Identification of a novel ion channel-related gene signature to predict prognosis and immune response of gastric cancer

Introduction

Worldwide, gastric cancer (GC) holds a position within the leading five malignancies based on mortality alongside incidence rates (1). Dissatisfying clinical results continue to characterize advanced GC patients, even with substantial developments realized in GC therapies encompassing immunotherapy, radiotherapy, chemotherapy, endoscopic therapy, and surgical treatment (2-5). Therefore, early detection of GC and the development of effective prognostic models are of critical importance (6,7). Although several prognostic risk models for GC, including the immune-related prognostic signature (8), the ferroptosis-related prognostic signature (9), and the angiogenesis-related prognostic signature of GC (10), have been developed, there remains a need for more effective biomarkers for the risk prediction and individualized treatment of GC.

Ion channels, as ion transport proteins on biological membranes, have been reported to exhibit abnormal expression in GC (11,12). Moreover, genes involved in ion channels and those regulating them have exhibited roles in altering fundamental cellular mechanisms within GC, encompassing cell cycle, apoptosis, invasion, proliferation, and migration (13-16). Studies further demonstrate that such genes contribute to the regulation of the tumor immune microenvironment (TIME) alongside resistance mechanisms against drugs in GC cells (17-20). Observations from these works imply that particular ion channel-related genes (ICRGs) exert effects on GC prognosis. Detection of ICRGs affecting GC outcomes holds potential to yield fresh biomarkers suited for initial detection and survival estimation, plus prospective sites for therapeutic strategies against GC. Notably, certain channel proteins are located at the cell exterior, permitting targeted agents to evade the necessity of crossing the plasma membrane, thus rendering them attractive options for advancing drug candidates (21). Several studies have established models for predicting the prognosis of prostate cancer, head and neck squamous cell carcinoma, and clear cell renal cell carcinoma based on ICRGs (22-24). Nonetheless, there is still a lack of research on prognostic models based on ICRGs in GC.

In our study, we identified ion channel-related prognostic genes in GC. Based on these genes, we established a risk model for predicting the prognosis of GC. Subsequently, we classified GC patients into high- and low-ICRGRS groups according to the risk score model and compared the survival status, gene mutation profiles, immune infiltration, and responses to drug treatments between the two groups. Finally, we verified the effects of CNGB3 on the proliferation, migration, invasion, intracellular Ca²⁺ concentration, and expression of major histocompatibility complex class I (MHC-I) in GC cells through in vitro experiments. We present this article in accordance with the MDAR and TRIPOD reporting checklists (available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-1-2874/rc).

Methods

Data collection and differential genes expression analysis

Figure 1 shows a flowchart of the study. Matrices depicting gene expression at the transcriptomic scale for normal tissues near GC samples were retrieved via The Cancer Genome Atlas (TCGA) database’s stomach adenocarcinoma (STAD) component (https://portal.gdc.cancer.gov/). Analytical procedures employing the DESeq2 module within R allowed for the discernment of differentially expressed genes (DEGs) associated with GC. This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments.

Figure 1 A flowchart of the study. GEO, Gene Expression Omnibus; ICRGs, ion channel-related genes; LASSO, least absolute shrinkage and selection operator; ROC, receiver operating characteristic; TCGA-STAD, The Cancer Genome Atlas-Stomach Adenocarcinoma.

Identification of ion channel-related hub genes

An ICRG set was obtained by searching the keyword “ion channel” in the GeneCards (https://www.genecards.org/) database and applying a filter based on the criterion of a relevance score exceeding 6. The DEGs in GC were intersected with the ICRG set to identify the ion channel-related DEGs in GC. Univariate Cox regression analysis and least absolute shrinkage and selection operator (LASSO) regression analysis were employed to identify 8 ion channel-related prognostic hub genes.

Construction and validation of the ICRG prognosis model

The model for evaluating prognostic risks was constructed by combining coefficients of individual genes with expression levels of the eight essential genes. Computation of the risk score adheres to this equation: ion channel-related gene risk score (ICRGRS)=∑i=18(coefficientgenei×expressiongenei). Gene expression information derived from GSE62254 (https://www.ncbi.nlm.nih.gov/geo/) in conjunction with the TCGA-STAD repository facilitated determination of risk values across all individuals. Subgrouping of individuals ensued into low-ICRGRS alongside high-ICRGRS categories dependent upon median risk thresholds. Utilization of receiver operating characteristic (ROC) curves, survival analysis, and risk factor analysis serves to confirm the risk score’s effectiveness in prognostic anticipation.

Analysis of independent prognostic factors

Univariate combined with multivariate Cox regression methodologies enabled recognition of standalone prognostic indicators relevant to GC. Leveraging those indicators allowed for the creation of a nomogram aimed at forecasting GC survival probabilities. ROC curves, calibration curves, and decision curve analysis (DCA) are employed to evaluate the effectiveness of the prediction model.

Gene mutation analysis

Mutation profiles pertaining to TCGA-STAD individuals assigned to low-ICRGRS alongside high-ICRGRS subgroups received scrutiny through application of the “maftools” package, thereby producing a chart outlining frequencies of diverse mutation categories.

Enrichment analysis

Examination of gene expression differences occurred through utilization of profile datasets derived from individuals within low-ICRGRS alongside high-ICRGRS subgroups. Sets of genes exhibiting differential expression then received scrutiny via Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway assessments coupled with Gene Ontology (GO) evaluations.

Immune infiltration analysis

Assessment of discrepancies in immune cell infiltration occurred through application of the CIBERSORT methodology across low-risk alongside high-risk subgroups for GC individuals derived from TCGA repositories. Evaluation via single sample Gene Set Enrichment Analysis (ssGSEA) facilitated contrasting of infiltration pertaining to immune functions. Moreover, comparisons involving diverse risk categories revealed differences in expression among immune checkpoints coupled with HLA family genes.

Immunotherapy response analysis

To anticipate immunotherapeutic responsiveness, evaluations of microsatellite instability (MSI), immunophenoscore (IPS), and Tumor Immune Dysfunction and Exclusion (TIDE) score occurred across low-risk alongside high-risk subgroups among GC individuals derived from TCGA repositories.

Drug sensitivity analysis

The “oncoPredict” package was used in combination with the Genomics of Drug Sensitivity in Cancer database to compare the differences in drug sensitivities between the two patient groups.

Cell culture

Cell lines encompassing GES-1, AGS and HGC27 cells were obtained from Shanghai Zhong Qiao Xin Zhou Biotechnology Co., Ltd. Cultivation of GES-1 and HGC27 occurs within RPMI-1640 medium (VivaCell, Shanghai, China), in opposition to the F12K medium (iCell, Shanghai, China) employed for AGS cells. Enrichment of these media involves incorporation of 10% fetal bovine serum (VivaCell, Shanghai, China).

RNA isolation and reverse transcription-quantitative polymerase chain reaction (RT-qPCR)

Extraction of RNA from entire cells (GES-1, AGS, HGC27) proceeded through application of TRIzol reagent (SparkJade, Jinan, China). The concentration and purity of RNA were measured using a Nanodrop spectrophotometer (Thermo Fisher Scientific, MA, USA). Utilization of a reverse transcription kit (SparkJade, Jinan, China) enabled implementation of reverse transcription. Employment of SYBR Green PCR Master Mix (Yesen, Shanghai, China) allowed RT-qPCR to occur via a quantitative PCR platform dependent upon fluorescence (Funglyn Biotech, Toronto, Canada). The relative mRNA expression levels of CACNA1I, PKD1L1, SLC24A2, EGF, LOX, AGT, CNGB3 and F5 were measured using β-actin as the internal control. Each experimental group consisted of three independent biological replicates, each with three further measurement replicates. Relative expression levels of the target genes were calculated using the 2^−ΔΔCq method (25). Primer sequences utilized throughout this examination appear within Table S1.

