TAGLN: a stiffness-driven prognostic biomarker for metastatic colorectal cancer Authors: Yanshuo Ye, Yiyang Jia, Wei Li Published: 2025 ABSTRACT TAGLN is identified as a key prognostic gene in colorectal cancer through TCGA-COAD analysis, with high expression correlating with poor survival and advanced tumor stage. Functionally, TAGLN overexpression promotes epithelial-mesenchymal transition and enhances cancer cell migration. Transcriptomic profiling reveals its involvement in extracellular matrix remodeling and cell adhesion pathways. Mechanistically, TAGLN expression is upregulated by stiff tumor-mimicking extracellular matrix and is correlated with key mediators of collagen crosslinking and EMT. Clinically, TAGLN exhibits progressive overexpression from normal epithelium to primary tumors and metastatic lesions. These findings establish TAGLN as a stiffness-responsive regulator of ECM remodeling and EMT, driving colorectal cancer metastasis and serving as a potential therapeutic target. Introduction Colorectal cancer (CRC) remains one of the most prevalent and deadly malignancies worldwide, with metastasis accounting for the majority of CRC-related deaths [1, 2]. Despite advances in early detection and treatment, the 5-year survival rate for metastatic CRC remains below 15%, highlighting the urgent need to identify novel molecular drivers of CRC progression. The tumor microenvironment, particularly the dynamically remodeled extracellular matrix (ECM), has emerged as a critical regulator of metastatic dissemination [3]. ECM stiffening [4], a hallmark of solid tumors [5], promotes malignant progression through biomechanical and biochemical signaling that triggers epithelial-mesenchymal transition (EMT) - a key developmental program co-opted by cancer cells to acquire migratory and invasive capabilities [6]. While several ECM-associated proteins have been implicated in CRC metastasis, the molecular mechanisms linking ECM remodeling to EMT activation remain incompletely understood, particularly regarding how cancer cells sense and respond to mechanical cues from their microenvironment. Transgelin (TAGLN), an actin-binding protein initially characterized in smooth muscle cells, has recently gained attention as a potential mediator of tumor progression [7, 8]. Although TAGLN has been reported to be dysregulated in various cancers [9], its functional role and regulatory mechanisms in CRC remain controversial and poorly understood. Preliminary evidence suggests TAGLN may participate in cytoskeletal reorganization and cell motility [10], yet its potential involvement in ECM-mediated mechano-transduction and EMT has not been fully explored. This knowledge gap is particularly significant given the established role of actin cytoskeleton remodeling in mechanical signal transduction and the increasing recognition of TAGLN as a stress-responsive protein [11]. Furthermore, while several studies have associated TAGLN expression with clinical outcomes in various malignancies [12], comprehensive analyses of its prognostic value across different CRC stages and its potential as a therapeutic target are lacking. Resolving these questions could provide critical insights into CRC pathogenesis and identify new avenues for therapeutic intervention. In this study, we present a comprehensive investigation of TAGLN’s clinical significance, biological functions, and mechanical regulation in CRC. Through integrated analysis of TCGA data and functional experiments, we demonstrate that TAGLN serves as a prognostic biomarker associated with advanced disease stages and poor survival outcomes. Our mechanistic studies reveal that TAGLN promotes EMT and enhances migratory capacity in CRC cells, while transcriptomic analyses uncover its role in regulating ECM organization pathways. Significantly, we establish that TAGLN expression is mechanically regulated by substrate stiffness and correlates with key mediators of collagen crosslinking and matrix remodeling. Using clinically relevant stiffness substrates and patient specimens, we provide evidence that TAGLN functions as a mechanoresponsive effector linking ECM stiffening to EMT activation in CRC. These findings not only elucidate a novel mechanobiological axis in CRC progression but also position