TIMER (http://timer.comp-genomics.org/) is a comprehensive resource for systematic analysis of immune infiltrates across diverse cancer types. This platform provides four modules for investigating the association between the immune infiltrates and genetic or clinical features and four modules for exploring cancer-related associations in The Cancer Genome Atlas (TCGA) cohorts (Li B. et al., 2016; Li et al., 2017; Li et al., 2020).

We used the Gene_DE module to investigate the mRNA expression of TRPM2 in different cancer types and matched normal tissues across all TCGA cancer types. The level of TRPM2 was expressed as log2 TPM, and the significance was determined by differential gene expression analysis using the edgeR software package.

We also used this platform to evaluate the correlation of the expression level of TRPM2 with tumor purity and tumor infiltration of B cells, CD4⁺ T cells, CD8⁺ T cells, neutrophils, macrophages, and dendritic cells. A heatmap of Spearman’s correlations between the level of TRPM2 and immune cell infiltration across diverse cancer types was also generated using the TIMER database. Furthermore, the correlation between TRPM2 expression and the gene markers of immune cells identified the potential subtypes of infiltrating immune cells.

GEPIA (http://gepia.cancer-pku.cn/) is an interactive web server for analyzing the RNA sequencing expression data of 9,736 tumors and 8,587 normal samples from TCGA and the Genotype-Tissue Expression (GTEx) project, respectively (Tang et al., 2017). The GEPIA database was utilized to evaluate the tumor/normal differential expression levels of TRPM2 and to conduct patient survival analysis based on the expression levels of TRPM2 across various cancer types. The GEPIA database utilizes log2(TPM + 1) for log scale with a |Log2FC| cutoff of 1 and a p-value cutoff of 0.01 in determining the differential expression of TRPM2, shown in Figure 1B.

The Kaplan–Meier plotter (http://kmplot.com/analysis/) can assess the effect of 54 k genes (mRNA, miRNA, and protein) on survival in 21 cancer types and discover and validate survival biomarkers (Gyorffy, 2021). The databases Gene Expression Omnibus (GEO), European Genome–Phenome Archive (EGA), and TCGA were used to assess the correlation between clinical outcomes and TRPM2 expression in different cancers and between TRPM2 expression and immune cell infiltration. We also calculated the hazard ratios (HRs) of 95% confidence intervals (CIs) and the log-rank p-value.

The TISIDB database (http://cis.hku.hk/TISIDB/index.php) is a portal for analyzing tumor and immune cell interactions that integrates multiple heterogeneous data types (Ru et al., 2019). Next, we analyzed the correlation between TRPM2 expression and clinical features, such as overall survival (OS), cancer stage, and cancer grade. Heatmap and dot plots showing the Spearman’s correlations between the expression of TRPM2 and the abundance of tumor immune-infiltrating cells across diverse cancer types were also generated using the TISIDB database.

The UALCAN database (http://ualcan.path.uab.edu/index.html) is available for online analysis of cancer omics data (TCGA, MET500, and CPTAC). This database allows users to identify biomarkers, perform pan-cancer gene expression analysis, obtain patient survival information, and analyze epigenetic regulation of gene expression (Chandrashekar et al., 2017). We used this database to validate the results of the correlation between TRPM2 expression and clinical features in the TISIDB database. A p < 0.05 indicated statistically significant differences.

The expression of TRPM2 was analyzed via the TIMER and GEPIA databases. The correlation between TRPM2 expression and survival prognosis, including OS, disease-free survival (DFS), and relapse-free survival (RFS) in pan-cancer, were analyzed in the GEPIA database and the Kaplan–Meier plotter platform. To compare the survival curves, we used the log-rank test to calculate the HRs, 95% CIs, and p-values in the Kaplan–Meier plotter and GEPIA. We used Spearman’s correlation analysis to evaluate the correlation between gene expression and immune cell infiltration in the TIMER and TISIDB databases. A p < 0.05 was considered statistically significant.

