Evolution, Function and Structure of the Splicing Factor PRPF39

Abstract

Beside 5’-capping and 3’-polyadenylation, splicing is one of the essential steps in the processing of most protein-coding genes in higher eukaryotes. It is catalyzed by the spliceosome, a large and dynamic RNA-protein molecular machine that encompasses five core components, the U1, U2, U4, U5 and U6 snRNPs. For each splicing reaction, a spliceosome is assembled anew in a stepwise manner. Spliceosomes must accurately recognize each splice site, as a single mistake can result in the production of a non-functional and potentially toxic protein. The crucial step of exon definition is facilitated by the U1 and U2 snRNP during early splicing. Cryo electron microscopy structures of early spliceosomal complexes in yeast have shown that the Prp39/Prp42 heterodimer is a crucial scaffolding subcomplex. It acts as a hub for multiple protein-protein interactions for example the contact between the U1 and the U2 snRNP, indicating that the Prp39/Prp42 heterodimer is important for the precise spatial positioning of the U1 and U2 snRNPs relative to each other. Interestingly there is no homolog for Prp42 in higher eukaryotes. PRPF39 is largely unstudied in higher eukaryotes and came to our attention because it is alternatively spliced in a differential manner in murine naïve vs. memory T-cells. I could show that an alternative exon is included in a differential manner. This can control PRPF39 expression by NMD in a tissue- and activation-dependent manner in mice and human, suggesting a role in adapting splicing efficiency to cell type specific requirements. Furthermore, I solved the crystal structure of murine PRPF39 at 3.3 Å resolution. The protein is largely α-helical and the structure shows the protein to be organized as a homodimer. Dimerization in solution could be confirmed with further biophysical assays. The mode of PRPF39 homodimerization is strikingly similar to heterodimerization of Prp39 and Prp42 in yeast. Structure guided point mutations could completely abolish dimerization and by using the monomeric PRPF39 mutants I could show that the monomer has a detrimental effect on splicing in vitro. Based on a structural comparison of murine PRPF39 and the yeast heterodimer, we performed a phylogenetic analysis showing, that organisms with a Prp39 homodimer have a substantially shortened U1 snRNA compared to organisms with a Prp39/Prp42 heterodimer. Our analysis indicates that a shortened U1 snRNA accompanied by a PRPF39 homodimer was crucial in the evolutionary development of more complex splicing. This observation is unexpected, as fewer splicing factors usually mirror lower splicing complexity and not the opposite. Taken together, my results reveal the structural and functional implications of murine PRPF39 on splicing. The data suggests, that a PRPF39 homodimer acts to substitute the Prp39/Prp42 heterodimer observed in yeast. Additionally, the reduction in RNA and protein complexity of the U1 snRNP may have been crucial in allowing highly complex and sophisticated splicing regulation across species.