Improved protein production and codon optimization analyses in Escherichia coli by bicistronic design

Abstract

SummaryDifferent codon optimization algorithms are available that aim at improving protein production by optimizing translation elongation. In these algorithms, it is generally not considered how the altered protein coding sequence will affect the secondary structure of the corresponding RNA transcript, particularly not the effect on the 5′‐UTR structure and related ribosome binding site availability. This is a serious drawback, because the influence of codon usage on mRNA secondary structures, especially near the start of a gene, may strongly influence translation initiation. In this study, we aim to reduce the effect of codon usage on translation initiation by applying a bicistronic design (BCD) element. Protein production of several codon‐optimized gene variants is tested in parallel for a BCD and a standard monocistronic design (MCD). We demonstrate that these distinct architectures can drastically change the relative performance of different codon optimization algorithms. We conclude that a BCD is indispensable in future studies that aim to reveal the impact of codon optimization and codon usage correlations. Furthermore, irrespective of the algorithm used, using a BCD does improve protein production compared with an MCD. The overall highest expression from BCDs for both GFP and RFP is at least twofold higher than the highest levels found for the MCDs, while for codon variants having very low expression from the MCD, even 10‐fold to 100‐fold increases in expression were achieved by the BCD. This shows the great potential of the BCD element for recombinant protein production.

Influence of the spacer region between the Shine–Dalgarno box and the start codon for fine‐tuning of the translation efficiency in Escherichia coli

“…, 2013; Nieuwkoop et al . , 2019). Comparison of the measured translation efficiency of mRNAs in our data set with that predicted by RBS Calculator (Salis et al , 2009; Salis, 2011) (Table S1, Predicted translation efficiency) demonstrated relatively good, although not ideal correlation ( r = 0.62).…”

Protein yield is tunable by synonymous codon changes of translation initiation sites

“…They found that weakly structured ribosome binding sites result in high red fluorescent protein levels. This was supported by recent studies using the endogenous folA and adk genes 48 and a dual-reporter system in E. coli 49 . These studies, and many others, support our finding that optimisation of the accessibility of translation initiation sites is a key to improve heterologous protein production.…”

“…Previous studies have used minimum free energy models to define the accessibility of a region of interest 27,48,49 . However, we have discovered that the opening energy is a better choice for modelling accessibility (see Fig 1A for example).…”

Molecular Cloning, Expression and Purification of G-CSF Isoform D, an Alternative Splice Variant of Human G-CSF

“…With using a well-expressed N-terminal protein fusion such as GST, G-CSF isoform D was successfully expressed. 22 We used GST as a fusion expression partner for two main reasons: First, GST as one of the most common fusion tags greatly enhances the expression of difficult-toexpress recombinant proteins by allowing the efficient initiation of translation. 23 High levels expression of the GST-fusion proteins in E. coli may result in the formation of inclusion bodies, as we observed in our experiment.…”