Expression and prognosis of NPC2 gene in pancarcinoma
Systematically examined the NPC2 gene's expression in all types of cancer, we discovered that it was expressed in all of them (Fig. 1A), and that the LGG, GBM, SKCM, THCG, and other cancers had significantly higher levels of expression (Fig. 1B). NCP2 gene was revealed to be a significant risk factor in GBM, HNSC, LGG, and STAD and to be negatively associated or even irrelevant in many cancers according to prognostic analysis (Fig. 1C).
We also investigated into the expression of the NPC2 gene in gliomas. In accordance with the pattern of TCGA patients and GTEx normal samples, the expression of NPC2 gene of GBM and LGG patients in the GSE68848 dataset was considerably greater than that of Normal patients (Fig. 1D and E). The significance of the NPC2 gene in glioma disease was made clear by all of these findings, so we performed additional research.
Expression analysis of NPC2 gene in different clinical subgroups of glioma
We examined differences in gene expression in various clinical categories to further investigate the potential association between the NPC2 gene and glioma. Analysis of the difference of NPC2 gene expression in different types of gliomas showed that the expression level of NPC2 gene was higher in GBM, rA, rAA and rGBM (Fig. 2A). The findings demonstrated that the expression of NPC2 gene was significantly higher in WHO IV than in WHO Ⅲ and Ⅱ patients (Fig. 2B), and patients with wild-type IDH mutation status under various grade subtypes had considerably higher NPC2 gene expression (Fig. 2C and D). Patients with various subtypes of the 1p/19q co-deletion status showed significantly varied gene expression (P < 0.001), and NPC2 gene was highly expressed in both WHO III and IV 1p/19q co-deletion Non-codle patients (Fig. 2E and F). Overall, patients with GBM IDH wildtype showed increased NCP2 gene expression (Fig. 2G). Additionally, male patients and patients older than 42 had greater gene expression levels (Fig. 2H-K). In the traditional subtypes, the sample numbers of patients in the high and low NPC2 groups was basically the same, indicating that NPC2-based patient grouping is an independent glioma typing strategy (Table 1).
Table 1
Correlation between NPC2 expression with clinicopathologic features of glioma
Characterstic | Low expression of NPC2 | High expression of NPC2 | p |
n | 152 | 152 | |
Gender,n(%) | | | 1 |
Female | 57 | 58 | |
Male | 95 | 94 | |
Age,n(%) | | | 5.53E−05 |
< 42 | 51(16.7%) | 87(28.6%) | |
>=42 | 101(33.3%) | 65(21.4%) | |
PRS_type | | | 0.04085 |
Primary | 100(32.9%) | 119(39.1%) | |
Recurrent | 37(12.2%) | 21(6.9%) | |
Secondary | 15(4.9%) | 12(3.9%) | |
Grade | | | 1.26E−13 |
WHO II | 10(6.25%) | 78(25.65%) | |
WHO III | 38(12.5%) | 35(11.51%) | |
WHO IV | 95(31.25%) | 39(12.82%) | |
IDH_mutation | | | 9.99E−13 |
Mutant | 50(16.44%) | 113(37.17%) | |
Wildtype | 102(33.55%) | 39(12.83%) | |
X1p19q_codel | | | 1.39E−07 |
Codel | 12(3.95%) | 50(16.44%) | |
Non-codel | 140(46.05%) | 102(33.55%) | |
Expression level of NPC2 and prognosis of glioma patients
The K-M survival curve revealed that patients with glioma, such as GBM and LGG, with high NPC2 expression had considerably worse prognoses (Fig. 3A-C). Univariate and multivariate COX analyses also confirmed that NPC2 was a significant risk factor in gliomas (Table 2 and Table 3). Additionally, patient K-M curves from the CGGA dataset demonstrated that low-expression NPC2 samples had considerably better prognoses than high-expression samples, supporting our findings (Fig. 3D).
