TNFAIP8L2- associated gene expression analysis data
The TIMER2 database was used to evaluate TNFAIP8L2 expression in TCGA samples from various kinds of cancers. As shown in Figure 1a, TNFAIP8L2 was expressed higher in breast invasive carcinoma (BRCA), esophageal carcinoma (ESCA, P < 0.001), glioblastoma multiforme (GBM, P < 0.01), head and neck squamous cell carcinoma (HNSC), kidney renal clear cell carcinoma (KIRC), kidney renal papillary cell carcinoma (KIRP), gastric carcinoma (STAD), and thyroid carcinoma (THCA) than in the corresponding control organizations. In contrast, colon adenocarcinoma (COAD, P < 0.001), lung adenocarcinoma (LUAD), lung squamous cell carcinoma (LUSC), pancreatic adenocarcinoma (PAAD, P < 0.01), and rectal adenocarcinoma (READ) were all lower than the corresponding control organizations.
As some TCGA databases lack control data for normal tissues, we further assessed differences in TNFAIP8L2 expression between tumor and normal tissues after integrating normal tissues from the GTEx dataset. The expression levels of acute myeloid leukemia (LAML), LGG, ovarian serous cystadenocarcinoma (OV), skin cutaneous melanoma (SKCM), and testicular germ cell tumor (TGCT) tissues were greater than those of normal tissues (Figure 1b, p < 0.05). However, we found no substantial alterations in other tumors, such as adrenocortical carcinoma (ACC), lymphoid neoplasm diffuse large B cell lymphoma (DLBC), sarcoma (SARC) and uterine carcinosarcoma(UCS).
In addition, the CPTAC dataset was used to examine TNFAIP8L2 protein levels in different tumors and normal tissues. There was no difference in TNFAIP8L2 total protein levels between breast, ovarian, colon, and lung adenocarcinomas and normal tissues. Protein expression was considerably higher in advanced tumor tissues of clear cells such as renal clear cell carcinoma (RCC), endometrial carcinoma (UCEC), HNSC, PAAD, and GBM (Figure 1c, all P < 0.001,) but significantly lower in hepatocellular carcinoma (HCC) tumor tissues (Figure 1c, P < 0.001) than in normal tissues.
In addition, the GEPIA2 tool indicated a correlation only between TNFAIP8L2 expression and the pathological classification of carcinomas such as SKCM, THCA, STAD, KIRC, kidney chromophobe (KICH), ESCA (Figure 1d, all P < 0.05), and not seen otherwise.
Survival analysis data
Using the TCGA and GEO databases, cancer cases were divided into low- and high-expression groups according to their TNFAIP8L2 expression profiles, and the relationship between TNFAIP8L2 expression and prognosis was explored in patients with diverse tumors. In the TCGA dataset, increased TNFAIP8L2 expression was connected with a poor prognosis in terms of total OS for LGG (P = 0.0036) and UVM (P = 0.0068) tumors, as shown in Figure 2a. The DFS analysis results (Figure 2b) revealed a link between high TNFAIP8L2 expression and poor prognosis in PRAD (P = 0.0074) among TCGA cases. Notably, decreased TNFAIP8L2 gene expression was correlated with a poor OS prognosis in BRCA (P = 0.0027), CESC (P = 0.0095), DLBC (P = 0.049), LUAD (P = 0.01), SARC (sarcoma) (P = 0.023), and SKCM (P = 0.0015) (Figure 2a), as well as with a poor DFS prognosis in ACC (P = 0.023) and CHOL (P = 0.025). (Figure 2b).
Genetic alteration analysis data
The cBioPortal tool was used to determine the status of TNFAIP8L2 gene alterations in different tumor tissues from the TCGA dataset. According to Figure 3a, the greatest prevalence of TNFAIP8L2 variations (about 11%) was observed in individuals with a predominance of LIHC “amplification.” All information about the kinds, locations, and case numbers of TNFAIP8L2 genetic alteration—including details regarding missense, truncating, and fusion mutations—is presented in Figure 3b. Two instances of stomach cancer and one of colorectal cancer were found at the R53Q/Q55Pfs*29 locus (Figure 3b). Missense mutations were the main type of TNFAIP8L2 gene mutations. The R53Q/Q55Pfs*29 site was observed in the 3D structure of the TNFAIP8L2 protein (Figure. 3c). Furthermore, we investigated a possible link between TNFAIP8L2 gene variations and survival prognosis in several cancer types. In BRCA, patients with TNFAIP8L2 gene alterations had DFS than those without gene alterations (P = 0.0291), but without significant differences in OS (P = 0.570), DSS (P = 0.905), or PFS (P = 0.381, Figure 3d).
