GC is a malignant neoplasm affecting the gastrointestinal tract, characterized by a relatively high incidence and mortality. A significant proportion of patients present at an advanced stage upon diagnosis of GC, resulting in a 5-year survival rate ranging from only 11–40%[7]. Therefore, there is an urgent need for sensitive and specific biomarkers for GC detection. In this study, bioinformatics methods for analyzing hub genes and pathways showed promise for providing new insights into the diagnosis, treatment, and prognosis of GC. Based on the GSE79973 expression profile from the GEO database, we identified 935 common differentially expressed genes and conducted an in-depth analysis using R and bioinformatics tools. Using the STRING online database, 9 important regulatory genes were ultimately screened out: COL1A1, COL1A2, COL4A1, COL4A2, COL6A3, FN1, ITGA5, THBS1, and TLR2. Based on TCGA database data, COL1A1, COL1A2, COL4A1, and TLR2 were found to potentially have good prognostic value for GC, with their high expression indicating a lower survival rate for GC. Therefore, they were considered hub genes for GC with prognostic value.
The PI3K signaling pathway plays a pivotal role in governing a multitude of fundamental cellular activities, encompassing proliferation, apoptosis, and migration of cells. With frequent alterations found in the PTEN/PI3K/AKT pathway in GC, this pathway is closely associated with the onset and progression of GC. The dysregulation in intricate networks comprising key proteins and signal cascades results in an imbalance between cell growth and apoptosis, thereby leading to tumor development. The PTEN/PI3K/AKT pathway is essential in determining the fundamental roles of cell death and survival. The pro-survival role of the Akt signaling pathway has been confirmed in previous studies using anti-tumor factors. Research has found that the overexpression of tumor suppressor gene growth inhibitor 3 in GC cells, in addition to inducing apoptosis, can also reduce the proliferation of G2/M phase cells and block the cell cycle. Dysfunction within the PI3K/AKT signaling pathway is implicated in this mechanism. Beyond promoting the onset of GC, abnormally activated PI3K signaling can also promote the progression of GC into a highly malignant type characterized by metastasis and chemotherapy resistance. Metastasis represents a primary cause of the recurrence and death in patients with GC and is a significant potential risk. Throughout this multistep process, many cell biological activities are controlled, including adhesion, migration, invasion, and angiogenesis.
The onset, progression, invasion, and metastasis of malignant tumors are frequently accompanied by alterations in the expression profiles of the extracellular matrix (ECM) and cell-surface receptors[8]. Collagen constitutes a prominent constituent of the ECM, serving as both an attachment site and a scaffold for cellular growth. It is able to induce proliferation, differentiation and migration of epithelial cells. Additionally, it holds significance in maintaining cell-cell adhesion, tissue integrity, and regeneration[9]. COL1A1 is a major component of collagen I; it is found in most connective tissues including cartilage. Some studies have shown that the overexpression of COL1A1 is associated with poor clinical outcomes, tumor aggressiveness, and metastasis. Furthermore, studies have demonstrated that GC cells synthesize and secrete COL1 at a rate markedly faster than fibroblasts.
