Dr Richard Obexer

Department of Chemistry

Manchester Institute of Biotechnology

The University of Manchester

Abstract

Biocatalysis, which harnesses enzymes to produce commodity chemicals and pharmaceuticals, is a rapidly expanding technology in the chemical industry. Apart from their high efficiencies and selectivities, enzymes are also particularly attractive catalysts as they can be engineered to adapt substrate scope, increase activity and stability under process conditions, or even unlock new chemistries. The introduction of directed evolution was a key development in enabling efficient enzyme engineering, as our quantitative understanding of enzyme catalysis remains limited. However, the success of directed evolution is strongly dependent on the number of analysed variants (throughput). Thus, various assay formats have been developed spanning throughputs from hundreds to billions of variants per round, including multi-well plate assays, in vivo selection systems, and microfluidic assays.

Here, I will present our recent work on the directed evolution of de novo designed enzymes, using a combination of multi-well plate assays and ultra-high throughput droplet-based microfluidics allowing us to screen up to 107 variants per day. Specifically, we have focused on recently designed de novo esterases that feature a catalytic triad, which is one of the most ubiquitous and sophisticated catalytic motifs in nature. By independently evolving the original serine-catalytic-triad enzyme alongside its cysteine-catalytic-triad counterpart, we performed a divergent evolution experiment that allows us to compare how two different catalytic nucleophiles shape the evolutionary trajectory of a multistep reaction within the same scaffold. Overall, we have so far been able to improve these designs by up to 50-fold and, importantly, we could demonstrate that the discovered mutations are exclusive to their respective catalytic nucleophile. Our results highlight that, despite the mechanistic similarity between serine and cysteine catalytic triads, the two active sites diverge onto distinct evolutionary trajectories from the earliest stages of evolution, and that mutations distal to the active site contribute to catalytic efficiency. These results, together with the broader exploration of the evolvability and sequence space of AI-designed de novo enzymes, provide new insights that will guide the design and engineering of the next generation of designer biocatalysts.

Biography

Following his PhD in under the supervision of D. Hilvert at ETH Zurich, Richard carried out postdoctoral research with H. Suga at the University of Tokyo, and subsequently with D. N. Woolfson at the University of Bristol. Richard started his independent career in 2023 at the Manchester Institute of Biotechnology at the University of Manchester, where he is currently a BBSRC Discovery Fellow. His research focus lies in protein engineering with new functions by directed evolution using ultra-high throughput techniques.