Preview

PDF, English Download (66MB) | Terms of use

Abstract

The 3’-untranslated regions (3’-UTRs) of protein coding transcripts play a well established role as cis regulators of RNA stability and protein translation. Recent findings also point to functions in the nucleus. This thesis reports on the analysis of a specific 3’-UTR, referred to here as CU-RNA, originating from the CDV3 gene (carnitine deficiency-associated gene expressed in ventricle 3) that was previously identified as a 3’-UTR with a putative nuclear function. In an initial characterisation of CU-RNA it was shown that it has nuclear functions due to the following criteria: (i) the transcript was able to rescue RNase A-induced chromatin aggregation, (ii) the transcript remained within the nucleus after microinjection, (iii) the endogenous transcript was present in the nucleus, (iv) by immobilizing the RNA at a specific genomic locus it was shown that locally tethered CU-RNA induces chromatin compaction, demonstrating its chromatin modifying activity. This chromatin recruitment system was subsequently used to investigate the mechanism of action of CU-RNA. It was found that CU-RNA derived fragments recruit enhancer of zeste homolog 2 (EZH2), the enzyme that sets trimethylation of histone H3 at lysine 27 (H3K27me3). By deletion analysis a CU-RNA sequence element, CU-RNA-T0, was identified that was necessary to also induce H3K27me3. The splicing regulators PTBP1 (polypyrimidine tract binding protein 1) and U2AF2 (U2 small nuclear RNA auxiliary factor 2), and the H3K27me3-promoting factor HNRNPK (heterogeneous nuclear ribonucleoprotein K) were found to specifically interact with CU-RNA-T0. In contrast, an alternative splicing variant of CDV3 that lacks the T0 element (CU-RNA-ΔT0) and did not induce H3K27me3, bound the splice factor HNRNPC (heterogeneous nuclear ribonucleoprotein C). It was therefore investigated whether the H3K27me3 inducing activity of CU-RNA-T0 affects splicing. Reducing the H3K27me3 level at the CDV3 locus by inhibition of EZH2 enzymatic activity indeed changed the ratio of the CDV3 transcript variant levels. Based on these results a model is proposed, in which the CU-RNA-T0 element regulates alternative splicing of its own transcript, the CDV3, via a feedback loop that establishes a specific chromatin signature. Thereby stable steady-state levels between the splicing variants are regulated that might be distinct with respect to their 3’-UTR sequence dependent cytoplasmic activities. This contribution to the understanding of the mechanism governing alternative splicing is of great importance to be able to develop novel drugs targeting deregulated alternative splicing in cancer.