A SPECTROELECTROCHEMICAL INVESTIGATION OF THE THERMODYNAMIC AND STRUCTURAL PROPERTIES OF THE 2-OXOGLUTARATE-DEPENDENT OXYGENASE, TAUD By Christopher Wayne John A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Chemistry – Doctor of Philosophy 2019 ABSTRACT A SPECTROELECTROCHEMICAL INVESTIGATION OF THE THERMODYNAMIC AND STRUCTURAL PROPERTIES OF THE 2-OXOGLUTARATE-DEPENDENT OXYGENASE, TAUD By Christopher Wayne John 2-Oxoglutarate (2OG)-dependent dioxygenases catalyze C-H activation while performing a wide range of chemical transformations making their method of action and thermodynamic properties of great interest to industrial synthesis. In contrast to their heme analogues, non-heme iron centers afford greater structural flexibility with important implications for their diverse catalytic mechanisms. Unfortunately, the non-heme and less accessible active sites of these enzymes makes it a challenge to study them. To counteract this issue, we develop a method that uses electrochemical mediators and combines normal pulse spectrovoltammetry (NPSV) with Fourier transform infrared (FTIR) for detection and subsequent global spectral regression analysis to resolve the structural and thermodynamic properties simultaneously. We develop comprehensive semiemipirical kinetic simulation models to investigate the thermodynamic and kinetic limitations of mediators/analyte interactions. These methods are first validated using methylene green and thionine acetate as mediators and myoglobin (Mb) as the analyte. Both the E½ and unbiased redox difference FTIR spectra of the Fe(II)/Fe(III) redox couple of Mb in reduction and oxidation NPSV modes were in good agreement with those reported earlier by independent techniques. The modeling effort yielded a flexible computational tool capable of quantitatively predicting the redox response in mediated electrochemical studies and defining its limitations. These methods are used to characterize an in situ structural model of the putative transient ferric intermediate of 2OG:taurine dioxygenase (TauD), demonstrating that the FeIII/II transition involves a substantial, fully reversible, redox-linked conformational change at the active site. This rearrangement changes the apparent redox potential of the active site between - 272 mV for reduction of the ferric state and 196 mV for oxidation of the ferrous state of the 2OG-Fe-TauD complex resulting in a maximal observed redox hysteresis in the wild type enzyme of 468 mV. Quantitative modeling of the transient redox response using two alternative reaction schemes across a variety of experimental conditions strongly supports the proposal for intrinsic protein reorganization as the origin of the experimental observations. We use H99A, D101Q, H255Q, and Y73I variants of TauD to investigate the structural origin of the redox- linked reorganization and the relative contributions of the active site residues to the dynamic tuning of the redox potential of TauD. Extended time-dependent redox titrations show that, in all cases, reorganization occurs as a multi-step process, with individual phases exhibiting different sensitivities to ligand substitutions. The H99A variant shows the largest net redox change relative to the wild type protein, suggesting that redox-coupled protonation of H99 is required for TauD to support highly positive potentials. The effect of the D101Q substitution suggests that changes in the metal coordination of the carboxylate group may be secondary to changes involving H99 and are required for the ensuing reorganization steps. The H255Q substitution inhibits the conformational change, providing evidence for its involvement in the structural rearrangement. An investigation of the pD sensitivity of wild type TauD exposes a protonation event at the active site of TauD most likely attributable to H99 or H255. Ultimately, we propose H99 is protonated in the ferrous form of TauD and forms a hydrogen bond with the protein backbone. Oxidation of the enzyme results in the loss of this hydrogen bond allowing movement in the H99-T100-D101 chain so that D101 can form a bidentate ligand with the ferric iron center. This dissertation is dedicated to my parents, Paul and Shari John. iv ACKNOWLEDGEMENTS First, I’d like to thank my advisor, Denis Proshlyakov, for all of his support and guidance. I’d also like to thank the members of my committee: Greg Swain for providing electrodes, Robert (Bob) Hausinger for allowing me to use his lab space for protein purification, and John McCracken for his help with EPR measurements and his amazing stories. Much appreciation goes out to the members of Bob’s lab. Celeste Warrel for her preliminary work on TauD that led to my thesis project, her help with protein purification, and for being a great friend. Salette Martinez for all of her help and advice. Caitlyn (Cait) Herr for providing me with samples of EFE. And both Cait and Joel Rankin for all of their help digging through freezers trying to find plasmids. Thank you to all of the members of my lab: Maggie Conway, Adam Fillion, Nathan Frantz, Emily Groth, Yan Levitsky, Artem Muchnik, Yegor Proshlyakov, Allison Stettler, and Kylee Voorhis. You all have been an amazing support system and have made grad school bearable. Thank you to all of the other friends I’ve made here in the Lansing area: Julia Busik, Dave Davis, Aaron and Chandra Irving, Kali Maisano, Trevor and Elizabeth Sutton. You all have been very welcoming and have made Michigan feel like home. Thank you to my entire family for your endless support and encouragement. Thank you to all of the members of the Hive. I am so lucky to be part of such an amazing group of friends. And finally, thank you Nala, for being the best dog the world has ever known. Your snuggles and licks keep me moving forward. v TABLE OF CONTENTS LIST OF TABLES ....................................................................................................................... viii LIST OF FIGURES ....................................................................................................................... ix