TAGLN as a potential therapeutic target for preventing metastatic dissemination. Our work bridges the gap between ECM biomechanics and intracellular signaling pathways, offering new perspectives on the mechanical regulation of cancer metastasis. Materials and methods Risk gene set and clinical data analysis The test dataset comprised RNA-seq expression profiles and clinical information from 338 COAD patients in TCGA. Gene expression was normalized using log2(TPM + 1). The least absolute shrinkage and selection operator (LASSO) Cox regression algorithm was employed for feature selection in the survival model, with validation performed via 10-fold cross-validation. Univariate and multivariate Cox proportional hazards regression analyses were further conducted to assess the association between gene sets and survival risk. All analytical methods were implemented in R (v4.0.3). The impact of gene expression on overall survival (OS) and disease-free survival (DFS) in COAD patients was assessed using the Gene Expression Profiling Interactive Analysis (GEPIA) database (http://gepia.cancer-pku.cn/) via log-rank testing. A hazard ratio (HR) > 1 was considered indicative of a risk factor. Cell culture Human CRC cell lines (HCT116, SW480, SW620, CACO2, and LOVO) were kindly provided by the Key Laboratory of Pathobiology, Ministry of Education, Jilin University. All cells were maintained in DMEM medium supplemented with 10% fetal bovine serum and 1% penicillin-streptomycin at 37 °C in a humidified atmosphere containing 5% CO2. Transfections The PLV3-CMV-TAGLN-Myc-GFP-Puro overexpression plasmid and corresponding empty vector control were purchased from MiaoLing Biotechnology Co., Ltd. For lentivirus production, 293 T cells were co-transfected with the target plasmid and packaging plasmids using Lipo2000 transfection reagent. After 48 h of transfection, the culture supernatant was collected and concentrated using a lentivirus concentration kit. HCT116 cells were then infected with the concentrated lentivirus in the presence of 8 µg/mL polybrene, followed by selection with 2 µg/mL puromycin to establish stable cell lines. For TAGLN knockdown, SW620 cells were transfected with siRNA sequence: 5’-GGTTTATGAAGAAAGCGCAGGAGCA-3’. Cell morphology Image acquisition was performed on a Nikon ECLIPSE Ti2 microscope with a 40× objective, yielding a pixel size of 0.325 μm × 0.325 μm. Cell morphology from bright-field images of adherent cells was analyzed using ImageJ (Fiji v1.54). Images were binarized using the Adjust Threshold module, and cells were automatically identified and their morphological descriptors were measured using the Analyze Particles module. The Aspect Ratio value from the results table, which indicates a more elongated, spindle-like shape as it increases, was used for statistical analysis. A minimum of 100 cells per group were analyzed randomly to ensure statistical reliability. Wound healing assay cells were seeded in 6-well plates at 5 × 10⁵ cells/well and cultured to confluence. A uniform scratch was created using a 200 µL pipette tip, followed by PBS washing to remove debris. Cells were maintained in serum-free medium to minimize proliferation effects. Wound closure was monitored at 0 and 24 h using phase-contrast microscopy, and migration distances were quantified using ImageJ by measuring the remaining wound area relative to the initial scratch. Quantitative PCR RNA extraction was carried out with the RNAsimple total RNA purification kit (TIANGEN, DP419) following the protocol provided by the manufacturer [13]. For cDNA synthesis, 2 µg of extracted RNA was converted using random hexamers (Takara, Japan). Quantitative PCR analysis was conducted to measure transcript abundance, employing methodology established in prior studies. Gene expression quantification was performed using the ΔΔCt approach, with β-Actin serving as the reference gene for normalization. The primers used were as follows: TAGLN, forward, GTCTGGCTGAAGAATGGCGTGATT, reverse, CCACCTGCTCCATCTGCTTGAAG. β-Actin, Forward, CACTCTTCCAGCCTTCCTTC, Reverse, GTACAGGTCTTTGCGGATGT. Western blot Cells were washed with ice-cold PBS and harvested into centrifuge tubes using cell scrapers. After lysis with RIPA buffer, the samples were centrifuged at 10,000 rpm for 10 