Identification and Validation of a Proliferation-Associated Score Model Predicting Survival in Lung Adenocarcinomas

Abstract

Aim. This study is aimed at building a risk model based on the genes that significantly altered the proliferation of lung adenocarcinoma cells and exploring the underlying mechanisms. Methods. The data of 60 lung adenocarcinoma cell lines in the Cancer Dependency Map (Depmap) were used to identify the genes whose knockout led to dramatical acceleration or deacceleration of cell proliferation. Then, univariate Cox regression was performed using the survival data of 497 patients with lung adenocarcinoma in The Cancer Genome Atlas (TCGA). The least absolute shrinkage and selection operator (LASSO) model was used to construct a risk prediction score model. Patients with lung adenocarcinoma from TCGA were classified into high- or low-risk groups based on the scores. The differences in clinicopathologic, genomic, and immune characteristics between the two groups were analyzed. The prognosis of the genes in the model was verified with immunohistochemical staining in 100 samples from the Department of Thoracic Surgery, Zhongshan Hospital, and the alteration in the proliferation rate was checked after these genes were knocked down in lung adenocarcinoma cells (A549 and H358). Results. A total of 55 genes were found to be significantly related to survival by combined methods, which were crucial to tumor progression in functional enrichment analysis. A six-gene-based risk prediction score, including the proteasome subunit beta type-6 (PSMB6), the heat shock protein family A member 9 (HSPA9), the deoxyuridine triphosphatase (DUT), the cyclin-dependent kinase 7 (CDK7), the polo-like kinases 1 (PLK1), and the folate receptor beta 2 (FOLR2), was built using the LASSO method. The high-risk group classified with the score model was characterized by poor overall survival (OS), immune infiltration, and relatively higher mutation load. A total of 9864 differentially expressed genes and 138 differentially expressed miRNAs were found between the two groups. Also, a nomogram comparing score model, age, and the stage was built to predict OS for patients with lung adenocarcinoma. Using immunohistochemistry, the expression levels of PSMB6, HSPA9, DUT, CDK7, and PLK1 were found to be higher in lung adenocarcinoma tissues of patients, while the expression of FOLR2 was low, which was consistent with survival prediction. The knockdown of PSMB6 and HSPA9 by siRNA significantly downregulated the proliferation of A549 and H358 cells. Conclusion. The proposed score model may function as a promising risk prediction tool for patients with lung adenocarcinoma and provide insights into the molecular regulation mechanism of lung adenocarcinoma.

Identifying M1-like macrophage related genes for prognosis prediction in lung adenocarcinoma based on a gene co-expression network

“…In this study, multivariate analysis confirmed the negative prognostic significance of PLK1 overexpression for cancer-specific overall survival [ 31 ]. PLK1 expression was found to be higher in the lung adenocarcinoma tissues of patients, which was consistent with survival prediction [ 32 ]. Similarly, PLK1 is associated with the aggressive progression and poor prognosis in lung squamous cell carcinoma patients [ 33 ].…”

Identification and validation of a 17-gene signature to improve the survival prediction of gliomas

“…Purkinje cell protein 4 (PCP4), an anti-apoptotic peptide that binds calmodulin, has been shown to promote the migration and invasion of human breast cells ( 34 ) and to benefit the prognosis of lung adenocarcinoma patients ( 35 ). Folate receptor beta 2 (FOLR2) is one of the prognostic genes associated with the breast cancer tumour microenvironment ( 36 ), and patients with high FLOR2 expression in lung adenocarcinoma have longer overall survival ( 37 ). Ribosomal protein L39 (RPL39) is a component of the 60S ribosome on the X chromosome (XQ24) ( 38 ).…”

CDK7 is a prognostic biomarker for non-small cell lung cancer

“…The authors stated a significantly poorer OS (p 0.036 and 0.003, respectively) in adNSCLC with high CDK7 protein expression. Bian et al (22) examined 100 samples of adNSCLC via immunohistochemistry and also found a significantly poorer OS for patients with high expression of CDK7 (p < 0.0001).…”

“…CDK7 score ≥ 2 was used as cut-off for dividing the samples into a high and low expression group. In another study dealing with CDK7 expression on adNSCLC group sizes are not explicitly stated (22).…”

“…However, for example, Jiang et al found an elevated CDK7 expression on OSCCs to be significantly associated with higher T-stage (p 0.009) (7) and Zhang et al stated that overexpression of CDK7 was significantly associated with worse tumor grade (p<0.01). Studies on protein expression of CDK7 on adNSCLC do not explicitly address its correlation with clinic-pathological parameters on which we focused (1,22).…”