Microarray gene expression profiling for identifying different responses to radiotherapy and chemoradiotherapy in patients with cervical cancer.

PURPOSE: Cervical cancer, which is treated by radiotherapy (RT) and chemoradiotherapy (CRT), has high morbidity and mortality in women. This study aimed to identify differences in gene response to CRT and RT.

MATERIALS AND METHODS: GSE3578 was downloaded from Gene Expression Omnibus including specimens from 20 RT-treated patients and 19 CRT-treated patients. Differentially expressed genes (DEGs) were identified using siggenes package in R. Protein-protein interaction (PPI) network was visualized by cytoscape. MCODE and cytoscape was used separately to mine and construct modules in the PPI network. Transcription factor (TF)-DEG and miRNA-DEG pairs were predicted and then visualized by cytoscape.

RESULTS: Total 22 upregulated and 181 downregulated genes were identified in CRT samples. Several functions were enriched for these DEGs. A module involving ZNF449 and ZNF673 was mined from the PPI network of downregulated genes. In the TF-DEG regulatory networks, downregulated GATA3 (which was modulated by SPI) was also a TF, as well as upregulated CDK6 was regulated by several TFs (e.g. GATA3). Hsa-miR-17, hsa-miR-34a, hsa-miR-124, hsa-miR-I185-2-3p, hsa- 1185-1-3p, and hsa-let-7f-2-3p were identified as key miRNAs in the miRNA-DEG regulatory network. Conclu- sion: CRT might cure cervical cancer by acting on those molecules that were more sensitive to CRT than CT.