Table 2
| p.value | HR |
PRS.type | 4.2e−13 | 0.35(0.26–0.46) |
Age | 0.00024 | 1.7(1.3–2.2) |
Gender | 0.66 | 0.94(0.72–1.2) |
Grade | 2.4e−20 | 0.18(0.12–0.26) |
IDH.mutation.status | 2.6e−13 | 2.8(2.1–3.7) |
X1p19q.codel.status | 1.2e−12 | 5.9(3.6–9.6) |
NPC2 | 1.1e−15 | 0.31(0.24–0.42) |
Table 3
Multivariate cox analysis
| p.value | HR |
PRS.typelow_PRS | 1.4e−07 | 0.44(0.33–0.6) |
Age > 42 | 7.8e−01 | 1(0.78–1.4) |
Gender Male | 8.3e−01 | 0.97(0.73–1.3) |
Gradelow_Grade | 3.4e−09 | 0.29(0.2–0.44) |
IDH.mutation.status whidtype | 1.1e−01 | 1.3(0.94–1.8) |
X1p19q.codel.status Non-codel | 3.6e−07 | 3.9(2.3–6.6) |
NPC2 Low_NPC2 | 4.6e−03 | 0.64(0.48–0.87) |
Analysis of prognostic value of NPC2 gene
After TCGA-glioma patients (AUC = 0.697) and patients with various clinical features were subjected to ROC curve analysis, it was discovered that NPC2 showed significant prognostic and predictive value in glioma patients (Fig. 4A-F). The AUC values of IDH mutation status, X1p/19q codel-status, PRS, age, Grade were 0.772, 0.764, 0.543, 0.644 and 0.745, respectively (Fig. 4B-F). In addition, the CGGC database results also confirm our findings (Fig. 4G).
Screening and functional enrichment analysis of differentially expressed genes
With screening parameters of p-adjust 0.05 and logFC | > 1, the R program "DESeq2" was utilized to screen for DEGs in the high and low NPC2 expression groups of TCGA-glioma samples (Fig. 5A). According to GO enrichment analysis, DEGs were particularly enriched in the regulation of trans-synaptic signaling, modulation of chemical, synaptic membrane, channel activity, and other processes (Fig. 5B). According to KEGG functional enrichment analysis, DEGs were primarily implicated in pathways such as retrograde endocannabinoid signaling and neuroactive ligand-receptor interaction (Fig. 5C). Figure 5D and 5E shows heat maps of NPC2 DNA methylation in GBM and LBB (using probes cg11740921, cg21915100) analyzed based on the MetSurv database.
Relationship between NPC2 gene and immune infiltration in glioma patients
We showed the relationship between the NPC2 gene and immune cells in Fig. 6A and B as a heat map and scatter diagram, respectively. The findings suggested that the NPC2 gene was substantially connected with Neutrophil, Macrophage, and DC in various forms of glioma. In instance, there were significant correlation between LGG and macrophage (0.81), as well as between macrophage and LGG (0.70) and GBM (0.70). Analysis of the immune cell level differences between the high and low NPC2 expression groups revealed that the levels of activated B cells and the other 14 immune cells were considerably greater in the high expression group(Fig. 6C).
Drug sensitivity analysis
This study analyzed the differences in drug sensitivity to anticancer drugs in patients with high-low NPC2 expression groups. These results indicated that the differential expression of NPC2 had a certain effect on the drug sensitivity of patients, indicating that NPC2 could provide some guidance for the drug use of patients with glioma (Fig. 7A - F).
Interaction gene set analysis
The String database was used to search out NPC2 interacting genes, including LDB3, GM2A, NUDT19, SLC9A3R2, GRN, NME2, CFB, NPC1 and RRAGA (Fig. 8A). And these genes were mainly related to DNA damage inhibit, hormone ER activate and other pathways in glioma (Fig. 8B). Moreover, genes NUDT19 and NME2 were significantly positively correlated with the prognosis of LGG (Fig. 8C). And the expression of SLC9A3R2, CFB and LDB3 genes were strong negative correlation with methylation level in LGG (Fig. 8D). By analyzing the correlation between NPC2 interacting gene set methylation and patient survival in glioma patients, it was found that the methylation of NPC2, NUDT19 and CFB was significantly correlated with the prognosis of LGG patients (Fig. 8E).
Validation of the expression of NPC2
The expression of NPC2 gene in cells and tissues verified by qRT-PCR showed that the expression of NPC2 gene in BRTBG cells was significantly higher than that of AVGP12, and the expression of NPC2 gene in cancer tissues was significantly higher (Fig. 9A and B).