Authors

Discussion

Gene expression in trypanosomes is regulated by post-transcriptional modulation of mRNA levels. The dominant mechanism that sets individual mRNA levels is codon use and a codon metric predicts the level of most mRNAs with considerable accuracy. Codon use determines the default half-life of an mRNA and it is likely that the extensive cohort of RNA binding proteins function by altering the levels of mRNAs and/or their rates of translation in response to stimuli including stresses, differentiation triggers, cell cycle transitions. The link between codon use and mRNA half-life is translation dependent and blocking the translation of an mRNA can increase its level more than 10-fold. How do variations in codon use lead to differences in mRNA stability? The major determinant is the speed of translation (codons translated/s) rather than the frequency of translation (initiations/s) and this is probably determined by the cognate tRNA availability. 

Once a decision is made to degrade an mRNA then a series of events occurs starting with 3’ to 5’ exonucleolytic removal of the polyA tail. The next step is removal of the 5’ cap followed by 5’ to 3’ exonucleolytic digestion of the mRNA. The key enzymes have been identified in trypanosomes: the 3’ polyA tail is removed by CAF1, a component of the NOT complex; the cap is removed by ALPH1; and the mRNA is degraded 5’ to 3’ by XRNA. Throughout this process RNA binding proteins such as polyA binding protein, cap binding protein, ribosomes, and other hangers on, have to be displaced to allow the nucleases access.

We would like to understand the molecular mechanism of the decision to degrade an individual mRNA based on codon use. The approach described here is a screen to identify proteins associated with the enzymes of mRNA degradation followed by phenotype analysis after depletion. First, protein complexes and interactions were identified by pulldowns using five different proteins. Second, additional loosely associated proteins were identified using proximity biotinylation by TurboID tagging of eight proteins. This screen was successful and amongst other findings it identified: (i) novel components of the NOT complex, (ii) a linear set of interactions: NOT complex->DHH1->SCD6->ALPH1->XRNA that is similar to the order of the steps in mRNA turnover, (iii) four further RNA helicases closely associated with mRNA turnover activities, and (iv) an association of the cap binding protein eIF4E1 and 4E-IP with most of the components in the mRNA turnover pathway.

We have begun the analysis of selected components by determining phenotype after depletion. To do this, we have developed an effective degron system with degradation of tagged proteins triggered by the addition of 5-PhIAA resulting in rapid (minutes) depletion of the target protein. This overcomes the problem of distinguishing primary and secondary mRNA phenotypes that is unavoidable when RNAi is used. The effect of depletion on mRNA was determined using RNAseq quantitation over a time course after the addition of 5-PhIAA. The assay was tested by preventing translation initiation by eIF2alpha depletion to test a prediction that mRNAs with a low codon score (geCAI) would be stabilised relative to those with a high score, this was indeed the case. The effect of depleting DHH1 was tested and this resulted in the selective degradation mRNAs with long ORFs. This is evidence that DHH1 provides a general protection against a degradation machinery that non-specifically targets ORFs, so that longer ORFs are more susceptible in the absence of DHH1. This is a novel finding and one that validates using the same approach with further candidates.