Identification of biomarkers for metastatic osteosarcoma based on DNA microarray data.

Osteosarcoma (OS) is a malignant bone tumor very often with pulmonary metastasis and is the main cause of OS mortality. The objective of this study was to screen for possible biomarkers of metastatic OS to explore the mechanisms of pulmonary metastasis of OS through network construction. GSE14359 was downloaded from the Gene Expression Omnibus database, which included 5 samples from conventional OS group with 2 replicates and 4 samples from OS pulmonary metastasis group in duplicate. Differentially expressed genes (DEGs) between two groups were identified by limma packages in R and classical t-test with the threshold of the false discovery rate (FDR) <0.05. The Database for Annotation, Visualization and Integrated Discovery (DAVID) were then used to perform functional annotation (FDR < 0.01). Differential coexpression network was constructed with subspace differential coexpression analysis (SDC), and genes with high degrees in the differential coexpression network were identified. A total of 1344 genes were screened as DEGs, including 677 up- and 667 down-regulated DEGs in the pulmonary metastasis of OS. Thirty-one significantly enriched functions were obtained, such as blood vessel morphogenesis, defense response, cell death and so on. DEGs with high degrees (brain-specific angiogenesis inhibitor 2 (BAI2), formin-like 1 (FMNL1), dual-specificity phosphatase 7 (DUSP7), transient receptor potential melastatin 2 (TRPM2), CBP80/20-dependent translation initiation factor (KIAA0427) and C120rf35) in the differential coexpression network were found. BAI2, FMNL1, DUSP7 and TRPM2 may be useful markers for predicting tumor metastasis and therapeutic targets for the treatment of OS patients with metastasis.