Dynamic structural fluctuations occurring over a broad range of timescales are now known to facilitate the catalytic function of enzymes, but there is less comprehensive experimental evidence linking fast-timescale, high frequency motions to the reaction coordinate. Interest in the role of such motions has recently surged and been the subject of intensive experimental efforts, in part due to the identification of enzymatic hydride tunnelling reactions. This mechanism involves transiently degenerate product and reactant states, which enable H-transfer to occur instantaneously without the need to surmount the activation barrier associated with traditional transition-state based models of enzyme catalysis. The primary gauge of tunnelling in enzyme-catalysed reactions is the identification of temperature dependent kinetic isotope effects (KIEs), i.e. the relative rates of a reaction where the transferred atom is substituted for an alternate isotope. The identification of temperature-, and also pressure-, dependent KIEs has resulted in the emergence of new models of describing enzymatic H-transfer. These invoke a role for fast-timescale protein motions that 'promote' transfer via tunnelling. A popular model system for studying enzymatic H-tunnelling reactions is Pentaerythritol tetranitrate reductase, which belongs to the Old Yellow Enzyme (OYE) family of ene-reductases. These nicotinamide coenzyme dependent oxidoreductases catalyse the stereospecific reduction of alpha/β-unsaturated alkene containing substrates. Here, the importance of donor-acceptor distances in determining the observed rate of PETNR reduction with NAD(P)H is probed via a detailed structural and kinetic analysis of site-directed variants. In addition, an investigation of distance-dependent Nuclear Overhauser effects via Nuclear Magnetic Resonance (NMR) spectroscopy is undertaken to assess active site organisation and measure donor-acceptor distances in PETNR-substrate complexes. A variable pressure NMR study reveals how NOE build- up is perturbed in high-energy conformers favoured as a result of the application of increased hydrostatic pressures. Recently there has been interest in exploiting the stereoselective properties of reactions catalysed by ene-reductase enzymes for use in biocatalytic reactions to produce industrially valuable compounds from renewable sources. The reactions of PETNR and additional OYE enzymes, Thermophilic old yellow enzyme and Xenobiotic reductase A, with both natural coenzymes and a set of synthetic Nicotinamide Coenzyme Biomimetics (NCBs) are also characterised. The NCBs represent affordable and fast-reacting alternatives to the physiological coenzymes. Reactions with NCBS are also shown to proceed via a tunnelling mechanism and furthermore, that enhanced donor-acceptor sampling correlates with the faster reactivity seen with these compounds.

How the fast (femtosecond-picosecond, fs-ps) protein dynamics contribute to enzymatic function has gained popularity in modern enzymology. With multiple experimental and theoretical studies developed, the most challenging part is to assess both the chemical step kinetics and the relevant motions at the transition state (TS) on the fast time scale. Formate dehydrogenase (FDH), which catalyzes a single hydride transfer reaction, is a model system to address this specific issue. I have crystallized and solved the structure of FDH from Candida boidinii (CbFDH) in complex with NAD+ and azide. With the guidance of the structure information, two active site residues were identified, V123 and I175, which could be responsible for the narrow donor-acceptor-distance (DAD) distribution observed in the wild type CbFDH. This thesis describes studies using kinetic isotope effects (KIEs) and their temperature dependence together with two-dimensional infrared spectroscopy on the recombinant CbFDH and its V123 and I175 mutants. Those mutants were designed to systematically reduce the size of their side chain (I175V, I175A, V123A, V123G and double mutant I175V/V123A), leading to broader distribution of DADs. The kinetic experiments identified a correlation between the DAD distribution and the intrinsic KIEs. The contribution of the fs-ps dynamics was examined via two-dimensional infrared spectroscopy (2D IR) by measuring the vibrational relaxation of TS analog inhibitor, aizde, reflecting the TS environmental motions. Our results provide a test of models for the kinetics of the enzyme-catalyzed reaction that invokes motions of the enzyme at the fs-ps time scale to explain the temperature dependence of intrinsic KIEs.