min. Protein concentrations in the supernatants were quantified using the BCA assay. Protein samples were mixed with 5× loading buffer and denatured. The proteins were then separated by SDS-PAGE and transferred onto PVDF membranes. Following blocking with 5% BSA for 1 h, the membranes were sequentially incubated with primary antibodies (overnight at 4 °C) and secondary antibodies (1 h at room temperature), and finally visualized using ECL chemiluminescence. Antibodies specific for TAGLN (Cat No.10493-1-AP, 1:1000), E-cadherin (Cat No. 20874-1-AP, 1:1000), Vimentin (Cat No. 10366-1-AP, 1:1000), β-actin (Cat No. 66009-1-Ig, 1:2000),anti-Mouse igG (Cat No. SA00001-1,1:5000) and anti-Rabbit igG (Cat No. SA00001-2, 1:5000) were all obtained from Proteintech (Wuhan, China). Images of protein bands were obtained using the Tanon Gel Imaging System (Tanon 4600), with at least three independent experiments conducted for each treatment condition for statistical analysis. The analysis module of the imaging system was used to calculate the grayscale value of each electrophoresis lane. The relative protein expression level refers to the grayscale value of the target protein divided by that of the housekeeping protein (Actin). Transcriptome sequencing and differential gene enrichment analysis Total RNA was extracted from HCT116-Vec and HCT116-TAGLN cells, followed by Illumina sequencing performed by Gene Denovo Biotechnology (Guangzhou). Enrichment analysis of the sequencing results was conducted using the Gene Denovo online platform. Preparation of polyacrylamide gel substrates The preparation of polyacrylamide (PA) gels substrates was described previously [14]. In brief, glass coverslips were functionalized with 3-aminopropyltriethoxysilane (APES) and crosslinked with glutaraldehyde. A mixture of acrylamide and bis-acrylamide at desired concentrations was degassed, followed by the addition of TEMED and ammonium persulfate to initiate polymerization. The solution was pipetted onto chloro-silanated glass slides and immediately covered with the APES-treated coverslips. After complete polymerization, the gel-coverslip constructs were washed with distilled water to remove any unreacted monomers. The gel surface was then activated with sulfo-SANPAH under UV light (365 nm) to enable covalent conjugation of 0.1 mg/ml collagen I. Following collagen coating, the gels were equilibrated in PBS for 24 h to ensure full swelling and stability, sterilized under UV light, and then used for cell seeding. Immunofluorescence Cells were fixed with 4% paraformaldehyde at room temperature for 30 min, followed by permeabilization and blocking with 0.1% Triton X-100 and 5% BSA for 20 min at room temperature. All samples were then incubated with primary antibodies overnight at 4 °C, and subsequently with Alexa Fluor-conjugated secondary antibodies for 1 h at room temperature. Nuclei were counterstained with 4’,6-diamidino-2-phenylindole (DAPI). Immunofluorescence images were acquired using an Olympus DP73 fluorescence microscope (Tokyo, Japan) with a 20× objective. Immunohistochemistry The CRC tissue microarray HcolA060CD01 was obtained from Shanghai Outdo Biotech Company. Following standard procedures, tissue sections were deparaffinized with xylene and rehydrated through an ethanol gradient. Antigen retrieval was performed using citrate buffer (pH 6.0) with heat treatment. Immunohistochemical staining was carried out using a ready-to-use kit (Maixin Biotechnology, Fuzhou, China) according to the manufacturer’s protocol [15]. After DAB development, sections were dehydrated through an ethanol series, cleared in xylene, and mounted with neutral balsam. The staining score was calculated based on both the intensity of staining and the proportion of positive cells. The staining intensity was graded as follows: 0 (negative), 1 (weak), 2 (moderate), and 3 (strong). The proportion of positive cells was scored as: 0 (< 5%), 1 (5%−25%), 2 (26%−50%), 3 (51%−75%), and 4 (> 75%). The final IHC score was obtained by multiplying the intensity by the proportion. Statistical analysis All experiments were independently repeated at least three times. Data are presented as the mean ± standard deviation (SD). Statistical analyses were performed using GraphPad Prism software (version 7.0). For comparisons between two groups, unpaired two-tailed Student’s t-tests were used. The F-test was applied to confirm the homogeneity of variances between the groups compared. If the variances were significantly different, Welch’s correction was applied instead. P < 0.05 was considered statistically significant. Results Identification of risk signature in the TCGA cohort Our objective was to identify risk genes associated with poor prognosis in CRC patients. Using data from the TCGA-COAD dataset, we first performed univariate Cox regression analysis (P < 0.01) to preliminarily screen genes significantly correlated with overall survival (OS). As shown in Fig. 1a-c, subsequently, LASSO regression was applied to further narrow down the candidate genes, ultimately retaining four risk genes (including TAGLN, MMP1, etc.). Although the predictive performance of the risk model was limited (Fig. 1d) (C-index = 0.566, 95% CI: 0.505–0.626), multivariate Cox analysis revealed that TAGLN was the only gene in the model exhibiting a risk-increasing trend (HR = 1.13, 95% CI: 0.999–1.28, P = 0.051), with its high expression showing a correlation with shorter patient survival (Fig. 1e-f). Therefore, we will further investigate the expression patterns and biological functions of TAGLN in CRC. Fig. 1Identification of risk signature in the TCGA cohort. (a) Coefficients of the LASSO analysis. (b) Cross-validation for tuning the parameter selection in the LASSO regression. (c) The genes involved in the signature and their coefficients. (d) nomogram to facilitate the prognosis prediction. (e) Forest plots of the Univariate Cox regression analyses in TCGA COAD cohorts. (f) Forest plots of the Multivariate Cox regression analyses in TCGA COAD cohorts Identification of risk signature in the TCGA cohort. (a) Coefficients of the LASSO analysis. (b) Cross-validation for tuning the parameter selection in the LASSO regression. (c) The genes involved in the signature and their coefficients. (d) nomogram to facilitate the prognosis prediction. (e) Forest plots of the Univariate Cox regression analyses in TCGA COAD cohorts. (f) Forest plots of the Multivariate Cox regression analyses in TCGA COAD cohorts Association between TAGLN and clinical characteristics of CRC We next analyzed the association between TAGLN expression and clinical characteristics in the TCGA-COAD cohort. We statistically analyzed the relationship between TAGLN expression and overall survival in a patient cohort with survival information (n = 270). High TAGLN expression was associated with significantly shorter overall survival and disease-free survival (Fig. 2a-b). In the total cohort (n = 438), TAGLN expression increased with higher tumor grade (Fig. 2c), though no statistically significant difference was observed across different T stages (Fig. 2d). Notably, TAGLN levels were significantly elevated in both N1 and N2 stage tumors (Fig. 2e). Importantly, TAGLN exhibited metastasis-related features, with markedly increased expression in tumors with distant metastasis (M1) or lymph node invasion (Fig. 2f-g). To evaluate the predictive value of TAGLN for patient survival, we performed receiver operating characteristic (ROC) curve analysis (Fig. 2h-i). While TAGLN showed limited predictive accuracy for 1-year and 5-year survival, it demonstrated strong performance in predicting 10-year survival outcomes (AUC = 0.842). Fig. 2Association between TAGLN and clinical characteristics of CRC. (a) Effect of TAGLN expression on overall survival of colorectal cancer patients. (b) Effect of TAGLN expression on disease free survival of colorectal cancer patients. (c) Expression of TAGLN in tumors of different stages. (d) Expression of TAGLN in tumors of different T stages. (e) Expression of TAGLN in tumors of different N stages. (f) Expression of TAGLN in tumors of different M stages. (g) Expression of TAGLN in tumors with lymphatic invasion or not. (h) Time-dependent ROC curves of TAGLN expression for 1-, 5-, and 10-year OS prediction. (i) Time-dependent ROC curves of TAGLN expression for 1-, 5-, and 10-year DFS prediction Association between TAGLN and clinical characteristics of CRC. (a) Effect of TAGLN expression on overall survival of colorectal cancer patients. (b) Effect of