MiRNAs are short regulatory RNAs, hundreds of which have so far been discovered in the human genome. Among the processes that are regulated by miRNAs are cell proliferation, cardiac and immune system development, lipid and glucose metabolism. Both oncogenic as well as tumor suppressor miRNAs are known, the best studied examples being the oncogenic cluster of miR-17 and the tumor suppressor let-7 miRNA. MiR-17 is part of a polycistronic cluster of co-expressed miRNAs that emerged through gene duplication, explaining the sequence similarity between cluster members, and has itself duplicated a number of times. MiR-17 is also part of a large family of miRNAs defined by their 5'-end sequence (also called "seed"), which is most important for target recognition. Members of this family are the miRNAs of the miR-302b-302d cluster, which is expressed in embryonic cells and is conserved between human and mouse, as well as those of the mouse-specific, embryonicallyexpressed miR-290-295 cluster. In a recent collaborative project with the group of Witek Filipowicz from the Friedrich Miescher Institute (Basel), we demonstrated that these mouse miRNAs regulate the de novo DNA methylation by targeting the transcript of retinoblastoma-like 2 (Rbl-2) protein, which negatively regulates the expression of de novo methylases. MiR-17 itself, whose 5'-end sequence is almost identical to that of the miR-290 cluster members, also targets Rbl-2 in adipocytes. Given that enhanced de novo methylation caused by disregulation of the Rbl-2 pathway has already been implicated in cancer, these findings may indicate that the described oncogenic properties of the miR-17 cluster are partly implemented through alterations in DNA methylation and chromatin structure.

Motivated by these results, we here propose to characterize in detail the transcripts and regulatory networks that are concertedly targeted by the members of the highly conserved miR-17-92 cluster. Towards this end, we can advantage of a mouse system that has already been established and was used in the collaborative project with the Filipowicz group, namely the Dicer-deficient mouse embryonic stem cells. In our previous study it was shown that these cells, that do not express endogeneous miRNAs, do incorporate exogeneous miRNAs into RNA-induced silencing complexes (RISC) that silence target mRNAs. Our approach is to transfect individual miRNAs into these Dicer-deficient mouse ESCs and to identify responding mRNAs by microarray profiling. Many studies have previously demonstrated that miRNAs have a detectable effect on the stability of their target mRNAs, and that this property can be used to identify miRNA targets. In a preliminary project, we started with a miR-17-92 cluster member whose 5' sequence is different from that of miR-17, namely miR- 19a. Its top candidate targets, identified on the basis of reproducible downregulation in triplicate experiments, include a number of cancer-related genes, consistent with the reported involvement of the entire cluster in oncogenesis. To quantify the effects of miRNAs on the level of target proteins, we have also established in our laboratory the pSILAC method (Baek et al., 2008; Selbach et al., 2008).

Building on these initial studies, we would now like to systematically study the functions of the four individual seed sequences represented in the miRNAs of the miR-17-92 cluster aiming to

1. identify and functionally validate cancer-related targets of individual members of the miR-17-92 miRNA cluster, and comparatively analyze the functions of co-expressed, very closely related miRNAs, each of which is evolutionarily conserved.

2. investigate the time-dependent regulatory cascades that are set in motion by the aberrant over-expression of this cluster of miRNAs, which appears to occur in malignancies. These studies will enable us to understand how the concerted action of a cluster of related and co-expressed miRNAs that are also broadly expressed under normal conditions, translates into oncogenesis, and may suggest future modalities of therapeutic intervention.