Regulation of IKKbeta by miR-199a affects NF-kappaB activity in ovarian cancer cells

Expression of microRNAs in diffuse large B cell lymphoma is associated with immunophenotype, survival and transformation from follicular lymphoma

MicroRNAs are naturally occurring small RNA species that regulate gene expression and are frequently abnormally expressed in cancers. However, the role of microRNAs in lymphoma is poorly understood. Therefore, we undertook a comprehensive study of microRNA expression in two of the most common lymphomas: diffuse large B-cell lymphoma (DLBCL) (n= 80) and follicular lymphoma (FCL) (n= 18) using microarrays containing probes for 464 human microRNAs. Unsupervised cluster analysis revealed distinct expression patterns between these two lymphomas and specific microRNA signatures (including members of the miR-17–92 cluster) were derived that correctly predicted lymphoma type in >95% of cases. Furthermore, we identified microRNAs in de novo DLBCL (n= 64) associated with germinal centre-like and non-germinal centre-like immunophenotypes, international prognostic index status and event-free survival in CHOP and rituximab (R)-CHOP treated patients. Despite the indolent nature of FCL a significant proportion of cases undergo high-grade transformation to more aggressive DLBCL. In order to see if transformation is associated with changes in microRNA expression we compared transformed DLBCL cases (n= 16) with de novo DLBCL, as well as FCL cases that underwent subsequent transformation (n= 7) with FCL cases that had not transformed at a median follow-up of 60 months (n= 11). Differential expression of 12 microRNAs correctly predicted >85% of transformed versus de novo DLBCL cases; six microRNAs (miR-223, 217, 222, 221 and let-7i and 7b) were found which could similarly predict or transformation in FCL (P< 0.05). These data suggest that microRNAs have potential as diagnostic and prognostic markers in these lymphomas and may be used to identify FCL patients at risk of high-grade transformation.

Pharmaco(epi)genomics in ovarian cancer

Ovarian cancer shows considerable variability in its chemoresponse, however, the prospect of individualized medicine holds high hopes for improving patient survival. The influence of interindividual genomic polymorphisms on drug response (pharmacogenomics) is well established, and a variety of candidate loci in ovarian tumors have been identified, including ERCC1, ABCB1 and p53 variants. Recently pharmacoepigenomic modulators of key genes and pathways, such as promoter methylation (MLH1 and BRCA1 genes) and microRNA regulation (PTEN/AKT and NF-κB pathways) have been implicated in ovarian cancer chemoresponse. Epigenomic studies have until now mainly focused on tumor-specific changes, although germ-line epigenetic change may also be of importance. However, assessing the relevance of these potential pharmaco(epi)genomic biomarkers in clinical trials requires well powered studies in homogeneous populations, with independent validation sets, to distinguish real associations from false-positives. In addition, the selection of one gene or locus as having sufficient phenotypic effect to impact on clinical outcome may be an oversimplification. Integrated approaches that identify stable pharmacogenomic and epigenomic patterns and their relationship with expression patterns and gene function will be increasingly necessary.