ИСТИНА

It is widely assumed that the secondary structure of 5’ UTR is a key feature which affects the efficiency of cap-dependent translation. Thus for long and highly structured 5’ UTRs alternative mechanisms, such as internal ribosomal entry, have been proposed. Here we demonstrate that the m7G-capped reporter mRNA with the 577 nt long 5’ UTR of Apaf-1 mRNA can be translated both in vivo and in vitro (in nuclease-untreated cytoplasmic extract from ascite cells Krebs-2) with efficiency comparable to that of 53 nt long 5’ UTR of beta-globin, in spite of the fact that this leader contains several highly structured domains which are believed to constitute an IRES. It should be noted that placement of the Apaf-1 5’ UTR in the intercistronic position of a bicistronic mRNA results in an extremely weak translation of the second cistron as compared with the respective monocistronic construct. When m7G-cap is replaced by non-functional A-cap for monocistronic mRNAs, the translation driven by the Apaf-1 5’ UTR is strongly reduced. Insertion of uAUG codons in the 5’ UTR of Apaf-1 results in a dramatic reduction of translation not only for m7G-capped, but also for A-capped monocistronic mRNAs. Consecutive deletions of the structural domains in the Apaf-1 leader shows only modest changes in translation, irrespective of whether monocistronic transcripts are m7G- or A-capped. Thus, these data strongly suggest that the 40S ribosome is capable of efficient scanning the 5’ UTR of Apaf-1 over its entire length and this is the case for both capped and uncapped RNAs. However, the intriguing fact found in this work is that in some cell lines the Apaf-1 5’ UTR is nevertheless more resistant (5-10 fold) to omission of the cap than other cellular 5’ UTRs used in this study. The mechanism which may determine the different cap-dependence of translation for various 5’ UTRs and which is alternative to the concept of cellular IRESs will be discussed.