Cellular mRNAs access second ORFs using a novel amino acid sequence-dependent coupled translation termination–reinitiation mechanism

Abstract

Polycistronic transcripts are rare in the human genome as unusual mechanisms are needed to translate the downstream ORFs, including leaky scanning, IRESs, or coupled termination–reinitiation mechanisms. Here the authors have devised an algorithm to identify mRNAs in the human transcriptome with two overlapping ORFs where a coupled termination–reinitiation mechanism might be relevant. Of the thousands of such transcripts identified, 22 of 24 were seen to express a protein from the second ORF suggesting that 3′ UTRs themselves have considerable coding potential. Five of these transcripts appeared to depend on a termination–reinitiation mechanism, and one of these depended on a specific aspartate-rich repeat peptide sequence at the carboxyl terminus of ORF1 for the coupling mechanism to be effective.

Abstract

Polycistronic transcripts are considered rare in the human genome. Initiation of translation of internal ORFs of eukaryotic genes has been shown to use either leaky scanning or highly structured IRES regions to access initiation codons. Studies on mammalian viruses identified a mechanism of coupled translation termination–reinitiation that allows translation of an additional ORF. Here, the ribosome terminating translation of ORF-1 translocates upstream to reinitiate translation of ORF-2. We have devised an algorithm to identify mRNAs in the human transcriptome in which the major ORF-1 overlaps a second ORF capable of encoding a product of at least 50 aa in length. This identified 4368 transcripts representing 2214 genes. We investigated 24 transcripts, 22 of which were shown to express a protein from ORF-2 highlighting that 3′ UTRs contain protein-coding potential more frequently than previously suspected. Five transcripts accessed ORF-2 using a process of coupled translation termination–reinitiation. Analysis of one transcript, encoding the CASQ2 protein, showed that the mechanism by which the coupling process of the cellular mRNAs was achieved was novel. This process was not directed by the mRNA sequence but required an aspartate-rich repeat region at the carboxyl terminus of the terminating ORF-1 protein. Introduction of wobble mutations for the aspartate codon had no effect, whereas replacing aspartate for glutamate repeats eliminated translational coupling. This is the first description of a coordinated expression of two proteins from cellular mRNAs using a coupled translation termination–reinitiation process and is the first example of such a process being determined at the amino acid level.