Expression Levels Of Knstrn In Human Normal Tissues And Cancers
KNSTRN expression was analyzed in 33 cancer datasets from the TCGA database. KNSTRN was significantly up-regulated in 26 types of cancer tissues out of the 33 (Fig. 1a). Among them, the KNSTRN expression was significantly higher in the HCC tissues compared to the normal liver tissues (p < 0.001, Fig. 1b). Furthermore, KNSTRN expression was significantly higher (p < 0.05) in the HCC tissues from the GSE55048 and the GSE56545 datasets (Fig. 1c-d). In addition, Receiver operating characteristic (ROC) curves were used to analyze the diagnostic value of KNSTRN. The area under the curve (AUC) of KNSTRN was 0.935, and the results suggested that KNSTRN might be a potential diagnostic biomarker (Fig. 1e).
Correlation Analysis Between Knstrn Expression And The Prognostic Value, Clinicopathological Characteristics
To better understand the correlation of KNSTRN expression in cancer, we used the clinical data and TCGA RNAseq (downloaded from UCSC Xena) to examine the prognosis of all TCGA cancer types. Applying this approach can be used to help uncover whether high KNSTRN expression influence cancer patient prognosis. Two prognostic parameters, consisting of overall survival (OS) and progress free interval (PFI), were included. For OS, high expression of KNSTRN in pancreatic adenocarcinoma (PAAD), lung adenocarcinoma (LUAD), ACC (Adrenocortical carcinoma), bladder urothelial carcinoma (BLCA), CESC (Cervical squamous cell carcinoma and endocervical adenocarcinoma), kidney renal papillary cell carcinoma (KIRP), LGG (Lower Grade Glioma), MESO (Mesothelioma), thymoma (THYM), UCEC (Uterine Corpus Endometrial Carcinoma), GBM/LGG (Glioma), OSCC (oral squamous cell carcinoma), ESCC (Esophageal Squamous Cell Carcinoma) and LIHC (Liver Hepatocellular Carcinoma) had unfavorable prognosis but thymoma (THYM) and ESCC (Esophageal Squamous Cell Carcinoma) patients with higher expression of KNSTRN indicated better prognosis (Fig. 2a). For PFI, increased expression of KNSTRN indicated poor prognosis in all affected tumors (Fig. 2b). By combination of OS and PFI, KNSTRN may be utilized as an unfavorable prognostic biomarker in patients with HCC. Detailed results revealed that elevated expression of KNSTRN was considerably associated with bad overall survival (HR = 1.48, 95%CI 1.05–2.09, P = 0.027) and progress free interval (HR = 1.41, 95%CI 1.05–1.89, P = 0.021) in HCC, as shown in Fig. 2c and Fig. 2d.
In addition, ROC curves were used to observe the predictive value of KNSTRN mRNA levels. Evaluating the AUC under the ROC curve was applied to predict the 1, 2, 5-year risk of HCC patients on OS (1-year, AUC = 0.687; 2-year, AUC = 0.621; 5-year, AUC = 0.634) (Fig. 2e). For PFI, KNSTRN expression also performed predictive effect on the risk of HCC patients (1-year, AUC = 0.655; 2-year, AUC = 0.638; 5-year, AUC = 0.659) (Fig. 2f).
We further assessed the expression of KNSTRN in various clinical phases of HCC, the results showed that KNSTRN expression level significantly related to the AFP, Histologic grade, Pathologic stage and T stage, as shown in Fig. 2g-j. Meanwhile, KNSTRN was higher expressed in the TP53-mutant group than that of TP53-Nonmutant group (Fig. 2k, P < 0.001). Thus, KNSTRN was involved in promoting the progress of HCC.
Knstrn Methylation Status And Its Prognostic Value
DNA methylation levels in the KNSTRN gene and the prognostic value of the CpG islands in the KNSTRN gene were evaluated using the MetSurv tool. The results showed 10 methylated CpG islands including cg08036289 and cg25710630 that showed elevated levels of DNA methylation (Fig. 3a). Correlation analysis indicated that expression of KNSTRN was significantly negatively correlated with its methylation (cg08036289) status (Fig. 3b). Furthermore, methylation levels of four CpG islands, namely, cg01301233(HR = 0.639), cg08036289(HR = 0.693), cg17805752(HR = 1.662), and cg25947121(HR = 1.979) were associated with the prognosis (p < 0.05) (Table 1). Elevated levels of KNSTRN methylation in these four CpG islands, especially cg08036289 were associated with a better overall survival of HCC patients compared to those with lower levels of CpG methylation in KNSTRN.
