The Expression and Clinical Features of FCHO1 in Different Database
Oncomine database was used to analyze the difference of FCHO1 mRNA expression in various tumors and our results showed that FCHO1 was significantly upregulated in colorectal cancer compared to the normal tissues (Figure 1A). Also, to explored FCHO1 expression in human cancers, we used TIMER database to analyze FCHO1 mRNA expression. The differential expression between the tumor and adjacent normal tissues for FCHO1 is shown in (Figure 1B). FCHO1 was highly-expressed in various cancers, including COAD. We identified that the expression level of FCHO1 was significantly increased in unpaired (Figure 1C) or paired (Figure 1D) COAD and normal tissues using TCGA database.
Next. we examined the correlation between FCHO1 and clinicopathological characteristic in COAD. The results of UALCAN showed that the expression of FCHO1 was significantly correlated with gender, histological subtypes, P53 mutation status, body weight, individual cancer stages, race, age and nodal metastasis status (Figure S1). In order to reflect the relationship between the accuracy of the predicted value and FCHO1 expression, The results of ROC revealed that FCHO1 had a favorable prediction accuracy for COAD, and the area under the ROC curve was 0.841 (95%CI:0.801-0.880) (figure 1E). In addition, GEIPA database was used to investigated the prognostic value in COAD, the results showed that high expression of FCHO1 in COAD predicted worse survival, including OS (HR=1.8 p=0.0022) (figureS1F), DFS (HR=1.6 p=0.043) (figureS1G). Analysis of protein of FCHO1expression level from the HPA database indicated that FCHO1 protein was obviously upregulated in COAD tissues(figureS1H). Collectively, these results indicated that FCHO1 was significantly overexpressed in COAD and may serve as a potential oncogene in COAD.
The Prognostic and Diagnostic Value of FCHO1 Expression in COAD
Clinical Characteristics and gene expression were downloaded from TCGA database. Based on the mean value of FCHO1 expression, about 454 patients with COAD were divided into high- and low-FCHO1 expression groups, but the expression of FCHO1 in COAD had no significance with clinical characteristics .Then we conducted Univariate Cox hazard regression analysis showed T stage (T3&T4), M stage(M1), Pathologic stage (Stage III and Stage IV) and N stage (N1 and N2) were significantly for COAD, Multivariate Cox hazard regression analysis showed that T stage (T3&T4), M stage(M1), Pathologic stage (Stage III and Stage IV) were significant with COAD, but the expression of FCHO1 in COAD had no significance.
Furthermore, to evaluate prognostic value of FCHO1 in clinical characteristics from TCGA database, the results showed that T stage (T3&T4), residual tumor R0 and N stage N2 were significantly related to OS for COAD (figure2A,2B,2C),but Pathologic stage (Stage III and Stage IV) and M stage (M1) had no significance (figure2D,2E). A nomogram for predicting probability of patients with 1-, 3- and 5-year overall survival (OS) was established to evaluate the prognostic value (figure 2G). Studies showed hypermethylation of the promoter CpG island silences tumor suppressor genes, such as Rb1, APC and DNA mismatch repair genes, in various cancer including CRC. Therefore, we hypothesized that hypermethylation of the promoter CpG island could inhibit gene expression, UALCAN was used to explore the promoter of FCHO1 methylation level in COAD, then we found that the methylation of FCHO1 in COAD tissues was remarkably decreased compared with normal tissues(figure2F). Also, the promoter of FCHO1 methylation level was significantly correlated with gender, histological subtypes, P53 mutation status, body weight, individual cancer stages, race, age and nodal metastasis status (figure S2). Taken together, these findings showed that FCHO1 may act as a promising diagnostic and prognostic biomarker in COAD.
Enrichment Analysis of FCHO1Co-Expression Gene in COAD
String and GeneMANIA were used to established genes-genes network across FCHO1 and its neighboring genes. By analyzing String and GeneMANIA, we acquired 30 FCHO1-related genes (10 genes for String and 20 for GeneMANIA) (Figure3A, 3B). these neighboring genes in GO analysis demonstrates that they are mainly involved in cytoplasm, plasma membrane, exosomes, receptor signaling complex scaffold activity, GTPase activator activity, signal transconduction, cell communication and transport (Figure 3C-3E). KEGG analysis for these genes showed that integrin family cell surface interactions, S1P1 pathway and Arf6 trafficking events (Figure 3F).
