Comprehensive Computational Assessment of SNAI1 and SNAI2 in Gastric Cancer: Linking EMT,Tumor Microenvironment,and Survival Outcomes

Abstract

Background:

Gastric cancer is aggressive with poor prognosis due to high invasion and metastasis rates, a hallmark of cancer. The Snail family (SNAI1 and SNAI2) drives EMT, enabling epithelial cells to gain migratory and invasive traits.

Methods:

We used “limma” package to identify genes with differential expression between high and low levels of SNAI1/SNAI2 in TCGA stomach adenocarcinoma dataset, intersecting these with cancer invasion and metastasis genes obtained from 5 databases. Using Cox regression analysis, we developed a risk score model and created a nomogram incorporating clinical data. The model’s prognostic accuracy was validated with survival and ROC analyses in both TCGA and GEO datasets. Additionally, we performed WGCNA and constructed a ceRNA network to investigate gene interactions, and used CIBERSORT analysis to evaluate immune cell composition in the tumor microenvironment.

Results:

We developed 5 and 9 risk signatures and nomograms incorporating clinical data. Survival analysis showed high-risk patients had worse overall survival than low-risk patients. WGCNA identified a lightyellow module associated with SNAI1 and SNAI2 expressions, emphasizing extracellular matrix organization. CeRNA network analyses found 6 common hub genes linked to SNAI1 and SNAI2. Immune profiling showed that SNAI1 expression was related to 8 types of immune cells, while SNAI2 was connected to 6, indicating their roles in influencing the tumor microenvironment.

Conclusion:

This study highlights the significant prognostic impact of SNAI1 and SNAI2 in stomach adenocarcinoma, linking their high expression to poorer survival and aggressive tumor behavior, while also identifying potential therapeutic targets through comprehensive computational analysis.

Keywords

stomach adenocarcinoma snail family risk score WGCNA CeRNA network CIBERSORT

Graphical Abstract

Introduction

Gastric cancer (GC), primarily adenocarcinoma, accounts for 90% to 95% of all stomach cancers and can be separated into cardia and non-cardia types.^1,2 Ethnicity, food, and illnesses like Helicobacter pylori (H. pylori) and human papillomavirus (HPV) all affect its incidence.^3-5 Gastric cancer was the fourth most common cause of cancer-related deaths worldwide in 2020 and the fifth most common type of cancer overall.^6,7

Although the epithelial-mesenchymal transition (EMT) is essential for tissue repair and embryonic development, it can also accelerate the growth of GC tumors. In this biological process, epithelial cells acquire the migratory and invasive characteristics of mesenchymal cells while losing their polarity and cell-cell adhesion.^8,9 Significant alterations occur in cells during EMT, such as the elevation of mesenchymal markers like vimentin and N-cadherin, which are controlled by transcription factors like SNAIL, TWIST, and ZEB, and the downregulation of epithelial markers, specifically E-cadherin. Increased tumor aggressiveness, metastasis, and resistance to treatment are linked to this process.^10-12 Understanding the role of EMT in GC can provide deeper insights into tumor development and support the development of early diagnosis markers and targeted therapies.¹³

EMT is intricately connected to the tumor immune microenvironment (TME), influencing and responding to changes in immune cell composition. EMT is linked to elevated levels of immunological checkpoints such as PD-L1 and immunosuppressive cells, specifically myeloid-derived suppressor cells. Snail, Zeb1, and Twist1 promote an invasive state in the carcinoma state by stimulating chemokine secretion, which attracts immunosuppressive cells. This creates a feedback loop that accelerates tumor progression and facilitates immune evasion.^14-17 Innate immune cells, such as neutrophils and macrophages, can also promote or inhibit the formation of tumors.^18,19 Designing treatments that increase anti-tumor immunity and enhance treatment results requires an understanding of the connection between EMT and the immunological landscape.²⁰

The Snail family, particularly Snail (SNAI1) and Slug (SNAI2), primarily contribute to mediating EMT. These zinc-finger transcription factors suppress the expression of epithelial markers while increasing mesenchymal characteristics by binding to E-box sites in target gene promoters. For example, in breast, gastric, colon, and hepatocellular carcinomas, aberrant activation of Snail and Slug is linked to enhanced tumor invasiveness and metastasis, as they facilitate the transition from immobile epithelial cells to motile mesenchymal cells.^21-23 SNAI1 is strongly controlled at both transcriptional and post-translational levels by a variety of growth factors and cytokines, promotes resistance to apoptosis and drug resistance, and increases the transcription of certain genes involved in development and cancer.²⁴ Furthermore, SNAI2 regulates cell migration during embryogenesis and tumor cell invasion. Its expression is often linked to poor prognosis in various cancers, as it promotes EMT and enhances stem cell-like properties in tumor cells, contributing to resistance against therapies and increased proliferation in certain cancer types.²⁵

Small, noncoding RNA molecules known as microRNAs (miRNAs) bind to target messenger RNAs (mRNAs) and either break them down or stop them from translating, thereby regulating gene expression. The competing endogenous RNA (ceRNA) hypothesis suggests that RNAs with common miRNA binding sites can regulate each other’s expression by competing for these sites. RNA abundance, location, binding strength, and structure influence ceRNA activity, and disturbances in these factors can disrupt ceRNA networks, potentially leading to diseases, notably cancer.^26-28 Certain long noncoding RNAs (lncRNAs) have the ability to increase the invasiveness of gastric cancer by upregulating the expression of transcription factors linked to EMT. Identifying prognosis-related ceRNA networks may progress diagnostic and therapeutic strategies for gastric cancer.^29-31

The complex interplay between SNAI1 and SNAI2 gene expression patterns remains a compelling area of investigation in stomach adenocarcinoma (STAD) research. Drawing upon multi-faceted computational systems biology, we designed a multilayered analytical framework that integrates several computational methods to illuminate the molecular consequences of SNAI1 and SNAI2 expression changes by stratifying samples into high and low expression cohorts for both SNAI1 and SNAI2. Our methodology weaves together several analytical techniques: differential gene expression (DGE) profiling to identify key transcriptional changes, weighted gene co-expression network construction to unveil coordinated gene behavior patterns, competing endogenous RNA network analysis to map regulatory interactions, and CIBERSORT deconvolution to characterize immune cell type populations.

Methods and Materials

Data Sources and Processing

We obtained clinical data and Level 3 RNA sequencing data from the STAD cohort in The Cancer Genome Atlas (TCGA), which included 412 STAD cases (Table 1, Supplemental Table S1) and 36 normal samples. We normalized the gene expression data using the trimmed mean of M-values (TMM) method in the “edgeR” package, followed by the voom transformation in the “limma” package in R to ensure accurate comparison of expression levels. We sourced genes related to cancer invasion and metastasis from several established databases such as CancerSEA (single-cell functional state atlas),³² CHG (cancer hallmark genes),³³ COSMIC (somatic mutation catalog),³⁴ GeneCards (comprehensive gene information),³⁵ and MSigDB (molecular signatures database).³⁶ In addition, we obtained microarray expression datasets GSE62254 and GSE15459 from the Gene Expression Omnibus (GEO) database, which included 300 and 200 gastric tumor samples, respectively. We calculated the mean expression levels of SNAI1 and SNAI2 from the 36 normal samples, which were used to stratify the 412 STAD samples into high and low expression groups. DGE analysis was subsequently conducted between these groups using the “limma” package, with an adjusted P-value cutoff of .05 to identify differentially expressed genes (DEGs).

Table 1.

Clinicopathological information of 412 STAD cases in TCGA database.