Table 1
Effects of methylation levels in the CpG sites of the KNSTRN gene on the prognosis of HCC patients.
CpG island | HR | p-value |
Body-cg08036289# | 0.693 | 0.042462946* |
Body;3'UTR-cg25710630# | 1.344 | 0.109162787 |
Body-cg05993386 | 0.814 | 0.274468019 |
1stExon;5'UTR-cg15674596 | 0.761 | 0.125774446 |
1stExon;5'UTR-cg17805752 | 1.662 | 0.01790666* |
Body-cg01301233 | 0.639 | 0.011333953* |
Body-cg26231582 | 0.779 | 0.172668362 |
Body-cg25947121 | 1.979 | 0.005027299* |
TSS200-cg04721087 | 0.853 | 0.383155881 |
TSS200-cg19947611 | 0.826 | 0.288088112 |
HCC, hepatocellular carcinoma; HR, hazard ratio; # indicating a high level of KNSTRN methylation; * P < 0.05. |
Enrichment Analysis Of Knstrn Gene Co-expression Network And Ppi Analysis In Lihc
To further obtain the biological significance of KNSTRN in HCC, the LinkedOmics database was utilized for the co-expression pattern of KNSTRN in TCGA-LIHC. As is plotted Fig. 4a, 5682 genes are positively correlated with KNSTRN and 2617 genes are significantly negatively correlated with KNSTRN (FDR < 0.05). Figure 4b and Fig. 4c show the heat maps of the top 50 genes positively and negatively associated with KNSTRN, respectively. See Supplementary File.1 for detailed descriptions of related genes.
We use R software package to perform Gene ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis of KNSTRN-related genes. Under the condition of adjusted P < 0.1, there are 222 biological process (BP), 35 cellular component (CC), 28 molecular functions (MF), and 10 KEGG. The bubble chart shows the first 5 pieces of information about GO and all information about KEGG. GO function annotation shows that KNSTRN-related genes are mainly involved in chromosome segregation, spindle, tubulin binding, ATPase activity (Fig. 4d). KEGG pathway analysis showed that KNSTRN-related genes are mainly involved in the Cell cycle, Cellular senescence and Progesterone-mediated oocyte maturation signaling pathway (Fig. 4e). Supplementary File.2 summarizes the GO and KEGG enrichment analysis details of KNSTRN-related genes.
To further understand the internal mechanism of the KNSTRN, the STRING database was utilized to perform the PPI network of KNSTRN. The analysis showed that KNSTRN is associated with SPAG5, CENPE, CCNB1, IQGAP1, CDC20, NUF2, BUB1B, CCNB2, CHMP1B and AURKB were 0.97, 0.922, 0.844, 0.838, 0.834, 0.833, 0.825, 0.802, 0.775 and 0.774 (Fig. 4f and Supplementary File.3).
Knstrn Expression Is Associated With Immune Signatures In Lihc
The infiltration status of 24 immune cell types in the HCC tissues was evaluated by ssGSEA. As shown in Fig. 5a, KNSTRN expression showed a negative correlation with the levels of CD8+ T cells (r = -0.199, p < 0.001), Cytotoxic cells (r = -0.254, p < 0.001), Dendritic cells (DCs) (r = -0.288, p < 0.001), Mast cells (r = -0.155, p < 0.001), Neutrophils (r = -0.288, p < 0.001), Plasmacytoid DC (pDCs) (r = -0.233, p < 0.001), gamma delta T cells (Tgd) (r = -0.183, p = 0.007), natural killer cells (NK) (r = -0.130, p = 0.012), and immature DC (iDC) (r = -0.119, p = 0.022). In contrast, the expression level of KNSTRN was positively correlated with the infiltrating levels of Th2 cells (r = 0.585, p < 0.001), T helper cells (r = 0.287, p < 0.001), and follicular helper T cell (TFH) (r = 0.128, p = 0.013).