Western blotting

Proteins derived from entire cells underwent extraction through a lysis buffer incorporating protease inhibitors (Solarbio, Beijing, China). Quantification of protein levels proceeded via the BCA kit (Beyotime Biotechnology, Shanghai, China). SDS-PAGE gel electrophoresis facilitated separation of identical quantities from protein specimens. Specimens of protein embedded within the gel were transferred to a PVDF membrane (Merck Millipore, Darmstadt, Germany). At ambient temperature, blocking of the membrane with 5% non-fat milk lasted one hour, after which primary antibody incubation occurred overnight at 4 ℃. Next, secondary antibody incubation followed at room temperature over one hour. Protein bands underwent development employing ECL reagents (SparkJade, Jinan, China) inside the ChemiScope 6200 Touch Imaging System (CLINX, Shanghai, China). Table S2 lists the primary and secondary antibodies.

Cell transfection

Delivery of small interfering RNAs (siRNAs) specific to CNGB3 into HGC27 and AGS cells utilized Lipofectamine 3000 (Thermo Fisher Scientific, Waltham, USA) functioning as the transfection reagent. The sequences of siRNA were listed in Table S3. Assessments via Western blotting substantiated the transfection proficiency.

Cell Counting Kit-8 (CCK-8) assay and colony formation assay

Seeding of HGC27 or AGS cells took place within 96-well plates at densities reaching 1,000 cells/well during implementation of the CCK-8 evaluation. Incorporation of 10 µL from the CCK-8 reagent (Dojindo, Kumamoto, Japan) ensued across wells, with subsequent determination of absorbance values at 450 nm employing a microplate reader (Bio-Rad, California, USA) over the course of days 1, 2, 3, and 4.

Implementation of the colony formation evaluation involved positioning 1,000 HGC27 or AGS cells inside 6-well plates. A 10-day incubation preceded cellular fixation via 4% paraformaldehyde, after which 1% crystal violet enabled staining. Enumeration through ImageJ software followed photographic capture of cellular colonies.

Wound healing assay

Positioning of HGC27 or AGS cells took place inside 6-well plates before incubation proceeded until confluency approximated 90%. A pipette tip enabled the formation of a wound through scratching, after which photographic documentation occurred immediately and at 24 hours subsequent to injury induction. ImageJ software facilitated subsequent quantification regarding cellular migratory speeds.

Transwell invasion assay

Placement of 5×10⁴ HGC27 or AGS cells, which had undergone maintenance in medium lacking serum, took place within the upper compartment of a transwell setup precoated using Matrigel diluted 1 to 8 (BD Biosciences, San Diego, USA), as the lower compartment received incorporation of 10% fetal bovine serum. Incubation over 48 hours preceded the removal of lingering cells inside the upper compartment, but cells that traversed toward the lower face of the upper compartment experienced fixation through 4% paraformaldehyde alongside staining via 1% crystal violet, followed by microscopic capture and tallying through ImageJ software.

Detection of Ca²⁺ level

Cellular Ca²⁺ levels were measured using a Fluo-4 AM Fluorescence Calcium Ion Detection Kit (Servicebio, Wuhan, China). Initially, the cells were washed with PBS, followed by incubation with the Fluo-4 AM detection solution at 37 ℃ for 30 min. After a subsequent wash with PBS, the Ca²⁺ levels in the cells were examined under a laser confocal microscope. The fluorescence intensity was detected using ImageJ software.

Statistical analysis

Utilization of R software (version 4.0.5; https://www.r-project.org/) in conjunction with GraphPad Prism (version 9.0; https://www.graphpad.com/) enabled execution of every statistical analysis. The number of repeated experiments involved is 3. Designation of statistical significance occurred at *P<0.05, **P<0.01, or ***P<0.001.

Results

Identification of ion channel-related hub genes in GC

Differential gene expression analysis was performed by contrasting GC samples against nearby normal samples within the TCGA-STAD dataset, resulting in the identification of a gene set containing 12,727 DEGs (available online: https://cdn.amegroups.cn/static/public/tcr-2025-1-2874-1.xls) (|LogFC|>1, adjusted P<0.05). By searching the keyword “ion channels” on GeneCards database and filtering genes with a relevance greater than 6,1090 ICRGs were obtained (available online: https://cdn.amegroups.cn/static/public/tcr-2025-1-2874-2.xls). The intersection of two gene sets was analyzed through Venn diagram visualization, revealing the identification of 380 ion channel-related DEGs in GC (Figure 2A). On this basis, 11 genes associated with prognosis can be screened out through univariate Cox regression analysis (Figure 2B). Eight ion channel-related hub genes were ultimately identified through LASSO regression analysis (Figure 2C,2D).

Figure 2 Detection of hub genes associated with ion channels within GC. (A) Venn plot depicting discernment of ion channel-related DEGs. (B) Prognostic genes underwent selection via univariate Cox regression methodology. (C) Graphical representation for parameters in LASSO regression. (D) Plot illustrating coefficients from LASSO regression. CI, confidence interval; DEGs, differentially expressed genes; GC, gastric cancer; HR, hazard ratio; ICRGs, ion channel-related genes; LASSO, least absolute shrinkage and selection operator.

Construction and validation of an ICRG signature for predicting the prognosis of GC

Based on the coefficients of 8 ion channel-related hub genes, we developed a risk score for assessing the prognosis of GC patients. The formula of the scoring system is as follows: ICRGRS = 0.0260 × Exp (CACNA1I) + 0.1292 × Exp (PKD1L1) + 0.1452 × Exp (SLC24A2) + 0.1610 × Exp (EGF) + 0.1490 × Exp (LOX) + 0.0337 × Exp (AGT) + 0.1611 × Exp (CNGB3) + 0.0913 × Exp (F5). For evaluating the prognostic utility of this scoring approach, cases of GC sourced from the TCGA-STAD assembly and GSE62254 underwent division into low- and high-ICRGRS subsets employing the ICRGRS median threshold in each instance. Mortality figures proved greater within the high-ICRGRS subset than in the low-ICRGRS counterpart across GSE62254 plus the TCGA collection (Figure 3A,3B). Elevated quantities of transcripts for the eight key genes emerged in the high-ICRGRS subset relative to the low one inside the TCGA database (Figure 3A). Within GSE62254, quantities of transcripts belonging to AGT, F5, CNGB3, SLC24A2, EGF, and LOX registered as increased in the high-ICRGRS subset over the low-ICRGRS one. Nonetheless, PKD1L1 transcripts showed no difference. Transcripts of CACNA1I were lower (Figure 3B). Assessments concerning survival uncovered diminished rates of overall survival (OS) and disease-specific survival (DSS) for the high-ICRGRS subset in contrast to the low-ICRGRS one among TCGA-STAD subjects (Figure 3C,3D). Kaplan-Meier curves for disease-free interval (DFI) and progression-free interval (PFI) demonstrated that a higher risk score was significantly correlated with worse prognosis in TCGA-STAD patients (Figure 3E,3F). Survival analysis of GSE62254 showed that patients in the high-ICRGRS subset had a poorer prognosis (Figure 3G). ROC curves of TCGA-STAD and GSE62254 demonstrated that the risk score exhibited effective prognostic stratification capabilities for patients with GC (Figure 3H,3I).