Clinicopathological variables	Number of cases	Percentage (%)
Stage
I	58	14.08
II	122	29.61
III	169	41.02
IV	39	9.47
Grade
G1	12	2.91
G2	148	35.92
G3	243	58.98
Gender
Male	267	64.81
Female	145	35.19

Creation of the Risk Score Model

We used the “Survival” package in R to perform univariate Cox regression analysis on genes that were common between the DEGs and the invasion and metastasis- related gene sets from the 5 databases. Next, we performed multivariate Cox regression analysis to identify genes that form the final risk signature. This risk signature underwent validation through Kaplan–Meier survival analysis, with patient risk scores calculated as the linear combination of gene expression levels (Exp) multiplied by their respective regression coefficients (β), as follows:

R i s k s c o r e = \sum_{i = 1}^{n} E x p i β i

Patients were then divided into high-risk and low-risk groups based on the median risk score.

Risk Signature Assessment and Nomogram Development

Kaplan–Meier survival curves were used to evaluate overall survival disparities between high and low risk groups. To assess the predictive efficacy of the risk signature, we conducted time-dependent receiver operating characteristic (ROC) analyses using the “survival ROC” package in R. These analyses were performed on the TCGA training set and validated on the GEO datasets. Nomograms, integrating 5-gene and 9-gene signatures derived from multivariate Cox analyses, were constructed using the “rms” package. Further, risk scores and other clinical covariates from the TCGA dataset informed the construction of additional nomograms.

Gene Set Enrichment Analysis

We conducted Gene Set Enrichment Analysis (GSEA) to determine the differential enrichment of Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways between high and low risk STAD groups. GSEA software (Java-based) facilitated this analysis, with the high and low risk groups delineated by the median risk scores.

Weighted Gene Co-expression Network Analysis (WGCNA)

WGCNA is a systems biology technique that characterizes gene correlation patterns across samples, identifies clusters of highly correlated genes (modules), and connects these clusters to external traits. It aids in network-based gene screening to discover potential biomarkers or therapeutic targets in diverse biological settings.³⁷ We used the “WGCNA” package in R to construct a scale-free co-expression network, where the network structure emphasizes strong correlations between gene pairs. We computed the variance of each gene in the preprocessed expression matrix, which was generated using the “limma” and “edgeR” packages with TCGA data. We selected genes with variance values above the third quartile threshold for further analysis. In order to incorporate 8635 genes out of 34 539 genes, the expression matrix was adjusted. These genes were then used as input for the WGCNA analysis. The process started with constructing a similarity matrix, which utilizes the absolute values of Pearson’s correlation coefficients between pairs of genes derived from expression data. We converted the similarity matrix into an adjacency matrix (aij) using a soft-thresholding power (β), emphasizing strong gene-gene connections and minimizes weaker correlations. The adjacency matrix was then transformed into a Topological Overlap Matrix (TOM), which measures how strongly genes are associated with one another. We performed hierarchical clustering on the TOM to group similar genes into network modules using the Dynamic TreeCut algorithm. The most representative genes, referred to as module eigengenes (MEs), were calculated as the first principal components, representing the overall gene expression levels within each module. We calculated module membership (MM) by determining the Pearson correlation between the expression profile of a gene and the ME, which represents the core expression pattern of the module. Additionally, gene significance (GS) was assessed to evaluate a gene’s relevance to other biological factors, with higher GS values indicating greater prognostic significance for patients. This systematic approach enables the construction and analysis of gene networks, providing valuable insights into gene interactions and their potential clinical implications. In addition, we performed enrichment analysis on genes in selected modules using the “enrichR” package.

Construction of the ceRNA Network

For the ceRNA network, we incorporated data on lncRNA-miRNA and mRNA-miRNA interactions from RNAInter³⁸ and miRTarBase³⁹ databases, respectively. We filtered these interactions based on DGE analysis results from the SNAI1 and SNAI2 high and low expression groups. We normalized the raw RNA-seq and miRNA-seq data from TCGA-STAD samples using the gdcVoomNormalization function in the GDCRNATools package, which was then used for ceRNA analysis. The gdcCEAnalysis function was used to construct ceRNA networks by integrating multiple analytical methods, including hypergeometric testing for shared miRNAs, Pearson correlation for lncRNA-mRNA associations and partial correlation analysis for network construction. This comprehensive approach allowed us to identify miRNAs correlated with both lncRNAs and mRNAs, pinpoint positively correlated lncRNA-mRNA pairs, and determine miRNAs with similar regulatory effects on lncRNAs and mRNAs. The resulting lncRNA-miRNA-mRNA interactions were visualized as a network using Cytoscape 3.10.1, providing a holistic view of the ceRNA landscape in STAD.

CIBERSORT Analysis

Cell-type Identification by Estimating Relative Subsets of RNA Transcripts (CIBERSORT), is a computational tool that analyzes the immune cell composition within the tumor microenvironment using gene expression data. It employs the LM22 gene signature, consisting 22 specific immune-related genes, to estimate the relative proportions of various immune cell types, including T cells, B cells, and macrophages. By using this technique, scientists can comprehend how these immune cells interact with tumor cells and affect the progression of cancer.^40,41 We utilized CIBERSORT (available at http://cibersortx.stanford.edu), to analyze the immune cell composition of each sample based on TCGA transcripts per million (TPM) data for STAD. The analysis was conducted with 500 permutations. We included only samples with a CIBERSORT output P-value <.05 for further analyses. To evaluate the association between the 22 immune cell types and the SNAI1/SNAI2 high and low expression groups, we employed the Wilcoxon rank-sum test.

Results

Differential Gene Expression Analysis of SNAI1 and SNAI2 in STAD

A schematic diagram (Figure 1) illustrates the methods and findings of our study. The average expression levels of SNAI1 and SNAI2 in 36 normal STAD samples were 2.4 and 3.6, respectively. Among 412 cancer samples, 303 with SNAI1 expression levels exceeding 2.4 were classified as the high expression group, while the remaining 109 with levels below 2.4 were categorized as the low expression group. Similarly, 155 cancer samples with SNAI2 expression levels above 3.6 were categorized as the high expression group, where the 257 samples with lower expression were classified as the low expression group. Utilizing the limma package with an adjusted P-value threshold of <.05, we identified 9587 DEGs between the high and low SNAI1 expression groups and 13 592 DEGs between the high and low SNAI2 expression groups (Figure 2).

Figure 1.

The comprehensive flow chart of the research process.

Figure 2.

Volcano plots of differentially expressed genes of SNAI1 (A) and SNAI2 (B) analyses.

Identification of Prognostic Signature Genes

We identified a total of 2417 genes associated with invasion and metastasis from 5 databases. We discovered 786 genes for SNAI1 and 1124 genes for SNAI2 by intersecting these genes with 9587 SNAI1-related and 13 592 SNAI2-related DEGs. We performed univariate Cox regression analysis on these gene sets using data from 366 STAD patients with complete clinical information. This analysis produced 57 significant genes from the SNAI1 set and 71 significant genes from the SNAI2 set (P < .05; Supplemental Table S2). Subsequently, we conducted multivariate Cox regression analysis on the significant genes identified in the univariate analysis. For the SNAI1 set, 5 genes emerged as significant prognostic markers (P < .05; Table 2). The SNAI2 set also showed 9 genes that were significant prognostic markers (P < .05; Table 3).

Table 2.

Multivariate cox regression analysis of 5 significant genes of SNAI1 analyses.

Ensembl gene ID	Coef	exp.coef.	se.coef.	z	Pr. . .z..	Gene symbol
ENSG00000162512	−0.59	0.55	0.30	−1.99	0.04	SDC3
ENSG00000108821	1.86	6.40	0.72	2.58	0.009	COL1A1
ENSG00000164692	−2.99	0.05	1.21	−2.47	0.01	COL1A2
ENSG00000099953	0.66	1.94	0.22	3.01	0.002	MMP11
ENSG00000188257	0.16	1.79	0.06	2.63	0.008	PLA2G2A

Table 3.