The mRNA level TRPM2 was analyzed in the TIMER (Figure 1A) and GEPIA databases (Figures 1B, C). The results showed that TRPM2 was highly expressed in a majority of cancers (Figure 1A). We also used the GEPIA database to validate the findings in the TIMER database and found that, compared to the corresponding normal tissues, the mRNA level of TRPM2 was significantly higher in most human tumors, including bladder urothelial carcinoma (BLCA), breast invasive carcinoma (BRCA), cervical squamous cell carcinoma and endocervical adenocarcinoma (CESC), cholangiocarcinoma (CHOL), colon adenocarcinoma (COAD), esophageal carcinoma (ESCA), head and neck squamous cell carcinoma (HNSC), kidney renal clear cell carcinoma (KIRC), kidney renal papillary cell carcinoma (KIRP), acute myeloid leukemia (LAML), ovarian serous cystadenocarcinoma (OV), pancreatic adenocarcinoma (PAAD), rectum adenocarcinoma (READ), skin cutaneous melanoma (SKCM), stomach adenocarcinoma (STAD), uterine corpus endometrial carcinoma (UCEC), and uterine carcinosarcoma (UCS) (Figures 1B, C). Moreover, the level of TRPM2 was significantly lower in glioblastoma multiforme (GBM) and brain lower-grade glioma (LGG) than that in normal tissue, while no significant differences were detected in sarcoma (SARC), testicular germ cell tumors (TGCT), adrenocortical carcinoma (ACC), diffuse large B-cell lymphoma (DLBC), kidney chromophobe (KICH), liver hepatocellular carcinoma (LIHC), lung adenocarcinoma (LUAD), pheochromocytoma and paraganglioma (PCPG), thyroid carcinoma (THCA), and thymoma (THYM) (Figures 1B, C).

Next, we investigated the prognostic value of TRPM2 in different cancers using GEPIA and the Kaplan–Meier plotter platform. In GEPIA, the Kaplan–Meier plots revealed that a high TRPM2 level was associated with a short OS and/or DFS in cancers, including KIRC (OS: HR = 1.7, p = 0.00048; DFS: HR = 1.1, p = 0.46), LGG (OS: HR = 1.6, p = 0.0069; DFS: HR = 1.4, p = 0.04), OV (OS: HR = 1.4, p = 0.005; DFS: HR = 1.2, p = 0.2), THYM (OS: HR = 11, p = 0.0063; DFS: HR = 2, p = 0.14), uveal melanoma (UVM) (OS: HR = 2.6, p = 0.036; DFS: HR = 1.8, p = 0.2), and LIHC (OS: HR = 1.4, p = 0.04; DFS: HR = 1, p = 0.87) (Figure 2).

Furthermore, using the Kaplan–Meier plotter, we found that the expression of TRPM2 was negatively correlated with prognosis in KIRC (OS: HR = 1.7, p = 0.00056; RFS: HR = 0.38, p = 0.19), OV (OS: HR = 1.51, p = 0.0024; RFS: HR = 0.81, p = 0.32), and KIRP (OS: HR = 0.4, p = 0.015; RFS: HR = 2.59, p = 0.0098) (Supplementary Figure S1). These findings indicated that TRPM2 was a prognostic factor for short OS in KIRC. In contrast, the expression level of TRPM2 was positively correlated with the prognosis of BLCA, READ, UCES, THCA, and STAD, as predicted by GEPIA and the Kaplan–Meier plotter (Supplementary Figure S2).

In order to examine the association between TRPM2 and clinical features, we assessed the multiple clinical prognostic values of TRPM2 in various cancers using the TIDISB database. Among the human cancers, KIRC was the most correlated cancer with regard to OS (Figure 3A), stage (Figure 3B), and tumor grade (Figure 3C). The association plots derived from the 533 KIRC cases in TCGA showed that the upregulation of TRPM2 expression was significantly associated with shorter OS, poor pathological stage, and tumor grade in KIRC (Figure 3D–F, respectively). Using the UALCAN database, we further validated the finding that the expression level of TRPM2 was negatively correlated with the prognosis of KIRC patients, in consideration of the tumor subtype (Figure 3G), cancer stage (Figure 3H), nodal metastasis (Figure 3I), and tumor grade (Figure 3J). The results demonstrated a significant correlation between high TRPM2 levels and these adverse clinicopathological parameters, which was consistent with those described above (Figure 2A). Therefore, we mainly focused on the function of TRPM2 in KIRC to examine its prognostic performance, clinical significance, and mechanism.

To elucidate whether TRPM2 is an independent risk factor for clinical outcomes of KIRC patients, univariate and multivariate Cox analyses were performed. In the univariate Cox analysis, the pathological stage, age, histologic grade, serum calcium, and TRPM2 expression were significantly correlated with OS (p < 0.001, p < 0.001, p < 0.001, p < 0.001, and p = 0.005, respectively) (Figure 4A). To exclude the confounder effect, the significant prognostic factors detected in the univariate analysis (Figure 4A) were evaluated further in the multivariate analysis (Figure 4B). Strikingly, multivariate Cox analysis revealed that the pathological stage, age, serum calcium, and TRPM2 expression were independent risk factors for OS (p = 0.039, p = 0.034, p = 0.023, and p = 0.001, respectively) (Figure 4B) of KIRC patients. Since calcium influx plays a critical role in tumor occurrence and development through TRPM2 channels, we considered serum calcium for the above Cox analysis. Interestingly, both univariate and multivariate analyses verified that serum calcium was an independent risk factor for the OS of KIRC patients (HR = 2.329, 95% CI = 1.125–4.824, p = 0.023) (Figure 4B). Based on the clinicopathological features and TRPM2, we also constructed a nomogram to predict the 1-, 5-, and 10-year OS rates using the Cox regression algorithm (Supplementary Figure S3).

