Directed evolution of artificial oxidoreductases.

  • Jack Steventon

Student thesis: Doctoral ThesisDoctor of Philosophy (PhD)


To minimise waste and maximise product yields, synthetic chemists have been designing catalysts for chemical reactions for centuries. But since primitive biological systems have existed, so have biological catalysts, termed enzymes. Natural enzymes have been subject to evolutionary pressures for millions of years driving the development of a diverse array of highly efficient, specialised catalysts. The field of protein design seeks to build enzymes from scratch as a test of our understanding of enzymes and to produce catalysts for novel, unnatural reactions. C45 is a 4-helix bundle with a prosthetic heme cofactor and is a de novo peroxidase with high catalytic efficiency for the turnover of many peroxidase substrates. It is an example of a maquette. Maquettes are a reduced complexity approach to protein design that aims to elucidate the minimum requirements for catalysis. This work aims to improve upon these catalytic properties and in doing so gain further insight into peroxidative catalysis. A peroxidase reaction couples the reduction of H2O2 to the oxidation of an organic substrate. For C45, the catalytic efficiency for the reduction of H2O2 is lower than that for the oxidation step, therefore, this property was targeted for improvement. To achieve this a technique named directed evolution (DE) was utilised. This process mimics the processes of molecular evolution by randomly altering a protein and selecting a protein variant with the desired change in function, in an iterative cycle until the desired outcome is achieved.

A laboratory evolution experiment requires the screening of protein libraries in high-throughput. Chapter 3 describes the development of a medium-throughput evolution protocol for the DE of peroxidases. This enabled the overexpression, purification and screening of protein variants, on a microplate scale. Liquid-handling robotics were used to increase the throughput and consistency of results. Chapter 4 lays the groundwork for the evolution of C45. epPCR and saturation mutagenesis libraries were constructed and checked for consistency and protein quality. An assay was then developed that yielded consistent results for detecting peroxidase activity in clarified cell-lysate.

C45 was evolved for enhanced peroxidase activity in chapter 5. epPCR libraries targeted the entire protein-coding gene and four rounds of evolution were performed. As a result, AP3.1 is 10 mutations away from C45 and displays random-coil secondary structure. Despite this the catalytic efficiency for the activation H2O2 was improved 1.8-fold over C45. Saturation mutagenesis libraries were developed targeting the distal heme face of C45. The greatest enhancement in catalytic efficiency was displayed by the point mutant F18R. The introduction of arginine into the active site is also found in natural peroxidases and is involved in H2O2 binding, activation, and catalytic intermediate stabilisation which correlates with the improved peroxide activation properties of C45_F18R.

C45 also acts as a carbene transferase (CT) in the asymmetric cyclopropanation of styrene. A low-throughput screen, of existing peroxidases, for carbene transferase activity, was per- formed in chapter 6. This was as a proof-of-concept experiment to ascertain whether this property of C45 could also be enhanced or modified. The evolved peroxidase AP3.2 was found to catalyse this reaction with 100 % ee for the S,S enantiomer, the opposite product to that of C45, with 48 % product conversion. This confirms that C45 is also suitable for CT evolution experiments.

In chapter 7 the incorporation of the unnatural amino acid N-methyl histidine (NMH) was trialled in the C-type maquette C45 and the B-type maquette m4D2. Unnatural amino acids are a way to modify and enhance proteins outside of the scope of what is possible with natural amino acids. An orthogonal tRNA synthetase/ tRNA pair, selective for NMH, was co-expressed with the desired maquette. NMH incorporation was successful only with m4D2 and due to low yields, NMH-m4D2 was not able to be characterised. However, it is possible to incorporate NMH into a maquette and future work will determine the effect of NMH on redox and catalytic properties.

The work presented shows how the already impressive catalytic properties of a de novo enzyme can be improved further. From the maquette platform, it is possible to evolve enzymes for different reactivities, and also develop powerful catalysts. Homogenous secondary structure is not fundamental to the peroxidative performance of C45-derived maquettes and the conserved features of natural peroxidases show similar function in manmade peroxidases. From here it is possible to develop a suite of specialist enzymes, implementing powerful catalysis, through the evolutionary framework provided.
Date of Award23 Jun 2020
Original languageEnglish
Awarding Institution
  • The University of Bristol
SupervisorAdrian J Mulholland (Supervisor) & J L R Anderson (Supervisor)


  • peroxidase
  • enzyme
  • evolution
  • high-throughput
  • artificial

Cite this