Multivariate cox regression analysis of 9 significant genes of SNAI1 analyses.

Ensembl gene ID	Coef	exp.coef.	se.coef.	z	Pr. . .z..	Gene symbol
ENSG00000005381	−2.52	0.08	0.97	−2.59	0.009	MPO
ENSG00000026508	0.56	1.76	0.19	3.03	0.002	CD44
ENSG00000099953	0.92	2.52	0.28	3.26	0.001	MMP11
ENSG00000108821	1.73	5.63	0.83	2.07	0.03	COL1A1
ENSG00000135318	0.42	1.51	0.17	2.44	0.01	NT5E
ENSG00000147246	0.71	2.03	0.27	2.66	0.007	HTR2C
ENSG00000162512	−0.9	0.4	0.4	−2.25	0.02	SDC3
ENSG00000185634	0.8	2.24	0.38	2.12	0.03	SHC4
ENSG00000197122	−1.02	0.36	0.37	−2.75	0.005	SRC

Development and Evaluation of Prognostic Models

We employed multivariate Cox regression analysis (using the Survival package in R) to construct 2 risk signatures for predicting overall survival (OS) in STAD patients. The first signature consisted of 5 genes, while the second comprised 9 genes. Based on median risk scores, we stratified 366 STAD patients into high- and low-risk groups. The OS differences between the 2 risk groups were then assessed using Kaplan-Meier survival analysis, which revealed a significant difference between the groups. Additionally, ROC analysis was carried out to examine the ability of the risk signatures to predict overall survival in STAD patients. The sensitivity and specificity of the risk signatures in predicting OS were assessed using the area under the curve (AUC) values (Figure 3A and B). Furthermore, within the TCGA dataset, only the Kaplan-Meier survival analysis and AUC at 5 years for SNAI1 expression failed to reach statistical significance. We validated the prognostic value of the risk signatures using independent GEO datasets. Significant OS differences between the high- and low-risk categories in the validation cohorts were verified by repeating the Kaplan-Meier survival analysis. Additionally, ROC analysis of these validation datasets demonstrated high sensitivity and specificity for the risk signatures (Figure 3C and D).

Figure 3.

Evaluation of risk signatures for predicting overall survival in STAD patients. (A) Left, Kaplan–Meier survival analysis of the high- and low-risk groups of SNAI1 analyses from TCGA. Right, AUC values from ROC analysis at 3 and 5 years. (B) Left, Kaplan–Meier survival analysis of the high- and low-risk groups of SNAI2 analyses from TCGA. Right, AUC values from ROC analysis at 3 and 5 years. (C) Similar to 2A from GEO. (D) Similar to 2B from GEO. (A) TCGA-SNAI1. (B) TCGA-SNAI2. (C) GEO-SNAI1. (D) GEO-SNAI2.

Nomogram Construction and Validation

We constructed nomograms based on multivariate Cox proportional hazards regression analyses using the “rms” package in R. The nomograms (Figure 4A and B) incorporated the 5-gene and 9-gene risk signatures to predict 3-year and 5-year survival probabilities in STAD patients. We also constructed nomograms that integrated the 5-gene and 9-gene signatures with clinical variables such as age, gender and cancer stage to predict 3-year and 5-year survival probabilities in STAD patients. The nomograms’ dependability was proved by the calibration curves, which showed a high degree of agreement between the actual and anticipated 3-year and 5-year survival probabilities (Figure 4C and D).

Figure 4.

Nomograms for predicting 3-year and 5-year survival probabilities in STAD patients. (A) Left, nomogram incorporating the 5-gene signature derived from multivariate cox proportional hazards regression of SNAI1 analyses. Right, calibration curves of the nomogram for predicting 3-year or 5-year overall survival (OS). (B) Nomogram incorporating the 9-gene signature derived from multivariate cox proportional hazards regression of SNAI2 analyses. (C) Nomogram integrating risk score with additional clinical data of SNAI1 analyses. (D) Nomogram integrating risk score with additional clinical data of SNAI2 analyses.

Gene Set Enrichment Analysis

We ran the Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis on the differentially expressed genes between the high-risk and low-risk groups for SNAI1 and SNAI2. The analysis showed that the high-risk SNAI1 phenotype was significantly enriched in pathways related to proteasome, homologous recombination, and base excision repair (Figure 5). Alternatively, the high-risk SNAI2 phenotype was enriched in focal adhesion, TGF-beta signaling pathway, and cancer-related pathways (Figure 6).

Figure 5.

GSEA analysis. (A) KEGG pathways of SNAI1 analyses indicate pathways including proteasome, homologous recombination and base excision repair. (B) Heatmap of GSEA analysis between high-risk and low-risk groups.

Figure 6.

GSEA analysis. (A) KEGG pathways of SNAI2 analyses indicate pathways including focal adhesion, TGF beta signaling pathway and pathways in cancer. (B) Heatmap of GSEA analysis between high-risk and low-risk groups.

Weighted Gene Co-expression Network Analysis (WGCNA)

We analyzed metastasis characteristics from the Tumor, Node, Metastasis (TNM) classification system and SNAI1/SNAI2 expression levels (categorized as high and low) as external traits in the WGCNA analysis. The initial expression matrix comprised 412 tumor samples. After excluding samples with incomplete metastasis information, we refined the dataset to 391 samples. Of these, 391 samples and 8635 selected genes served as input for the WGCNA analysis. Using the average clustering method, we clustered the 391 samples and identified 2 outlier samples from the clustering tree, which were subsequently removed. Following outlier removal, we performed sample clustering again with updated information for subsequent WGCNA analyses (Figure 7A and B).

Figure 7.

WGCNA analysis of metastasis characteristics and SNAI1/SNAI2 gene expression. (A) Dendrogram of samples and heat map of metastasis and SNAI1 expression traits. (B) Dendrogram of samples and heat map of metastasis and SNAI2 expression traits. (C) The soft threshold in the scale-free topology fitting of SNAI1 analyses. (D) The soft threshold in the scale-free topology fitting of SNAI2 analyses. (E) Correlation between module eigengenes and metastasis and SNAI1 expression traits. (F) Correlation between module eigengenes and metastasis and SNAI2 expression traits.

We established a scale-fee network and identified modules strongly correlated with SNAI1 and SNAI2 expression levels. The soft thresholding power was set to 3 for both SNAI1 and SNAI2 modules (Figure 7C and D). We determined the resulting modules, designating non-clustering modules as gray. The analysis identified 15 gene modules for both SNAI1 and SNAI2 analyses. Among these, ME-lightyellow demonstrated the strongest correlation with both SNAI1 (r = .57, P = 1e-34) and SNAI2 (r-.65, P = 3e-47) expression groups (Figure 7E and F). Gene Ontology (GO) enrichment analysis using the “enrichR” package revealed significant biological processes, notably extracellular matrix (ECM) organization (Figure 8).

Figure 8.

GO enrichment analysis of lightyellow module genes of SNAI1 (A) and SNAI2 (B) analyses.

Construction and Analysis of ceRNA Networks

The DGE analysis of SNAI1 and SNAI2 expression groups identified 4025 and 5841 differentially expressed mRNAs, respectively. We filtered the mRNA-miRNA interactions from miRTarBase to retain only those corresponding to mRNAs identified as DEGs. Similarly, DGE analysis for SNAI1 and SNAI2 expression groups revealed 91 and 143 differentially expressed lncRNAs, respectively. We then filtered lncRNA-mRNA interactions from RNAInter to include only those related to the differentially expressed lncRNAs. We used these filtered interactions as input for the GDCRNATools package during the network analysis phase. The ceRNA network analysis (Supplemental Table S3) identified 223 edges and 118 nodes associated with SNAI1, while for SNAI2, 587 edges, and 255 nodes were identified (Supplemental Table S4). We visualized the ceRNA networks for SNAI1 and SNAI2 using these edges and nodes in Cytoscape (Figure 9A and B, Supplemental Figures S1 and S2). Further analysis of the networks using Cytoscape’s “Analyze Network” tool allowed us to identify the top 10 genes with the highest degree values, which we classified as hub genes (Tables 4 and 5).

