Impact of Extracellular Matrix-Related Genes on the Tumor Microenvironment and Prognostic Indicators in Esophageal Cancer: A Comprehensive Analytical Study

Esophageal cancer is a major global health challenge with a poor prognosis. Recent studies underscore the extracellular matrix (ECM) role in cancer progression, but the full impact of ECM-related genes on patient outcomes remains unclear. Our study utilized next-generation sequencing and clinical data from esophageal cancer patients provided by The Cancer Genome Atlas, employing the R package in RStudio for computational analysis. This analysis identified significant associations between patient survival and various ECM-related genes, including IBSP, LINGO4, COL26A1, MMP12, KLK4, RTBDN, TENM1, GDF15, and RUNX1. Consequently, we developed a prognostic model to predict patient outcomes, which demonstrated clear survival differences between high-risk and low-risk patient groups. Our comprehensive review encompassed clinical correlations, biological pathways, and variations in immune response among these risk categories. We also constructed a nomogram integrating clinical information with risk assessment. Focusing on the TENM1 gene, we found it significantly impacts immune response, showing a positive correlation with T helper cells, NK cells, and CD8+ T cells, but a negative correlation with neutrophils and Th17 cells. Gene Set Enrichment Analysis revealed enhanced pathways related to pancreatic beta cells, spermatogenesis, apical junctions, and muscle formation in patients with high TENM1 expression. This research provides new insights into the role of ECM genes in esophageal cancer and informs future research directions.


Introduction
Esophageal cancer (EC) is the sixth most common malignancy globally, characterized by a poor prognosis and high invasiveness [1].Over 95% of EC cases are either squamous cell carcinomas or adenocarcinomas.Squamous cell carcinoma is more prevalent in developing countries, whereas adenocarcinoma is more common in developed countries [2].Te early symptoms of EC are often not apparent, leading to signifcant treatment delays [3].A routine pathological biopsy performed under an endoscope is the most common diagnostic method.Although some patients may beneft from early surgery, recurrences and distant metastases can occur during subsequent adjuvant therapy [4].In-depth studies of the tumor microenvironment can enhance the understanding of tumor genesis and progression, facilitating the discovery of therapeutic targets [5].
Tumor cells reside in a complex microenvironment known as the tumor microenvironment (TME).Te extracellular matrix (ECM), a fundamental component of the TME, consists of various proteins secreted by cells, providing structural support and mediating cell interactions [6].Te abnormal ECM in the TME can afect the biological behaviors of cancer cells in multiple ways.According to a comprehensive review by Gilkes et al., changes in ECM content can directly infuence its biological properties, contributing to cancer metastasis by afecting tumor cell heterogeneity [7].Chaki et al. found that the interaction of Nck adapter proteins with downstream kinase 1 facilitates ECM degradation and cancer progression [8].DiGiacomo et al. used a fbroblast-derived ECM scafold for cell culture and discovered that the ECM scafold signifcantly decreases the sensitivity of ER + breast cancer cells to ER-targeted therapy, a condition that can be reversed by the binding of FGF2 to FGFR1 [9].Te ECM is also regulated by immune cells.Haj-Shomaly et al. revealed that CD8+ Tcells can induce ECM remodeling and cancer metastasis in paclitaxeltreated mice [10].Tian et al. demonstrated that the microsome proteins derived from cancer cells AGRN, SER-PINB5, and CSTB can promote pancreatic cancer metastasis and are associated with poor prognosis [11].Additionally, the ECM and other cells in the TME can create a robust barrier around cancer cells in solid tumors, reducing the efectiveness of immunotherapy [12].Drugs targeting the ECM can disrupt collagen fber arrangement, enhancing immune cell infltration and the efcacy of therapeutic drugs.[13].
A wealth of publicly available high-throughput datasets facilitates secondary data analyses and research.In this study, we investigated the role of ECM-related genes in EC.Using various algorithms and analyses, we identifed several ECM-related genes-IBSP, LINGO4, COL26A1, MMP12, KLK4, RTBDN, TENM1, GDF15, and RUNX1-as significantly associated with patient survival.We developed a prognostic signature that efectively diferentiates between high-and low-risk patient groups in terms of survival outcomes.Detailed analyses, including clinical correlation, biological enrichment, and immune infltration, were performed to delineate the distinctions between these groups.Additionally, we combined clinical data and risk scores to construct a nomogram that exhibited superior predictive performance.Notably, TENM1 was selected for further investigation.Immunohistochemistry results revealed that TENM1 protein levels were reduced in EC tumor tissues.Moreover, immune infltration analysis demonstrated a positive correlation of TENM1 with T helper cells, NK cells, and CD8+ T cells and a negative correlation with neutrophils and T17 cells.Gene Set Enrichment Analysis (GSEA) showed that pathways related to pancreas beta cells, spermatogenesis, apical surface, and myogenesis were upregulated in patients with high levels of TENM1.

