Figure 5. Improved luciferase activity when using an optimized Kozak consensus. A: HEK 293T/17, Y-79, and WERI-RB-1 cells were transfected with dual-luciferase constructs that were identical but for the Kozak consensus of the Renilla luciferase. One construct contained the original Kozak consensus (GCCGCCACC) and the other the retina-derived Kozak consensus (ACCGAGACC) at positions −9 to −1 of the Renilla coding sequence. The remaining construct sequence, including the firefly expression cassette, was identical for both plasmids. B: Luciferase activity was assessed 48 h post-transfection with Renilla luciferase activity presented relative to firefly activity. Data were confirmed to be normally distributed (Appendix 1), and a two-way analysis of variance (ANOVA) identified a significant influence of the Kozak sequence on Renilla luciferase activity (p<0.0001, n = 24 from four independent biological replicates with six technical replicates of each). **p<0.004, *p<0.03. Error bars represent the standard error of the mean (SEM).