23 Therefore, we next compared the transcriptional levels of miR-492 with that of KRT19 and its pseudogene. Upon PLAG1 knockdown, a strong decrease in miR-492 levels was not associated with changes in transcriptional activity of KRT19 (Fig. 1). However, miR-492 induction by PLAG1 overexpression resulted in a moderate induction of KRT19 in HepT1, but not in HUH6 cell clones. Interestingly, PLAG1 modulation was accompanied

by a strong anticorrelation between miR-492 expression and the pseudogene of KRT19. In order to identify Maraviroc mouse the regulatory network and putative direct targets of miR-492 we generated clones that stably express the precursor of miR-492 (pMif-miR-492) and a control vector (pMif-control) in the HUH6 and HepT1 cell lines. pMif-miR-492 clones of both cell lines exhibited enhanced miR-492 expression levels compared to control clones (Fig. 3A) up to 15-fold in HepT1 and up to 4.4-fold in HUH6. Whole genome expression analysis identified 194 genes with significant (adjusted P-value ≤ 0.05) differential expression (Supporting Table 2). Because overexpression of miR-492 must be expected to repress the mRNA targets that are regulated by direct binding interaction of the miR sequence, we prioritized genes being strongly down-regulated and predicted by at least one target prediction program as potential HIF cancer direct targets for quantitative confirmation (Fig. 3B). Significant

down-regulation was confirmed for HSD3B1 (3 beta-hydroxysteroid dehydrogenase), the transcription factor TCF21, the liver-related enzymes ST6GAL1 (ST6 beta-galactosamide alpha-2,6-sialyltranferase 1), BAAT (bile acid coenzyme A: amino acid N-acyltransferase), and GDA (guanine deaminase), the tumor suppressor gene CDKN2A (cyclin-dependent kinase inhibitor 2A), the liver protein ALB (albumin), as well as the proapoptotic gene BID (BH3 interacting domain death agonist) at least in one HB cell line (Fig.

3B). However, MG-132 NINJ2 (ninjurin 2) expression could not be confirmed to be down-regulated in both miR-492 overexpressing cell lines. Furthermore, we confirmed a strong suppressing effect of miR-492 on the liver tumor-related genes AFP and CASP4 (caspase-4). These genes, however, are not direct targets of miR-492. Annotation and enrichment analyses of functional categories revealed that miR-492 up-regulated genes were most significantly enriched in gene clusters involved in developmental processes, anatomical structure development, and cell communication, whereas suppressed genes were most significantly overrepresented in the functional clusters of metal binding, extracellular space occurrence, and developmental processes (Supporting Table 3). Taken together, these data point to an important influence of miR-492 on a range of genes that are involved in liver metabolism and extracellular structures.