Construction And Validation Of Phage- Assisted Evolution System Of Protein-protein Interaction

Protein-protein interactions (PPI) comprise core function of proteins. To optimize PPI or even to create new PPI via directed evolution are important goals of protein engineering. In conventional directed evolution, target genes need to be manually separated in each round of selection, followed by manual construction of mutant libraries for the next round. Only a few rounds (or generations) of directed evolution can be executed. In the phage-assisted continuous evolution (PACE) system (Esvelt et al, Nature,472(7344),499-503), such manual interventions have been omitted. The target gene carried by the phage can be evolved at a rate of several to a few tens of minute per generation. This makes PACE a technique of great potential for protein engineering. Theoretically, PACE can be applied to the directed evolution of various types of protein function. Still, its applications so far have been limited to RNA polymerases, which directly carries out transcription. In principle, PACE can also be applied to the directed evolution of PPI, provided that the strength of PPI can be converted to regulated reproduction M13 phage. In the current thesis, we constructed and verified a core module for using PACE to evolve PPI. This module employs a bacterial two-hybrid system to couple target PPI to the reproduction of M13 phage. PACE is a relatively complex system, whose running needs several modules to work together in a coordinated manner. In this regard, this thesis also contains work on the improvement of a mutagenesis module as well as on the testing and debugging of a continuous flow and selection platform. The work presented by the current thesis may form the basis of a complete workflow to evolve PPI with PACE. The contents of individual chapters are summarized below.In Chapter Ⅰ, I give brief overviews of PPI and directed evolution. Then I introduce the working mechanism of PACE, emphasizing the advantage of PACE relative to conventional directed evolution.In Chapter Ⅱ, I report the construction and verification of a bacterial two-hybrid system. It couples the propagation of M13 phage to PPI. This system is consist of a phage plasmid and an accessory plasmid. In the phage plasmid, the coding gene of protein PⅢ has been deleted. In addition, a target gene that encodes one of the proteins participating in the PPI has been introduced. The accessory plasmid contains the rest of the two-hybrid system as well as a gene that encodes PⅢ and is expected to be regulated by PPI. The key parameters here include the expression strengths of the upstream promoter of PⅢ, with and without PPI, respectively. If the promoter is too strong, it may cause leaking expression of PⅢ, leaving the phage-uninfected bacteria to resist the infection by phage; if the promoter is too weak, the phage-infected bacteria may not be able to produce enough PⅢ to support phage propagation. A range of promoters from the Standard Biological Parts contained in the iGEM part library have been tested. An appropriate promoter have been identified. After that, I used the interaction between ras-raf to prove that the modified phage can propagate at and only at the presence of the target PPI. Furthermore, both the expression level of PⅢ and the rate of phage reproduction are dependent on the strength of PPI.To implement the complete workflow of PACE locally, a range of tasks involving platform setup, testing and debugging need to be finished. In Chapter III, I report the work I have done in this regard. One is to tune up the performance of the mutagenesis plasmid. The role of that plasmid is to introduce high mutation rate. Initial assay of such a plasmid obtained from another group suggested that the mutation rate could not reach expectation. Analysis suggested that this may have to do with its suboptimal response to the arabinose inducer. I re-engineered this plasmid to make it more sensitive to inducer. We have also executed the complete workflow of PACE a number of times. Through these we found that a key factor for PACE is the balance between the respective rates of bacterial reproduction, of phage propagation, and of volume substitution. The feasible parameter window for a proper balance may be quite narrow. As results, our trials with different culture conditions and different volume substitution rate have not yet been able to maintain a constant density of phage in the system. Instead, the number of phages decreased continuously. As trials by real experiments take long and are costly, to locate feasible parameters more precisely, I built a mathematical model of the process and wrote a program to carry out simulations by the model. This model may guide the selection of parameters in future experiments.