Open-Accessed Data
Collection.Genomic and clinical data for EC patients were sourced from Te Cancer Genome Atlas (TCGA) program, specifcally the TCGA-ESCA project.Individual expression profles (STAR-COUNTS) and clinical data were accessed via the TCGA-GDC program.For accurate probe annotation, the latest human genomic annotation fle (GRCh38.p13) was downloaded from the ENSEMBL database.We excluded genes with a median expression value below 0.1 to ensure robust data quality.To address the skewness in gene expression data, we transformed the expression matrix using the log2 scale after adding a pseudo-count of 1. Mutation data for the genome were also retrieved from the TCGA database.Data preprocessing and analysis of diferentially expressed genes were conducted using the Limma package, following specifed thresholds.Te tumor stemness index, mRNAsi, was obtained from the supplementary information of a prior study [14].Due to the limited availability of normal tissue samples in the TCGA database, additional normal tissue data from the GTEx database were included in the analysis.Te baseline characteristics of the enrolled patients are presented in Table 1.

Protein Interaction
Network.Potential interactions among coding proteins were explored using the STRING database, with search parameters specifcally set for "Homo sapiens" to ensure species-specifc relevance [15].

Biological Enrichment Analysis.
Biological enrichment analysis was performed using the ClueGO plugin within the Cytoscape software, focusing on signifcantly enriched terms (P < 0.01) to facilitate efective visualization and interpretation [16].Gene ontology (GO) analysis was conducted using the R package clusterProfler in the RStudio environment, applying flter criteria of "P value <0.05" and "q value <0.05" [17].Furthermore, GSEA was employed to delineate biological diferences between two groups, referencing specifc pathway sets [18].

Clinical and Prognosis
Analysis.Prognostic factors were initially identifed using univariate Cox regression analysis with a signifcance level set at P < 0.05.Te identifed genes were further refned using LASSO regression analysis to optimize the selection of prognostic variables [19].Subsequently, a multivariate Cox regression model was developed to ascertain independent prognostic factors.We constructed a predictive model using the formula: "Risk score � (Coef_i × Expression_i) for each signifcant gene i," thereby enabling precise risk stratifcation.We also examined clinical correlations with patient characteristics such as gender, clinical stage, and TNM classifcation.Kaplan-Meier survival curves and receiver operating characteristic (ROC) curves were utilized to evaluate the prognostic accuracy of the model.

Establishment of Nomogram.
A nomogram was developed to quantitatively predict patient survival using the rms package in RStudio.Te predictive performance of the nomogram was evaluated through calibration curves, comparing predicted survival probabilities with observed outcomes.

Exploration of TME.
Te relative abundance of immune and stromal cells in the TME of EC patients was analyzed using the estimate package in R. Additionally, various algorithms including CIBERSORT, XCELL, EPIC, MCPCOUNTER, QUANTISEQ, TIMER, and ssGSEA algorithm were employed to evaluate immune cell infltration levels in the EC TME [20].
2.7.Specifc Drug Sensitivity.Sensitivity to immunotherapy in EC patients was determined using the Tumor Immune Dysfunction and Exclusion (TIDE) algorithm [21].

Genetics Research
Sensitivity to targeted drugs was assessed using data from the Genomics of Drug Sensitivity in Cancer database.[22].
2.8.Single-Cell Analysis.Single-cell RNA sequencing data analysis was performed to explore the cellular heterogeneity and the specifc expression patterns of TENM1 in EC.We utilized the TISCH database, a comprehensive resource for tumor-infltrating single-cell transcriptomics.Tis platform allowed us to conduct an online analysis to identify the specifc cell types expressing TENM1 within the tumor microenvironment of EC patients.