TAGLN expression on disease free survival of colorectal cancer patients. (c) Expression of TAGLN in tumors of different stages. (d) Expression of TAGLN in tumors of different T stages. (e) Expression of TAGLN in tumors of different N stages. (f) Expression of TAGLN in tumors of different M stages. (g) Expression of TAGLN in tumors with lymphatic invasion or not. (h) Time-dependent ROC curves of TAGLN expression for 1-, 5-, and 10-year OS prediction. (i) Time-dependent ROC curves of TAGLN expression for 1-, 5-, and 10-year DFS prediction Construction of HCT116 cell line stably overexpressing TAGLN We evaluated TAGLN expression in several common colorectal cancer cell lines. TAGLN was highly expressed in more aggressive cell lines, such as SW620 and LoVo, but showed lower expression in the weakly invasive HCT116 cells (Fig. 3a). To investigate the functional role of TAGLN, we generated stable TAGLN-overexpressing HCT116 cell lines via lentiviral transduction, with successful transfection confirmed by PCR and Western blot (Fig. 3b-d). Interestingly, control cells infected with empty lentivirus retained their typical cobblestone-like epithelial morphology (Fig. 3e), whereas TAGLN-overexpressing cells adopted a more spindle-shaped appearance—a potential indicator of EMT. Therefore, we examined the effect of TAGLN overexpression on the EMT markers E-cadherin (E-cad) and vimentin. As predicted, TAGLN caused a downregulation of E-cad and an increase in Vimentin, indicating a tendency towards EMT(Fig. 3f). Next, we used SW620 cells with high expression of TAGLN for gene knockdown. Similarly, TAGLN knockdown resulted in an increase in E-cadherin and a decrease in Vimentin (Fig. 3g). These findings reinforce the association between TAGLN and EMT in CRC. Fig. 3Construction of HCT116 cell line stably overexpressing TAGLN. (a) The expression of TAGLN in common colorectal cancer cell lines. (b) Transfection effect of lentivirus carrying GFP. (c) Identification of TAGLN mRNA levels in stable transfected cell lines. (d) Identification of TAGLN protein levels in stable transfected cell lines. (e) Morphological changes of HCT116 cells after overexpression of TAGLN. (f) The protein expression of EMT markers in TAGLN overexpression cell line. (g) The protein expression of EMT markers in TAGLN knock down SW620 cell line. Scale = 100 μm, * p<0.05 Construction of HCT116 cell line stably overexpressing TAGLN. (a) The expression of TAGLN in common colorectal cancer cell lines. (b) Transfection effect of lentivirus carrying GFP. (c) Identification of TAGLN mRNA levels in stable transfected cell lines. (d) Identification of TAGLN protein levels in stable transfected cell lines. (e) Morphological changes of HCT116 cells after overexpression of TAGLN. (f) The protein expression of EMT markers in TAGLN overexpression cell line. (g) The protein expression of EMT markers in TAGLN knock down SW620 cell line. Scale = 100 μm, * p<0.05 The downstream signal network regulated by TAGLN To investigate the functional network of TAGLN, we performed transcriptome sequencing on TAGLN-overexpressing HCT116 cells and control vector-transfected cells. As shown in Fig. 4a, TAGLN overexpression resulted in 212 differentially expressed genes (DEGs). GO enrichment analysis revealed that these DEGs were significantly associated with ECM organization and cell adhesion pathways (Fig. 4b). GSEA further demonstrated that TAGLN upregulates multiple ECM-related pathways (Fig. 4c). Notably, TAGLN also activated a colorectal cancer-associated gene set in GSEA (Fig. 4d), consistent with its role as a risk factor in CRC. Importantly, TAGLN overexpression promoted pathways related to epithelial cell migration (Fig. 4e-f). To validate these findings functionally, we conducted a wound-healing assay to assess the impact of TAGLN on HCT116 cell migration. As shown in Fig. 4g, TAGLN significantly enhanced the migratory capacity of HCT116 cells, further supporting its association with CRC metastasis. These findings compel us to consider the mechanistic links connecting TAGLN with ECM remodeling and colorectal cancer progression. Fig. 4The downstream signal network regulated by TAGLN. (a) Volcano plot reflected DEGs between HCT116-Vec and HCT116-TAGLN groups. (b) GO enrichment analysis