KEY TO ABBREVIATIONS ....................................................................................................... xii CHAPTER 1: INTRODUCTION ....................................................................................................1 C-H ACTIVATION .....................................................................................................................2 CYTOCHROMES P450 ..............................................................................................................2 2OG-DEPENDENT OXYGENASES ..........................................................................................4 THERMODYNAMICS OF C-H ACTIVATION ........................................................................6 THERMODYNAMIC SIGNIFICANCE OF THE DEPROTONATED F3 INTERMEDIATE .8 DEVELOPING AN IN SITU MODEL OF THE F3 INTERMEDIATE .....................................9 DISSERTATION OUTLINE .....................................................................................................11 REFERENCES ...........................................................................................................................13 CHAPTER 2: FOURIER TRANSFORM INFRARED SPECTROVOLTAMMETRY AND QUANTITATIVE MODELING OF ANALYTES IN KINETICALLY CONSTRAINED REDOX MIXTURES ....................................................................................................................17 INTRODUCTION ......................................................................................................................18 EXPERIMENTAL PROCEDURES ..........................................................................................19 SAMPLE PREPARATION ...............................................................................................19 NPSV MEASUREMENTS ................................................................................................20 DETERMINATION OF SOLUTION KINETICS ............................................................21 DATA ANALYSIS AND SIMULATIONS ......................................................................22 RESULTS ...................................................................................................................................22 DISCUSSION ............................................................................................................................40 CONCLUSION ..........................................................................................................................43 REFERENCES ...........................................................................................................................45 CHAPTER 3: STRONGLY COUPLED REDOX-LINKED CONFORMATIONAL SWITCHING AT THE ACTIVE SITE OF THE NON-HEME IRON-DEPENDENT DIOXYGENASE, TAUD ..............................................................................................................51 INTRODUCTION ......................................................................................................................52 EXPERIMENTAL PROCEDURES ..........................................................................................53 SAMPLE PREPARATION ...............................................................................................53 SPECTROSCOPIC MEASUREMENTS ..........................................................................54 RESULTS ...................................................................................................................................55 DISCUSSION ............................................................................................................................69 CONCLUSION ..........................................................................................................................74 vi REFERENCES ...........................................................................................................................75 CHAPTER 4: STRUCTURAL ORIGIN OF THE LARGE REDOX-LINKED REORGANIZATION IN THE 2-OXOGLUTARATE DEPENDENT OXYGENASE, TAUD ..81 INTRODUCTION ......................................................................................................................82 EXPERIMENTAL PROCEDURES ..........................................................................................84 PROTEIN PURIFICATION AND SPECTROELECTROCHEMICAL MEASUREMENTS ...........................................................................................................84 CHEMICAL REDOX TITRATIONS ...............................................................................85 RESULTS ...................................................................................................................................85 DISCUSSION ............................................................................................................................95 CONCLUSION ..........................................................................................................................99 REFERENCES .........................................................................................................................101 CHAPTER 5: EVIDENCE FOR PROTONATION EVENTS AT THE ACTIVE SITE OF TAUD ..........................................................................................................................................105 INTRODUCTION ....................................................................................................................106 EXPERIMENTAL PROCEDURES ........................................................................................107 SAMPLE PREPARATION .............................................................................................107 NPSV MEASUREMENTS ..............................................................................................107 RESULTS .................................................................................................................................107 DISCUSSION ..........................................................................................................................112 