Te Expression Pattern of ECM-Related Genes in EC and
Teir Biological Role.Te overall workfow of this study is presented in Figure S1.To account for the diferences between cancerous and normal tissues, we initially investigated the expression patterns of ECM-related genes in EC.Our fndings revealed that 91 ECM genes were downregulated, while 109 genes were upregulated in EC tumor tissue.GO analysis indicated that these ECM-related genes are involved in processes such as glycosaminoglycan binding (GO: 0005539), endopeptidase activity (GO: 0004175), extracellular structure organization (GO: 0043062), collagen catabolic process (GO: 0030574), cell-substrate adhesion (GO: 0031589), collagen metabolic process (GO: 0032963), basement membrane (GO: 0005604), collagen trimer (GO: 0005581), Golgi lumen (GO: 0005796), extracellular matrix organization (GO: 0030198), endoplasmic reticulum lumen (GO: 0005788), laminin complex (GO: 0043256), and extracellular matrix disassembly (GO: 0022617) (Figure S2A).ClueGO analysis further demonstrated that these ECMrelated genes were predominantly enriched in organ growth, chondrocyte diferentiation, glycosaminoglycan catabolic process, skeletal system development, regulation of cell adhesion, cell-substrate adhesion, and extracellular matrix organization (Figure S2B).

Te Genomic Diference in Diferent EC Patients.
Genomic diferences can lead to varied cell behaviors.Consequently, we aimed to elucidate the prognostic variations from a genomic standpoint.A positive correlation between the risk score and tumor mutational burden (TMB) was observed, suggesting that patients with high-risk scores might exhibit progressive genomic mutations (Figure 3(a)).However, no signifcant correlations were found between microsatellite instability (MSI) and mRNAsi (Figures 3(b) and 3(c)).Additionally, while a negative correlation was evident between the risk score and immune score, such correlations were absent between the stromal score and ESTIMATE score (Figures 3(d), 3(e) and 3(f )).

Immunotherapy and Drug Sensitivity.
We then sought to explore the diferences in immunotherapy response and drug sensitivity among patient groups.However, the expression of key immune checkpoints showed no signifcant diferences between high-and low-risk patients (Figure S3).Additionally, there was no statistically signifcant correlation between the risk score and the TIDE score, suggesting that the risk score does not signifcantly infuence EC immunotherapy outcomes (Figure 3(g)).Interestingly, a slight correlation was observed between immune dysfunction and the risk score (Figure 3(h)).Drug sensitivity analysis revealed that patients in the low-risk group may be more responsive to AKT inhibitors and erlotinib (Figures 3(i), 3(j), 3(k), 3(l), 3(m), 3(n), 3(o), and 3(p)).

Biological Enrichment and Immune Microenvironment
Analysis.Te progression and malignant behavior of cancer are infuenced by various pathways and cascade reactions.
Biological enrichment studies showed that pathways related to pancreas beta cells, coagulation, peroxisomes, IL6/JAK/ STAT3 signaling, and oxidative phosphorylation were activated in high-risk patients (Figure 4(a), Hallmark).GSEA based on the Kyoto Encyclopedia of Genes and Genomes (KEGG) revealed that pathways associated with maturityonset diabetes of the young, DNA replication, the citrate (TCA) cycle, base excision repair, and sphingolipid metabolism were enriched in these patients (Figure 4

Further Exploration of TENM1 in EC.
TENM1 has not been previously reported in the literature.Consequently, TENM1 was selected for further analysis in EC.Prognostic analysis revealed that TENM1 had no signifcant impact on overall survival, disease-free survival, or progression-free survival in patients (Figures 6(a), 6(b), 6(c)).However, the number of samples may afect these outcomes; thus, these results should be interpreted with caution.ssGSEA demonstrated a positive correlation between TENM1 and T helper cells, NK cells, and CD8+ T cells, while it showed a negative correlation with neutrophils and T17 cells (Figure 6(d)).GSEA revealed that pathways related to pancreas beta cells, spermatogenesis, apical surface, and myogenesis were upregulated in patients with elevated TENM1 levels (Figure 6(e)).Single-cell analysis showed that TENM1 was mainly expressed in malignant and fbroblasts in EC microenvironment (Figures S4A-S4D).

Expression Level of TENM1 in EC Cells
. Furthermore, we tried to detect the mRNA and protein expression level of TENM1 in EC cells.We found that there was no signifcant diference between EC cells and normal cells (HET-1A vs. EC9706, KYSE150, YES2) (Figures 7(a) and 7(b)).

