The role of mRNA 5'-noncoding and 3'-end sequences on 40S ribosomal subunit recruitment, and how RNA viruses successfully compete with cellular mRNAs to ensure their own protein synthesis.

Since the elaboration of the scanning model to explain eukaryotic translation initiation, alternative hypotheses have gained support. Cap and 5' end-independent recruitment of the 40S ribosomal subunit conferred by the presence of an internal ribosome entry segment (IRES) in the 5'UTR of the mRNA is widely accepted, and has been formally and definitively proven for a picornavirus. However, the mechanism of IRES function remains essentially a black box. Using the complex viral IRESes as model systems, approaches taken to shed light on the mystery include systematic comparisons and molecular genetic analyses. The hypothesis that actively translated mRNAs are circular, rather than linear, molecules is based on rather indirect evidence. This model has invoked a revision of the image of 40S ribosomal subunit recruitment, to include recycling from the mRNA 3'- to the 5'-end in addition to true de novo 5'-end directed entry. Biochemical and genetic studies are used to define the network of interactions necessary for efficient ribosome recruitment. This has lent weight to the concept of mRNA 5'-3' cross-talk and clarified the mechanics of how this enhances translation efficiency. These refinements and revisions to the model of translation initiation form the core of this review, with current knowledge being considered from the perspective on how host-cell translation could yield to selective viral translation via the phenomenon of translational shut-off.