With the modular nature of our project, we had many moving parts which required extensive engineering, from plasmid design to phage engineering. One of these components was expression of our metal-binding proteins and encapsulins in an E. coli model system. To first optimise expression of each of these proteins, we aimed to express them individually. In our first iteration of this component, we designed and cloned our plasmids containing genes encoding various metal-binding proteins and encapsulins, a process of its own, and transformed BL21(DE3) E. coli cells. However, we were unable to isolate our target proteins from these transformed cells nor confirm overexpression. Considering our metal binding proteins contain high levels of cysteine residues which may complicate their folding and stability in a non-specialised host, we considered whether using a different E. coli strain better suited to expression of such high-cysteine proteins might yield better results. Learning from our first cycle, we decided to try transforming Shuffle E. coli cells, which are characterised by their ability to express correctly-folded high-cysteine proteins1. In this second cycle, our protein expression was more successful. In light of these results, we proceeded with a third cycle of engineering, this time aiming to co-express metal-binding proteins and encapsulins.
Another component of our system involves using engineering phages to deliver genes encoding metal-binding proteins and encapsulins into bacteria. An important component to this arm of our project was the design of an efficient CRISPR/Cas9 system which could select for recombinant phages. However, we did not know which genes in our chosen phages were critical to their replication. Guided by literature on other B. subtilis phages, we designed guide RNAs (gRNAs) targeting various hypothetical proteins in the Φ29 genome. We tried to clone our designed gRNAs into a Cas9 plasmid, but this step posed an unexpected challenge. In our first attempt at an infusion transformation to construct our plasmid, we observed no transformed colonies. We learned that issues with infusion transformation could occur when using very short fragments of nucleic acids, such as our gRNAs. In our second cycle of transformation, we tried to optimise our protocol, increasing the ratio of gRNA. We observed some transformed colonies, but still very little success overall. Sequencing plasmids from these transformed colonies, we found they did contain the target plasmid, indicating the issue in our previous protocol was indeed the low ratio of gRNA. Utilising this information, we went through a third cycle of infusion transformation where we further increased the ratio of gRNA which resulted in much more success, allowing us to move on to the next step of testing our gRNAs.