Discussion
EC remains a signifcant global health threat [23].For earlystage disease, surgical resection is the preferred treatment option.Nonetheless, there is still a high risk of postoperative recurrences and metastases [24].In cases of advanced stages or recurrence, chemotherapy is commonly employed, although its benefts are somewhat limited [24].Additionally, the adverse efects of chemotherapeutic drugs can partly hinder the successful treatment of EC.In the current biological era, advancements have facilitated disease understanding and the identifcation of novel therapeutic targets.Terefore, the identifcation of biomarkers that can guide the diagnosis and treatment of EC is crucial.
In this study, we explored the roles of ECM-related genes in EC.Using a series of algorithms and analyses, we identifed several ECM-related genes-IBSP, LINGO4,    Genetics Research COL26A1, MMP12, KLK4, RTBDN, TENM1, GDF15, and RUNX1-that are signifcantly associated with patient survival.We established a prognostic prediction signature that diferentiates between high-and low-risk groups, refecting varied survival outcomes.To elucidate the differences between these groups, we performed clinical correlation, biological enrichment, and immune infltration analyses.Furthermore, we integrated clinical data with risk scores to develop a nomogram that demonstrates enhanced predictive accuracy.Notably, the gene TENM1 was selected for in-depth analysis.Immunohistochemistry revealed that TENM1 protein levels were downregulated in EC tumor tissues.Immune infltration analysis indicated positive correlations of TENM1 with T helper cells, NK cells, and CD8+ T cells, and negative correlations with neutrophils and T17 cells.GSEA showed that pathways related to pancreas beta cells, spermatogenesis, apical surface, and myogenesis were upregulated in patients with elevated TENM1 levels.
Our study identifed the ECM-related genes IBSP, LINGO4, COL26A1, MMP12, KLK4, RTBDN, TENM1, GDF15, and RUNX1 as signifcantly associated with patient survival.Several of these genes have been implicated in various cancers.For instance, in breast cancer, Wu et    cell death [33].We also observed increased genomic instability in high-risk patients, a well-known cancer hallmark.Tis heightened instability often leads to more aggressive cancer behavior.Correlation analyses showed that risk scores were positively associated with Tregs and resting mast cells.Generally, Tregs contribute to creating an inhibitory immune microenvironment.Wang et al. demonstrated that CCL20, secreted by colon cancer cells, enhances chemotherapy resistance by promoting Treg infltration [34].Similarly, Li et al. discovered that a specifc formula reduces breast cancer metastasis by inhibiting Treg diferentiation and infltration, which is induced by TAM/CXCL1 [35].Drug sensitivity analysis revealed that patients classifed as low-risk exhibited greater sensitivity to AKT inhibitors VIII and erlotinib compared to their high-risk counterparts.Tese low-risk patients likely possess more stable genomic characteristics, which may infuence drug sensitivity, although the underlying mechanisms remain unclear.Prior research suggests that genomic features can impact erlotinib's efcacy; for example, Lu et al. identifed mutations in lung cancer that modulate the drug's response [36].Similarly, Cai et al. reported that genetic alterations in breast cancer could reduce sensitivity to PI3Kα inhibitors [37].Additionally, we observed diferences in specifc pathways between high and low-risk patients, some of which have been previously associated with erlotinib response.For instance, Karaca et al. demonstrated a link between the Wnt/β-catenin signaling pathway and erlotinib's promotive efects in endometrial cancer cells [38].Tese biological variances likely contribute to the heightened sensitivity of low-risk patients to AKT inhibitors VIII and erlotinib.12 Genetics Research Despite the rigorous nature of our analysis, this study has several limitations.First, our research sample predominantly consists of individuals from Western populations, which may introduce racial bias and limit the generalizability of our fndings.Second, the presence of incomplete clinical data may lead to inherent biases, although we anticipate that more comprehensive clinical characteristic data in the future will enhance the reliability of our fndings.Tird, our validation was limited to the protein level of TNEM1 in EC.Future studies should explore additional molecules to broaden our understanding.

Figure 3 :Figure 4 :Figure 5 Figure 5 :Figure 6 :
Figure 3: Immunotherapy and drug sensitivity analysis.Notes.(a-c) Correlation of risk score and TMB, MSI, and mRNAsi; (d-f ) correlation of risk score and immune score, stromal score, and estimate score; (g) correlation between TIDE score and risk score; (h) correlation between risk score and immune dysfunction; (i-p) drug sensitivity analysis between high-and low-risk groups.

Figure 7 :Figure 6 :
Figure 7: Expression level of TENM1 in EC cells.Notes.(a) Te mRNA level of TENM1 in EC and normal cells; (b) the protein level of TENM1 in EC and normal cells.

Table 1 :
Baseline information of enrolled patients.
al. reported that IBSP, secreted from ER + breast cancer cells,