Figure 3 Development and verification of a gene signature involving ICRGs for predicting GC prognosis. (A,B) Assessment contrasting gene expression quantities from eight hub genes alongside mortality across low- and high-ICRGRS subgroups within TCGA-STAD (A) as well as GSE62254 (B) datasets. (C-F) TCGA-STAD individuals categorized into high-ICRGRS alongside low-ICRGRS subgroups feature Kaplan-Meier plots depicting OS (C), DSS (D), DFI (E), and PFI (F). (G) Among individuals from GSE62254, evaluations regarding survival occurred for low-ICRGRS alongside high-ICRGRS subgroups. (H,I) Individuals from TCGA-STAD (H) and GSE62254 (I) featuring high-ICRGRS alongside low-ICRGRS subgroups display ROC plots. AUC, area under the curve; DSS, disease-specific survival: DFI; FPR, false positive rate; GC, gastric cancer; ICRGRS, ion channel-related gene risk score; ICRGs, ion channel-related genes; OS, overall survival; PFI, progression-free interval; ROC, receiver operating characteristic; TCGA-STAD, The Cancer Genome Atlas-Stomach Adenocarcinoma; TPR, true positive rate.

The ICRG signature is an independent prognostic factor of GC

Univariate together with multivariate Cox prognostic evaluations were carried out on cases drawn from TCGA-STAD to determine the independent prognostic capacity of ICRGRS in GC. Outcomes highlighted the autonomous prognostic functionality of this scoring system within GC contexts (Figure 4A,4B). A nomogram based on the risk score and clinical prognostic factors was established to predict the survival of GC patients (Figure 4C). ROC curves (Figure 4D), calibration curves (Figure 4E), and DCA (Figure 4F-4H) demonstrated that the nomogram possessed strong predictive performance for estimating 1-, 3-, and 5-year survival probabilities in patients with GC.

Figure 4 The ICRGRS is an independent prognostic factor of GC. (A,B) Among individuals sourced from TCGA-STAD, prognostic assessments took place via univariate (A) alongside multivariate (B) Cox regression techniques. (C) Nomogram of predicting the 1-, 3-, and 5-year survival probabilities in patients with GC. (D-H) Scrutiny of the nomogram’s forecasting competence involved ROC curves (D), calibration curves (E), and DCA (F-H). AUC, area under the curve; CI, confidence interval; DCA, decision curve analysis; FPR, false positive rate; GC, gastric cancer; HR, hazard ratio; ICRGRS, ion channel-related gene risk score; M, metastasis; N, node; ROC, receiver operating characteristic; T, tumor; TCGA-STAD, The Cancer Genome Atlas-Stomach adenocarcinoma; TPR, true positive rate.

Gene mutation analysis in GC patients with high- and low-ICRGRS

To explore the association between risk scores and tumor gene mutations in patients with GC, gene mutation analyses were performed separately in the high-ICRGRS and low-ICRGRS groups of TCGA-STAD patients. Findings highlighted the dominance of missense variants constituting the mutation form most commonly detected throughout the two categories (Figure 5A,5B). Additionally, single nucleotide polymorphisms (SNPs) represented the most frequently observed variant type in both groups (Figure 5A,5B). C>T transitions represented the most prevalent single nucleotide variations (SNVs) observed in both patient groups (Figure 5A,5B). However, notable differences were observed between the two groups with respect to the foremost five mutated genes. The results indicated that the top five mutated genes in the high-risk group were MUC16, TP53, TTN, LRP1B, and CSMD3, while in the low-risk group, the most frequently mutated genes were ARID1A, SYNE1, MUC16, TTN, and LRP1B (Figure 5A,5B). Notably, TP53 mutations were more prevalent in the high-risk group, whereas ARID1A mutations were more frequent in the low-risk group, suggesting that different mutation patterns may be associated with distinct prognostic outcomes.

Figure 5 Scrutiny of genetic mutations across GC-afflicted individuals categorized into low-ICRGRS alongside high-ICRGRS subgroups. (A) Examination involving mutations among genes in persons affected by GC who display high-ICRGRS. (B) Examination involving mutations among genes in persons affected by GC who display low-ICRGRS. GC, gastric cancer; ICRGRS, ion channel-related gene risk score.

Enrichment analysis of DEGs in GC patients with high- and low-ICRGRS

We conducted an analysis of DEGs among TCGA-STAD patients categorized into high-ICRGRS and low-ICRGRS subgroups, which resulted in discovering 831 DEGs (available online: https://cdn.amegroups.cn/static/public/tcr-2025-1-2874-3.xls). GO analysis indicated substantial enrichment of such genes within biological process (BP) categories including detection of chemical stimulus involved in sensory perception, detection of chemical stimulus involved in sensory perception of smell, cell-cell adhesion via plasma-membrane adhesion molecules, anion transmembrane transport, homophilic cell adhesion via plasma membrane adhesion molecules, and neuropeptide signaling pathway (Figure 6A). For cellular component (CC), predominant enrichment occurred in ion channel complex, transporter complex, cation channel complex, synaptic membrane, and transmembrane transporter complex (Figure 6A). Regarding molecular function (MF), primary enrichment was observed in glutamate receptor activity, cation channel activity, G protein-coupled peptide receptor activity, metal ion transmembrane transporter activity, ion channel activity, and passive transmembrane transporter activity (Figure 6A). Assessment through KEGG pathways for the DEGs showed that the leading ten pathways encompassed steroid hormone biosynthesis, chemical carcinogenesis-DNA adducts, metabolism of xenobiotics by cytochrome P450, complement and coagulation cascades, GABAergic synapse, glutamatergic synapse, retinol metabolism, bile secretion, retrograde endocannabinoid signaling, and neuroactive ligand-receptor interaction (Figure 6B). These findings suggest that the high-risk group may exhibit dysregulated ion transport and synaptic signaling, which could contribute to the malignant progression of GC.

Figure 6 Enrichment analysis of DEGs in GC patients with high- and low-ICRGRS. (A) GO assessments and (B) KEGG pathway evaluations applied to DEGs among GC individuals exhibiting high- and low-ICRGRS. BP, biological process; CC, cellular component; DEGs, differentially expressed genes; GC, gastric cancer; GO, Gene Ontology; ICRGRS, ion channel-related gene risk score; KEGG, Kyoto Encyclopedia of Genes and Genomes; MF, molecular function.

Immune infiltration analysis of GC patients in high- and low-ICRGRS groups

We assessed immune infiltration features by evaluating variations in immune cell infiltration across high-ICRGRS and low-ICRGRS subgroups in TCGA-STAD patients. Outcomes showed that infiltration levels for eosinophils, resting natural killer (NK) cells, memory resting CD4⁺ T cells, M2 macrophages, and monocytes appeared elevated within the high-ICRGRS subgroup, while CD8⁺ T cells, activated NK cells, and T follicular helper cells presented decreased infiltration in comparison with the low-ICRGRS subgroup (Figure 7A). When immune function profiles underwent assessment, type II interferon and para-inflammatory responses proved augmented in the high-ICRGRS group, but T cell co-stimulation and antigen-presenting cell co-inhibition activities emerged as heightened in the low-ICRGRS group (Figure 7B). Evaluation regarding immune checkpoints indicated that CEACAM19 and CD28 displayed notably increased levels in the high-ICRGRS group relative to the low-ICRGRS group (Figure 7C). For expression of genes linked to HLA-DRB4, HLA-DRB3, HLA-DQB2, HLA-G, and HLA-E exhibited considerably greater levels in the high-ICRGRS group than in the low-ICRGRS group (Figure 7D).

