Mechanisms and Determinants of RNA Turnover: Plant IRESs and Polycistrons for Metabolic Engineering

There is a strong need for tools that allow multiple transgenes to be expressed in genetically engineered plants. For the last 30 years it has been believed that nearly all eukaryotic mRNAs were monocistronic, with ribosomes entering at the 5' end and scanning through the 5'UTR to the first AUG codon. It is now clear that perhaps 3% of vertebrate and yeast mRNAs utilize IRESs (Internal Ribosome Entry Sites) within their 5'UTRs to promote the internal entry of ribosomes to mRNAs and subsequent translation of protein without scanning. The working hypothesis behind this proposal is that IRES sequences function in plants and can be used to engineer the efficient co-expression of multiple proteins from polycistronic transcripts. Our goal was to translate multiple proteins from single polycistroic mRNAs. We cloned four IRESs from the following sources: CrTMV (plant virus), EMCV (human encephalomyocarditis virus), eIF4G (human), and c-myc (human). All four IRES were cloned into a specially designed test vector with the strong constitutive ACT2 actin regulatory sequences and flanked by multicloning sites for two reporter genes. These four IRESs were tested in three different test systems with strong paired reporter activities: two fluorescent proteins, two mercury resistance enzymes, and two biosynthetic enzymes making thiolpeptides. All of the four IRES constructs with the fluorescent protein reporter genes were tested for transient expression after particle gun bombardment of tobacco BY2 cells. Three of the IRESs gave reasonable activity (10%-40%) for the second cistron fluorescent reporter (DsRFP) relative to the first cistron reporter (GFP). As a control, translational blocking sequence placed at the 5' end of duplicate constructs had little effect on activity from the second cistron, but blocked the first cistron. These initial positive data lead us to examine the four IRES constructs with three pairs of reporters in hundreds of transgenic Arabidopsis plants. All constructs expressed the first cistron (GFP, merB, and ECS) at high levels, but none expressed the second cistron (DsRFP, merA, GS) at significant levels. Our results from several different reporter systems did not support continuing this examination of IRES activity in plants