UBE2V2 was up-regulated in LUAD tissues compared to normal samples
The 497 LUAD and 54 normal samples obtained from the TCGA network platform were analyzed by R (v 3.5.2). We compared the mRNA of UBE2V2 between LUAD and adjacent normal samples. The result predicted that the UBE2V2 mRNA in LUAD was remarkably higher than normal tissues (P<0.001) (Fig. 1a, b). To further validate the above result of the TCGA, the protein level of UBE2V2 was tested by Western blotting in 10 pairs of fresh LUAD and normal adjacent tissues. The results revealed UBE2V2 protein expression was remarkably up-regulated in LUAD (Fig. 1c). We inferred that UBE2V2 might participate in the development of LUAD.
Correlations between expression of UBE2V2 and clinical parameters
To begin with, by using the "ggplot2" package to explore that overexpression of UBE2V2 was obviously associated with Gender (P=0.0031, Age (P=0.0019) and T classification (P=0.041) based on TCGA (Fig. 2a-f). Then, ICH analysis revealed the association between UBE2V2 expression and clinicopathological characteristics. The low level of UBE2V2 expression was 41 samples (45.05%) in total, and the other 50 (54.95%) samples showed high level of UBE2V2 expression. The result showed that UBE2V2 was correlated with four factors: Gender (P=0.021), Tumour Stage (P=0.042), Differentiated degree(P=0.015) and Lymph node metastasis (P=0.021) (Table 1).
UBE2V2 predicted poor overall survival in LUAD patients
In TCGA, “survival” package was used to draw survival curves. The overall survival (OS) analysis illustrated that high UBE2V2 expression had a worse prognosis than low UBE2V2 expression in LUAD patients (P=0.047) (Fig.3a). Through GEPIA, the same survival curve trend was proven (P<0.001) (Fig. 3b). The Percent Survival of ICH also verified the above results: high expression of UBE2V2 tended to have a worse prognosis in LUAD (P=0.0132) (Fig. 3c). Meanwhile, the univariate analysis based on TCGA forecasted that Stage (HR: 1.65; 95% CI: 1.42–1.92; P<0.01), T classification (HR: 1.53; 95% CI: 1.25–1.87; P<0.01), N classification (HR: 1.69; 95% CI: 1.41–2.03; P<0.01), and UBE2V2 expression (HR:1.46; 95%CI: 1.07–2; P=0.014) were remarkably related with overall survival (Table 2). And the multivariate analysis, which depicted as a forest boxplot, forecasted that UBE2V2 expression (P =0.012) (Fig. 3e) was an independent prognosis factor for LUAD. The same results were verified via ICH. In the ICH, univariate analysis discovered the up-regulation of UBE2V2 especially affected the survival of LUAD patients in all variables (HR: 2.002, 95%CI: 1.139-3.517, P=0.016) (Table 3). And multivariate analysis indicated UBE2V2 (HR: 1.816, 95% CI: 1.046‑3.315, P=0.042) could be considered as an independent prognostic factor in LUAD (Table 3). In summary, UBE2V2 might be predicted as an adverse molecular marker for malignant progression of LUAD based on the TCGA and ICH analysis.
Diagnostic value of UBE2V2 expression in LUAD
So as to evaluate whether UBE2V2 had diagnostic value in LUAD, 497 tumour samples and 54 normal samples from the TCGA database were used to generate receiver operating characteristic (ROC) curves. The "pROC" package was used to calculate the area under the curve, which is 0.772. This analysis indicated that UBE2V2 had a considerable diagnostic value in LUAD (Fig. 3d).
LUAD cell lines selection and transfection efficiency
For further molecular mechanism and phenotype research, two LUAD cell lines were selected from four LUAD cell lines to apply to the next experiment. The UBE2V2 protein expression levels in A549 and SPCA1 cells were relatively high compared to H1299 and H1650 cells via Western blotting (Fig. 4a, b). Therefore, A549 and SPCA1 cells were selected to be transfected with UBE2V2-shRNAs and Con-shRNA lentiviruses. Western blotting detected the transfection efficiency of lentivirus. The results showed UBE2V2 expression was down-regulated with three different shUBE2V2 primer sequences, and the expression of UBE2V2 was most significantly decreased in shRNA-2 (Fig. 5 a, b and 6 a, b).