Figure 7 Immune infiltration analysis and immunotherapy response analysis of GC patients in high- and low-ICRGRS groups. (A) Examinations of infiltration via immune cells and (B) evaluations of immune functionalities for GC individuals within high-ICRGRS alongside low-ICRGRS categories from TCGA repositories. Expression of immune checkpoints (C) and HLA-related genes (D) in GC patients with high- and low-ICRGRS groups in TCGA database. (E) Contrasts involving IPS, (F) TIDE values, and (G) MSI signatures across GC individuals displaying low-ICRGRS alongside high-ICRGRS categories within TCGA repositories. ns, not significant; *, P<0.05; **, P<0.01; ***, P<0.001. APC, antigen-presenting cell; CCR, C-C chemokine receptor; GC, gastric cancer; HLA, human leukocyte antigen; ICRGRS, ion channel-related gene risk score; IFN, interferon; IPS, immunophenoscore; MHC, major histocompatibility complex; MSI, microsatellite instability; NK, natural killer; TCGA, The Cancer Genome Atlas; TIDE, Tumor Immune Dysfunction and Exclusion.

Immunotherapy response analysis of GC patients in different risk groups

To delve deeper into the relationship linking risk scores with immunotherapy outcomes, an IPS assessment was performed. Outcomes indicated that individuals within the low-ICRGRS subgroup displayed superior responsiveness toward inhibitors of PD-1 and CTLA-4 immune checkpoints (Figure 7E). When TIDE scores were examined, elevated values emerged in the high-ICRGRS subgroup (Figure 7F). Evaluation through MSI revealed greater MSI within the low-ICRGRS subgroup (Figure 7G). These findings suggest that those classified in the low-ICRGRS category derive enhanced benefits from immunotherapeutic approaches.

Drug sensitivity analysis of GC patients with different risk scores

We examined differences in responsiveness to anticancer medications among individuals categorized into high-ICRGRS and low-ICRGRS subgroups. Findings suggested that members of the low-ICRGRS subgroup could derive enhanced advantages from savolitinib, palbociclib, nilotinib, irinotecan, erlotinib, dactinomycin, oxaliplatin, camptothecin, and 5-fluorouracil relative to their high-ICRGRS counterparts (Figure 8A). In contrast with the low-ICRGRS subgroup, elevated susceptibility toward ZM447439, RO-3306, NU7441, BMS-754807, and JQ1 appeared in those from the high-ICRGRS category (Figure 8B).

Figure 8 Drug sensitivity analysis of GC patients with different risk scores. The drugs to which the low-ICRGRS (A) and high-ICRGRS (B) groups exhibited greater sensitivity. ns, not significant; *, P<0.05; **, P<0.01; ***, P<0.001; ****, P<0.0001. GC, gastric cancer; IC₅₀, half maximal inhibitory concentration; ICRGRS, ion channel-related gene risk score.

Knockdown of CNGB3 suppressed the proliferation, migration, and invasion of GC in vitro

We utilized RT-qPCR for assessing variations in expression among eight hub genes between gastric epithelial cells (GES-1) and GC cells (HGC27, AGS), with the goal of confirming ICRGs’ effects on GC within cellular settings. The findings revealed that SLC24A2, EGF, and LOX exhibited markedly higher expression in HGC27 cells than in GES-1 cells, while AGS cells did not display any significant change. Conversely, the expression of CACNA1I, PKD1L1, and AGT was notably lower in both AGS and HGC27 cell lines compared to GES-1. In AGS cells, F5 expression was enhanced, but no substantial alteration was seen in HGC27. Moreover, substantial elevation in CNGB3 expression occurred within HGC27 and AGS cells as opposed to GES-1 (Figure 9A). Since CNGB3 expression was elevated in both GC cell lines compared to gastric mucosal epithelial cells, we chose CNGB3 to examine its effect on GC cell function. siRNA directed against CNGB3 was utilized so that its role in GC cells could undergo additional scrutiny. Confirmation of CNGB3 knockdown took place via Western blotting in HGC-27 and AGS cells (Figure 9B,9C). The original images for western blots were in Figure S1. Colony formation along with CCK-8 assays indicated notable diminishment of proliferative potential in GC cells upon CNGB3 knockdown (Figure 9D-9G). As shown through wound healing assays, CNGB3 suppression caused considerable impairment to migratory properties of GC cells (Figure 9H,9I). Transwell invasion assays further confirmed a significant decrease in the invasion potential upon CNGB3 knockdown (Figure 9J,9K). Collectively, these findings indicate that CNGB3 knockdown inhibits the malignant phenotype of GC in vitro.

Figure 9 Knockdown of CNGB3 suppressed the proliferation, migration, and invasion of GC in vitro. (A) Analyses through RT-qPCR occurred for expression of eight hub genes associated with ion channels across cell lines including AGS, GES-1, and HGC27. (B,C) Verification of knockdown proficiency via western blotting. (D,E) CCK-8 evaluations alongside (F,G) colony formation assays were implemented to probe proliferative potentials in transfected GC cells (staining method: crystal violet staining; The image used for colony formation assays was taken from a six well plate and was not enlarged.). (H,I) Examinations via wound healing methodologies addressed migratory attributes of transfected GC cells. Scale bar: 200 μm (magnification: 40×). (J,K) Transwell invasion assays were conducted to analyze the invasion ability of transfected GC cells. Scale bar: 100 μm (staining method: crystal violet staining; magnification: 100×). ns, not significant; *, P<0.05; **, P<0.01; ***, P<0.001. CCK-8, Cell Counting Kit-8; GC, gastric cancer; RT-qPCR, reverse transcription-quantitative polymerase chain reaction.

Knockdown of CNGB3 suppressed the intracellular Ca²⁺ concentration and upregulated the expression of MHC-I in GC cells

As CNGB3 encodes a subunit of cyclic nucleotide-gated (CNG) ion channels, its knockdown may disrupt intracellular ion homeostasis, particularly Ca²⁺ flux (26). We measured cytosolic Ca²⁺ concentrations. As the results in Figure 10A,10B, it demonstrated that the Ca²⁺ concentrations decreased with the CNGB3 knocking down. To evaluate whether CNGB3 knockdown alters tumor immunogenicity, we have examined the expression of immune-related molecules, such as MHC-I on GC cells. Figure 10C-10E demonstrated that the CNGB3 knocking down could increase the expression of MHC-I. These investigations will help elucidate CNGB3 contributes to immune evasion and targeting it could synergize with immunotherapy.

Figure 10 Knockdown of CNGB3 suppressed the intracellular Ca²⁺ concentration and upregulated the expression of MHC-I in GC cells. (A,B) Intracellular Ca²⁺ levels following CNGB3 gene knockdown. Confocal illustration of intracellular Ca²⁺ imaging (A) and quantification (B). Scale bar: 100 μm (staining method: Fluo-4 AM staining; magnification: 100×). (C-E) Expression level of MHC-I molecules in cells after CNGB3 gene knockdown. **, P<0.01; ***, P<0.001. GC, gastric cancer; MHC-I, major histocompatibility complex class I.

Discussion

A prognostic model for risk evaluation in GC was constructed utilizing ICRGs that affect prognosis. External alongside internal validation procedures verified the predictive capability of this model. A multivariate analysis, incorporating the ICRGRS with clinical characteristics of GC patients, identified the ICRGRS as an independent predictor of GC outcomes. Furthermore, development of a nomogram took place for predicting survival likelihoods in patients by incorporating clinical variables along with ICRGRS, which additionally validated its robust forecasting potential.