of DEGs. (c) Representative GSEA analysis of DEGs. (d) GSEA analysis for Colorectal Cancer pathway. (e) GSEA analysis for Positive regulation of endothelial cell migration pathway. (f) GSEA analysis for Positive regulation of blood vessel endothelial cell migration pathway. (g) Wound-healing assay to assess the impact of TAGLN on HCT116 cell migration The downstream signal network regulated by TAGLN. (a) Volcano plot reflected DEGs between HCT116-Vec and HCT116-TAGLN groups. (b) GO enrichment analysis of DEGs. (c) Representative GSEA analysis of DEGs. (d) GSEA analysis for Colorectal Cancer pathway. (e) GSEA analysis for Positive regulation of endothelial cell migration pathway. (f) GSEA analysis for Positive regulation of blood vessel endothelial cell migration pathway. (g) Wound-healing assay to assess the impact of TAGLN on HCT116 cell migration TAGLN expression is regulated by ECM stiffness TAGLN has been implicated in ECM remodeling and cell migration, leading us to naturally consider its potential role in tumor progression, where the ECM often enhances migration and invasion by inducing EMT. To explore this, as shown in Fig. 5, we further examined the correlation between TAGLN and key factors involved in collagen crosslinking (a critical contributor to ECM remodeling and stiffening), ECM degradation (hydrolytic enzyme-related genes), and EMT drivers in the TCGA-COAD cohort (n = 438). Our analysis revealed a strong positive correlation between TAGLN and LOX, LOXL2, and PLOD2, suggesting an association between TAGLN and collagen stiffening. Additionally, TAGLN showed significant positive correlations with matrix metalloproteinases (MMPs), including MMP2, MMP9, and MMP14. Notably, key EMT-inducing transcription factors such as TGFB1, SNAI1, and TWIST1 also exhibited strong correlations with TAGLN. These findings support the hypothesis that TAGLN may serve as a molecular bridge linking ECM remodeling and stiffening to the activation of EMT in tumor cells. Fig. 5Correlation analysis of TAGLN with genes involving in collagen cross-linking, matrix remodeling, and EMT driving. (a) Correlation analysis of TAGLN with collagen cross-linking genes. (b) Correlation analysis of TAGLN with matrix remodeling genes. (c) Correlation analysis of TAGLN with EMT driver genes Correlation analysis of TAGLN with genes involving in collagen cross-linking, matrix remodeling, and EMT driving. (a) Correlation analysis of TAGLN with collagen cross-linking genes. (b) Correlation analysis of TAGLN with matrix remodeling genes. (c) Correlation analysis of TAGLN with EMT driver genes Increased stiffness is a hallmark of tumor ECM remodeling and a well-established driver of EMT in cancer cells. To investigate whether TAGLN expression is influenced by substrate stiffness, we cultured HCT116 cells on fibronectin-coated PA hydrogels with varying stiffness levels (Fig. 6a). Gels mimicking the stiffness of normal colon tissue (~ 500 Pa) were defined as soft substrates, while those resembling the rigidity of primary colorectal tumors (~ 1500 Pa) were classified as medium substrates. A gel stiffness of 4000 Pa, used to mimic advanced tumors, was defined as the stiff substrate. As shown in Fig. 6b, c, both the mRNA and protein levels of TAGLN increased significantly with increasing substrate stiffness. Immunofluorescence analysis revealed that the stiff substrate significantly enhanced the spreading area of HCT116 cells and upregulated TAGLN expression (Fig. 6d). Quantification of Paxillin, a key focal adhesion protein, revealed a significant increase in the number of paxillin-positive foci per cell on stiffer substrates (Fig. 6e), consistent with enhanced cellular spreading and adhesion that facilitates migration. Fig. 6TAGLN expression is regulated by ECM stiffness. (a) Schematic diagram of construction process of PA gel substrates with different stiffness. (b) Identification of TAGLN mRNA levels. (c) Identification of TAGLN protein levels. (d) Immunofluorescence detection of TAGLN and Paxillin expression. (e) Quantification of paxillin-positive foci per nucleus. Scale = 50 μm. * p<0.05 TAGLN expression is regulated by ECM stiffness. (a) Schematic diagram of construction process of PA gel substrates with different stiffness. (b) Identification of