Investigating application of mRNA elongation control models to the design of specific target region sequence variants and the impact on protein synthesis.

During disease, or under cell stress, post-transcriptional programmes are initiated to orchestrate an appropriate cellular response through the regulation and control of protein synthesis. One such avenue for regulation is during translation. Traditionally, much of translational control was thought to occur during

the initiation stage, but recent studies suggest the elongation stage is also a central regulatory node. It has been shown that message specific, elongation control is central to disease response mechanisms in both tumorigenesis and neurogenerative disorders. However, it is unclear to what extent factors such as codon usage and the relative abundance of their cognate tRNA impact this regulation. Transcripts can experience ribosomal pausing (stalling) during elongation which may be attributed to heightened dependence on eEF2 translocation, but is also due to tRNA abundance. The university of Kent has developed a computational elongation programme which can model these interactions along a transcripts sequence and can identify regions decoding at a slow rate resulting in ribosomal stalling and potentially a mark of elongation control. Here, we selectively modified these areas for 3 mRNA transcripts by substituting canonical codons with synonymous codons with different decoding rates in order to generate multiple transcripts with differing decoding times. We found that optimising these areas, by using synonymous codons that correspond to abundant cognate tRNA, had a major impact on unoptimised transcripts at the 5' region immediately following the start codon and can be used to tune protein production. Conversely, diosoptimising these regions lead to decreases in protein production compared to native sequences, and in some cases completely prevented translation. slow regions identified further downstream of the start codon were also shown to impact protein production when modified but less reliably. Other codon alterations that weren't fully optimal nor disoptimal lead to a variety of impacts on protein production showcasing the complicated nature of elongation control. Differences were also observed for secreted proteins which may be affected by alteration made at the 5' end of their mRNA transcript more so than non-secretory proteins. Mildly hypothermic conditions also heightened differences in protein production for many sequences compare to protein production of their native sequences. These results assist in confirming the elongation model's accuracy in predicting slow regions involved in elongational control and modifications achieved in this manner could be broadly applied to biotherapeutic production as well as incorporated into medical treatments such as gene therapy or mRNA vaccines.