1 The expression of ARMCX genes in pan-cancers
The expression data of ARMCX genes were obtained from the TCGA database. Next, the software of R language was utilized to perform the differential expression analysis. Differential expression of ARMCX genes (including ARMCX1, ARMCX2, ARMCX3, ARMCX4, ARMCX5, ARMCX6, and ARMCX7P) exists in the 18 TCGA tumors and corresponding normal tissues (including BLCA, BRCA, CHOL, COAD, ESCA, GBM, HNSC, KICH, KIRC, KIRP, LIHC, LUAD, LUSC, PRAD, READ, STAD, THCA, UCEC) (Figures 1A-1G). Of course, It is worth noting that 18 cancers were selected when the normal samples were greater than 5. Subsequently, we also use R language to calculate the expression of ARMCX genes in 18 cancers. Results showed that ARMCX1, ARMCX2, and ARMCX3 were higher expression than ARMCX4, ARMCX5, ARMCX5, and ARMCX7P (Figure 1H). In addition, the logFC value of ARMCX genes was displayed (Figure 1I).
2 Survival analysis of ARMCX genes in pan-cancers
The results of survival analysis are shown in Figure 2A. We found that the low expression of ARMCX1 was related to poor survival in KIRC, LGG, MESO, PAAD, SARC, and STAD. The high expression of ARMCX1 was related to poor survival in CESC. The low expression of ARMCX2 was associated with a worse prognosis in KIRC, MESO, and PAAD. The high expression of ARMCX2 was associated with a worse prognosis in KIRP, LGG, and STAD. The low expression of ARMCX3 has a better prognosis in STRC. The low expression of ARMCX3 has a worse prognosis in KIRC, LUAD, MESO, and UCEC. The low expression of ARMCX4 has a worse prognosis in LGG and PAAD. The low expression of ARMCX5 was associated with poor survival in ACC and MESO. The high expression of ARMCX5 was associated with poor survival in ESCA and SARC.
The low expression of ARMCX6 had a better prognosis in LGG, MESO, and SARC. The high expression of ARMCX6 had a better prognosis in LUAD and UVM. The low expression of ARMCX7P is related to poor survival in KICH. The HIGH expression of ARMCX7P is related to poor survival in KIRP, LIHC, and READ. In addition, we also performed the Cox analysis of pan-cancer, and the result is shown in Figure 2B.
3 The expression of ARMCX genes is associated with immune cell filtration and stemness in pan-cancers
Firstly, we explore the correlation between the expression of ARMCX genes and immune subtype, including wound healing (immune C1), IFN-gamma dominant (immune C2), inflammatory (immune C3), lymphocyte depleted (immune C4), immunologically quiet (immune C5), TGF-beta dominant (immune C6). Results unfolded that the expression of ARMCX genes was significantly different in the immune subtype (Figure 3A). Moreover, we found that most of the expression of ARMCX genes was negatively related to estimate (Figure 3B), immune (Figure 3C), and stromal scores (Figure 3D). Results revealed that the content of immune and stromal was low in 33 tumors, and the content of tumor cells was high in 33 tumors. Furthermore, we also found that most of the expression of ARMCX genes was closely negatively associated with stemness scores, including RNAss (Figure 3E) and DNAss (Figure 3F). Results uncovered that the higher expression of ARMCX genes, the lower the tumor stemness score index, the lower activity of tumor stem cells, and the lower degree of differentiation in 33 tumors.
4 The expression of ARMCX genes is associated with immune cell filtration and stemness in PAAD
Furthermore, we focus on the roles of ARMCX genes in PAAD. We only found that ARMCX1 and ARMCX4 were closely associated with the immune subtype and the expression of ARMCX1 and ARMCX4 was significantly different in the immune subtype (Figure 4A). In addition, we also found that significant difference exists in RNAss, DNAss, stromal, immune, and estimate scores (Figure 4B). Results discovered that most of the expression of ARMCX genes was negatively stemness scores, including RNAss and DNAss, which indicated that the higher expression of ARMCX genes, the lower the index of the tumor stemness score, the lower activity of tumor stem cells, and lower degree of differentiation in PAAD. We also uncovered that most of the expression of ARMCX genes was positively related to estimate, immune and stromal scores, which revealed that the content of immune and stromal was high and the content of tumor cells was low in PAAD.
5 Identification and analysis of ARMCX1 based on the TCGA
To further explore the mechanism of ARMCX genes in PAAD, we use the GEIPA online database. Based on the module of survival map of the GEIPA online database, we only found that ARMCX1 and ARMCX2 were closely associated with OS of PAAD patients (Figure 5A), and ARMCX1 was closely associated with DFS of PAAD patients (Figure 5B). Thus, we select ARMCX1 as our object in this study. Based on the TCGA database, we perform the differential expression analysis of ARMCX1 in the 4 normal pancreas and 179 tumor tissues. We only found that the expression of ARMCX1 is higher in the normal than in the tumor tissue. However, the result does not show statistical significance (Figure 5C). We believe that the results are due to a lack of normal samples. Thus, RT-PCR was utilized to test the expression of ARMCX1. Results uncovered that the expression is highly expressed in the normal pancreas cell line than in the PAAD cell lines (including AsPC-1, PANC-1, MIA PaCa-2, Capan-2, and BxPC-3) (Figure 5D). Subsequently, we performed the survival and independent prognosis analysis, and we discovered that the low expression of ARMCX1 is closely associated with poor OS (Figure 5E), and ARMCX1 is an independent prognosis factor of PAAD patients (Figures 5F, 5G). Moreover, we analyzed the clinical correlation between the expression of ARMCX1 and age (Figure 5H), gender (Figure 5I), and stage (Figure 5J). Results uncovered that a significant difference exists in stage Ⅰ and stage Ⅱ.