Figure 9.

ceRNA networks of SNAI1 (A) and SNAI2 (B) analyses.

Table 4.

ceRNA hub genes of SNAI1 analyses.

Degree	Gene
23	hsa-mir-302e
23	hsa-mir-520e
23	hsa-mir-520b
13	hsa-mir-3666
13	hsa-mir-4295
13	KCNQ1OT1
11	H19
10	ARGFX
9	hsa-mir-320a
9	hsa-mir-133b

Table 5.

ceRNA hub genes of SNAI2 analyses.

Degree	Gene
68	hsa-mir-520b
68	hsa-mir-520e
68	hsa-mir-302e
27	hsa-mir-4295
27	hsa-mir-3666
26	hsa-mir-4644
24	hsa-mir-4306
24	hsa-mir-6088
24	hsa-mir-4770
20	hsa-mir-320a

Immune Cell Profile Association With SNAI1 and SNAI2 Expression

We analyzed 22 immune cell types defined in the LM22 signature matrix of CIBERSORTx across 412 tumor samples from TCGA, using TPM data as input (Supplemental Figure S3). CIBERSORTx analysis yielded significant results (P < .05) for 410 samples, which were retained for subsequent analyses. We implemented Wilcoxon rank sum tests to determine the relationship between immune cell composition and SNAI1/SNAI2 expression. For SNAI1, 8 immune cell types showed close relations (P < .05) to high or low expression levels: memory B cells, activated dendritic cells, M0 and M2 macrophages, activated and resting mast cells, neutrophils, and resting NK cells (Figure 10). For SNAI2, 6 immune cell types showed close relations to high or low expression levels (P < .05): M0, M1. and M2 macrophages, activated mast cells, neutrophils and CD8+ T cells (Figure 11). These findings imply that by modifying the quantity of particular immune cell types, SNAI1 and SNAI2 expression levels may modify the tumor immune microenvironment.

Figure 10.

The significant relationship between 8 immune cells and high and low expression of SNAI1.

Figure 11.

The significant relationship between 6 immune cells and high and low expression of SNAI2.

Discussion

Expression levels of SNAI1 and SNAI2 promote EMT, a critical process in cancer metastasis. Our goal in this work was to clarify the predictive relevance of SNAI1 and SNAI2 expression groups with high and low levels in STAD. Our findings indicate that elevated expressions of these genes are closely associated with poorer overall survival outcomes. To support these observations, we employed advanced analytical techniques that allowed us to assess risk and survival probabilities effectively.

We initiated our study with a detailed DGE analysis, revealing significant differences in SNAI1 and SNAI2 expression across patient cohorts. Cox regression analysis confirmed these findings and identified a gene signature linked to invasion and metastasis related genes that intersected with the DEGs of SNAI1 and SNAI2. We divided patients into groups with high and low risk by computing risk scores based on SNAI1 and SNAI2. Kaplan-Meier survival analysis showed that high-risk patients had significantly lower survival probabilities over time, highlighting the association between elevated SNAI1 and SNAI2 expressions and aggressive tumor behavior. We further validated the predictive capability of our model using time-dependent ROC curves, achieving high AUC values that demonstrate its effectiveness in distinguishing patient risk levels. The observed lack of statistical significance in the Kaplan-Meier survival analysis and the 5-year AUC for SNAI1 expression within the TCGA dataset warrants a detailed consideration of the inherent biases and characteristics intrinsic to the utilized data. The discrepancy in these specific analytical outcomes, when juxtaposed with findings from other datasets (eg, GEO), likely stems from fundamental differences in cancer sample composition. These variations may encompass disparities in patient demographics, the distribution of tumor stages, prior treatment regimens, all of which can exert a considerable influence on the observed relationships between gene expression and clinical endpoints. Furthermore, the intrinsic heterogeneity and plasticity of cancer, manifesting as inter- and intra-tumoral variability, represent a critical biological layer contributing to the divergence in analytical findings across distinct cohorts. Consequently, the unique molecular and clinical profiles represented within the TCGA dataset may not fully capture the spectrum of SNAI1′s prognostic relevance observed in other contexts. Future investigations employing meta-analytical approaches or integrated analyses across diverse datasets are warranted to disentangle the specific contributions of dataset-specific biases and the underlying biological variability in shaping the observed associations. Finally, to enhance clinical utility, we developed nomograms as practical tools for clinicians that combined these risk scores with other clinical factors including stage, age, and gender. The calibration plots illustrated a close alignment between predicted survival probabilities and actual observed outcomes, emphasizing the reliability of our nomogram in clinical practice.

Furthermore, we provided a comprehensive biological rationale for our findings by employing a range of sophisticated methodologies: WGCNA, ceRNA network analysis, and CIBERSORT. Each of them contributed uniquely to our understanding, collectively enhancing our insight into the complex underlying mechanisms at play. This multifaceted approach allowed us to delve deeper into the biological processes, offering a robust justification for the observed correlations (Summarized in Figure 12).

Figure 12.

High expression levels of SNAI1 and SNAI2 are associated with increased risk, supported by specific biological pathways and immune cell interactions.

WGCNA is a robust systems biology technique that facilitates the identification of co-expressed gene modules and their associations with clinical traits. By integrating WGCNA into our study, we gained invaluable insights into the biological pathways linked to elevated expressions of SNAI1 and SNAI2. Through this analysis, the lightyellow module was strongly correlated with both SNAI1 and SNAI2. This module underscored critical biological processes, including ECM organization, the formation of external encapsulating structures, and collagen fibril organization, thereby illuminating the intricate mechanisms through which these genes may influence tumor behavior.

Our findings emphasize the crucial role of ECM organization and external encapsulating structures in EMT. The ECM, a complex network surrounding cells, provides structural support and biochemical signals necessary for cellular functions. During EMT, changes in ECM composition disrupt cell adhesion, enhancing cellular motility and allowing tumor cells to detach and invade surrounding tissues.^42-44 The arrangement of collagen fibrils within the ECM is vital for its mechanical properties. Well-aligned collagen enhances the tensile strength and stiffness of ECM, supporting tissue structure and influencing cellular behaviors such as migration, proliferation, and differentiation in response to the mechanical characteristics of the matrix.^45-47

Our multivariate Cox regression analysis further elucidated the connections between the lightyellow module and tumor behavior by identifying 3 shared genes SDC3, MMP11 and COL1A1 as potential biomarkers for poor prognosis in STAD. MMP11 is a member of the family of enzymes known as matrix metalloproteinases (MMPs), which break down elements of the extracellular matrix. MMP11 has been linked to the development of several malignancies, including gastric cancer, and is especially engaged in tissue remodeling. Increased tumor aggressiveness and metastasis are associated with its expression, which is frequently elevated in malignant tissues.^48-50 COL1A1 which encodes type I collagen, the most abundant collagen in the human body, can influence the TME by modifying ECM composition and stiffness, which aids cancer cell migration along collagen tracks.^51-57 SDC3, a cell surface proteoglycan, enhances cell adhesion and migration while interacting with growth factors in the ECM. This interaction promotes tumor progression by modulating the TME.^58-60 Together, these findings reveal a regulatory network where SNAI1 and SNAI2, along with MMP11, COL1A1, and SDC3, promote EMT and enhance the invasive capacity of STAD cells. Targeting this network could effectively inhibit tumor invasion and metastasis, potentially improving therapeutic outcomes for STAD patients.