CACNA1I, classified among the gated calcium ion channels, was previously documented to influence the aggressive development observed in multiple neoplasms (27). Whether CACNA1I plays a carcinogenic or tumor-suppressive role in breast cancer remains controversial (28,29). Consistent with our findings, previous study has shown that high expression of CACNA1I is associated with a poor prognosis in GC (30). PKD1L1, known as a component of a calcium-permeant ion channel, is involved in key cellular signaling processes (31). To date, no studies have been published regarding its potential role in tumorigenesis. SLC24A2, as a potassium-dependent sodium-calcium exchanger, has been found to be significantly upregulated and correlated with poor outcome in GC. Additionally, it may serve as a candidate hub gene that bridges type 2 diabetes and the development of cancer (32). Extensive reports have covered the participation of EGF within GC contexts. Downstream signaling pathways get triggered by EGF after attachment to EGFR as its receptor, resulting in augmented invasion, migration, and proliferation coupled with diminished apoptosis among GC cells (33). Additionally, EGF has been associated with the development of drug resistance (34). Currently, therapeutic agents targeting the EGF-EGFR axis are under active investigation. Close associations between heightened LOX expression and poor clinical prognoses alongside metastatic tumor events in GC have emerged from various investigations (35). LOX not only facilitates cancer cell dissemination but also enhances chemoresistance (36,37). The efficacy of LOX-targeted therapeutic agents is under evaluation. Various examinations have probed functions associated with AGT across GC scenarios (38,39). AGT induces advancement in metastasis alongside proliferation of cancer cells, but GC xenografts display augmented responsiveness to 5-fluorouracil following its knockdown (40). Association with adverse prognostic outcomes in GC pertains to CNGB3, responsible for encoding the CNG ion channel’s beta subunit (41). F5 overexpression has been linked to adverse prognosis in GC patients. A study suggests that F5 enhances tumor spread while simultaneously reducing immune cell infiltration by engaging the ITGA2-extracellular matrix signaling pathway (42). These genes formed the basis for the creation of a prognostic risk model in GC via our research. In addition, malignant traits within GC cells received augmentation from CNGB3, as evidenced by functional assays.

Our KEGG enrichment analysis of DEGs between the high- and low-ICRGRS groups revealed significant enrichment in pathways closely related to ion channel function and neuronal signaling, including neuroactive ligand-receptor interaction, GABAergic synapse, glutamatergic synapse, and retrograde endocannabinoid signaling. This suggests that the high-risk phenotype is not merely defined by the expression of individual ion channel genes but reflects a broader dysregulation of ion channel-mediated signaling networks. Notably, the enrichment of synaptic pathways in GC tissues points to a phenomenon of neuronal reprogramming or ion channel remodeling, which may drive tumor progression by modulating intracellular calcium flux and membrane potential.

We integrated mutation data to explore the genetic landscape associated with ICRGRS. The high-ICRGRS group exhibited a distinct mutation profile, with TP53, MUC16, TTN, LRP1B, and CSMD3 as the top five mutated genes. In contrast, the low ICRGRS-group showed a higher frequency of ARID1A mutations. TP53 mutations are well-established drivers of genomic instability and aggressive tumor behavior, and their enrichment in the high-risk group provides a genetic basis for the poor prognosis observed (43). ARID1A, a chromatin remodeling gene frequently mutated in MSI-high GC, is associated with better immunotherapy response and may partly explain the favorable immune profile of the low-risk group (44).

The TIME has been shown to play a critical role in tumor prognosis (45). Tumor-associated macrophages originate via differentiation from monocytes and penetrate into the tumor microenvironment (46). M2 macrophages facilitate progression in GC, aid in evading tumor immunity, and correlate with poor prognostic outcomes among GC-affected individuals, as evidenced by investigations (47,48), and their enrichment in the high-risk group is consistent with the poor prognosis observed in these patients. The transition between activation and rest of CD4⁺ memory T cells is an important part of the TIME. The infiltration of resting CD4⁺ memory T cells in GC portends a poorer prognosis (49). In non-lymphoid-derived solid organ tumors, an elevated frequency of T follicular helper cells is frequently associated with improved prognosis (50,51). Significant contributions to GC’s anti-tumor immune responses come from CD8⁺ T cells alongside NK cells (52-54). The accumulation of resting CD4⁺ memory T cells and resting NK cells, along with the reduction of activated CD8⁺ T cells and T follicular helper cells, further supports an immunosuppressive and functionally exhausted immune landscape in the high ICRGRS group. This immune profile is consistent with an “immune-desert” or “immune-excluded” phenotype, which is typically associated with poor response to immunotherapy. In contrast, the low ICRGRS group displayed features of a “hot” or inflamed tumor microenvironment, characterized by higher infiltration of activated immune cells and enhanced T cell co-stimulation activity. The high ICRGRS group showed significantly higher expression of CEACAM19 and CD28. CEACAM19, a member of the carcinoembryonic antigen family, has been implicated in immune evasion and may serve as a potential immune checkpoint. Notably, higher expression for HLA-DRB4, HLA-DRB3, HLA-DQB2, HLA-G, and HLA-E occurs in GC samples derived from the high-ICRGRS subgroup relative to the low-ICRGRS counterpart. HLA-G and HLA-E are known to play critical roles in immune evasion by inhibiting NK cell and T cell-mediated cytotoxicity, which may contribute to the immunosuppressive phenotype and poor prognosis in the high-risk group (55-57). The enrichment of these immunosuppressive cells and molecules in the high ICRGRS group aligns with its poorer survival. Nonetheless, clarification awaits regarding functions played by HLA-DRB4, HLA-DRB3, and HLA-DQB2 within GC. Suppression of anti-tumor immune activities potentially underlies the adverse prognostic influence exerted by prognostic genes associated with ion channels in GC, according to our supposition.

This study also compared the IPS between the high-ICRGRS group and the low-ICRGRS group. Markedly diminished IPS values emerged for the high-ICRGRS subgroup in contrast with the low-ICRGRS equivalent. Evaluation involving TIDE scores across both subgroups indicated augmented values within the high-ICRGRS category. Enhanced precision in forecasting outcomes for neoplastic cases treated via anti-CTLA-4 alongside anti-PD-1 approaches is shown by TIDE scores relative to tumor mutational burden (TMB) or PD-L1 concentrations (58). Inadequate reactions toward treatments blocking immune checkpoints manifest in those GC cases featuring augmented TIDE values (59,60). Positive reactions to therapies inhibiting immune checkpoints arise in neoplasms designated MSI-high when set against MSI-low variants (61,62). The high ICRGRS group exhibited higher TIDE scores and lower MSI, indicating a higher likelihood of immune escape and reduced responsiveness to immune checkpoint inhibitors. In contrast, the low ICRGRS group showed higher IPS and MSI, suggesting better potential responses to immunotherapy. These findings collectively indicate that the ICRGRS not only reflects an immunosuppressive TME but also serves as a predictive marker for immunotherapy response in GC patients.

Our drug sensitivity analysis revealed that the low ICRGRS group exhibited significantly lower half maximal inhibitory concentration (IC₅₀) values for several conventional chemotherapeutic agents, including oxaliplatin, 5-fluorouracil, irinotecan, and camptothecin. These agents are cornerstone drugs in GC chemotherapy. Oxaliplatin and 5-fluorouracil are components of the XELOX (capecitabine and oxaliplatin) and FOLFOX (folinic acid​, 5-fluorouracil​and ) regimens, which are standard first-line or adjuvant chemotherapy for advanced GC (63,64). Irinotecan and its active metabolite Camptothecin are used in second-line or later settings, often in combination with other agents (65). The finding that low ICRGRS patients are more sensitive to these drugs aligns with their better prognosis and suggests that the ICRGRS may help identify GC patients who are more likely to benefit from conventional chemotherapy. These findings support the potential utility of ICRGRS in guiding chemotherapy selection.

