Investigation of a novel interacting partner of the mRNA 5’ methylguanosine capbinding protein eIF4E

Abstract

The canonical model for initiating the translation of capped mRNAs entails the eIF4F complex (eIF4E, which binds the cap; eIF4G, which is a scaffold adapter that helps circularise the mRNA through its binding to eIF4E and the 3’end-bound PABP; eIF4A, which is an RNA helicase). Accumulating evidence has unveiled alternate modes of translational initiation of capped mRNAs that either operate alongside the canonical model or in conditions when the eIF4F complex is inactivated such as under stress. These mechanisms are either cap-dependent or independent and they might target specific mRNAs based on characteristic features in their sequences. The eIF3g subunit possesses a short linear motif (SLiM) that bears some resemblance to the eIF4E binding site. We hypothesised that eIF3g binds to eIF4E and mediates the translation of a subset of mRNAs. Using the NanoBiT protein-protein interaction assay, we provide in vivo evidence that eIF3g engages in a novel weak interaction with the cap-binding protein eIF4E1. Mutagenesis analysis revealed that this interaction is mediated, at least in part, by the short linear motif (SLiM) in eIF3g and the conserved tryptophan 73 in eIF4E1. Polysomal profiling and puromycylation analysis in siRNA knockdown and rescue expression experiments elucidated that disrupting this interaction interfered with mRNA translation initiation at the global level. Furthermore, it impaired the normal proliferation of the cells. Using CRISPR/Cas9 gene editing technology, we generated a novel HEK 293 cell line which expresses a modified eIF3g in which the SLiM amino acid sequence was replaced with alanine residues. This presumably interferes with eIF3g association with eIF4E1. Similar to our findings with siRNA and rescue experiments, this cell line displayed a major downregulation in the translational initiation process and its proliferation was slightly slower. We performed genome-wide translational profiling to examine whether some transcripts are specifically translated by eIF3g and eIF4E binding. No such mRNAs were returned by this analysis. However, even though none of them overlapped, this analysis returned some as yet unvalidated candidates from the siRNA knockdown and rescue experiments and many candidates from cells with a homozygous knock-in of the SLiM alanine mutation generated by CRISPR/Cas9 whose translational efficiency might be influenced by the novel interaction. Using our findings, coupled with previous structural studies, we propose a model in which the novel binding between eIF3g and eIF4E1 might be taking place during the translational initiation phase.