcESCs, highlighting its role in cytoskeletal remodeling during ectopic endometriosis. To confirm CRMP4’s role in the polymerization of G-actin to F-actin, we employed ultracentrifugation and a G-actin/F-actin separation kit to assess the ratios. after CRMP4 modulation [ 33 ]. The results demonstrated that CRMP4 significantly promotes G-actin polymerization into F-actin. The G-actin/F-actin interconversion is fundamental to cytoskeletal remodeling and influences the nuclear translocation factor translocation by modulating G-actin levels, thereby activating downstream factors that alter cell function. According to the reports of the existing studies, we can infer that CRMP4 drives actin polymerization via dual pathways: directly, by forming C-terminal tetramers that crosslink F-actin (Kapp ≈ 730 µM⁻¹) and stabilize newly polymerized filaments [ 45 ]; and indirectly, by sequestering RhoA to suppress RhoA-ROCK-LIMK signaling, reduce cofilin phosphorylation, and expand the F-actin pool [ 29 , 46 ]. In endometriotic stromal cells these dual actions converge to amplify actin remodeling and invasion, positioning CRMP4 upstream of the globally dysregulated RhoA-downstream actin-binding proteins observed in endometriotic lesions [ 47 ]. In recent years, the roles of serum response factor (SRF) and its co-activator myocardin-related transcription factor (MRTF) in the progression of various cancers have garnered significant attention [ 48 ]. MRTF, as a cofactor of SRF, dissociates from actin monomers upon polymerization, enters the nucleus and activates SRF target genes, including ATF3, FOSB, JUNB, and ZFP36 [ 49 , 50 ].To assess MRTF/SRF pathogenic role in EMs, we measured their expression by WB, RT-qPCR and immunofluorescence. The findings revealed that MRTF expression was elevated in in EU group, while SRF protein and mRNA levels were significantly higher in both in EU and CU groups versus CON group. Additionally, RT-qPCR showed ATF3, FOSB, JUNB, and ZFP36 mRNA were significantly upregulated in ectopic and in situ EM tissues, suggesting that MRTF/SRF may play crucial roles in EMs progression. Nuclear actin polymerization effectively regulates the nuclear localization and activity of MRTF-A [ 35 ]; actin polymerization facilitates MRTF-A and F-actin interaction, activating SRF-dependent gene expression [ 34 ]. In EM-EcESCs, this study found CRMP4 expression influenced MRTF nuclear translocation. Recent studies have illuminated cytoskeletal remodeling in endometriosis, yet our work identifies a distinct mechanistic axis. Gong et al. demonstrated that KAT14 cooperates with SRF to promote fibrosis in ovarian endometrioma [ 51 ], while Marquardt et al. showed that SRF loss in eutopic endometrium triggers inflammatory fibrosis [ 52 ]. In sharp contrast, our study reveals that CRMP4 functions as an upstream regulator that drives actin polymerization-dependent MRTF nuclear translocation to enhance invasive migration of ectopic stromal cells—a malignant-like behavior mechanistically independent of fibrotic progression. Furthermore, Raja Xavier et al. linked PlGF-Rac1 signaling to actin polymerization but did not explore downstream MRTF/SRF activation [ 53 ], a gap our findings directly address. Arendt et al. comprehensively reviewed actin-binding proteins in endometriosis yet did not identify CRMP4 as a master regulator of the pro-invasive MRTF/SRF axis [ 47 ]. Collectively, these comparisons position our work as the first to delineate a CRMP4-centric pathway specifically governing ectopic stromal cell invasion, offering a therapeutic target orthogonal to fibrosis-driven mechanisms. To test if CRMP4 acts via actin polymerization, we treated CRMP4-modulated ihESC lines with Jasplakinolide or Cytochalasin D, assessing MRTF nuclear/cytoplasmic expression and translocation. RT-qPCR results indicated a positive correlation between the expression levels of the MRTF/SRF target genes ATF3, FOSB, JUNB, and ZFP36 and CRMP4 expression. Collectively, these experimental results suggest that CRMP4 promotes actin polymerization, facilitating MRTF nuclear translocation, SRF activation and downstream target-gene expression, thereby promoting the migration and invasion of EM-EcESCs. In the mouse model of endometriosis, CRMP4 inhibition significantly downregulated the mRNA levels of MRTF/SRF-dependent target genes FOSB, JUNB, ATF3, and ZFP36. Notably, both FOSB and JUNB function as transcription factors of activator protein 1 (AP-1), a complex that regulates gene expression in response to various stimuli, involving cytokines, growth factors, and multicellular responses such as stress response, differentiation, proliferation, and apoptosis [ 49 ]. AP-1 transcriptional regulation during inflammation specifically depends on JUNB [ 54 ]. Previous literature has reported that FOSB is upregulated in ectopic EMs tissues, correlating with elevated 17β-estradiol and local MMP9 [ 55 ]. Furthermore, the RNA-binding protein TTP, encoded by ZFP36, acts as a post-transcriptional regulator of the inflammatory response by binding to and degrading the mRNA of various cytokines [ 56 ]. Thus, these target genes play crucial roles in the occurrence and development of EMs.