The ceRNA network also emerged as a crucial modulator in our analysis, particularly in its influence on EMT. Transcription factors SNAI1 and SNAI2 are essential for EMT by suppressing epithelial markers like E-cadherin and encouraging mesenchymal phenotypes. CeRNAs, particularly lncRNAs, function as molecular sponges for miRNAs that regulate SNAI1 and SNAI2.⁶¹ For instance, lncRNA MALAT1 can sequester miRNAs, leading to increased SNAI1 and SNAI2 activity and facilitating cancer cell invasion.⁶² Similarly, the miR-203/SNAI2 axis has shown relevance in various cancers, where miR-203 downregulation enables SNAI2-driven tumor progression.²⁷ This network is further involved in therapeutic resistance, as observed with miR-34 and miR-200 in colorectal cancer, highlighting potential targets to overcome resistance mechanisms.⁶³

The ceRNA network, which regulates gene expression by sequestering miRNAs, plays a crucial role in modulating immune cell functions and tumor-immune interactions in gastrointestinal cancers. In gastric cancer, ceRNAs influence the immune microenvironment, affecting cancer progression.⁶⁴ They regulate immune checkpoints like PD-L1, promote EMT,⁶⁵ modulate tumor-associated macrophages (TAMs)⁶⁶ and T-cell functions.⁶⁷ For instance, lncRNAs such as SNHG14 regulate PD-L1 and EMT,^68,69 while MALAT1 influences TAM polarization⁶⁶ and T-cell exhaustion and infiltration.⁷⁰ These interactions highlight the potential of targeting ceRNA networks for therapeutic interventions in gastrointestinal cancers.

A number of miRNAs that were shared by the SNAI1 and SNAI2 networks were found by our ceRNA network study, including hsa-mir-520b, hsa-mir-520e, hsa-mir-4295, hsa-mir-3666, hsa-mir-320a, and hsa-mir-302e. These miRNAs may serve as regulatory nodes in the network, influencing (silencing) the expression of their target genes involved in invasion and metastasis. MiR-520b plays a significant role in regulating tumor growth across various cancers,^71-82 primarily by targeting genes that influence cell proliferation and migration. It inhibits cancer stemness in head and neck cancer by targeting CD44.⁸³ MiR-520e controls the development of colorectal cancer cells via the Wnt/β-catenin pathway⁸⁴ and selectively targets Cyclin D1 in breast cancer.^85,86 MiR-4295 promotes tumor growth by downregulating tumor suppressor genes like BTG1 and IRF1 in bladder cancer and osteosarcoma, respectively.^87,88 Additionally, it contributes to gastric cancer progression by downregulating PTEN and is implicated in pancreatic ductal adenocarcinoma by decreasing Glypican-5 levels.^89,90 MiR-3666 acts as a tumor suppressor in several cancers by inhibiting cell growth and invasion.^91,92 In colorectal cancer, it is downregulated and may target SATB2.⁹³ The miR-320 family suppresses EMT and enhances apoptosis across multiple cancers.⁹⁴ MiR-320a specifically targets RAB14 to inhibit gastric cancer progression⁹⁵ and miR-302e influences colorectal cancer cell viability by inhibiting the JAK-STAT pathway.⁹⁶

Finally, the CIBERSORT analysis allowed us to evaluate the immune cell composition within the TME, revealing how high levels of SNAI1 and SNAI2 may modulate immune responses. We found significant correlations between these gene expressions and various immune cell types, notably an increase in M0 and M2 macrophages, activated mast cells, and neutrophils associated with high SNAI1 and SNAI2 levels. This indicates that these genes not only promote cancer progression through EMT but also influence the TME by affecting specific immune populations.^97-99 M2 macrophages promote tumor growth and metastasis and are well-known for their immunosuppressive and tissue-remodeling functions.¹⁰⁰ Activated mast cells enhance angiogenesis and tumor progression,¹⁰¹ while neutrophils contribute to inflammation, fostering a supportive environment for metastasis.¹⁰² Integrating CIBERSORT analysis into our risk assessment model emphasizes the importance of understanding the immune landscape dynamics of tumor behavior and patient prognosis.

The interplay between genes SDC3, MMP11, COL1A1 and immune cell populations significantly influences cancer invasion and metastasis in gastrointestinal cancers such as STAD. SDC3 promotes ECM remodeling and recruits M2 macrophages, which secrete immunosuppressive cytokines.⁵⁹ MMP11 facilitates EMT and metastasis by interacting with SNAI1 and SNAI2 and promoting M2 macrophage differentiation.^103-106 In addition to promoting the migration and invasion of cancer cells, COL1A1 is important in regulating the immunological response in the TME, especially affecting the actions of immune cells including T cells and macrophages.^107,108 Understanding these interactions can provide insights into cancer progression and potential therapeutic strategies.

Conclusion

In conclusion, our study delineates the critical prognostic role of SNAI1 and SNAI2 in STAD. Elevated expression levels of these genes are closely tied to poorer survival outcomes, primarily through driving EMT and aggressiveness in tumor behavior. Through applying multi-faceted computational systems biology, we identified a robust gene signature linked to invasion and metastasis. We validated our findings using time-dependent ROC curves and developed a clinically applicable nomogram for risk assessment. Furthermore, we gained novel insights into the underlying biological and immune mechanisms in GC.

This multifaceted analytical approach not only advances our understanding of STAD progression but also identifies potential therapeutic targets for clinical intervention.

Supplemental Material

sj-docx-4-cix-10.1177_11769351251352892 – Supplemental material for Comprehensive Computational Assessment of SNAI1 and SNAI2 in Gastric Cancer: Linking EMT, Tumor Microenvironment, and Survival Outcomes

Supplemental material, sj-docx-4-cix-10.1177_11769351251352892 for Comprehensive Computational Assessment of SNAI1 and SNAI2 in Gastric Cancer: Linking EMT, Tumor Microenvironment, and Survival Outcomes by Maryam Kalantari-Dehaghi, Hasan Rahimi-Tamandegani and Modjtaba Emadi-Baygi in Cancer Informatics

Supplemental Material

sj-docx-5-cix-10.1177_11769351251352892 – Supplemental material for Comprehensive Computational Assessment of SNAI1 and SNAI2 in Gastric Cancer: Linking EMT, Tumor Microenvironment, and Survival Outcomes

Supplemental material, sj-docx-5-cix-10.1177_11769351251352892 for Comprehensive Computational Assessment of SNAI1 and SNAI2 in Gastric Cancer: Linking EMT, Tumor Microenvironment, and Survival Outcomes by Maryam Kalantari-Dehaghi, Hasan Rahimi-Tamandegani and Modjtaba Emadi-Baygi in Cancer Informatics

Supplemental Material

sj-docx-6-cix-10.1177_11769351251352892 – Supplemental material for Comprehensive Computational Assessment of SNAI1 and SNAI2 in Gastric Cancer: Linking EMT, Tumor Microenvironment, and Survival Outcomes

Supplemental material, sj-docx-6-cix-10.1177_11769351251352892 for Comprehensive Computational Assessment of SNAI1 and SNAI2 in Gastric Cancer: Linking EMT, Tumor Microenvironment, and Survival Outcomes by Maryam Kalantari-Dehaghi, Hasan Rahimi-Tamandegani and Modjtaba Emadi-Baygi in Cancer Informatics

Supplemental Material

sj-docx-7-cix-10.1177_11769351251352892 – Supplemental material for Comprehensive Computational Assessment of SNAI1 and SNAI2 in Gastric Cancer: Linking EMT, Tumor Microenvironment, and Survival Outcomes