The high ICRGRS group showed increased sensitivity to several targeted agents, including ZM447439 (Aurora kinase inhibitor) (66), RO-3306 (CDK1 inhibitor) (67), NU7441 (DNA-PK inhibitor) (68), BMS-754807 (IGF-1R/IR inhibitor) (69), and JQ1 (BET inhibitor) (70). As a BET inhibitor, JQ1 has shown antitumor activity in GC preclinical models by suppressing MYC-driven oncogenic pathways (71). ZM447439 and RO-3306 target mitotic regulators (Aurora kinase A and CDK1), which are frequently overexpressed in aggressive GC subtypes (72,73). Preclinical studies have demonstrated that CDK1 inhibition suppresses GC proliferation and enhances chemosensitivity (73). As a DNA-PK inhibitor, NU7441 may potentiate the effects of DNA-damaging agents. However, its application in GC remains largely preclinical, and further validation is needed (74). BMS-754807 has shown antitumor activity in GC cell lines, but clinical development has been limited due to metabolic side effects and modest efficacy in unselected populations (75). These observations highlight the potential for ICRGRS to guide patient stratification in future clinical trials of targeted therapies.

CNGB3 encodes the beta subunit of CNG ion channels, which are critical for regulating calcium and sodium influx in response to intracellular cyclic nucleotides (26). Although CNG channels are best characterized in sensory transduction (76), emerging evidence suggests their expression in non-excitable cells, including cancer cells, where they may modulate calcium signaling, proliferation, and migration. In our study, knockdown of CNGB3 markedly suppressed GC cell proliferation, migration, invasion, and cytosolic Ca²⁺ concentrations, suggesting that CNGB3 may promote malignant phenotypes through ion channel-mediated signaling. Calcium signaling is a well-established driver of cancer cell proliferation and motility (77). These findings suggest that CNGB3 may contribute to sustained Ca²⁺ influx, thereby activating downstream pathways such as MAPK/ERK, or PI3K/AKT, which are known to promote GC progression. MHC-I serves as a vital target in immunotherapy since it might play a substantial role in activating T cells to fight against tumor cells (78). The down-regulation of MHC-I is of great significance in the immune escape of cancer cells (79). Tumors with high MHC-I expression demonstrate a more favorable prognosis (79). Our research indicates that the knockdown of CNGB3 in GC cells can up-regulate the expression of MHC-I, which suggests that CNGB3 may also influence the progression of GC by regulating immunity.

Certain constraints emerged within this investigation too. Construction of the scoring framework occurred based exclusively upon information sourced from public repositories, since integration of clinical specimens for sequencing-mediated confirmation was absent. Examination thereafter remained restricted toward CNGB3’s contribution to oncogenic characteristics exhibited by GC cells, but evaluations concerning effects on functions from the additional seven genes associated with prognosis were omitted. Complete delineation awaits regarding foundational molecular processes through which CNGB3 influences operations in GC cells.

Conclusions

In conclusion, this study established a risk model for predicting the prognosis of GC based on 8 ion channel-related prognostic genes. The model is expected to offer novel insights into the accurate prediction of GC outcomes and the potential effectiveness of immunotherapy. In addition, cellular experiments demonstrated that the knockdown of CNGB3 significantly inhibited the proliferation, migration, and invasion of GC cells, decreased the intracellular Ca²⁺ concentration, and upregulated the expression of MHC-I in these cells, which provided a novel possible target for the treatment of GC.

Acknowledgments

We would like to thank the TCGA and GEO databases.

Footnote

Reporting Checklist: The authors have completed the MDAR and TRIPOD reporting checklists. Available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-1-2874/rc

Data Sharing Statement: Available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-1-2874/dss

Peer Review File: Available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-1-2874/prf

Funding: This study was supported by National Natural Science Foundation of China (No. 81870453).

