Fig. 3

Potential molecular mechanisms in glioma predicted by bio-information systems. a a heatmap for top 10 differentially expressed miRNAs in glioma according to the data on GSE65626 microarray, in which the abscissa indicates sample number and the ordinate indicates miRNAs; the histogram in right top is the color scale, and each rectangle corresponds to a sample expression value (has, Homo sapiens); b a Venn diagram of the intersections between the downstream miRNAs of NCK1-AS1 predicted on RNA22 (https://cm.jefferson.edu/rna22/) and the lowly-expressed miRNAs in microarray GSE62656; c binding sites between NCK1-AS and has-miR-138-2-3p predicted on RNA22; d intersection of the target genes of miR-138-2-3p predicted on several online systems (RNA22, miRDB, mirDIP, miRWalk and DIANA); e above intersect genes compared with the highly expressed genes in the glioma microarrays GSE50161 and GSE35493; f-g TRIM24 expression in GSE50161 and GSE35493; h TRIM24 expression in the TCGA-GBM database (T: tumor; N, normal); i, binding sites between miR-138-2-3p and TRIM24 predicted on RNA22; j-k, expression of miR-138-2-3p, and mRNA expression of TRIM24 and β-catenin in glioma and normal brain tissue samples using RT-qPCR (j) and protein expression of TRIM24 and β-catenin by western blot analysis, respectively, * compared to the normal group, p < 0.05; L-M, binding relation between NCK1-AS1 and miR-138-2-3p, and between miR-138-2-3p and TRIM24 detected using dual luciferase reporter gene assay. * compared to the mimnic NC group, p < 0.05. Measurement data were measured using mean ± SD; differences between two groups were analyzed using the unpaired t test, while that among multiple groups were analyzed using one-way ANOVA; Repetition = 3