Figure 4.

Testing the IRES activity of the five selected 5′UTRs using bicistronic reporter constructs. (A) The IRES activity of each 5′UTR in βgal/UTR/CAT bicistronic construct measured as CAT protein levels relative to β-galactosidase activity. All values are relative to the empty vector, which is set to 1. The 5′UTR of ATF4 and the empty vector are negative controls. (B) Testing for cryptic promoter activity within the UTRs. Promoter activity was tested in bicistronic constructs in which the CMV promoter that drives expression of the bicistronic message was excised. These constructs were transfected into 293 cells and CAT and β-galactosidase expression was determined. A monocistronic plasmid in which CAT gene expression is driven by the CMV promoter was used as a positive control, as well as the bicistronic constructs of AQP4 and ELG1 that contain the CMV promoter. The neomycin phosphotransferase II (Neo PTII) gene, which also resides on the transfected plasmids, was used as a transfection control. ELISA measurements of Neo PTII protein levels are presented as O.D. values. (C) The AQP4 and ELG1 bicistronic messages were tested for spurious splicing events using quantitative RT-PCR of amplicons within the βgal and CAT ORFs. The fold differences of the CAT and βgal amplicons was calculated as 2^{−[Ct(βgal) – Ct(CAT)]} and was plotted relative to the ATF4 negative splicing control, which has no IRES activity. Values represent the mean of nine biological samples and error bars represent ±S.E.M. of nine biological samples. (D) Monocistronic UTR-CAT constructs with and without a hairpin [ΔG = −60 kcal/mol as calculated by RNAStructure (45)] that was used to halt ribosome scanning were co-transfected with a βgal expressing plasmid. Each set of values is normalized to its non-hairpin constructs and β-galactosidase activity.