Efficient translation initiation directed by the 900-nucleotide-long and GC-rich 5' untranslated region of the human retrotransposon LINE-1 mRNA is strictly cap dependent rather than internal ribosome entry site mediated

Retrotransposon L1 is a mobile genetic element of the LINE family that is extremely widespread in the mammalian genome. It encodes a dicistronic mRNA, which is exceptionally rare among eukaryotic cellular mRNAs. The extremely long and GC-rich L1 5' untranslated region (5'UTR) directs synthesis of numerous copies of RNA-binding protein ORF1p per mRNA. One could suggest that the 5'UTR of L1 mRNA contained a powerful internal ribosome entry site (IRES) element. Using transfection of cultured cells with the polyadenylated monocistronic (L1 5'UTR-Fluc) or bicistronic (Rluc-L1 5'UTR-Fluc) RNA constructs, capped or uncapped, it has been firmly established that the 5'UTR of L1 does not contain an IRES. Uncapping reduces the initiation activity of the L1 5'UTR to that of background. Moreover, the translation is inhibited by upstream AUG codons in the 5'UTR. Nevertheless, this cap-dependent initiation activity of the L1 5'UTR was unexpectedly high and resembles that of the beta-actin 5'UTR (84 nucleotides long). Strikingly, the deletion of up to 80% of the nucleotide sequence of the L1 5'UTR, with most of its stem loops, does not significantly change its translation initiation efficiency. These data can modify current ideas on mechanisms used by 40S ribosomal subunits to cope with complex 5'UTRs and call into question the conception that every long GC-rich 5'UTR working with a high efficiency has to contain an IRES. Our data also demonstrate that the ORF2 translation initiation is not directed by internal initiation, either. It is very inefficient and presumably based on a reinitiation event.