MRNA degradation in the control of gene expression in yeast
The pathways for eukaryotic mRNA translation and degradation are composed of numerous interconnected elements. I have characterized the general cellular roles of factors involved in mRNA degradation pathways and their part in the control of gene expression in the budding yeast Saccharomyces cerevisiae. Xrn1p is the non-essential cytoplasmic 5´ exoribonuclease required for rapid mRNA turnover. xrn1 mutants have been associated with wildly disparate phenotypes, including karyogamy and meiotic recombination. I identified two translation mutations (in eIF4E and eIF2B) from a synthetic lethal screen, supporting my assertion that aberrant gene expression contributes to xrn1 pleiotropy. I then demonstrated that mutations in capping enzyme and in eIF4G genetically interact with xrn1. My results contradict the currently held model that stabilization of mRNAs by deletion of XRN1 should suppress the inviability of upstream translation mutations. I conclude that the accumulation of messages in an xrn1 mutant is in fact lethal in combination with particular defects in translation. A previous xrn1 synthetic lethal screen identified SKI2 and SKI3. My screen also identified mutations complemented by SKI4/CSL4, SKI6/RRP41, RRP46, and SKI8. Debate persists over the primary role of the Ski proteins. PolyA-minus mRNA is not translated efficiently in wild-type eukaryotic cells, but is translated efficiently in ski mutants, perhaps due to altered translational specificity. However, as the SKI genes are required for 3´ mRNA degradation, it is possibly a consequence of inhibition of 3´ mRNA decay. I show that Ski2p, Ski3p and Ski8p form a stable complex and that Ski2p and Ski3p are cytoplasmic, not nuclear as previously maintained. To further distinguish between the Ski models and to directly assess the heretofore unexamined effects of 3´ mRNA degradation on translation in non-mutant cells, I show that an RNA can be translated efficiently in wild-type cells without polyA or Pab1p when 3´ degradation is blocked in cis. In addition, this enhanced expression phenocopies a ski mutant. Hence, functional redundancy is the simplest model to explain synthetic lethality between xrn1 and ski mutations: mRNA decay is an essential process and lethality consequently arises from inhibition of both 5´ and 3´ pathways for mRNA degradation.