In addition, we divided 374 tumor samples into two groups based on KNSTRN expression, with 187 samples in the high-expression group and 187 samples in the low-expression group. We used the TIMER database to analyze immune infiltration in LIHC with different KNSTRN expression levels (Fig. 5b-c). The results showed that the infiltration levels of CD8 + T cells, Cytotoxic cells, dendritic cells (DCs), immature DC (iDC), mast cells, Neutrophils, and Plasmacytoid DC (pDCs) in LIHC patients with high KNSTRN expression were significantly lower than those with low KNSTRN expression. In contrast, compared with the low expression group, the infiltration of T helper cells, follicular helper T cell (TFH), and Th2 cells increased in the high expression group of KNSTRN (P < 0.05).
The TIMER tool was used to analyze the correlation between KNSTRN and immune markers of various immune cells in HCC. The results showed that the expression of KNSTRN was significantly correlated with the immune markers CD19, CD79A and KRT20 of B cells in HCC (P < 0.001, Table 2). We also analyzed a variety of T cells with different functions, such as CD8 + T cells, Tfh cells, Th1 cells, Th2 cells, Th17 cells, and Treg. The results after adjustment of tumor purity showed that the expression level of KNSTRN was significantly correlated with most of the immune markers of different T cells in HCC. including CD8B, CD25, CD183, CD185, CD278, CD360, IL23R, CD196 and the details are in the Table 2. It indicates that KNSTRN may be involved in the T cell immune response in HCC.
Table 2
Correlation analysis between KNSTRN and biomarkers of immune cells in HCC.
Description | Gene markers | Cor | P |
TAM | CCL2 | 0.172684115 | 0.001281692 |
CD68 | 0.220295047 | 3.66E-05 |
IL10 | 0.299383055 | 1.42E-08 |
Monocyte | CD86 | 0.369072841 | 1.42E-12 |
CSF1R(CD115) | 0.219770068 | 3.83E-05 |
M1-Macrophage | NOS2 | 0.064632958 | 0.231150943 |
M2-Macrophage | CD163 | 0.170170045 | 0.001511499 |
VSIG4 | 0.2000583 | 0.000183791 |
CSF1R(CD115) | 0.219770068 | 3.83E-05 |
DCs | ITGAX(CD11C) | 0.412117649 | 1.41E-15 |
CD1C | 0.156991091 | 0.003460685 |
NRP1 | 0.266451719 | 5.11E-07 |
Neutrophils | CCR7 | 0.175168576 | 0.001086553 |
ITGAM(CD11b) | 0.368704729 | 1.50E-12 |
CD59 | 0.185140579 | 0.000547686 |
Th1 | STAT1 | 0.388529912 | 7.07E-14 |
STAT4 | 0.313003873 | 2.80E-09 |
TBX21 | 0.166513087 | 0.00191368 |
CD4 | 0.320109865 | 1.16E-09 |
IFNG(IFN-g) | 0.265276296 | 5.76E-07 |
Th2 | GATA3 | 0.279114842 | 1.36E-07 |
STAT6 | 0.206907115 | 0.000108322 |
CXCR4 | 0.398449824 | 1.41E-14 |
Table 2 continued. Correlation analysis between KNSTRN and biomarkers of immune cells in HCC. |
Description | Gene markers | Cor | P |
Th2 | CCR4 | 0.326269095 | 5.34E-10 |
Treg | FOXP3 | 0.305346433 | 7.05E-09 |
CCR8 | 0.4664058 | 4.90E-20 |
STAT5B | 0.312617726 | 2.94E-09 |
TGFB1 | 0.274360461 | 2.25E-07 |
NKs | KIR2DL1 | 0.025079244 | 0.642498226 |
KIR2DS4 | 0.115780341 | 0.031558788 |
KIR3DL1 | 0.051755686 | 0.33782438 |
B cell | CD19 | 0.240184996 | 6.44E-06 |
CD79A | 0.198538563 | 0.000206183 |
KRT20 | 0.273816454 | 2.38E-07 |
CD8 + T Cell | CD8B | 0.190422106 | 0.00037555 |
CD25(IL2RA) | 0.32205797 | 9.11E-10 |
Tfh | CD183(CXCR3) | 0.292671826 | 3.06E-08 |
CD185(CXCR5) | 0.252698015 | 1.99E-06 |
CD278(ICOS) | 0.341886247 | 6.81E-11 |
Th17 | CD360(IL21R) | 0.339112691 | 9.90E-11 |
IL23R | 0.191527421 | 0.000346592 |