Supplemental material, sj-docx-7-cix-10.1177_11769351251352892 for Comprehensive Computational Assessment of SNAI1 and SNAI2 in Gastric Cancer: Linking EMT, Tumor Microenvironment, and Survival Outcomes by Maryam Kalantari-Dehaghi, Hasan Rahimi-Tamandegani and Modjtaba Emadi-Baygi in Cancer Informatics

Supplemental Material

sj-pdf-1-cix-10.1177_11769351251352892 – Supplemental material for Comprehensive Computational Assessment of SNAI1 and SNAI2 in Gastric Cancer: Linking EMT, Tumor Microenvironment, and Survival Outcomes

Supplemental material, sj-pdf-1-cix-10.1177_11769351251352892 for Comprehensive Computational Assessment of SNAI1 and SNAI2 in Gastric Cancer: Linking EMT, Tumor Microenvironment, and Survival Outcomes by Maryam Kalantari-Dehaghi, Hasan Rahimi-Tamandegani and Modjtaba Emadi-Baygi in Cancer Informatics

Supplemental Material

sj-pdf-2-cix-10.1177_11769351251352892 – Supplemental material for Comprehensive Computational Assessment of SNAI1 and SNAI2 in Gastric Cancer: Linking EMT, Tumor Microenvironment, and Survival Outcomes

Supplemental material, sj-pdf-2-cix-10.1177_11769351251352892 for Comprehensive Computational Assessment of SNAI1 and SNAI2 in Gastric Cancer: Linking EMT, Tumor Microenvironment, and Survival Outcomes by Maryam Kalantari-Dehaghi, Hasan Rahimi-Tamandegani and Modjtaba Emadi-Baygi in Cancer Informatics

Supplemental Material

sj-pdf-3-cix-10.1177_11769351251352892 – Supplemental material for Comprehensive Computational Assessment of SNAI1 and SNAI2 in Gastric Cancer: Linking EMT, Tumor Microenvironment, and Survival Outcomes

Supplemental material, sj-pdf-3-cix-10.1177_11769351251352892 for Comprehensive Computational Assessment of SNAI1 and SNAI2 in Gastric Cancer: Linking EMT, Tumor Microenvironment, and Survival Outcomes by Maryam Kalantari-Dehaghi, Hasan Rahimi-Tamandegani and Modjtaba Emadi-Baygi in Cancer Informatics

Footnotes

ORCID iDs

Maryam Kalantari-Dehaghi

Modjtaba Emadi-Baygi

Author Contributions

Maryam Kalantari-Dehaghi: Data curation, Formal analysis, Investigation, Methodology, Software, Validation, Visualization, Writing—original draft, and Writing—review and editing. Hasan Rahimi-Tamandegani: Data curation, Formal analysis, Investigation, Methodology, Project administration, Software, Validation, Writing—original draft, and Writing—review and editing. Modjtaba Emadi-Baygi: Conceptualization, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Supervision, Validation, Writing—original draft, and Writing—review and editing.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported in part by Shahrekord University [1403].

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Supplemental Material

Supplemental material for this article is available online.

References

Zhou

Huang

Feng

Teng

Exploring TCGA database for identification of potential prognostic genes in stomach adenocarcinoma. Cancer Cell Int. 2020;20:264-312.

Ilic

Epidemiology of stomach cancer. World J Gastroenterol. 2022;28(12):1187-1203.

Machlowska

Baj

Sitarz

Maciejewski

Sitarz

Gastric cancer: epidemiology, risk factors, classification, genomic characteristics and treatment strategies. Int J Mol Sci. 2020;21(11):4012.

Iwu

Iwu-Jaja

CJ.

Gastric cancer epidemiology: current trend and future direction. Hyg. 2023;3(3):256-268.

Mamun

Younus

Rahman

MH.

Gastric cancer-epidemiology, modifiable and non-modifiable risk factors, challenges and opportunities: an updated review. Cancer Treat Res Commun. 2024;41:100845.

Morgan

Arnold

Camargo

, et al. The current and future incidence and mortality of gastric cancer in 185 countries, 2020-40: a population-based modelling study. med. 2022;47:101404.

Mousavi

Ilaghi

Elahi Vahed

Nejadghaderi

SA.

Epidemiology and socioeconomic correlates of gastric cancer in Asia: results from the GLOBOCAN 2020 data and projections from 2020 to 2040. Sci Rep. 2025;15(1):6529.

Huang

R-L

A-M.

Epithelial-mesenchymal transition in gastric cancer. Am J Transl Res. 2015;7(11):2141-2158.

Xie

Wang

Liu

Epithelial-mesenchymal transition orchestrates tumor microenvironment: current perceptions and challenges. J Transl Med. 2025;23(1):386.

10.

Derynck

Weinberg

RA.

EMT and cancer: more than meets the eye. Dev Cell. 2019;49(3):313-316.

11.

Fontana

Mestre-Farrera

Yang

Update on epithelial-mesenchymal plasticity in cancer progression. Annu Rev Pathol Mech Dis. 2024;19(1):133-156.

12.

Tomecka

Kunachowicz

Górczyńska

, et al. Factors determining epithelial-mesenchymal transition in cancer progression. Int J Mol Sci. 2024;25(16):8972.

13.

Peng

Wang

Fang

Wang

Tong

Role of epithelial-mesenchymal transition in gastric cancer initiation and progression. World J Gastroenterol. 2014;20(18):5403-5410.

14.

Taki

Abiko

Ukita

, et al. Tumor immune microenvironment during epithelial–mesenchymal transition. Clin Cancer Res. 2021;27(17):4669-4679.

15.

Romeo

Caserta

Rumio

Marcucci

The vicious cross-talk between tumor cells with an EMT phenotype and cells of the immune system. Cells. 2019;8(5):460.

16.

Fedele

Melisi

Permissive state of EMT: the role of immune cell compartment. Front Oncol. 2020;10:587.

17.

Mullins

RDZ

Pal

Barrett

Heft Neal

Puram

. Epithelial-mesenchymal plasticity in tumor immune evasion. Cancer Res. 2022;82(13):2329-2343.

18.

Liu

Ali

Zhang

Cui

The role of innate immune cells in the tumor microenvironment and research progress in anti-tumor therapy. Front Immunol. 2022;13:1039260.

19.

Han

, et al. Innate immune cells in tumor microenvironment: a new frontier in cancer immunotherapy. Iscience, 2024;27:110750.

20.

Wang

Zhang

, et al. The pan-cancer landscape of crosstalk between epithelial-mesenchymal transition and immune evasion relevant to prognosis and immunotherapy response. NPJ Precis Oncol. 2021;5(1):56.

21.

Debnath

Huirem

Dutta

Palchaudhuri

Epithelial-mesenchymal transition and its transcription factors. Biosci Rep. 2022;42(1):BSR20211754.

22.

Barrallo-Gimeno

Nieto

MA.

The snail genes as inducers of cell movement and survival: implications in development and cancer. Development. 2005;132:3151-3161.

23.

Zhou

BP.

Snail: more than EMT. Cell Adh Migr. 2010;4(2):199-203.

24.

Dong

Epigenetic regulation and post-translational modifications of SNAI1 in cancer metastasis. Int J Mol Sci. 2021;22(20):11062.

25.

Pulkka

O-P

Nilsson

Sarlomo-Rikala

, et al. SLUG transcription factor: a pro-survival and prognostic factor in gastrointestinal stromal tumour. Br J Cancer. 2017;116(9):1195-1202.

26.

Wang

Hou

, et al. The emerging function and mechanism of ceRNAs in cancer. Trends Genet. 2016;32(4):211-224.

27.

