21BBSRC-NSF/BIO - Evolving quantum mechanical tunnelling in enzymes

H-transfer reactions involving hydride, hydrogen atom or proton transfers are ubiquitous in enzyme-catalysed reactions. It is now well established that these reactions can occur via a mechanism involving some degree of quantum mechanical tunnelling (QMT), and thus are a feature of quantum biology. The QMT contribution is highly variable and an open question remains whether there has been evolutionary pressure to select for QMT during enzyme catalysis. Theoretically, it is possible to increase the rate of an enzyme-catalysed H-transfer reaction by increasing the tunnelling contribution, so this could offer a fitness advantage. However, typical strategies to increase the tunnelling contribution are also likely to increase the rate of classical (over-the-barrier) H-transfer.
In this proposal we aim to combine experimental studies of H-tunnelling in enzyme catalysed reactions with directed (laboratory) evolution (DE) and computational chemistry. We will study two unrelated enzymes in parallel:
(i), alcohol dehydrogenase (ADH) is a well-studied model hydride-tunnelling and industrial biocatalysis enzyme.
(ii), MBHase is a de novo Morita-Baylis-Hillman enzyme, that we have recently developed through DE.
DE will be performed using high- and medium-throughput screening methods that select for activity with both protiated and deuterated substrates (so either H- or D- transfer occurs) in order to allow parallel evolution of enzymes evolved specifically for H- and D- transfer. Selected variants will be experimentally characterised using additional kinetics methods (stopped-flow pre-steady state rate constants and kinetics isotope effects, and temperature dependencies) and by X-ray crystallography to solve structures where possible. These selected variants will also be investigated using computational chemistry, which will allow the QMT contribution to be determined. This approach will allow a new and more comprehensive approach to the exploration of the relationships between H- and D- transfer kinetics and QMT during enzyme evolution, and will establish whether selecting for improved D-transfer kinetics provides new avenues for improving enzyme performance.
We will take advantage of our recent breakthroughs in correlating computed QMT contributions with experimental kinetics isotope effects (ratio of H- and D-transfer kinetics), in evolving MBHase and in the development of the free energy QM/MM computational methods of characterising enzyme-catalysed reaction chemistry. To the best of our knowledge, this is the first attempt at using DE to probe QMT in enzymes, or to use deuterium kinetics as selection during DE experiments. A recent paper showed improvement in a laboratory evolved enzyme for deuteration of short chain acids, but selection was performed with protium, not deuterium. The project is thus both timely and novel and will provide new insight into the role of QMT during evolution of H-transfer enzymes. Further, as efficient routes to selective deuteration of pharmaceuticals is currently a hot topic, and many industrially-important enzymes catalyse H-transfers, the methodology also promises to lead to a new approach to enzyme optimisation for applications in synthesis and industrial biocatalysis.

It is now well established that H-transfer reactions can occur via a mechanism involving some degree of quantum mechanical tunnelling (QMT). An open question remains whether there has been evolutionary pressure to select for QMT during enzyme catalysis. Theoretically, it is possible to increase the rate of an H-transfer reaction by increasing the tunnelling contribution, so this could offer a fitness advantage. However, typical strategies to increase the tunnelling contribution are also likely to increase the rate of classical H-transfer. In this proposal we aim to combine experimental studies of H-tunnelling in enzyme catalysed reactions with directed evolution (DE) and computational chemistry. We will study two unrelated enzymes:
(i), alcohol dehydrogenase (ADH), a model hydride-tunnelling and industrial biocatalysis enzyme.
(ii), MBHase, a de novo Morita-Baylis-Hillman enzyme that we recently developed through directed evolution (DE).
We will develop new methodology for the parallel evolution of enzymes evolved specifically for H- and D- transfer, using separate selection with protiated and deuterated substrates. Selected variants will be experimentally characterised using additional kinetics methods (pre-steady state rate constants, KIEs, temperature dependencies, etc) and by X-ray crystallography to solve structures where possible. These variants will also be investigated using computational chemistry: QM/MM calculations coupled with classical and quantum Transition Path Sampling (TPS) and TPS free energy computations. This will allow the QMT contribution to be determined and then mapped to (laboratory) evolution trajectories for ADH and MBHase to determine how QMT contributes to rate improvements generated through laboratory selection. Selection on deuterated substrates will also allow us to establish whether selecting for improved D-transfer kinetics provides new avenues for improving enzyme performance, e.g. for improvement of industrial biocatalysts.