Analysis of the RNA phosphotransferase and tRNA splicing activities of human Clp1.

摘要

tRNA splicing is an essential process that requires the excision of an intron within the pre-tRNA and the subsequent ligation of the resulting two tRNA halves. In Saccharomyces cerevisiae the tRNA splicing pathway has been well characterized and can be described as occurring in three steps; cleavage, healing and sealing. The cleavage step involves excision of the intron to generate two tRNA halves; a 5' half containing a 2',3'-cyclic phosphodiester and a 3' half containing a 5'-OH. The subsequent steps of the pathway are performed by the multifunctional tRNA ligase (Trl1) protein. Its kinase and cyclic phosphodiesterase (CPDase) activities heal the ends of the tRNA halves to make them suitable substrates for its sealing activity. The CPDase activity of Trl1 cleaves the 2',3'-cyclic-phosphodiester to generate a 2'-PO4, 3'-OH end. The kinase activity phosphorylates the 5'-OH to generate a 5'-PO4 end. The ligase activity of Trl1 then seals the two ends to generate a 2'-PO4, 3'-5' phosphodiester at the newly formed spliced junction. In the final step, the 2'-PO4 is removed by the 2'-phosphotransferase Tpt1.;In mammalian cells, two types of tRNA splicing pathways have been described. In addition to a yeast-type pathway, a second pathway has been described that does not require the healing of tRNA ends before they are sealed. A longstanding challenge in the tRNA splicing field has been to identify the mammalian protein catalysts that perform tRNA splicing. Here, I present an in vivo assay for the identification and characterization of new tRNA splicing proteins. By using this assay as a tool, I present data which indicates that the human Clp1 (hClp1) kinase can perform tRNA splicing in S. cerevisiae. I also show that the yeast homolog (yClp1) displays no apparent kinase activity in vitro and cannot function as a tRNA splicing enzyme in vivo. Furthermore, a potential yClp1 phosphotransferase activity is not required for cell viability given that "active site" mutants can function in lieu of wild type yClp1 in vivo. Finally, I perform a structure-activity analysis of hClp1 to identify protein features that are important for its RNA phosphotransferase activity. The instructive finding is that hClp1 resembles T4 Pnk and the yeast and plant tRNA ligases with respect to its active site constituents.