TAGLN mRNA levels. (c) Identification of TAGLN protein levels. (d) Immunofluorescence detection of TAGLN and Paxillin expression. (e) Quantification of paxillin-positive foci per nucleus. Scale = 50 μm. * p<0.05 TAGLN expression is associated with CRC metastasis Finally, as shown in Fig. 7, we examined TAGLN expression in clinical CRC specimens. Using a tissue microarray containing normal colon tissue, primary colorectal cancer, and metastatic lesions, we performed immunohistochemical staining and observed a distinct expression pattern: TAGLN was nearly undetectable in normal colon epithelium, weakly expressed in primary tumors, and strongly upregulated in metastatic foci. These clinical findings further support our hypothesis that TAGLN promotes CRC metastasis and is associated with poor prognosis. Fig. 7TAGLN expression in clinical colorectal cancer specimens. (a) TAGLN staining in normal colon tissue. (b) TAGLN staining in colorectal cancer tissue. (c) TAGLN staining in metastatic colorectal cancer tissue. (d) Statistical analysis of staining scores for TAGLN in different stages of colorectal cancer progression. Scale=200um. * p<0.05. TAGLN expression in clinical colorectal cancer specimens. (a) TAGLN staining in normal colon tissue. (b) TAGLN staining in colorectal cancer tissue. (c) TAGLN staining in metastatic colorectal cancer tissue. (d) Statistical analysis of staining scores for TAGLN in different stages of colorectal cancer progression. Scale=200um. * p<0.05. Discussion The present study establishes TAGLN as a mechanosensitive regulator of CRC progression that bridges ECM remodeling with EMT activation. Our clinical data demonstrate that elevated TAGLN expression correlates with advanced tumor stage, metastatic potential, and poor patient survival, positioning it as a promising prognostic biomarker. Notably, the stepwise increase in TAGLN expression from normal epithelium to primary tumors and metastatic lesions suggests its involvement in CRC dissemination. These findings align with emerging evidence implicating actin-binding proteins in cancer progression, but extend current knowledge by revealing TAGLN’s specific association with lymph node and distant metastasis. While previous studies reported conflicting roles for TAGLN in different cancer types [16], our integrated approach provides compelling evidence for its pro-metastatic function in CRC. The particularly strong association with 10-year survival (AUC = 0.842) suggests TAGLN may serve as a long-term prognostic indicator, potentially identifying patients who would benefit from more aggressive surveillance or therapy. TAGLN is rarely expressed in epithelial cells but is highly expressed in fibroblasts [17]. As such, TAGLN has been widely recognized as a marker for fibroblasts, particularly cancer-associated fibroblasts (CAFs) [18]. Importantly, emerging evidence underscores its functional significance in the tumor microenvironment. TAGLN-positive fibroblasts have been identified as a specific CAF subtype that promotes metastasis, drives EMT, and contributes to an immunosuppressive microenvironment, correlating with poor prognosis and immunotherapy resistance in CRC [19]. Furthermore, TAGLN is validated as a CAF-specific marker upregulated in CRC tissues, highlighting its potential as a candidate for stromal-targeting therapies [17]. Early studies on TAGLN suggested a potential tumor-suppressive role in colon cancer [20, 21], as its expression was downregulated in colon cancer tissues compared to normal colon tissue. However, this observation may simply reflect the initial expansion of cancer cells and the displacement of stromal cells in the ECM [22]. In advanced tumors, TAGLN expression is significantly upregulated and correlates with poorer patient survival, clearly indicating its pro-tumorigenic function. The ECM in progressive tumors undergoes extensive remodeling and becomes considerably stiffer than in early-stage tumors [23, 24]. Although the number of cancer-associated fibroblasts increases, our research reveals that even in epithelial-derived tumor cells, a stiff ECM can induce TAGLN expression, which is associated with the acquisition of a mesenchymal-like phenotype. Therefore, we propose that TAGLN functions as an ECM-responsive gene that