Studying splicing and the formation of mRNPs

Splicing is a fundamental step in eukaryotic gene expression. During this process, introns are excised from pre-mRNAs and exons are ligated to form a continuous reading frame. As a consequence of splicing, a number of proteins are deposited on the mRNA product. These proteins can influence a number of downstream processes of the mRNA metabolism, such as export from the nucleus, translation efficiency and stability in the cytoplasm. It becomes evident that splicing not only plays an important role in the maturation of an mRNA in the nucleus, but also influences distant downstream processes affecting the ultimate fate of mRNAs. The exon junction complex (EJC) is assembled ~20-24 nucleotides upstream of splice junctions in a splicing-dependent, but sequence-independent manner the presence of ATP (Ballut et al., 2005). Therefore, its assembly can occur in the absence of the physiological deposition machinery, the spliceosome. However, under these experimental conditions, its physiological assembly hierarchy was not determined. One of the goals of this thesis was to study EJC deposition during splicing. The stepwise assembly of EJC during splicing, in parallel to the spliceosome assembly was investigated. We used an in vitro splicing system that faithfully recapitulates the splicing-dependent deposition of EJC proteins. We found that eIF4A3 and MAGOH-Y14 formed a pre-EJC before exon ligation and that deposition of eIF4A3, MAGOH and Y14 required splicing, while Barentsz and UPF3b did not require splicing to bind at the EJC. These results show that the minimal EJC (eIF4A3 and MAGOH-Y14) assembles on mRNA prior to exon ligation, while Barentsz and UPF3b join the complex later, after the completion of splicing to form an NMD-competent core EJC. While studying the splicing-mediated deposition of EJCs, we noticed that MAGOH mutants lacking the interaction with PYM subtly but reproducibly co-immunoprecipitated more spliced MINX mRNA than wild type MAGOH. Therefore, we hypothesized that EJCs containing such mutants are more stable than EJCs with wild type MAGOH. To directly test if PYM influences the EJC, we purified recombinant PYM (rPYM) from bacterial expression cultures and added increasing amounts of rPYM to splicing reactions containing FLAG-tagged EJC proteins. The investigation led to the following observations: PYM reduced the amount of EJCs bound to spliced mRNAs and that the N-terminus of PYM was required for EJC disassembly. In addition, it was shown that PYM dissociated assembled EJCs but did not inhibit EJC assembly. These results reveal that PYM is an EJC disassembly factor in vitro that antagonizes important EJC functions. A number of protein complexes assemble on the mRNA in the nucleus, such as EJC and TREX. Evidence exists in the literature that EJC and TREX are connected. Some of the TREX complex components have been included into the EJC components (Gatfield et al., 2001; Le Hir et al., 2001b; Le Hir et al., 2000). It has yet to be demonstrated whether TREX complex deposition is EJC-dependent. We investigated the TREX complex deposition using the in vitro splicing system described before. The role of DDX39, a UAP56 paralogue protein, was addressed for the first time and compared to UAP56. The results revealed that TREX deposition was splicing-independent and in particular that the recruitment of UAP56 and DDX39 was EJC-independent, while ALY/REF interacted with the EJC. In addition, functional characterization of ALY/REF and UAP56 revealed that the ALY/REF residues 204-241 were required for binding to the cap fragment, but not to the EJC fragment and that UAP56 mutants did not precipitate unspliced pre-mRNA or spliced mRNA fragments. The splicing-independent TREX complex deposition suggests that splicing is not necessary for export, but it may enhance export of spliced mRNAs