CD196(CCR6) | 0.286880271 | 5.84E-08 |
* P < 0.05; ** P < 0.01; *** P < 0.001. |
Concurrently, we used TIMER database to investigate whether the expression of KNSTRN is related to the level of common immunosuppressive cells and cancer associated fibroblast (CAF) in LIHC. As shown in Fig. 5d, KNSTRN expression showed a positive correlation with the levels of T cell regulatory cells (Tregs, r = 0.37, p < 0.001), Macrophages-M2 (r = 0.339, p < 0.001), Myeloid-derived suppressor cells(MDSC, r = 0.576, p < 0.001), and cancer associated fibroblast (CAF, r = 0.304, p < 0.001). Furthermore, positive correlations between the expression levels of KNSTRN and the expression of gene markers in exhausted T cells, such as PDCD1(PD-1, r = 0.282, p < 0.001), CD274(PDL-1, r = 0.207, p < 0.001), CTLA4(r = 0.271, p < 0.001), HAVCR2 (TIM-3, r = 0.259, p < 0.001), LAG3(r = 0.233, p < 0.001), and TOX (r = 0.161, p = 0.002) were observed in HCC (Fig. 5e-j). Moreover, the expression of KNSTRN also was positively correlated with TP53 gene that was an unambiguous tumor suppressor gene in LIHC patients (Fig. 5k).
Knstrn Related Cerna Network Construction In Lihc
There is growing evidence that the competing endogenous RNA (ceRNA) network plays a key role in a variety of human cancers. To ascertain whether KNSTRN was modulated by some ncRNAs, we first predicted upstream miRNAs that could potentially bind to KNSTRN and finally found 6 miRNAs, including miR-22-3p, miR-136-5p, miR-411-5p, miR-411-5p, miR-1270, and miR-676-3p. To improve visualization, a miRNA-KNSTRN regulatory network was established using cytoscape software (Fig. 6a). Based on the action mechanism of miRNA in regulation of target gene expression, there should be negative correlation between miRNA and KNSTRN. KNSTRN was significantly negatively correlated with hsa-miR-22-3p (r=-0.383, P < 0.001; Fig. 6b). Finally, the expression and prognostic value of hsa-miR-22-3p in HCC were determined. As presented in Fig. 6c and Fig. 6d, hsa-miR-22-3p was markedly down-regulated in HCC and its upregulation was positively linked to patients, prognosis. All these findings suggest that hsa-miR-22-3p might be the most potential regulatory miRNA of KNSTRN in HCC.
Next, the upstream lncRNAs of hsa-miR-22-3p were predicted using starBase database and miRnet. A total of 52 possible lncRNAs were forecasted. Identically, to improve visualization, a lncRNA-miR-22-3p regulatory network was constructed by cytoscape software (Figure S1). According to the competing endogenous RNA (ceRNA) hypothesis, lncRNA could increase mRNA expression by competitively binding to shared miRNAs. Therefore, there should be negative correlation between lncRNA and miRNA or positive correlation between lncRNA and mRNA. As shown in Table S1, among all the 52 lncRNAs, 27 were negatively correlated with hsa-miR-22-3p, and 20 were postively correlated with KNSTRN among the above 27 lncRNAs (Table S2). Subsequently, the prognostic values of the 20 lncRNAs in HCC were assessed. As suggested in Fig. 7a and Fig. 7b, only HCC patients with higher expression of MIR4435-2HG and FGD5-AS1 possessed poorer OS (HR = 1.88; 1.52, all P < 0.05). Taking expression analysis, survival analysis, and correlation analysis into consideration Fig. 7c-f, MIR4435-2HG and FGD5-AS1 might be the two most potential upstream lncRNAs of hsa-miR-22-3p/ KNSTRN axis in HCC.