In addition, research has discovered that the α1 chain of type I collagen (COL1A1) activates the Rac1-GTP, p-JNK, and RhoA-GTP pathways through the WNT/planar cell polarity pathway. RhoGTPases and the JNK pathway relay signals from the cell surface frizzled gene and ROR2/RYK co-receptor to the cell nucleus, a crucial process in the metastasis of tumor cells[10]. Chemotherapy is deemed the cornerstone of treatment for patients with metastatic GC. In addition to standard first-line and second-line treatments, further chemotherapy results have not effectively prolonged the survival of patients. Hence, there exists a clinical imperative to explore novel therapeutic approaches to manage the progression of GC and improve the prognosis for patients with metastatic GC. Apatinib significantly contributes to various tumor treatments owing to its favorable side effects and enhanced structure. However, a primary practical limitation of anti-angiogenic treatment strategies is the inevitable development of drug resistance[11]. Yet, the molecular mechanisms underlying apatinib resistance against GC remain unclear. As the most abundant matrix protein in the tumor stroma, COL1 can promote tumor progression by fostering cancer cell growth, invasion, metastasis, and resistance to anti-tumor drugs[12]. The interaction between COL1 and resistance against cisplatin in ovarian cancer has been confirmed by KW von Rekowski et al[13]. Moreover, KEGG PEA showed that COL1A1 regulates carboplatin resistance in ovarian cancer cells via "ECM-receptor interaction" and "focal adhesion" pathways[14]. Thus, COL1A1 may contribute to the metastatic process of GC. Prior research has revealed that the COL1A2 gene influences the proliferation, differentiation, adhesion, and metastasis of cells by encoding the most abundantly expressed COL1 in the fibrillar collagen family. Studies have indicated that the COL1A2 overexpression in GC is attributed to the coordination between EP300 and TWIST1[15]. In addition, a similar mechanism has been noted in multiple myeloma cells, where the proliferation of cancer cells is promoted by the SP1/EP300 complex through the regulation of the IQGAP1 transcription[16]. Type IV collagen α1 (COL4A1) significantly contributes to tumor invasion by inducing tumor budding. Heterotrimers, comprising COL4A1 and COL4A2, represent one of the predominant constituents in nearly all basement membranes. Therefore, COL4A1 or COL4A2 mutations are multifunctional[17], yet the molecular mechanism of COL4A2 in GC remains elusive. Endothelial TLR2 promotes angiogenesis by recruiting pro-angiogenic immune cells and assumes a pathological role in pro-inflammatory diseases.
Fibroblasts that are involved in cancers serve as pivotal factors in their malignant progression, with COL6A3 primarily expressed in these cells. Knockout experiments have validated the involvement of COL6A3 in the proliferation and invasion of CRC cells[18]. Gene ontology (GO) annotations showed that co-expressed genes of FN1 are mainly involved in the organization of extracellular structures, metabolism of collagens, integrin-mediated signaling pathways, migration of substrate-dependent cells, peptide cross-linking, artery agenesis, and the migration of muscle cells[19]. Furthermore, genomics studies have suggested the involvement of FN1 in alterations of GC immune checkpoints and macrophage biomarkers[19]. Within the tumor microenvironment, immune cell infiltration has been shown to exert a pivotal impact on the progression of cancers[20, 21]. FN1 expression is essential in the invasion and migration of tumors. Studies have demonstrated a correlation between the expression of FN1 and the extent of GC-related immune infiltration. The expression of FN1 exhibits a close association with tumor-infiltrating immune cells, such as macrophages, Treg cells, NK cells, CD8+ T cells, and dendritic cells[22, 23]. Integrin subunit α 5 (ITGA5) primarily participates in "integrin-mediated signaling pathway", "leukocyte migration", and "cell-substrate adhesion"[24]. Thrombospondin 1 (THBS1) promotes metastasis by inducing the exhaustion of cytotoxic T cells and disrupting vascularization[25].
This study, through GEO and TCGA data mining, identified differential expression of COL1A1, COL1A2, COL4A1, and TLR2 in tumor and non-tumor tissues, which is intricately linked to the onset and progression of GC. Moreover, the ROC curve analysis has confirmed its biological significance as a major advantage in prognostic prediction for GC. These investigations have also, for the first time, unveiled the prognostic value of these four genes in patients with GC, offering novel avenues for subsequent disease diagnosis and treatment research.
Inevitably, our study has certain limitations. To begin with, as a retrospective study, additional prospective studies are required to validate our findings. Secondly, further comprehensive experiments are needed to validate the reliability of the mechanism analysis. Therefore, extensive studies will carried out in the future to elucidate the mechanistic relevance of these genes to the progression and prognosis of GC.