Yang

Liu

Yang

Gao

ceRNAs in cancer: mechanism and functions in a comprehensive regulatory network. J Oncol. 2021;2021(1):1-12.

28.

Zhang

, et al. ceRNA in cancer: possible functions and clinical implications. J Med Genet. 2015;52(10):710-718.

29.

Liu

Petinrin

, et al. Uncovering the ceRNA network related to the prognosis of stomach adenocarcinoma among 898 patient samples. Biochem Genet. 2024;62:4770-4790.

30.

Hou

Wang

Luo

Identification of a potential competing endogenous RNA (ceRNA) network in gastric adenocarcinoma. J Gastrointest Oncol. 2023;14(2):1019-1036.

31.

Liu

Xue

, et al. Competitive endogenous RNA is an intrinsic component of EMT regulatory circuits and modulates EMT. Nat Commun. 2019;10(1):1637.

32.

Yuan

Yan

Zhang

, et al. CancerSEA: a cancer single-cell state atlas. Nucleic Acids Res. 2019;47(D1):D900-D908.

33.

Zhang

Huo

Xie

, et al. CHG: a systematically integrated database of cancer hallmark genes. Front Genet. 2020;11:29.

34.

Tate

Bamford

Jubb

, et al. COSMIC: the catalogue of somatic mutations in cancer. Nucleic Acids Res. 2019;47(D1):D941-D947.

35.

Stelzer

Rosen

Plaschkes

, et al. The GeneCards suite: from gene data mining to disease genome sequence analyses. Curr Protoc Bioinformatics. 2016;54(1):1.30. 1-1.30. 33.

36.

Liberzon

Birger

Thorvaldsdóttir

, et al. The Molecular Signatures Database (MSigDB) hallmark gene set collection. Cell Syst. 2015;1(6):417-425.

37.

Langfelder

Horvath

WGCNA: an R package for weighted correlation network analysis. BMC Bioinformatics. 2008;9:1-13.

38.

Kang

Tang

Q, J

, RNAInter v4.0: RNA interactome repository with redefined confidence scoring system and improved accessibility. Nucleic Acids Res. 2022;50(D1):D326-D332.

39.

Huang

H-Y

Lin

Cui

, et al. miRTarBase update 2022: an informative resource for experimentally validated miRNA-target interactions. Nucleic Acids Res. 2022;50(D1):D222-D230.

40.

Newman

Liu

Green

, et al. Robust enumeration of cell subsets from tissue expression profiles. Nat Methods. 2015;12(5):453-457.

41.

Chen

Khodadoust

Liu

Newman

Alizadeh

AA.

Profiling tumor infiltrating immune cells with CIBERSORT. Methods Mol Biol. 2018;1711:243-259.

42.

Scott

Weinberg

Lemmon

CA.

Mechanochemical signaling of the extracellular matrix in epithelial-mesenchymal transition. Front Cell Dev Biol. 2019;7:135.

43.

Tzanakakis

Kavasi

Voudouri

, et al. Role of the extracellular matrix in cancer-associated epithelial to mesenchymal transition phenomenon. Dev Dyn. 2018;247(3):368-381.

44.

Yuan

Zhang

, et al. Extracellular matrix remodeling in tumor progression and immune escape: from mechanisms to treatments. Mol Cancer. 2023;22(1):48.

45.

Sun

The mechanics of fibrillar collagen extracellular matrix. Cell Rep Phys Sci. 2021;2(8):100515.

46.

Orgel

San Antonio

Antipova

Molecular and structural mapping of collagen fibril interactions. Connect Tissue Res. 2011;52(1):2-17.

47.

Ricard-Blum

The collagen family. Cold Spring Harb Perspect Biol. 2011;3(1):a004978.

48.

Ran

Liao

Zhang

HH.

The paradoxical role of matrix metalloproteinase-11 in cancer. Biomed Pharmacother. 2021;141:111899.

49.

Huang

H-C

Shiu

, et al. The impact of matrix metalloproteinase-11 polymorphisms on colorectal cancer progression and clinicopathological characteristics. Diagnostics. 2022;12(7):1685.

50.

Molière

Jaulin

Tomasetto

Dali-Youcef

Roles of matrix metalloproteinases and their natural inhibitors in metabolism: insights into health and disease. Int J Mol Sci. 2023;24(13):10649.

51.

Devos

Zoidakis

Roubelakis

Latosinska

Vlahou

Reviewing the regulators of COL1A1. Int J Mol Sci. 2023;24(12):10004.

52.

Down-regulation of COL1A1 inhibits tumor-associated fibroblast activation and mediates matrix remodeling in the tumor microenvironment of breast cancer. Open Life Sci. 2023;18(1):20220776.

53.

Wang

, et al. Microenvironment remodeled by tumor and stromal cells elevates fibroblast-derived COL1A1 and facilitates ovarian cancer metastasis. Exp Cell Res. 2020;394(1):112153.

54.

Sun

Kan

, et al. COL1A1: a novel oncogenic gene and therapeutic target in malignancies. Pathol Res Pract. 2022;236:154013.

55.

Jin

Xue

Unveiling collagen’s role in breast cancer: insights into expression patterns, functions and clinical implications. Int J Gen Med. 2024;17:1773-1787.

56.

Liu

Xue

Zhong

, et al. Prognostic prediction, immune microenvironment, and drug resistance value of collagen type I alpha 1 chain: from gastrointestinal cancers to pan-cancer analysis. Front Mol Biosci. 2021;8:692120.

57.

Fujisaki

Futaki

Epithelial-mesenchymal transition induced in cancer cells by adhesion to type I collagen. Int J Mol Sci. 2022;24(1):198.

58.

Hillemeyer

Espinoza-Sanchez

Greve

, et al. The cell surface heparan sulfate proteoglycan syndecan-3 promotes ovarian cancer pathogenesis. Int J Mol Sci. 2022;23(10):5793.

59.

Prieto-Fernández

Egia-Mendikute

Bosch

, et al. Hypoxia promotes syndecan-3 expression in the tumor microenvironment. Front Immunol. 2020;11:586977.

60.

Czarnowski

Syndecans in cancer: a review of function, expression, prognostic value, and therapeutic significance. Cancer Treat Res Commun. 2021;27:100312.

61.

Gao

The ceRNA network regulates epithelial-mesenchymal transition in colorectal cancer. Heliyon. 2023;9(3):e14143.

62.

Hussein

Valinezhad

Adel

Munirathinam

MALAT-1 is a key regulator of epithelial–mesenchymal transition in cancer: a potential therapeutic target for metastasis. Cancers. 2024;16(1):234.

63.

Kim

Song

Choi

, et al. Competing endogenous RNA of snail and Zeb1 UTR in therapeutic resistance of colorectal cancer. Int J Mol Sci. 2021;22(17):9589.

64.

Zhao

Liu

, et al. MicroRNA-140 suppresses Helicobacter pylori-positive gastric cancer growth by enhancing the antitumor immune response. Mol Med Rep. 2019;20(3):2484-2492.

65.

Zabeti Touchaei

Vahidi

. MicroRNAs as regulators of immune checkpoints in cancer immunotherapy: targeting PD-1/PD-L1 and CTLA-4 pathways. Cancer Cell Int. 2024;24(1):102.

66.

Mohapatra

Pioppini

Ozpolat

Calin

GA.

Non-coding RNAs regulation of macrophage polarization in cancer. Mol Cancer. 2021;20:24-15.

67.

Chen

Feng

Kahlert

, et al. Construction of ceRNA networks associated with CD8 T cells in breast cancer. Front Oncol. 2022;12:883197.

68.

Jiang

Pan

Chen

Wang

Zhu

The role of lncRNAs and circRNAs in the PD-1/PD-L1 pathway in cancer immunotherapy. Mol Cancer. 2021;20:116-117.