The features of TIME are emerging as critical biomarker in predicting the prognosis of patients and the efficacy of treatment. Recent studies have suggested a high immune infiltration level in KIRC with the highest overall T cells, CD8⁺ T cells, T helper 1 (Th1) cell, dendritic cells, neutrophils, and cytotoxic cells (Vuong et al., 2019). Therefore, we analyzed the correlation of the level of TRPM2 with the immune infiltration level in various cancer types. The results showed that the expression of TRPM2 was significantly positively correlated with B cells (r = 0.277, p = 1.51e−09), CD8⁺ T cells (r = 0.226, p = 1.84e−06), CD4⁺ T cells (r = 0.438, p = 6.24e−23), macrophages (r = 0.459, p = 8.76e−25), neutrophils (r = 0.502, p = 1.22e−30), and dendritic cells (r = 0.518, p = 1.32e−32) in KIRC (Figure 5A). Moreover, the level of TRPM2 showed a positive correlation with the infiltration of B cells (r = 0.152, p = 4.66e−02), CD8⁺ T cells (r = 0.138, p = 7.28e−02), CD4⁺ T cells (r = 0.397, p = 9.28e−08), macrophages (r = 0.393, p = 1.02e−07), neutrophils (r = 0.413, p = 1.93e−08), and dendritic cells (r = 0.395, p = 8.66e−08) in PAAD (Figure 5A). However, TRPM2 was not correlated with B cells (r = 0.023, p = 6.63e−01), CD8⁺ T cells (r = −0.049, p = 3.49e−01), CD4⁺ T cells (r = −0.096, p = 6.73e−02), macrophages (r = 0.031, p = 5.58e−01), neutrophils (r = −0.118, p = 2.49e−02), and dendritic cells (r = −0.042, p = 4.21e−01) in BLCA (Figure 5A).

Then, the TISIDB database was used to further explore the correlations between the level of TRPM2 and the 28 tumor immune-infiltrating cell subtypes. The results showed that the level of TRPM2 was associated with 25 immune cell subtypes in KIRC (Figure 5B). Notably, myeloid-derived suppressor cells (MDSCs) (r = 0.843, p < 2.2e−16), T follicular helper (Tfh) cells (r = 0.74, p < 2.2e−16), and macrophages (r = 0.689, p < 2.2e−16) displayed relative strong correlation with TRPM2 expression in KIRC (Figure 5C). Others, such as immature B cells (r = 0.667, p < 2.2e−16), Th1 (r = 0.66, p < 2.2e−16), effector memory CD8⁺ T cells (r = 0.637, p < 2.2e−16), activated dendritic cells (r = 0.641, p < 2.2e−16), activated CD4 T cells (r = 0.546, p < 2.2e−16), Tregs (r = 0.617, p < 2.2e−16), activated B cells (r = 0.64, p < 2.2e−16), natural killer (NK) cells (r = 0.532, p < 2.2e−16), NK T cells (r = 0.584, p < 2.2e−16), and mast cells (r = 0.539, p < 2.2e−16), were moderately correlated with TRPM2 (Figure 5B). In addition, TRPM2 showed a weak correlation with immune cells of activated CD8⁺ cells (r = 0.198, p = 5.68e−05), CD56dim (r = 0.161, p = 0.00112), and CD56bright (r = 0.133, p = 0.00737) in BLCA (Figure 5D). In addition, the correlation between the level of TRPM2 and the degree of immune infiltration in various cancers were analyzed using the TIMER database and displayed as a heatmap (Supplementary Figure S4). These findings strongly indicated that TRPM2 serves as a major tumor immune infiltration regulator in KIRC.