The relationship between protein structure, motions, and catalytic activity is an evolving perspective in enzymology. An interactive approach, where experimental and theoretical studies examine the same catalytic mechanism, is instrumental in addressing this issue. We combine various techniques, including steady state and pre-steady state kinetics, temperature dependence of kinetic isotope effects (KIEs), site-directed mutagenesis, X-ray crystallography, and quantum mechanics/molecular mechanics (QM/MM) calculations, to study the catalytic mechanisms of thymidylate synthase (TSase). Since TSase catalyzes the last step of the sole intracellular de novo synthesis of thymidylate (i.e. the DNA base T), it is a common target for antibiotic and anticancer drugs. The proposed catalytic mechanism for TSase comprises a series of bond cleavages and formations including activation of two C-H bonds: a rate-limiting C-H→C hydride transfer and a faster C-H→O proton transfer. This provides an excellent model system to examine the structural and dynamic effects of the enzyme on different C-H cleavage steps in the same catalyzed reaction. Our experiments found that the KIE on the hydride transfer is temperature independent while the KIE on the proton transfer is temperature dependent, implying the protein environment is better organized for H-tunneling in the former. Our QM/MM calculations revealed that the hydride transfer has a transition state (TS) that is invariable with temperature while the proton transfer has multiple subsets of TS structures, which corroborates with our experimental results. The calculations also suggest that collective protein motions rearrange the network of H-bonds to accompany structural changes in the ligands during and between chemical transformations. These computational results not only illustrate functionalities of specific protein residues that reconcile many previous experimental observations, but also provide guidance for future experiments to verify the proposed mechanisms. In addition, we conducted experiments to examine the importance of long-range interactions in TSase-catalyzed reaction, using both kinetic and structural analysis. Those experiments found that a remote mutation affects the hydride transfer by disrupting concerted protein motions, and Mg2+ binds to the surface of TSase and affects the hydride transfer at the interior active site. Both our experiments and computations have exposed interesting features of ecTSase that can potentially provide new targets for antibiotic drugs targeting DNA biosynthesis. The relationship between protein structure, motions, and catalytic activity learned from this project may have general implications to the question of how enzymes work.

The present dissertation seeks to deepen our understanding of hydrogen transfers and especially C-H bond activations in enzymes. Hydrogen transfers are ubiquitous in chemistry and biology and a thorough understanding of how they occur and what factors influence them will facilitate developments in biomimetic catalysis, rational drug design, and other fields. A particular difficulty with H-transfers is the importance of nuclear quantum effects to the reaction, particularly tunneling. The overall scope of the work here aims to examine how experimental kinetic isotope effects (KIEs) can be interpreted with a particular type of tunneling model, referred to as Marcus-like models, to yield a semi-quantitative picture of the physical mechanisms of H-transfers. Previous work had used this kind of model to qualitatively interpret experimental data using a combination of intuition and generalized theories. The work here examines these theories in quantitative detail, testing and calibrating our intuition in the context of several experimental systems. The first chapter of research (ch. II) focusses on the temperature dependence of primary KIEs and how these experiments can be quantitatively interpreted as a probe for certain kinds of enzyme or solvent dynamics. The subsequent chapters (ch. III-VI) focus on the use of secondary KIEs to determine the detailed structures of tunneling ready states (TRSs) and how the dynamics of H-tunneling affect those structures. These chapters focus primarily on the TRS of the enzyme alcohol dehydrogenase, but by examining an uncatalyzed analogue to that reaction (ch. VI), the work gains some insight about similarities and differences between catalyzed and uncatalyzed reactions. In summary, the work uncovers some principles of catalysis, not just the mechanism of a catalyzed reaction. The mechanism of C-H activation presented here provides an elegant solution to problems that have been vexing to accommodate within traditional models. This work constitutes some initial steps in making Marcus-like models quantitatively useful as a supplement or even replacement for traditional models of reactivity.

A series of para-substituted phenylethylamine analogues bound to the enzyme aromatic amine dehydrogenase have been simulated using quantum mechanical electronic structure calculations and molecular mechanical molecular dynamics simulations. Trends have been verified connecting bond dissociation energy (and thus driving force) to observed rate constants and activation enthalpy. Trends have been identified in connecting statistics drawn from molecular dynamics simulations and the temperature dependence of the kinetic isotope effect, notably that as the temperature dependence of the kinetic isotope effect increases the flexibility of the promoting vibration decreases. This is explained as being more effected by thermal energy put into the system, and therefore more affected by temperature.

京都大学 / 0048 / 新制・課程博士 / 博士(理学) / 甲第18448号 / 理博第4008号 / 新制||理||1578(附属図書館) / 31326 / 京都大学大学院理学研究科化学専攻 / (主査)教授 林 重彦, 教授 寺嶋 正秀, 教授 松本 吉泰 / 学位規則第4条第1項該当 / Doctor of Science / Kyoto University / DGAM