Comparative analysis of hepatocellular carcinoma and cirrhosis gene expression profiles.

Gene expression data of hepatocellular carcinoma (HCC) was compared with that of cirrhosis (C) to identify critical genes in HCC. A total of five gene expression data sets were downloaded from Gene Expression Omnibus. HCC and healthy samples were combined as dataset HCC, whereas cirrhosis samples were included in dataset C. A network was constructed for dataset HCC with the package R for performing Weighted Gene Co­expression Network Analysis. Modules were identified by cluster analysis with the packages flashClust and dynamicTreeCut. Hub genes were screened out by calculating connectivity. Functional annotations were assigned to the hub genes using the Database for Annotation, Visualization and Integration Discovery, and functional annotation networks were visualized with Cytoscape. Following the exclusion of outlier samples, 394 HCC samples and 47 healthy samples were included in dataset HCC and 233 cirrhosis samples were included in dataset C. A total of 6 modules were identified in the weighted gene co­expression network of dataset HCC (blue, brown, turquoise, green, red and yellow). Modules blue, brown and turquoise had high preservation whereas module yellow exhibited the lowest preservation. These modules were associated with transcription, mitosis, cation transportation, cation homeostasis, secretion and regulation of cyclase activity. Various hub genes of module yellow were cytokines, including chemokine (C­C motif) ligand 22 and interleukin­19, which may be important in the development of HCC. Gene expression profiles of HCC were compared with those of cirrhosis and numerous critical genes were identified, which may contribute to the progression of HCC. Further studies on these genes may improve the understanding of HCC pathogenesis.

Clinicopathological and prognostic implications of the miR-200 family in patients with epithelial ovarian cancer.

The aim of the present study was to investigate the association of the expression of members in the miR-200 family with clinicopathological characteristics and their impacts on overall survival in patients with epithelial ovarian cancer (EOC). Expression levels of members in the miR-200 family, including miR-200a, miR-200b, miR-200c, miR-141, and miR-429, were detected by using miRNA qRT-PCR and in situ hybridization. Associations of their expression with clinicopathological factors and overall survival were statistically evaluated. Among five members in the miR-200 family, the expression levels of miR-200a, miR-200b and miR-200c were significantly higher in EOC tissues than those in normal surface ovarian epithelium tissues, in line with the findings ofin situ hybridization analysis. In addition, tumors with high miR-200a and miR-200 bexpressionwere both more likely to have advanced stage (both P=0.006) and higher grade (P=0.01 and 0.02), whilehighmiR-200 cexpression was onlysignificantly associated with advanced stage disease (P=0.01). Moreover, univariate analysis showed that the patients with high miR-200a, miR-200b and miR-200c expression all correlated with shorter overall survival in EOC patients (all P<0.001). Multivariate statistical analysis further identified miR-200a, miR-200b and miR-200c asindependent prognostic factorsfor EOC (all P=0.01). In conclusion, these findings suggest that miR-200a, miR-200b and miR-200c overexpression may promote the aggressive tumor progression and be recognized as reliable markers to predict the survival in patients with EOCs. The three miRNAs could be attractive therapeutic targets in patients with advanced-stage EOCs.