contributes to colon cancer metastasis. While some studies have employed stiffer environments to investigate the effects of mechanical force on tumor behavior, our study remained within a stiffness range that does not reach the physiological extreme. Given that advanced tumors or tumor cores can develop severe fibrosis with stiffness potentially up to several tens of kilopascals, we recommend that future work incorporates a broader spectrum of substrate stiffness. Our mechanistic investigations reveal that TAGLN orchestrates a pro-metastatic program through dual regulation of ECM remodeling and EMT. The observed correlations between TAGLN and collagen crosslinking enzymes (LOX [25]/LOXL2 [26, 27]/PLOD2 [28]), matrix metalloproteinases (MMPs) [29], and EMT transcription factors [30–32] suggest it functions as a central node connecting matrix modification to cellular plasticity. This is further supported by transcriptomic analyses showing TAGLN overexpression enriches ECM organization and cell adhesion pathways. The stiffness-dependent upregulation of TAGLN provides critical insight into how mechanical cues from the tumor microenvironment are transduced into biochemical signals that drive EMT. Importantly, our demonstration that stiff substrates simultaneously induce TAGLN and vimentin expression offers a plausible mechanism for how progressive ECM stiffening during tumor evolution could promote metastatic dissemination. These findings complement recent work showing mechanical forces can directly activate EMT programs, but identify TAGLN as a specific mechano-transducer in this process. The spindle-like morphology of TAGLN-overexpressing cells and their enhanced migratory capacity further support its role in facilitating metastasis, consistent with reports of TAGLN promoting invasion in other cancers. Beyond its role as a mechanosensitive effector [33], emerging evidence suggests that TAGLN expression is regulated by key oncogenic signaling pathways, particularly the TGF-β/Slug axis. In bladder cancer, a progression-dependent correlation exists between TGF-β and TAGLN, where TGF-β‑mediated cell migration is critically dependent on TAGLN expression [34]. Furthermore, TAGLN promotes EMT and metastasis by significantly upregulating the EMT‑transcription factor Slug, indicating its position downstream of TGF-β and upstream of key EMT drivers. Concurrently, study [35] identified that TAGLN forms a positive feedback loop with the mechanosensitive kinase Src. Src activation upregulates TAGLN expression, while TAGLN‑mediated cytoskeletal remodeling in turn sustains Src activation, thereby amplifying the mechanotransduction signal. This regulatory network is consistent with findings in fibrotic models, where stiff substrates directly upregulate smooth muscle genes including TAGLN and ACTA2 [36]. Collectively, these insights position TAGLN at the nexus of biochemical and biomechanical signaling pathways, converging to drive cytoskeletal remodeling, EMT, and cancer progression. The clinical implications of our study are twofold. First, TAGLN’s association with metastatic features suggests its potential utility as a diagnostic marker for identifying high-risk patients who may benefit from closer monitoring or adjuvant therapy. Second, the mechano-regulation of TAGLN expression points to novel therapeutic opportunities targeting the TAGLN-ECM-EMT axis. Pharmacological inhibition of LOX/LOXL2 [37, 38], which showed strong correlation with TAGLN, could potentially disrupt this pro-metastatic cascade. Our findings also raise intriguing questions about whether TAGLN expression could serve as a readout for treatment response to ECM-modifying therapies. Future studies should explore whether TAGLN knockdown can reverse stiffness-induced EMT and whether its inhibition can prevent metastasis in vivo. While our work focused on CRC, the conserved role of ECM remodeling in solid tumors suggests TAGLN may have broader relevance across cancer types. A key remaining challenge is understanding how TAGLN’s known actin-binding function integrates with its newly identified role in mechano-transduction. Addressing this could reveal fundamental principles about how mechanical signals are converted into transcriptional programs that drive metastasis.