(562a) Re-Engineering Multicloning Sites for Function and Convenience
Metabolic engineering efforts require well-tuned expression of heterologous genes. Often, this task is accomplished through the use of plasmids and synthetic genetic circuits intended to express controlled levels of any gene of interest. These expression cassettes usually contain pre-defined multiple cloning sites (MCSs) which enable the facile expression and cloning of recombinant genes. In all synthetic biology applications, multiple cloning sites are thought to be benign, non-interacting elements that exist for mere convenience. However, a promoter element is usually placed upstream of the MCS, resulting in additions to the 5’ untranslated region which may influence downstream gene expression in eukaryotes. Here, we demonstrate that mRNA secondary structure in common MCSs can significantly inhibit protein translation in multiple transcriptional contexts, especially for short, codon-optimized genes. Furthermore, we show that this inhibition is highly dependent on cloning location, resulting in drastic, nonmonotonic changes in protein production as different restriction sites are used. To mitigate this effect, we developed a model of structure-based translation inhibition which can predict the effects of mRNA secondary structure on protein expression in eukaryotes, in particular, the yeast Saccharomyces cerevisiae. This model enables the rational re-engineering of multicloning sites to allow consistently high levels of translation regardless of gene context. We demonstrate this utility through experimental validation of these improved genetic elements. Several novel multicloning sites were developed for use in conjunction with a wide range of promoters which show significant improvement over commonly used constructs, validating the predictive ability of our model. Moreover, these experiments indicate that rational, context-specific optimization of biological parts represents both an attainable and desirable goal for improvement of biosynthetic parameters. Finally, these results demonstrate that synthetic biology parts interact, and thus cannot be designed independently.