Conflicts of Interest: Both authors have completed the ICMJE uniform disclosure form (available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-1-2874/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Bray F, Laversanne M, Sung H, et al. Global cancer statistics 2022: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin 2024;74:229-63. [Crossref] [PubMed]
2. Sundar R, Nakayama I, Markar SR, et al. Gastric cancer. Lancet 2025;405:2087-102. [Crossref] [PubMed]
3. Wang G, Huang Y, Zhou L, et al. Immunotherapy and targeted therapy as first-line treatment for advanced gastric cancer. Crit Rev Oncol Hematol 2024;198:104197. [Crossref] [PubMed]
4. Yasuda T, Wang YA. Gastric cancer immunosuppressive microenvironment heterogeneity: implications for therapy development. Trends Cancer 2024;10:627-42. [Crossref] [PubMed]
5. Sexton RE, Al Hallak MN, Diab M, et al. Gastric cancer: a comprehensive review of current and future treatment strategies. Cancer Metastasis Rev 2020;39:1179-203. [Crossref] [PubMed]
6. Lopes C, Pereira C. Advances towards gastric cancer screening: Novel devices and biomarkers. Best Pract Res Clin Gastroenterol 2025;75:102009. [Crossref] [PubMed]
7. Luzko I, Moreira L, Bornschein J. Screening for and surveillance of premalignant conditions of the stomach. Best Pract Res Clin Gastroenterol 2025;75:101978. [Crossref] [PubMed]
8. Huo J, Wu L, Zang Y. Development and Validation of a Robust Immune-Related Prognostic Signature for Gastric Cancer. J Immunol Res 2021;2021:5554342. [Crossref] [PubMed]
9. Tang X, Yu Y, Liu N, et al. Identification of ferroptosis-related subtypes, characteristics of TME infiltration and development of prognostic models in gastric cancer. Int Immunopharmacol 2024;130:111610. [Crossref] [PubMed]
10. Luo J, Liang M, Ma T, et al. Identification of angiogenesis-related subtypes and risk models for predicting the prognosis of gastric cancer patients. Comput Biol Chem 2024;112:108174. [Crossref] [PubMed]
11. Lim HYG, Yada S, Murakami K, et al. AQP5: A functional gastric cancer stem cell marker in mouse and human tumors. Science 2025;390:eadr2428. [Crossref] [PubMed]
12. Huang DX, Zhou QZ, Luo HM, et al. PIEZO2 in tumors: from mechanobiological switches to activity-targeted therapies. J Exp Clin Cancer Res 2025;45:6. [Crossref] [PubMed]
13. Chen J, Zhang M, Ma Z, et al. Alteration and dysfunction of ion channels/transporters in a hypoxic microenvironment results in the development and progression of gastric cancer. Cell Oncol (Dordr) 2021;44:739-49. [Crossref] [PubMed]
14. Zhang WJ, Luo HL, Liu JP, et al. P2X7 receptor promotes the growth and metastasis of gastric cancer by activating P13/AKT/GSK-3 beta signaling (experimental research). Int J Surg 2025;111:3752-66. [Crossref] [PubMed]
15. Yang Y, Gu X, Weng W, et al. SUMOylation-induced membrane localization of TRPV1 suppresses proliferation and migration in gastric cancer cells. Cell Commun Signal 2024;22:465. [Crossref] [PubMed]
16. Zhang X, Zong R, Han Y, et al. Novel benzoylurea derivative decreases TRPM7 channel function and inhibits cancer cells migration. Channels (Austin) 2024;18:2396339. [Crossref] [PubMed]
17. Longo V, Mazzone P, Calice G, et al. CLIC2 regulates immunosuppression and macrophage differentiation in genomically stable gastric cancer. Biol Direct 2025;20:89. [Crossref] [PubMed]
18. Chen B, Liu X, Yu P, et al. H. pylori-induced NF-κB-PIEZO1-YAP1-CTGF axis drives gastric cancer progression and cancer-associated fibroblast-mediated tumour microenvironment remodelling. Clin Transl Med 2023;13:e1481.
19. Laurino S, Russi S, Sabato C, et al. The inhibition of SLC8A1 promotes Ca(2+)-dependent cell death in Gastric Cancer. Biomed Pharmacother 2025;182:117787. [Crossref] [PubMed]
20. Shiozaki A, Katsurahara K, Kudou M, et al. Amlodipine and Verapamil, Voltage-Gated Ca(2+) Channel Inhibitors, Suppressed the Growth of Gastric Cancer Stem Cells. Ann Surg Oncol 2021;28:5400-11. [Crossref] [PubMed]
21. Stokłosa P, Borgström A, Kappel S, et al. TRP Channels in Digestive Tract Cancers. Int J Mol Sci 2020;21:1877. [Crossref] [PubMed]
22. Luo Y, Liu X, Li X, et al. Identification and validation of a signature involving voltage-gated chloride ion channel genes for prediction of prostate cancer recurrence. Front Endocrinol (Lausanne) 2022;13:1001634. [Crossref] [PubMed]
23. Han Y, Shi Y, Chen B, et al. An ion-channel-gene-based prediction model for head and neck squamous cell carcinoma: Prognostic assessment and treatment guidance. Front Immunol 2022;13:961695. [Crossref] [PubMed]
24. Zhu Z, Lei Z, Qian J, et al. The Ion Channel-Related Gene Signatures Correlated With Diagnosis, Prognosis, and Individualized Treatment in Patients With Clear Cell Renal Cell Carcinoma. Front Pharmacol 2022;13:889142. [Crossref] [PubMed]
25. Livak KJ, Schmittgen TD. Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods 2001;25:402-8. [Crossref] [PubMed]
26. Gerhardt MJ, Priglinger SG, Biel M, et al. Biology, Pathobiology and Gene Therapy of CNG Channel-Related Retinopathies. Biomedicines 2023;11:269. [Crossref] [PubMed]
27. Phan NN, Wang CY, Chen CF, et al. Voltage-gated calcium channels: Novel targets for cancer therapy. Oncol Lett 2017;14:2059-74. [Crossref] [PubMed]
28. Ragab Ibrahim FAE, Naser Hussein ZU, Yousef AI, et al. Insights on possible interplay between epithelial-mesenchymal transition and T-type voltage gated calcium channels genes in metastatic breast carcinoma. Heliyon 2022;8:e10160. [Crossref] [PubMed]
29. Jayathirtha M, Neagu AN, Whitham D, et al. Investigation of the effects of downregulation of jumping translocation breakpoint (JTB) protein expression in MCF7 cells for potential use as a biomarker in breast cancer. Am J Cancer Res 2022;12:4373-98.
30. Wang CY, Lai MD, Phan NN, et al. Meta-Analysis of Public Microarray Datasets Reveals Voltage-Gated Calcium Gene Signatures in Clinical Cancer Patients. PLoS One 2015;10:e0125766. [Crossref] [PubMed]
31. Juan T, Ribeiro da Silva A, Cardoso B, et al. Multiple pkd and piezo gene family members are required for atrioventricular valve formation. Nat Commun 2023;14:214. [Crossref] [PubMed]
32. Zhang D, Qian C, Wei H, et al. Identification of the Prognostic Value of Tumor Microenvironment-Related Genes in Esophageal Squamous Cell Carcinoma. Front Mol Biosci 2020;7:599475. [Crossref] [PubMed]
33. Patrad E, Khalighfard S, Amiriani T, et al. Molecular mechanisms underlying the action of carcinogens in gastric cancer with a glimpse into targeted therapy. Cell Oncol (Dordr) 2022;45:1073-117. [Crossref] [PubMed]
34. Lin L, Li X, Pan C, et al. ATXN2L upregulated by epidermal growth factor promotes gastric cancer cell invasiveness and oxaliplatin resistance. Cell Death Dis 2019;10:173. [Crossref] [PubMed]
35. Jia Z, Gao J, Wang Y, et al. Clinicopathological and prognostic value of lysyl oxidase expression in gastric cancer: a systematic review, meta-analysis and bioinformatic analysis. Sci Rep 2022;12:16786. [Crossref] [PubMed]
36. Liu H, Sun X, Dong B, et al. Systematic Characterisation and Analysis of Lysyl Oxidase Family Members as Drivers of Tumour Progression and Multiple Drug Resistance. J Cell Mol Med 2025;29:e70536. [Crossref] [PubMed]
37. Nai A, Zeng H, Wu Q, et al. lncRNA/miR-29c-Mediated High Expression of LOX Can Influence the Immune Status and Chemosensitivity and Can Forecast the Poor Prognosis of Gastric Cancer. Front Cell Dev Biol 2021;9:760470. [Crossref] [PubMed]
38. Gao W, Yang M. Identification by Bioinformatics Analysis of Potential Key Genes Related to the Progression and Prognosis of Gastric Cancer. Front Oncol 2022;12:881015. [Crossref] [PubMed]
39. Jin P, Ji X, Bai J, et al. Identification of prognosis and therapy related intratumoral microbiome and immune signatures in gastric cancer. Front Immunol 2025;16:1622959. [Crossref] [PubMed]
