Heavy enzymes: Probing fast dynamics in enzyme catalysis by mass modulation

A central paradigm in biochemistry is that protein function is defined by structure. However, in solution proteins are inherently dynamic molecules, exhibiting motions on timescales ranging from bond stretches through to slow domain motions and normal mode vibrations. An important open question in enzymology remains the role of such dynamics, and whether motions/vibrations on timescales faster than turnover can couple to chemical steps during catalysis. There has been good progress in establishing the role of slower dynamics such as loop opening/closing during enzyme turnover using NMR approaches, but direct evidence for the coupling of faster (sub-nanosecond) dynamics to chemistry remains illusive and controversial and is based largely on the anomalous temperature dependencies of kinetic isotope effects (KIE; e.g. the ratio of rate constants: kH/kD). The role of such fast dynamics remains an important question, as motions on similar timescales to chemistry have the potential to profoundly affect the reaction outcome, and thus offer a means to control (enzyme) reactivity.

Stable isotopically-labelled proteins (typically with 2H, 13C and/or 15N) have been exploited as an experimental tool for many years, particularly by the NMR and vibrational spectroscopy communities. The implicit assumption has generally been that isotopic labelling does not significantly perturb protein function. However, it was demonstrated that several isotopically labelled 'heavy' enzymes have measurably slower reaction kinetics. These data were interpreted in terms of the 'Born-Oppenheimer approximation', where increased protein mass (due to labelling) alters bond vibrational frequencies without affecting electrostatic properties of the enzyme. In this case, these results suggest that the lower frequency of (fast) bond vibrations in the 'heavy enzymes' may lead to a reduction in conformational sampling and thus chemical barrier crossing; the rate of reaction is proportional to the rate of barrier crossing.

We recently extended the 'heavy enzyme' approach to study vibrational coupling in the Old Yellow enzyme (OYE) pentaerythritol tetranitrate reductase (PETNR). We showed that the temperature dependence of the KIE, which is often used as definitive evidence of protein environmental coupling, is significantly increased in 'heavy' PETNR. This strongly suggests that vibrational coupling can be enhanced by isotopic labelling of proteins. Clearly, the 'heavy enzyme' methodology can be used as a powerful tool to study enzyme coupling and dynamics, but important questions remain. Mass perturbation will affect all vibrations within the protein, so experimental observation of the timescale(s) of any relevant vibrational coupling between protein and chemical coordinate is highly desirable in order to firmly establish the theoretical origin of the 'heavy enzyme' effect. Further, a computational study of a 'heavy' DHFR enzymes suggests that an increased dynamic coupling to the chemical coordinate is detrimental to DHFR catalysis. It is now timely to also consider whether the dynamic coupling of enzyme motions to the chemical coordinate is generally optimised (e.g. by evolution) and thus whether this could be exploited to enhance reactivity or 'drug' enzyme targets.

By combining our unique variable temperature and pressure KIE measurements with the 'heavy' enzyme method, we will study the H transfer reactions during both halves of the catalytic cycles of two homologous OYEs, PETNR and morphinone reductase (MR). Selective cofactor and amino acid labelling we will identify (if present) networks of vibrationally-coupled residues, while molecular dynamics simulations and ultrafast spectroscopies will be used to establish the timescales of such coupled vibrations.

In solution proteins are inherently dynamic molecules, exhibiting motions on timescales ranging from bond stretches (~1000s /cm; fs) through to slow domain motions and normal mode vibrations (<1 /cm; ms). An important open question in enzymology remains the role of such dynamics, and whether motions/vibrations on timescales faster than turnover can couple to chemical steps during catalysis (i.e. to the reaction coordinate). There has been good progress in establishing the role of ms-ns dynamics such as loop opening/closing during enzyme turnover using NMR approaches, but direct evidence for the coupling of faster (sub-ns) dynamics to chemistry remains illusive and controversial and is based largely on the anomalous temperature dependencies of kinetic isotope effects. The role of fast dynamics remains an important question, as motions on similar timescales to chemistry (ps-fs; specifically, the time required to traverse the transition state) have the potential to profoundly affect the reaction outcome, and thus offer a means to control (enzyme) reactivity.
Recently, the use of isotopically labelled 'heavy' enzymes has provided new experimental evidence for the direct vibrational coupling of the protein to the chemical barrier. By combining our unique variable temperature and pressure kinetic isotope measurements with the 'heavy' enzyme method, we propose to establish whether dynamic coupling of protein motions to chemistry is likely to be a general feature of enzyme catalysis. We will study the H transfer reactions during both halves of the catalytic cycles of two homologous Old Yellow enzymes (OYEs) pentaerythritol tetranitrate reductase (PETNR) and morphinone reductase (MR). Selective cofactor and amino acid labelling we will identify (if present) networks of vibrationally-coupled residues, while MD simulations and ultrafast (fs-ns) spectroscopies will establish the timescales of such vibrations.

Catalysis and enzymes are central to life systems. Our understanding of catalysis underpins the exploitation of enzymes in industrial biotechnology (IB), through rational structure-based redesign, and for therapeutic targeting of enzymes by the pharmaceutical industrial. The beneficiaries will initially be academics: enzymologists, physical chemists, biochemists and those interested in (bio)chemical catalysis. However, should new physical models of enzyme catalysis and protein dynamics emerge from our work on the dynamical control of Old Yellow enzymes, this will have profound effects on how we (re)design enzymes for use in IB, and how we target proteins and enzymes therapeutically. This will lead to longer-term knock-on impact on the drug discovery and (industrial) biocatalysis sectors, and ultimately wider society, who benefit from these industries.

Industrial biotechnology (IB) has the potential to offer complete solutions to major societal challenges such as health, energy supply and food production/security. This will be driven through integration of biological processes and major changes to chemical production, agriculture and pharmaceuticals. There has never been a more pressing time to enable informed, predictive design of enzyme catalysts, which requires detailed knowledge of the chemistry that drives enzyme reactions across all relevant distance and time scales. The transition from screening/evolution to more predictive (in silico) design requires study of protein dynamics linked to computational simulations. Our programme will develop and exploit new experimental capabilities (the 'heavy enzyme' methodology and ultrafast spectroscopy) that will help shape this transition.

We do not anticipate an immediate commercial impact, given our research is still at the fundamental stage. However, the enzymes (MR and PETNR) used in this study are of established importance in industrial biocatalysis, so a deeper understanding of their function is likely to lead to longer-term improvements in their control and reactivity; ultimately leading to the design of better biocatalysts. As appropriate, we will seek to protect any IP arising from this work at the earliest stage. Our strategy for translating the technology is to establish IP protection through UMIP (Manchester's IP office). We will communicate our progress through networking events with external stakeholders.

We also plan to describe our research to the wider community, making use of the University of Manchester Open Days, Science Spectacular at the Manchester Museum, and the Royal Society Summer Science Exhibition as well as engagement with local schools. We will also make use of a range of internet-based communication tools, including regular updates to websites and podcasts.

Staff working on this project will gain new skills in the integration of enzymology, (ultfa)fast reaction kinetics and computational chemistry. These skills will equip the individuals with a rare multidisciplinary skillset, and enable them to make a valuable and practical contribution to the continued growth of this activity in the UK. Training will be in new quantitative biophysical methods, computational simulations and kinetic methods. The project will also lead to the development of unique methodology and instrument infrastructure within the UK academic sector. This unique environment and skill base will enable us to train and translate programmes involving protein dynamics to other protein systems and to encourage the adoption of these approaches by the wider community.