Testing of internal translation initiation via dicistronic constructs in yeast is complicated by production of extraneous transcripts

Introduction of sequences of interest into an intercistronic spacer (ICS) of dual reporter plasmids is the main experimental set-up used to identify and study internal ribosome entry sites (IRESs). We studied internal initiation of translation in yeast using the dicistronic approach. Three viral sequences and a polylinker-derived reference sequence were inserted into the ICS of a dual reporter plasmid upstream of the firefly luciferase gene, luc. LUC expression was taken as a putative indication of internal translation initiation from the studied sequences. Interestingly, all sequences mediated 3′ LUC expression. However, northern blot analysis revealed that in addition to the dicistronic mRNAs, transcripts containing only the LUC-encoding sequences were produced from the plasmids. Electroporation studies with in vitro synthesized mRNAs showed that expression from the 3′ cistron of the dicistronic mRNAs was below the level of detection. This suggested that the observed LUC expression from yeast transformed with the dicistronic expression plasmids did not originate from dicistronic messages. Deletion of the promoter increased 3′ LUC expression. Similarly, repression of transcription prevented 5′ cistron expression whereas 3′ LUC expression was stimulated. These results suggested that the observed LUC expression did not result from partially degraded or spliced dicistronic RNAs but rather from transcripts synthesized from cryptic promoters. Despite high LUC activity, northern blot analysis detected few transcripts from yeast transformed with the promoterless constructs. Therefore, our data indicate that the functional assay is more revealing than RNA analysis in the case of very sensitive reporter genes generally used in IRES studies. Furthermore, our studies show that there is a clear need for detailed analysis prior to concluding the mechanism of 3′ cistron expression.