We assessed the correlation between the expression of TRPM2 and the level of tumor-infiltrating immune cell gene markers in KIRC and BLCA using the TIMER database. The results showed that the level of TRPM2 in KIRC tissues was strongly associated with immune markers of B cells, CD8⁺ T cells, dendritic cells, M1/M2 macrophages, monocytes, neutrophils, general T cells, T-cell exhaustion, tumor-associated macrophages (TAMs), Th1, Th2, and Tregs, but not with NK, Tfh, and Th17 cells (Table 1). A major factor determining tumor progression over time is the overall proportion and property of T cells within the TIME (Binnewies et al., 2018). Notably, the level of TRPM2 was significantly correlated with various subtypes of T cells based on cell markers, including CD8⁺ T-cell markers (CD8A and CD8B), T-cell (general) markers (CD3D, CD3E, and CD2), exhausted T-cell markers (CTLA4, GZMB, LAG-3, and PDCD 1), Th1 markers (TBX21, STAT4, STAT1, and IFNG), Th2 markers (GATA3 and STAT5A), Treg markers (FOXP3 and CCR8), and with neutrophil markers (ITGAM and CCR7), dendritic cell markers (ITGAX, HLA-DPA1, HLA-DRA, and HLA-DPB1), and B-cell markers (CD79A and CD19) in KIRC (Table 1). Also, a significant correlation was established between the level of TRPM2 and the expressions of marker genes in different subsets of macrophages, including M1 macrophage markers (IRF5), M2 macrophage markers (MS4A4A, VSIG4, and CD163), and TAMs (IL10 and CD68), in KIRC (Table 1). However, only nine immune cell marker genes were significantly associated with the levels of TRPM2 in BLCA. Furthermore, the expression of TRPM2 was not markedly related to most marker genes of CD8⁺ T cells, NK cells, Th2, and Th17 cells in BLCA. These findings revealed that TRPM2 is involved in the regulation of tumor immune infiltration in KIRC.

TRPM2 is involved in the regulation of tumor immune infiltration in KIRC (Figure 5 and Table 1). Moreover, upregulated TRPM2 indicated poor prognosis in KIRC/PAAD patients, but better prognosis in BLCA patients. Thus, we hypothesized that TRPM2 may affect the prognosis of patients partially through the regulation of immune infiltration. Analysis of the Kaplan–Meier plotter revealed that the level of TRPM2 was associated with OS in KIRC, PAAD, and BLCA patients based on the enrichment of different immune cells. In KIRC, a high TRPM2 level in enriched CD4⁺ memory T cells, CD8⁺ T cells, macrophages, and Tregs indicated poor prognosis of patients with KIRC (Figure 6A). However, no significant correlation was established between the expression of TRPM2 and OS in enriched B cells, mesenchymal stem cells, Th1 cells, and Th2 cells (Figure 6A). On the other hand, in PAAD, a high expression of TRPM2 was related to poor prognosis only in CD8⁺ T cells and enriched B cells, but showed no statistically significant difference (Figure 6B). The clinical samples for Th1 cell enrichment were too few to analyze in PAAD. In BLCA, low-TRPM2 expression groups showed short OS in enriched CD4⁺ memory T cells, CD8⁺ T cells, Tregs, mesenchymal stem cells, NK T cells, Th2 cells, Th1 cells, B cells, and macrophages (Figure 6C). The above results suggested that TRPM2 may affect the prognosis of patients by potentially regulating the infiltration of immune cells in KIRC.

To gain in-depth insights into the molecular mechanism of TRPM2 as a prognostic marker in KIRC, the gene expression correlation was calculated with Pearson’s correlation analysis to reveal the genes associated with TRPM2 in KIRC patients. Subsequently, we conducted GO and KEGG analysis of the TRPM2-related genes in KIRC. Pearson’s correlation coefficients between TRPM2 and the genes of interest were calculated, and only genes with |r| > 0.5 and p < 0.05 were subjected to GO classification and KEGG pathway enrichment. The top 5 GO categories are shown in Figure 7. In KIRC, the above genes were enriched in several biological processes, such as T-cell activation, regulation of T-cell activation, lymphocyte activation, leukocyte cell–cell adhesion, and lymphocyte differentiation. For molecular function (MF), these genes were mainly associated with cytokine receptor activity, MHC protein binding, MHC protein complex binding, IgG binding, and cytokine binding. Figure 7 also shows the most significant KEGG pathways, namely, hematopoietic cell lineage, Staphylococcus aureus infection, leishmaniasis, osteoclast differentiation, and cell adhesion molecules. In addition, GO cell component (CC) analysis indicated that TRPM2 plays a key role in the regulation of T-cell activation on the external side of the plasma membrane, secretory granule membrane, tertiary granule membrane, plasma membrane receptor complex, and tertiary granule membrane in KIRC.

Conversely, in BLCA, TRPM2-correlated genes (|r| > 0.5 and p < 0.05) failed to enrich in the GO and KEGG analyses. We expanded the range of TRPM2-correlated genes (|r| > 0.3 and p < 0.05) and found that only the CC was enriched, indicating that TRPM2 may function mainly in the preribosome, lamellipodium, large subunit precursor of preribosome, and tertiary granule membrane (Supplementary Figure S5).