Understanding and Tuning the Catalytic Bias of Hydrogenase

Abbas Abou Hamdan (1) , Sébastien Dementin (1) , Pierre-Pol Liebgott (1) , Oscar Gutiérrez-Sanz (2) , Pierre Richaud (3, 4) , Antonio L de Lacey (5) , Marc Rousset (1) , Patrick Bertrand (1) , Laurent Cournac (6) , Christophe Léger (1)

1 BIP - Bioénergétique et Ingénierie des Protéines
2 CSIC - Catalytic Spectroscopy Laboratory
3 BVME - Biologie végétale et microbiologie environnementale - UMR7265
4 EBMP - Environnement, Bioénergie, Microalgues et Plantes
5 ICP - instituto de Catalysis y Petroleoquimica
6 BCMPB - Biologie cellulaire et moléculaire des plantes et des bactéries

When enzymes are optimized for biotechnological purposes, the goal often is to increase stability or catalytic efficiency. However, many enzymes reversibly convert their substrate and product, and if one is interested in catalysis in only one direction, it may be necessary to prevent the reverse reaction. In other cases, reversibility may be advantageous because only an enzyme that can operate in both directions can turnover at a high rate even under conditions of low thermodynamic driving force. Therefore, understanding the basic mechanisms of reversibility in complex enzymes should help the rational engineering of these proteins. Here, we focus on NiFe hydrogenase, an enzyme that catalyses H2 oxidation and production, and we elucidate the mechanism that governs the catalytic bias (the ratio of maximal rates in the two directions). Unexpectedly, we found that this bias is not mainly determined by redox properties of the active site, but rather by steps which occur on sites of the proteins that are remote from the active site. We evidence a novel strategy for tuning the catalytic bias of an oxidoreductase, which consists in modulating the rate of a step that is limiting only in one direction of the reaction , without modifying the properties of the active site. The four Michaelis parameters (two maximal rates and two values of Km) which characterize the forward and reverse reactions of a one-substrate one-product enzyme are related to each other and to the equilibrium constant of the reaction by the Haldane equation. 1 The forward and reverse maximal rates sometimes differ so much that certain enzymes were designated as "one-way enzymes". 2 The origin of such kinetic asymmetry, referred to as "cat-alytic bias", has rarely been investigated. Jencks proposed that directionality may result from the destabilization of the enzyme-substrate complex, which would decrease the energy required to reach the transition state in the forward direction. 2 This ""Circe effect"" is controversial 3 and has found no echo in the case of oxidoreductas-es, whose directionality is always discussed by comparing the potential of the substrate/product redox couple with the potential of either the active site or the redox centers of the electron transfer (ET) chain, when there is one (see examples in supplementary information). In trying to explain the catalytic bias from a single property of the enzyme (the potential of either the active site or the ET chain), one implicitly assumes that a single redox step, ET either between the substrate and the active site or to/from an electron relay, determines both maximal rates. However, the catalytic cycle of oxidore-ductases involves various steps (substrate binding, product release, proton and electron transfers, active-site chemistry) and it may occur that the rate limiting step (rls) is not the same when the enzyme works forward or backward. Two different steps may define the two maximal rates, and their ratio. Demonstrating that this can occur requires that the rls be defined in both directions in a series of variants that exhibit different catalytic preferences. Hereafter, we do so by characterizing a series of Desulfovibrio fructosovorans (Df) NiFe hydrogenase mutants. Previously, we have shown that the WT enzyme catalyzes H2 production and oxidation at similar maximal rates, and that narrowing the substrate channel using site directed mutagenesis has no effect on the maximal rate of H2 oxidation. 4,5 Here we show that these mutations slow H2 production up to 100-fold. In redox titra-tions, the active site of the mutants that have little reduc-tive activity behaves like in the WT enzyme. We use a novel method based on the isotope-exchange assay to determine the rates of H2 release from the active site to the solvent, and we conclude that this step limits H2 production whereas H2 entry does not determine the maximal rate of H2 oxidation. Conversely, the previous observation 5 that modifying the electron transfer chain selectively slows H2 oxidation shows that electron transfer limits the rate of H2 oxidation but not H2 production. 6 This is the first demonstration, on a specific example, that slowing a step that is rate limiting only when the enzyme works in one direction is a general mechanism for biasing the enzyme in the other direction, independently of the redox properties of the cofactors. We prepared the enzyme samples as described previously. 5 H2 oxidation rates were measured using a spec-trophotometric assay at pH 8, 30°C, with 50mM oxidized methyl viologen (MV), under one atm. of H2. 5 The Km for H2 were measured electrochemically. 5 The voltammo-grams in fig 2 were obtained with the enzyme bound to a rotating graphite electrode. 7 H2 production was followed using mass spectrometry (MS), in a solution initially saturated with Ar, with a saturating concentration of re

Domaines: Biochimie [q-bio.BM]

Abbas Abou Hamdan, Sébastien Dementin, Pierre-Pol Liebgott, Oscar Gutiérrez-Sanz, Pierre Richaud, et al.. Understanding and Tuning the Catalytic Bias of Hydrogenase. Journal of the American Chemical Society, 2012, 134 (20), pp.8368-8371. ⟨10.1021/ja301802r⟩. ⟨hal-01977597⟩