One of the earliest engineered self-selecting systems (SSS), phage-assisted continuous evolution (PACE), has been used to evolve modified proteases, improved Bt toxins, Cas9 variants, and soluble antibodies. However, due to the difficulty of setting up a continuous system, PACE has commonly been implemented as a discretized system, which has proved simpler to assemble and produces results on timescales similar to PACE. Overall, we hoped to implement and improve this discretized system, which we term PREDCEL+ in homage to the 2017 Heidelberg iGEM team, in order to support a wider array of standardized selection schema that can be used to evolve a wider net of proteins.
Like PACE, PREDCEL+ relies on M13 bacteriophage to carry a library of mutant genes. The replication of these viruses is tied to the fitness of the mutant gene such that the most fit gene mutants produce more offspring, which eventually become the majority of the viral population. Conventional PACE and PREDCEL creates this linkage by creating a biocircuit whose output is M13 gene III (gIII), which encodes a phage protein (pIII) involved with phage entry into the cell. However, even low amounts of gIII expression render cells resistant to phage infection, complicating matters. We decided to use a different phage protein involved in mediating phage entry, pVI (encoded by gVI), for our selection based on findings that it does not render cells resistance to further infection, is unstable in the absence of pIII, and leads to higher phage production when expressed at higher rates.
One roadblock encountered early-on when using PACE was that phage may not be able to propagate in culture if they all encode mutants poor at performing the task to be evolved. This leads to ‘phage washout,’ where evolution halts due to the inability of the phage to replicate. This issue was solved by using a genetic element termed a ‘drift cassette’, which can inducibly express gIII or gVI when needed. Typically, drift is induced earlier in the evolution and reduced gradually to zero as the evolution progresses. However, it is unclear how much or when drift should be induced without periodically conducting plaque assays to visualize the amount of phage at any given moment. The Stanford 2019 iGEM team thus developed an ‘infection reporter’ based on GFP fluorescence output to report on what fraction of cells at any given moment in time are infected with phage. GFP expression is driven by the phage shock promoter (Ppsp), which responds to filamentous phage infection (specifically, expression of M13 gIV, which causes pores in the E. coli membrane that allow phage progeny to leave the cell) in addition to osmotic shock, organic solvents, and blockage of secA machinery used for protein export. We used two sequences for Ppsp in our designs - one described by the 2018 FSU iGEM team (part BBa_K2832003) and the other the wild-type Ppsp in plasmid pDB023f1 from Prof. David Liu’s lab (part BBa_K3258000).
To further simplify the cloning of PREDCEL+ and limit the possibility of phages containing mutant proteins propagating outside of engineered host cells, we split our system into three plasmids, which we term the helper plasmid (HP), accessory and mutagenesis plasmid (APMP), and the selection plasmid (SP). The components of each are detailed in the figure below:
To test our designed PREDCEL+ architecture, we decided to recapitulate the evolution done on the T7 RNA Polymerase to allow it to recognize the T3 RNA polymerase promoter conducted in the seminal PACE paper1. We thus used a more specific APMP and SP. All plasmids used are outlined below:
Novel Selection Schema Using PREDCEL+
Once validated, we aimed to use PREDCEL+ to test out a standardized selection schema for biosensor evolution. Making use of the discretized nature of PREDCEL, we thought it easy to transfer between multiple populations of bacteria in such a way that the gene to evolve is selected to be fit in all of them. We thought we could make use of this advantage to select for biosensors by negatively and positively selecting for and against binding, respectively, in the presence and absence of the desired molecule to sense13. As a proof of concept, we decided to evolve the lacI repressor to respond to new substrates and to adjust the dynamic range of the lacI IPTG dose response. To this end, we developed two APMPs - one for positive selection and the other for negative selection. All plasmids used for the evolution are presented below:
Experiments and Results
While we were unable to conduct an evolution using our PREDCEL+ system, we were able to confirm that our co-transformed plasmids produce enough protein to yield viable phage. S2060 E. coli cells pre-transformed with the APMP and HP plasmids were made competent and transformed with the SP (Alternative versions encoding either an alternate start codon in front of gIV or a weak RBS in front of gIV). Plaques formed, indicating M13 phages are being made.
We also were able to characterize our infection reporter designs and showed that GFP expression occurs after infection as expected - see our characterization data on BBa_K2832003. We also found that the infection reporter was at times unreliable, so more extensive characterization work is currently underway (e.g. to determine periods of growth in which the infection reporter is most reliable). Updates on this and our ongoing efforts to conduct our proof of concept and biosensor evolutions using PREDCEL+ will be provided on the relevant Biobrick parts pages.