Abstract: OBJECTIVE:To screen for key genes affecting prognosis of cervical cancer (CC),and to explore expression and prognostic values of AURKA in cervical squamous cell carcinomas (CSCC). METHODS:mRNA expression profiles and clinical data from GEO database of CC patients were downloaded,and the differentially expressed genes (DEGs) between cervical and normal cervical tissues were screened by bioinformatics technology. PPI network and survival analyses showed that AURKA was a potential biomarker associated with CC staging and prognosis. GEPIA2 and HPA databases were used to verify expressions of AURKA mRNA and its encoded protein in both cervical tissues,and in different stages of cancer. Kaplan-meier method was used to analyze whether the high and low expression of AURKA (median truncated value) was significantly correlated with survival in patients with cervical cancer. Subsequently,quantitative real-time PCR (qPCR) and Enzyme-linked immunosorbent assay (ELISA) were used to detect the mRNA and protein levels of AURKA in serum of CSCC patients,cervical intraepithelial neoplasia patients and healthy female subjects. RESULTS:GEPIA2 database analyses showed high expression of AURKA mRNA in cervical cancer tissues (P<0.05). Further analysis of the relationship with cervical cancer stages showed that the expression levels in different cancer stages were significantly different (P<0.05),and the levels increased with the progression of tumor staging. Compared with normal cervical tissues,expression of the AURKA protein was also increased using the HPA database analyses,and the difference was statistically significant (P<0.05). Kaplan-Meier survival analyses showed that patients with higher AURKA gene expressions had shorter survival times. qPCR and ELISA showed that the relative expressions of AURKA mRNA in CSCC patients and cervical intraepithelial neoplasia were significantly higher than that in healthy subjects,the difference was statistically significant (P<0.05). Expressions of serum AURKA protein in CSCC patients was significantly higher than that in patients with cervical intraepithelial neoplasia and health subjects (P<0.05). Results from ROC curve and χ² test showed that the diagnostic value of AURKA mRNA was high. The FIGO staging,lymph node metastasis,tumor size and muscle layer infiltration depth of the serum AURKA mRNA high expression group were higher than those in the AURKA mRNA low expression group (P<0.05). CONCLUSION:Expressions of AURKA mRNA and protein in the serum of CSCC patients were elevated,and the combination of the two tests had potential application value for the diagnosis of CSCC and the monitoring of prognosis.

Key words: AURKA, cervical cancer, bioinformatics, tumor diagnosis, prognostic monitoring