40. Zhu L, Ma M, Zhang L, et al. System Analysis Based on Lipid-Metabolism-Related Genes Identifies AGT as a Novel Therapy Target for Gastric Cancer with Neoadjuvant Chemotherapy. Pharmaceutics 2023;15:810. [Crossref] [PubMed]
41. Chen S, Wang Z. Integration of mult-omics and nucleotide metabolism reprogramming signature analysis reveals gastric cancer immunological and prognostic features. Cancer Cell Int 2024;24:212. [Crossref] [PubMed]
42. Gao L, Han Q, Ma C, et al. Multi-omic analysis reveals the role of coagulation factor family genes and their predictive value for immune checkpoint inhibitors efficacy in gastric cancer. Int J Biol Macromol 2025;322:146600. [Crossref] [PubMed]
43. Cai HQ, Zhang LY, Fu LM, et al. Mutational landscape of TP53 and CDH1 in gastric cancer. World J Gastrointest Surg 2024;16:276-83. [Crossref] [PubMed]
44. Lu S, Duan R, Cong L, et al. The effects of ARID1A mutation in gastric cancer and its significance for treatment. Cancer Cell Int 2023;23:296. [Crossref] [PubMed]
45. Oya Y, Hayakawa Y, Koike K. Tumor microenvironment in gastric cancers. Cancer Sci 2020;111:2696-707. [Crossref] [PubMed]
46. Zhou Z, Yang Z, Wang J, et al. Research progress on tumour-associated macrophages in gastric cancer Oncol Rep 2021;45:35. (Review). [Crossref] [PubMed]
47. He Y, Hong Q, Chen S, et al. Reprogramming tumor-associated macrophages in gastric cancer: a pathway to enhanced immunotherapy. Front Immunol 2025;16:1558091. [Crossref] [PubMed]
48. Gambardella V, Castillo J, Tarazona N, et al. The role of tumor-associated macrophages in gastric cancer development and their potential as a therapeutic target. Cancer Treat Rev 2020;86:102015. [Crossref] [PubMed]
49. Sun Y, Liu L, Fu Y, et al. Metabolic reprogramming involves in transition of activated/resting CD4(+) memory T cells and prognosis of gastric cancer. Front Immunol 2023;14:1275461. [Crossref] [PubMed]
50. Gutiérrez-Melo N, Baumjohann D. T follicular helper cells in cancer. Trends Cancer 2023;9:309-25. [Crossref] [PubMed]
51. Cui C, Craft J, Joshi NS. T follicular helper cells in cancer, tertiary lymphoid structures, and beyond. Semin Immunol 2023;69:101797. [Crossref] [PubMed]
52. Xia M, Wang B, Wang Z, et al. Epigenetic Regulation of NK Cell-Mediated Antitumor Immunity. Front Immunol 2021;12:672328. [Crossref] [PubMed]
53. Sukri A, Hanafiah A, Kosai NR. The Roles of Immune Cells in Gastric Cancer: Anti-Cancer or Pro-Cancer? Cancers (Basel) 2022;14:3922. [Crossref] [PubMed]
54. Ding JT, Yang KP, Zhou HN, et al. Landscapes and mechanisms of CD8(+) T cell exhaustion in gastrointestinal cancer. Front Immunol 2023;14:1149622. [Crossref] [PubMed]
55. Morinaga T, Iwatsuki M, Yamashita K, et al. Dynamic Alteration in HLA-E Expression and Soluble HLA-E via Interaction with Natural Killer Cells in Gastric Cancer. Ann Surg Oncol 2023;30:1240-52. [Crossref] [PubMed]
56. Chen QY, Zhou WJ, Zhang JG, et al. Prognostic significance of the immune checkpoint HLA-G/ILT-4 in the survival of patients with gastric cancer. Int Immunopharmacol 2022;109:108798. [Crossref] [PubMed]
57. Vaquero-Yuste C, Juarez I, Molina-Alejandre M, et al. Hampered CD8 + ILT2 + T cell activation by HLA-G suggests a new immune checkpoint in gastric adenocarcinoma. Gastric Cancer 2026;29:83-96. [Crossref] [PubMed]
58. Jiang P, Gu S, Pan D, et al. Signatures of T cell dysfunction and exclusion predict cancer immunotherapy response. Nat Med 2018;24:1550-8. [Crossref] [PubMed]
59. Li X, Dong H, Chen L, et al. Identification of N7-methylguanosine related subtypes and construction of prognostic model in gastric cancer. Front Immunol 2022;13:984149. [Crossref] [PubMed]
60. Sun LD, Zhang LL, Wan Z, et al. TNFRSF12A expression in stomach adenocarcinoma and its preliminary role in predicting immunotherapy response. Front Immunol 2025;16:1578068. [Crossref] [PubMed]
61. Puliga E, Corso S, Pietrantonio F, et al. Microsatellite instability in Gastric Cancer: Between lights and shadows. Cancer Treat Rev 2021;95:102175. [Crossref] [PubMed]
62. Wu X, Gu Z, Chen Y, et al. Application of PD-1 Blockade in Cancer Immunotherapy. Comput Struct Biotechnol J 2019;17:661-74. [Crossref] [PubMed]
63. Li N, Qiu M, Zhang Y, et al. A randomized phase 2 study of HLX22 plus trastuzumab biosimilar HLX02 and XELOX as first-line therapy for HER2-positive advanced gastric cancer. Med 2024;5:1255-1265.e2. [Crossref] [PubMed]
64. Ramaswamy A, Bhargava P, Dubashi B, et al. Docetaxel-oxaliplatin-capecitabine/5-fluorouracil (DOX/F) followed by docetaxel versus oxaliplatin-capecitabine/5-fluorouracil (CAPOX/FOLFOX) in HER2-negative advanced gastric cancers. JNCI Cancer Spectr 2024;8:pkae054. [Crossref] [PubMed]
65. Sakai D, Kadowaki S, Kawabata R, et al. Randomized Phase III Trial of Ramucirumab Beyond Progression Plus Irinotecan in Patients With Ramucirumab-Refractory Advanced Gastric Cancer: RINDBeRG Trial. J Clin Oncol 2025;43:2196-207. [Crossref] [PubMed]
66. Josen T, Yuki R, Saito Y, et al. The Aurora B inhibitor ZM-447439 induces caspase-independent necrosis-like death in v-Src oncogene-expressing cells via accumulation of extra-lysosomal cathepsin B. Exp Cell Res 2026;454:114840. [Crossref] [PubMed]
67. Huang Y, Fan Y, Zhao Z, et al. Inhibition of CDK1 by RO-3306 Exhibits Anti-Tumorigenic Effects in Ovarian Cancer Cells and a Transgenic Mouse Model of Ovarian Cancer. Int J Mol Sci 2023;24:12375. [Crossref] [PubMed]
68. Luo C, Fang C, Zou R, et al. Hyperglycemia-induced DNA damage response activates DNA-PK complex to promote endothelial ferroptosis in type 2 diabetic cardiomyopathy. Theranostics 2025;15:4507-25. [Crossref] [PubMed]
69. de Billy E, Pellegrino M, Orlando D, et al. Dual IGF1R/IR inhibitors in combination with GD2-CAR T-cells display a potent anti-tumor activity in diffuse midline glioma H3K27M-mutant. Neuro Oncol 2022;24:1150-63. [Crossref] [PubMed]
70. Yasukawa Y, Hattori N, Iida N, et al. SAA1 is upregulated in gastric cancer-associated fibroblasts possibly by its enhancer activation. Carcinogenesis 2021;42:180-9. [Crossref] [PubMed]
71. Ba M, Long H, Yan Z, et al. BRD4 promotes gastric cancer progression through the transcriptional and epigenetic regulation of c-MYC. J Cell Biochem 2018;119:973-82. [Crossref] [PubMed]
72. Lv C, Liu Y, Zhang H. The Oncogenic Role of AURKA in Gastric Cancer: Mechanisms, Pathways, and Clinical Relevance. Carcinogenesis 2025;bgaf054.
73. Al-Mathkour M, Chen Z, Poveda J, et al. CDK1 drives SOX9-mediated chemotherapeutic resistance in gastric cancer. J Exp Clin Cancer Res 2025;44:284. [Crossref] [PubMed]
74. Geng W, Tian D, Wang Q, et al. DNA-PKcs inhibitor increases the sensitivity of gastric cancer cells to radiotherapy. Oncol Rep 2019;42:561-70. [Crossref] [PubMed]
75. Carboni JM, Wittman M, Yang Z, et al. BMS-754807, a small molecule inhibitor of insulin-like growth factor-1R/IR. Mol Cancer Ther 2009;8:3341-9. [Crossref] [PubMed]
76. Dai G, Varnum MD. CNGA3 achromatopsia-associated mutation potentiates the phosphoinositide sensitivity of cone photoreceptor CNG channels by altering intersubunit interactions. Am J Physiol Cell Physiol 2013;305:C147-59. [Crossref] [PubMed]
77. Romito O, Guéguinou M, Raoul W, et al. Calcium signaling: A therapeutic target to overcome resistance to therapies in cancer. Cell Calcium 2022;108:102673. [Crossref] [PubMed]
78. Ke ZB, Chen JY, Xue YT, et al. Mechanical signal modulates prostate cancer immune escape by USP8-mediated ubiquitination-dependent degradation of PD-L1 and MHC-1. Cell Death Dis 2025;16:413. [Crossref] [PubMed]
79. Wang H, Luo X, Yang B, et al. XPR1 promotes ovarian cancer growth and regulates MHC-I through autophagy. Genes Dis 2025;12:101507. [Crossref] [PubMed]