69.

Jiang

Zhao

, et al. The role of NcRNAs to regulate immune checkpoints in cancer. Front Immunol. 2022;13:853480.

70.

Erber

Herndler-Brandstetter

Regulation of T cell differentiation and function by long noncoding RNAs in homeostasis and cancer. Front Immunol. 2023;14:1181499.

71.

Zhang

X-D

Zhang

Q-F

Liu

Effects of miR-520b on migration and invasion of cervical cancer cells through negative regulation of SOS1. Int J Clin Exp Med. 2019;12(10):12046-12056.

72.

Liu

Sun

, et al. GATA6 suppresses migration and metastasis by regulating the miR-520b/CREB1 axis in gastric cancer. Cell Death Dis. 2019;10(2):35.

73.

Xiao

Zhou

, et al. MicroRNA-520b functions as a tumor suppressor in colorectal cancer by inhibiting defective in cullin neddylation 1 domain containing 1 (DCUN1D1). Oncol Res. 2018;26(4):593-604.

74.

Wang

Pang

Zuo

Zhang

MicroRNA-520b suppresses proliferation, migration, and invasion of spinal osteosarcoma cells via downregulation of frizzled-8. Oncol Res. 2017;25(8):1297-1304.

75.

Zhou

Gao

Zhang

, et al. miR-520b inhibits proliferation, migration and invasion in gallbladder carcinoma by targeting RAB22A. Arch Med Sci. 2021;17(2):481-491.

76.

Cui

Liu

Wan

, et al. miR-520b inhibits the development of glioma by directly targeting MBD2. Am J Cancer Res. 2017;7(7):1528-1539.

77.

Zhang

Lang

Zou

, et al. miR-520b promotes breast cancer stemness through Hippo/YAP signaling pathway. Onco Targets Ther. 2019;12:11691-11700.

78.

Zhang

Ning

, et al. miR-520b/e regulates proliferation and migration by simultaneously targeting EGFR in gastric cancer. Cell Physiol Biochem. 2016;40(6):1303-1315.

79.

Zhang

Zhu

, et al. MLK3 is a newly identified microRNA-520b target that regulates liver cancer cell migration. PLoS One. 2020;15(3):e0230716.

80.

Zhang

Role of miR-520b in non-small cell lung cancer. Exp Ther Med. 2018;16(5):3987-3995.

81.

Liu

Luo

Fang

FGF16 regulated by miR-520b enhances the cell proliferation of lung cancer. Open Med. 2021;16(1):419-427.

82.

Guan

Cai

Sun

Upregulation of miR-520b promotes ovarian cancer growth. Oncol Lett. 2017;14(3):3155-3161.

83.

Y-C

Cheng

Lee

, et al. miR-520b as a novel molecular target for suppressing stemness phenotype of head-neck cancer by inhibiting CD44. Sci Rep. 2017;7(1):2042.

84.

Zhang

, et al. MicroRNA-520e targets AEG-1 to suppress the proliferation and invasion of colorectal cancer cells through Wnt/GSK-3β/β-catenin signalling. Clin Exp Pharmacol Physiol. 2020;47(1):158-167.

85.

Long

, et al. MiR-520e regulates cell proliferation, apoptosis and migration in breast cancer. Oncol Lett. 2016;12(5):3543-3548.

86.

Liang

Yao

CyclinD1 is a new target gene of tumor suppressor miR-520e in breast cancer. Open Med. 2019;14(1):913-919.

87.

Nan

Wang

, et al. MiR-4295 promotes cell growth in bladder cancer by targeting BTG1. Am J Transl Res. 2016;8(11):4892.

88.

Cheng

Huang

Duan

Zhuang

ZL.

MiR-4295 promotes cell proliferation, migration and invasion of osteosarcoma through targeting interferon regulatory factor 1. Oncol Lett. 2020;20(5):260-261.

89.

Lei

Gan

Yang

, et al. MiR-4295 promotes the malignant progression of gastric cancer via targeting PTEN. Comb Chem High Throughput Screen. 2022;25(11):1897-1906.

90.

Yuan

Zhang

Cao

Yang

High expression of microRNA-4295 contributes to cell proliferation and invasion of pancreatic ductal adenocarcinoma by the down-regulation of Glypican-5. Biochem Biophys Res Commun. 2018;497(1):73-79.

91.

MicroRNA-3666 inhibits breast cancer cell proliferation by targeting sirtuin 7. Mol Med Rep. 2017;16(6):8493-8500.

92.

Chen

Cao

MicroRNA-3666 suppresses cell growth in head and neck squamous cell carcinoma through inhibition of PFKFB3-mediated Warburg effect. Onco Targets Ther. 2020;13:9029-9041.

93.

Yang

Xia

Zhou

miR-3666 suppresses cellular proliferation and invasion in colorectal cancer by targeting SATB2. Mol Med Rep. 2018;18(6):4847-4854.

94.

Liang

Tang

MicroRNA 320, an anti-oncogene target miRNA for cancer therapy. Biomedicines. 2021;9(6):591.

95.

Liu

Shao

Xing

MiR-320a serves as a negative regulator in the progression of gastric cancer by targeting RAB14. Mol Med Rep. 2017;16(3):2652-2658.

96.

Chen

Song

Nie

, et al. CXCL1 regulated by miR-302e is involved in cell viability and motility of colorectal cancer via inhibiting JAK-STAT signaling pathway. Front Oncol. 2020;10:577229.

97.

Fang

Zhang

Sun

Analysis of the effect of SNAI family in breast cancer and immune cell. Front Cell Dev Biol. 2022;10:906885.

98.

Tang

Sui

Weng

Liu

SNAIL1: linking tumor metastasis to immune evasion. Front Immunol. 2021;12:724200.

99.

Pioli

Whiteside

Weis

JH.

SNAI2 and SNAI3 transcriptionally regulate cellular fitness and functionality of T cell lineages through distinct gene programs. Immunobiology. 2016;221(5):618-633.

100.

Liu

Geng

Hou

New insights into M1/M2 macrophages: key modulators in cancer progression. Cancer Cell Int. 2021;21:389-397.

101.

Saxena

Singh

Tumor associated mast cells: biological roles and therapeutic applications. Anat Cell Biol. 2020;53(3):245-251.

102.

Carnevale

Di Ceglie

Grieco

, et al. Neutrophil diversity in inflammation and cancer. Front Immunol. 2023;14:1180810.

103.

Radisky

DC.

Matrix metalloproteinase-induced epithelial-mesenchymal transition in breast cancer. J Mammary Gland Biol Neoplasia. 2010;15:201-212.

104.

Pan

Dai

Wei

, et al. Matrix metalloproteinase 11 promotes migration and invasion of colorectal cancer by elevating slug protein. Int J Med Sci. 2024;21(11):2170-2188.

105.

Fang

Tao

Lino-Silva

MMP11 as a prognostic biomarker correlated with immune infiltrates in pancreatic adenocarcinoma. J Gastrointest Oncol. 2024;15(3):1224-1244.

106.

Bai

Huo

Fang

Shang

MMP11 is associated with the immune response and immune microenvironment in EGFR-mutant lung adenocarcinoma. Front Oncol. 2023;13:1055122.

107.

Zhang

Liu

Wang

, et al. COL1A1 is a potential prognostic biomarker and correlated with immune infiltration in mesothelioma. Biomed Res Int. 2021;2021(1):5320941.

108.

Rømer

AMA

Thorseth

M-L

Madsen

. Immune modulatory properties of collagen in cancer. Front Immunol. 2021;12:791453.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.05 MB

0.04 MB

0.11 MB

0.12 MB

0.01 MB

0.04 MB

0.34 MB