ИСТИНА

Reconstruction of initiation complexes from individual components (ribosome subunits, Met-tRNA, purified initiation factors and mRNA) is the most promising method for studying the molecular mechanisms of translational initiation in mammals. However until now the applications of this fruitful approach to the investigation of a cap-dependent translation was limited to a small set of mRNA species due to failure to assemble 48S complexes with mRNAs containing any GC-rich secondary structures in 5'-UTR (e.g. mRNA for beta-actin, hsp70, late adenovirus mRNA, and much more). Nevertheless, such mRNAs are successfully translated in cell lysate, so the question arises if there is any unknown initiation factor(s) in lysate allowing them to initiate translation. Here we report that initiation of translation on such mRNAs does not require any additional components besides canonical ones, but initiation factor eIF4B is needed in rather greater amount then in the case of mRNAs with unstructured 5'-UTRs, and eIF4B preparation must be fresh as this factor is extremely unstable in purified state and looses its activity during long storage even in deep frost. In addition, some technical modifications of the original 48S reconstruction protocol were found to be necessary to improve 48S assembly. Furthermore, for the first time we studied quantitative requirements of mRNA to eIF4B and also to the other components of RNA helicase complex (eIF4A and eIF4F) in real translation initiation system on the set of closely related RNA transcripts. eIF2 requirements were analyzed as well and were found to be also increased in the case of mRNAs with structured 5'-UTR, allowing us to propose a model in which eIF2 assists 40S subunit to bind to mRNA's 5'-leader and to scan toward the initiation codon.