DNA methylation analysis
Abnormal DNA methylation is closely correlated with tumorigenesis, development, and cellular carcinogenesis. Moreover, alterations in DNA methylation levels and the degree of methylation of specific genes can be used as diagnostic indicators of tumors (16). The DNA methylation levels of TNFAIP8L2 were compared between normal and primary tumor tissues using the UALCAN and TCGA databases. According to the UALCAN database, TNFAIP8L2 methylation expression levels were considerably lower in BLCA, BRCA, CHOL, KIRP, KIRC, LIHC, LUAD, LUSC, PCPG, PAAD, PRAD, READ, and UCEC tumor tissues than in normal tissues (Figure 4a). Moreover, TNFAIP8L2 methylation expression levels were significantly increased in HNSC and SKCM tumor tissues (Figure 4a).
Immune infiltration analysis data
TNFAIP8L2 is an immune checkpoint regulator of inflammation and metabolism closely connected to tumorigenesis, progression, or metastasis(17-19). Therefore, in this investigation, we used various algorithms to investigate the possible connection between various degrees of immune cell infiltration, CD8+ T- cell, macrophage, and TNFAIP8L2 gene expression in various kinds of TCGA tumors. We observed a statistically significant positive correlation between TNFAIP8L2 expression and infiltrative levels of cancer-associated fibroblasts in TCGA tumors of BLCA, BRCA, BRCA-LumA, COAD, ESCA, HNSC, HNSC-HPV-, KIRP, LUAD, LUSC, PAAD, PCPG, PRAD, READ, STAD, THCA, THCA tumors in TCGA (Figure 5a). In addition, the algorithm was used to obtain scatter data for the above tumors (Figure 5b). For example, based on the MCPCOUNTER algorithm, the TNFAIP8L2 expression level in BLCA was positively correlated with the infiltration level of cancer-associated fibroblasts (Figure 5b, cor = 0.404, P = 7.32e-16). Figure 5c shows a positive association between TNFAIP8L2 expression and macrophages in ACC, COAD, KIRP, KIRC, KICH, HNSC-HPV-, LUAD, MESO, OV, READ, PRAD, PCPG, and UCEC. Furthermore, we noticed a positive correlation between TNFAIP8L2 expression and CD8+ T cells in CESC, KIRC, SKCM, and UVM tumor tissues (Figure 5d).
Enrichment analysis of TNFAIP8L2-related partners
To gain deeper insights into the underlying molecular processes through which the TNFAIP8L2 gene contributes to tumor growth, we performed KEGG and GO enrichment analyses to search for TNFAIP8L2 target-binding proteins and genes associated with TNFAIP8L2 expression. The STRING tool revealed 18 TNFAIP8L2 binding proteins, as corroborated by experimental data. Figure 6a shows the proteins' interaction network. Next, we used the GEPIA2 method in combination with TCGA data to identify the top 100 genes related to TNFAIP8L2 expression. Figure 6b shows that TNFAIP8L2 expression levels were positively associated with ARHGDIB (R = 0.62), IGSF6 (R = 0.65), RGS19 (R = 0.64), S1PR4 (R = 0.67), and GMFG (R = 0.71) gene expression levels. The corresponding heatmap data also showed that TNFAIP8L2 was positively linked with the above five genes in most cancer types (Figure 6c). A Venn diagram was further used to analyze the interrelationships between the two groups, which had a common member: RAC2 (Figure 6d).
In this research, we integrated the two databases and conducted KEGG and GO enrichment studies on the combined results. In Figure 6e, the KEGG analysis revealed that osteoclast differentiation and the Fc gamma R−mediated phagocytosis may involve in the connection between TNFAIP8L2 and tumor progression. Furthermore, the GO enrichment study revealed that most of these genes correlate with neutrophil-mediated immunity in the BP category, actin filament in the CC category, and superoxide-generating NADPH oxidase activator activity binding in the MF category (Figure 6f).