This research reveals the significant role of CRMP4 in the progression of endometriosis. CRMP4 promotes the polymerization of globular actin (G-actin) into filamentous actin (F-actin), providing the morphological basis for the amoeboid movement of stromal cells. Additionally, it facilitates the nuclear translocation of MRTF, which further activates SRF and upregulates the expression of its target genes, including FOSB, JUNB, ATF3, and ZFP36, thereby enhancing the migration and invasion capabilities of stromal cells. Thus, targeting CRMP4 may offer a promising approach to inhibit the robust migratory and invasive states of endometrial stromal cells by regulating their cytoskeletal remodeling, ultimately improving the treatment outcomes for endometriosis.

Endometriosis(EMs) is a common chronic gynecological condition with an incidence rate of 10% to 15% among women of reproductive age, affecting nearly 200 million women worldwide [ 1 – 3 ]. Current treatment options for EMs, medical or surgical therapy, remains suboptimal: radical resection risks irreversible infertility, conservative surgery is followed by high recurrence, and prolonged endocrine suppression induces hypo-estrogenism and osteoporosis that compromise adherence [ 4 ]. The exact mechanisms underlying EMs are not fully understood. Although considered a benign condition, EMs disseminates to multiple pelvic and extra-pelvic organs, including the brain, mimicking malignant behavior [ 5 ]. A widely accepted theory proposed by Sampson, known as the “retrograde menstruation theory,” suggests that endometrial cells and tissue fragments can flow backward through the fallopian tubes into the pelvic cavity and subsequently implant and grow. Retrograde menstruation is necessary but insufficient, as only 10% of women with reflux develop lesions [ 6 ]. Additionally, other hypotheses, such as genetic factors, immune imbalance, stem cell transformation, and coelomic epithelial metaplasia, have not comprehensively explained the disease’s etiology [ 7 , 8 ]. Endometrial cells located ectopically in EMs exhibit behaviors similar to malignant tumor cells, displaying abnormal adhesion, colonization, migration, and invasive growth. Recent studies have indicated that cell migration plays a crucial role in the adhesion, implantation, and differentiation of endometrial tissue within the abdominal cavity [ 9 – 11 ]. The cytoskeleton is a dynamic network composed of highly ordered and interconnected filamentous protein polymers, including actin filaments, microtubules, and intermediate filaments. Actin filaments are directly involved in forming and maintaining tissue morphology while also serving as a source of cellular movement [ 12 ]. Recent research has shown that the occurrence and progression of EMs are accompanied by cytoskeletal remodeling [ 13 ]. Cell migration and invasion are crucial steps in the formation of ectopic lesions, and the cytoskeleton and its binding proteins are fundamental to this process, with many molecules involved in its precise regulation. However, the key steps and regulatory mechanisms by which endometrial cells induce migration and invasion through cytoskeletal remodeling remain unclear. In our transcriptomic profiling of ectopic lesions revealed enrichment of cytoskeletal regulators, including CRMP4. Recent tumour data indicate that CRMP4 can either restrain or promote invasion depending on context: its loss accelerates prostate-cancer dissemination [ 14 ], whereas high expression improves neuroblastoma survival [ 15 ]. In urothelial carcinoma, CRMP4 over-expression enhances motility and predicts poor prognosis, an effect linked to actin-network rewiring and metabolic reprogramming [ 16 ]. Therefore, the expression patterns of CRMP4 may vary across different tumors, and its biological functions may be mediated through multiple mechanisms. Whether CRMP4 operates as a pro- or anti-invasive factor in endometriosis remains untested. Pseudopodia are cellular structures formed by the polymerization of free actin (G-actin). The influence of the cytoskeleton on cellular biological functions can be mediated through the regulation of cell morphology, as well as by concurrently affecting the expression of transcription factors that control cell migration, invasion, and adhesion through the regulation of G-actin and polymerized actin (F-actin) [ 17 – 21 ]. Among these factors, Myocardin-related transcription factor A (MRTF-A) and Yes-associated protein (YAP) can translocate to the nucleus and become activated under conditions of actin polymerization, promoting the transcription of downstream target genes and thereby enhancing tumor cell proliferation, invasive migration, and metastasis [ 22 – 24 ]. Although YAP activation is documented in endometrial stromal cells [ 25 ], nuclear MRTF signalling remains unstudied in endometriosis. Elucidating how cytoskeletal remodelling couples to MRTF-driven transcription is therefore essential in the context of endometriosis. Our research indicates that CRMP4 is up-regulated in ectopic endometrial stromal cells (EcESCs). It drives G-actin polymerisation, enlarging lamellipodia and focal adhesions and simultaneously triggering MRTF nuclear import. Subsequent MRTF/SRF signalling amplifies pro-migration gene expression, accelerating invasion. We propose that inhibiting CRMP4 expression could suppress actin cytoskeleton polymerization, thereby reducing the migratory and invasive characteristics associated with endometriosis. Based on these findings, CRMP4 shows promise as a target for the treatment and diagnosis of endometriosis.

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