Knockdown of UBE2V2 suppressed the migration in A549 and SPCA1 cells
We applied Transwell analysis to investigate whether knockdown of UBE2V2 affected the migration ability of LUAD cells. The results demonstrated that knockdown of UBE2V2 (especially shRNA-2) remarkably inhibited the migration of A549 and SPCA1 cells from the upper chamber to the lower chamber in the 24-well plate (Fig. 5 a, g and 6 a, g).
Knockdown of UBE2V2 inhibited EMT process in SPCA1 and A549 cells
As we known, tumour metastasis involves many replication processes, including the loss of polarity on the top-bottom surface of epithelial cells, the absence of tight junctions and adhesion junctions between cells and the degradation of extracellular matrix and basement membrane. To investigate whether UBE2V2 was related to epithelial-mesenchymal transition (EMT) in LUAD, EMT-related proteins including E-cadherin, N-cadherin, Vimentin and MMP2 were studied via Western blotting. The results showed that the expression of cell epithelial protein E-cadherin protein was significantly increased, while the expression of proteins with characteristics of mesenchymal cells such as Vimentin, N-cadherin and MMP2 protein was obviously reduced after Knockdown of UBE2V2 (especially shRNA-2) in A549 and SPCA1 cells (Fig. 7a and 8a).
Knockdown of UBE2V2 suppressed the proliferation and induced apoptosis in A549 and SPCA1 cells
In order to explore whether UBE2V2 affected the proliferation and apoptosis of LUAD, a series of experiments were used in our research. Colony formation assay showed that the number of colony cells with knockdown of UBE2V2 were clearly reduced compared to shCon cells (Fig. 7a, b). Through flow cytometry analysis, we detected the proportion of healthy cells, early/late apoptotic cells and necrotic cells in knockdown of UBE2V2 cells. The results observed a remarkable increase in the percentage of early/late apoptotic cells (especially shRNA-2) in A549 and SPCA1 cells compared to shCon cells (Fig.7c, d). Western blotting analysis displayed that the proliferation marker PCNA remarkably reduced in LUAD cells after knockdown of UBE2V2 (Fig. 4 c, d, e).
Knockdown of UBE2V2 led to cell-cycle arrest in A549 and SPCA1 cells
Flow cytometry was used to analyze whether knockdown of UBE2V2 affects the cell cycle in LUAD cells. After knockdown of UBE2V2, cell replication stagnated in the early stage of DNA synthesis (G1 phase): the proportion in the GI phase increased significantly, yet the proportion was remarkably diminished in the DNA synthesis phase (S phase) (Fig.7e, 7f).
GSEA investigation of UBE2V2
The signaling pathway of UBE2V2 involved in the development of LUAD was obtained by GSEA. GSEA showed remarkably differences (P-value < 0.050, FDR < 0.050) though enrichment of KEGG pathways in sample high expression of UBE2V2. Based on the normalized enrichment score (NES), the enrichment signaling pathway with the most significant phenotypic enrichment was selected. KEGG pathway analysis identified five pathways that had a positive correlation with UBE2V2 expression: DNA replication, cell cycle, mismatch repair, ubiquitin mediated proteolysis and nucleotide excision repair. The five pathways with the strongest negative correlation were arachidonic acid metabolism, allograft rejection, hematopoietic cell lineage, leukocyte trans endothelial migration, intestinal immune network for IgA production, as shown in Fig. 8a. These results indicated that the pathways regulating DNA damage repair and immune microenvironment, which are critically important in LUAD patients.
The association between UBE2V2 and TIICs in LUAD
The association between UBE2V2 expression and the level of infiltration of different immune cells was studied through the "Gene" panel of TIMER (Fig. 8b). The results revealed that UBE2V2 was positively bound up with tumour purity in LUAD (r=0.105, P=0.019) and CD8 + T cells (r=0.095, P=0.036), but negatively bound up with B cells (r=−0.245, P<0.0001), CD4 + T cells (r=−0.238, P<0.0001), macrophages (r=−0.106, P=0.019), and dendritic cells (r=−0.188, P<0.0001).