To further characterize the potential function of FCHO1 gene, we explored FCHO1-positively or negatively associated genes in COAD using LinkedOmics. The top 50 FCHO1 positively or negatively associated genes were shown (Figure 4A, 4B). Besides, we used LinkedOmics database to analyze these significant genes in GO analysis. The bar bubble graph revealed the top 10 messages for biological process (BP), cellular component (CC) and molecular function (MF) enrichment analysis, respectively. (Figure 4C-E). For BP, they were mainly enriched in metabolic process, biological regulation and response to stimilus (Figure 4C). For CC, they mainly enriched in membrane, nucleus, and membrane-enclosed lumen (Figure 4D). For MF, these genes mainly enriched in protein binding, ion binding, and nucleic acid binding (4F). Moreover, we conducted GSEA for FCHO1co-expression genes using NGSEA datebase. Results showed that they were significantly linked with MicroRNAs in cancer, Oxidative phosphorylation, Cell cycle, Notch signaling pathway, VEGF signaling pathway and p53 signaling pathway (Figure 4H-L). Collectively, studies may help us to understand the complicate molecular mechanism of FCHO1 in COAD.
The Relationship of FCHO1 with Tumor Hypoxia and Anaerobic Metabolism
Hypoxia is a common feature of tumor microenvironment and plays an important role in tumorigenesis and aggressiveness. Some studies showed hypoxia had been verified to be involved in glycolysis[16]and immune cell infiltration[17] in colorectal cancer. At the same time, Increased studies indicated that glycolysis played an important role in the progression of COAD[18, 19]. GSEA analysis of co-expressed genes suggests FCHO1 was linked with oxidative phosphorylation, which was closely related to glycolysis. Moreover, we hypothesis whether tumor hypoxia and glycolysis associated with FCHO1 expression in COAD. To analyzed TCGA-COAD date sets to investigate hypoxia-related genes and glycolysis-related genes with FCHO1 expression in COAD (Figure 5A, B). The results of correlation analysis showed FCHO1 expression was positively correlated with hypoxia-related genes (NFKB1, VEGFA and VEGFB), while FCHO1expression was negatively correlated with HIF1A and KRAS. Also, FCHO1 expression was positively correlated with glycolysis-related genes (ENO1, HK1, G6PD, PGK1, PKM and SLC2A1), while FCHO1 expression was negatively correlated with SLC2A1. Based on TIMER, Starbase database to explore these positive relate gene whether correlated with FCHO1 expression. The result revealed that hypoxia-related genes (NFKB1, VEGFB) was simultaneously positively correlated with FCHO1 expression in TIMER, Starbase database (Figure5C-F). Also, glycolysis-related genes (HK1, G6PD and SLC2A1) was positively associated with FCHO1 expression in TIMER, Starbase database. These results suggested that FCHO1 may involve in metabolism reprogramming and hypoxic microenvironment in COAD.
Correlation between FCHO1 and immune infiltration in COAD
Previous studies indicated that immune cells infiltrationplayed an important role in the progression of COAD[17, 20]. By analyzing TIMER database to analyze the correlation between FCHO1 expression and immune infiltrating cells in COAD. we found that FCHO1 was positively associated with the expression of CD4+ T cell, while negatively correlated with the expression levels of CD8+ T cells (Figure 6A). Also, we identified that FCHO1 CNV was closely associated with the degree of infiltration of B cell, CD8+ T cell, and dendritic cell (figure 6B). We further analyzed the correlation between FCHO1and immune cell infiltration level verified by ssGSEA in TCGA-STAD using Spearman correlation. Results showed high FCHO1 expression was significantly positively correlated with NK CD56 bright cells, Th17 cells, pDC and NK cells (Figure 6C). The enrich score of NK CD56 bright cells, Th17 cells, pDC and NK cells were positively associated with high expression levels of FCHO1(Figure 6D). The correlation between FCHO1 expression and these immune cells were showed in Figure 6E-H using Spearman correlation.
To further evaluate the relationship between FCHO1 and various immune infiltrating cells of COAD. we used TIMER, StarBase, and TCGA to explore the association between FCHO1 and immune marker genes of several immune cells. The result showed that the expression of FCHO1 was positively associated with B cell, CD8+ T Cell, Tfh, Th1, Treg, T cell exhaustion, M1 Macrophage and Natural killer cell. Moreover, PDCD1 and LAG3 is immune checkpoints in tumor. By analyzing TIMER, Starbase database, PDCD1 and LAG3 expression was positively correlated with FCHO1 expression (Figure 7). These results suggested that tumor immune cells may participate in FCHO1-induced progression of COAD.