6 GSEA analysis of ARMCX1 in TCGA database
To explore the mechanism of ARMCX1 influencing the PAAD progression, we perform the gene set enrichment analysis (GSEA). Firstly, PAAD samples were divided into two groups, high and low, based on the median expression of ARMCX1. GSEA software was utilized to perform GSEA analysis. The result of the top 5 is shown in Figure 6. The calcium signaling pathway, complement and coagulation cascades, dilated cardiomyopathy, neurotrophin signaling pathway, and vascular smooth muscle contraction were enriched in the high expression group based on the KEGG (Figure 6A). The base excision repair, DNA replication, proteasome, ribosome, and spliceosome were enriched in the low expression group based on the KEGG (Figure 6B). The collagen trimer, intrinsic component of postsynaptic membrane, intrinsic component of synaptic membrane, postsynaptic membrane, and synaptic membrane were enriched in the high expression group based on the GO (Figure 6C). The ribosome biogenesis, large ribosomal subunit, ribosomal subunit, ribosome and structural constituent of ribosome were enriched in the low expression group based on the GO (Figure 6D).
7 Methylation analysis of ARMCX1
DNA methylation leads to the silencing of gene expression. The study showed that abnormal methylation influences cancer progression 19. In our study, we verified that the expression of ARMCX1 is downregulated. Thus, we speculated that the downregulation is closely associated with DNA methylation. From the box plots (Figure 7A), we found that the methylation level is higher in the tumor than in the normal group.
Furthermore, we explore the correlation between the expression of ARMCX1 and the methylation level of the ARMCX1 promoter. Based on the Pearson and Spearson curves (Figure 7B), we found that the expression of ARMCX1 is negatively related to the methylation level of the ARMCX1 promoter. In addition, we also found that high methylation level is closely associated with poor OS, DFI, and PFI in PAAD patients (Figure 7C). Therefore, we concluded that ARMCX1 promoter methylation leads to the low expression of ARMCX1 in PAAD and is associated with poor OS, DFI, and PFI in PAAD patients.
8 Immune cell infiltration analysis based on CIBERSORT algorithm, TIMER, TISIDB databases
To observe the correlation between the expression of ARMCX1 and immune cell infiltration, we conducted the Immune cell infiltration analysis based on the CIBERSORT algorithm, TIMER, and TISIDB databases. Based on the CIBERSORT algorithm, we found that regulatory T cells (Tregs), M2 macrophages, resting mast cells, and neutrophils were regarded as differential immune cells based on the median of ARMCX1 expression (Figure 8A). In addition, we also found that resting mast cells (Figure 8B) and M2 macrophages (Figure 8C) were positively related to the expression of ARMCX1. Plus, plasma cells (Figure 8D) and regulatory T cells (Tregs) (Figure 8E) were negatively related to the expression of ARMCX1. Thus, we speculated that ARMCX1 is closely associated with regulatory T cells (Tregs), M2 macrophages, and resting mast cells. Subsequently, based on the TIMER database, we discovered that the expression of ARMCX1 is closely B cell, CD8 T cell, CD4 T cell, macrophage, neutrophil, and dendritic cell (Figure 8F). Furthermore, according to the TISIDB database, we explore the correlation between the expression of ARMCX1 and 28 TILs as well as immunoinhibitors. From the heatmap (Figure 9A), there was a correlation between the expression of ARMCX1 and 28 TILs in multiple cancers. In addition, we found that the expression of ARMCX1 is positively correlated with activated CD4 T cell, CD56dim natural killer cell, and type 17 T helper cell (Figure 9B); the expression of ARMCX1 is negative correlated with activated B cell, eosinophil, immature dendritic cell, immature B cell, macrophage, mast cell, myeloid-derived suppressor cell, memory B cell, natural killer cell, natural killer T cell, plasmacytoid dendritic cell, effector memory CD4 T cell, T follicular helper cell, gamma delta T cell, type 1 T helper cell and regulatory T cell (Figure 9C). Furthermore, from the heatmap (Figure 9D), we observed that the expression of ARMCX1 is closely associated with most immunoinhibitors in multiple cancers. We found that the expression of ARMCX1 is positively correlated with IL10RB, LGALS9, and PVRL2 (Figure 9E); the expression of ARMCX1 is negatively connected with CD244, CD160, CD274, CSF1R, HAVCR2, KDR, IL10, TGFBR1, ADORA2A, BTLA, CD96 and PDCD1LG2 (Figure 9F).
All in all, firstly, based on the pan-caner analysis of ARMCX genes, we found that ARMCX genes play vital roles in survival, immune infiltration, and stemness in multiple cancers. Next, we focus on the PAAD. We also observed that ARMCX genes play vital roles in immune infiltration and stemness in PAAD. Furthermore, ARMCX1 was selected as our object in PAAD. Moreover, we perform the associated analysis of ARMCX1 in PAAD. Lastly, we concluded that ARMCX1 is the low expression, and the low expression is closely associated with poor OS. ARMCX1 is an independent prognosis factor for PAAD patients. In addition, the downregulation of ARMCX1 is also closely associated with the hypermethylation of the ARMCX1 promoter. Lastly, through conducting the immune cell infiltration analysis, we found that the expression of ARMCX1 is related to some immune cells and immunoinhibitors. Although most studies were based on bioinformatics analysis, we believe our study will contribute to PAAD treatment, medical progress, and scientific advancement.