Translation control in Trypanosoma brucei by mRNA cap-binding proteins and their interaction partners

Gene expression in Trypanosoma brucei lacks regulatory mechanisms at the level of RNA polymerase II transcription. The initiation step of translation is therefore a critical point for controlling gene expression. Despite being unicellular eukaryotes, the parasites are equipped with a large repertoire of EIF4E cap-binding proteins to adjust the gene expression program according to the different environments encountered in the bloodstream of a mammalian host and the tsetse fly vector, where they multiply as so-called bloodstream forms (BSFs) and procyclic forms (PCFs), respectively. EIF4E3 and EIF4E4 are considered canonical translation initiation factors that associate with EIF4G4 and EIF4G3, respectively, to form EIF4F-like complexes that initiate bulk translation. Loss of any of the two proteins is not compatible with normal cellular life at both BSF and PCF stages. EIF4E1 and EIF4E2 act independently of EIF4G proteins, and are thus hypothesized to have regulatory functions. EIF4E1 was previously shown to act as a translational repressor by interacting with 4EIP, while EIF4E2 associates with a homolog of the histone mRNA stem-loop-binding protein, SLBP2, in PCFs. Possible functions of the two smallest members, EIF4E5 and EIF4E6, which do in turn interact with different EIF4G proteins, point towards a role in cellular integrity and motility, as evidenced by knockdown experiments in PCFs. The aim of this work was to dissect the roles of the non-canonical EIF4E proteins in BSFs and/or PCFs, as well as during differentiation processes. Previously obtained data suggested that the functions are unlikely to be identical, but the level of redundancy was to be addressed in this study. Loss-of-function phenotypes were assessed, protein interaction partners were analyzed by quantitative mass spectrometry and yeast-2-hybrid experiments, and associated mRNAs were examined by qPCR and RNA sequencing. Despite accumulating evidence in Leishmania, a translation-promoting function of EIF4E1 without 4EIP could not be confirmed. EIF4E1 was shown to associate with similar mRNA subsets in the presence and absence of 4EIP in PCFs, where it is essential. In contrast, its binding partner 4EIP is required for BSF to PCF differentiation, which proceeds normally without EIF4E1. The mechanisms of action of EIF4E1/4EIP are reminiscent of 4EHP/GIGYF2 complexes in mammals, which initiate translation-coupled mRNA decay. Accordingly, 4EIP-dependent associations with the terminal uridylyl transferase 3 (TUT3) and the CAF-NOT deadenylation complex were uncovered, but the uridylase activity of TUT3 was not required for 4EIP-dependent repression. EIF4E2 was found to serve an essential role in the regulation of S-phase mRNAs in BSFs, which are known to be stabilized by the RNA-binding protein PUF9. There is evidence that the latter interacts directly with EIF4E2/SLBP2 complexes. EIF4E2 further appeared to be involved in allelic exclusion, which ensures expression of a single variant surface glycoprotein (VSG) variant at a time. The latter covers the BSF in a dense glycoprotein coat and constitutes approximately 10% of total cellular protein. Loss of EIF4E2 resulted in promiscuous expression of numerous, otherwise silenced VSG genes, which could be rescued by reconstitution of EIF4E2 expression. EIF4E6 was revealed to be dedicated to promoting VSG mRNA translation by cooperating with activating MKT1 complexes. The latter were previously shown to be directed to target mRNAs by RNA-binding proteins. This was accomplished by the F-box-containing protein CFB2 in the case of EIF4E6. Depletion of EIF4E6 led to a global translational shutdown in BSFs. Evidence in support of EIF4E3 providing additional translational initiation at VSG mRNAs was obtained, likely to help the parasite cope with the high demand for VSG proteins. Collectively, the results presented highlight the independent roles of the different cap-binding proteins in T. brucei, and loss of a particular cap-binding protein can generally not be compensated for by others at particular life cycle stages