CONCLUSION ........................................................................................................................113 REFERENCES .........................................................................................................................114 CHAPTER 6: CONCLUSION AND FUTURE DIRECTIONS .................................................116 INTRODUCTION ....................................................................................................................117 NPSV AND COMPUTATIONAL METHODS ......................................................................118 INSIGHTS FROM TAUD EXPERIMENTS ...........................................................................118 FUTURE DIRECTIONS ..........................................................................................................122 CONTINUED STUDIES OF pD SENSITIVITY IN TAUD ..........................................122 RESONANCE RAMAN STUDIES OF AQUO LIGANDS ...........................................122 OTHER STRUCTURAL STUDIES................................................................................123 FERRYL IRON STUDIES ..............................................................................................124 REFERENCES .........................................................................................................................126 vii LIST OF TABLES Table 2.1. Empirical kinetic properties of the redox mediators .....................................................37 Table 3.1. Reduction and oxidation potentials of TauD calculated from experimental NPSV profiles ...........................................................................................................................................58 Table 3.2. List of mediators used during NPSV measurements with their thermodynamic and kinetic properties ............................................................................................................................61 Table 3.3. List of other mediators tested for NPSV of TauD ........................................................62 Table 3.4. Calculated oxidation and reduction potentials of profiles generated using model [1] .63 Table 4.1. Apparent initial rate constants of the oxidation (kOx) or reduction (kRd) of 2OG-Fe- TauD by various mediators ............................................................................................................88 Table 5.1. pD sensitive modes (cm-1) and tentative assignments ................................................109 viii LIST OF FIGURES Figure 1.1. CYP450 mechanism ......................................................................................................3 Figure 1.2. Catalytic cycle of TauD including the CYP450 inspired hydroxyl radical rebinding route (black) and the proposed alkoxide intermediate route (blue). ................................................5 Figure 1.3. Bordwell thermodynamic cycle represented by CYP450 and 2OG-dependent oxygenase mechanisms ....................................................................................................................6 Figure 1.4. Structures and observed pKa of Cmp II in HRP and CYP450, and the F3 state of TauD ................................................................................................................................................8 Figure 1.5. Catalytic cycle of TauD (black) with emphasis on hydrogen atom transfer (HAT) .....9 Figure 1.6. Plot of E½ versus solution pH showing boundaries for pKa transitions .......................11 Figure 2.1. Optically transparent thin layer electrochemical (OTTLE) cell ..................................21 Figure 2.2. Potential profile of NPSV ............................................................................................22 Figure 2.3. Deconvolution of NPSV FTIR data ............................................................................25 Figure 2.4. Resolution of spectra and Nernstian population profiles of mediators .......................26 Figure 2.5. NPSV/GSR resolution of redox transition of Mb........................................................27 Figure 2.6. Sensitivity of NPSV to mediator concentration and pulse width ................................28 Figure 2.7. Kinetic model for the heterogeneous mediated electrolysis of an analyte in solution 29 Figure 2.8. Modeling pre-equilibrium NPSV changes in the mediated reaction ...........................34 Figure 2.9. Comparison NPV concentration profiles of a single analyte calculated by the complete and reduced models (Fig. 2.7) ........................................................................................35 Figure 2.10. Sensitivity of the unmediated redox transition in Mb to NPSV pulse width ............35 Figure 2.11. Determination of klim and of MG at high concentrations ....................................36 Figure 2.12. klim values for MG determined at various concentrations ..........................................38 Figure 2.13. Concentration dependence of mediator-limited NPSV response ..............................36 ix *elk Figure 2.14. Determination of homogenous bimolecular rates of reactions between mediators and Mb. .................................................................................................................................................39 Figure 2.15. Cumulative effect of mediator mixtures ....................................................................40 Figure 2.16. The role of mediator in determining the minimum NPSV pulse duration ................43 Figure 3.1. NPSV transitions in Fe-TauD, 2OG-Fe-TauD, and taurine-2OG-Fe-TauD ...............57 Figure 3.2. NPSV of redox transitions in 2OG-TauD with Zn or Fe bound..................................58 Figure 3.3. Titration of 70 µM 2OG-Fe(II)-TauD into 100 µM FCN ...........................................59 Figure 3.4. Chemical models for the mediated electrochemistry of TauD ....................................60 Figure 3.5. Effect of thermodynamic properties of the analyte on the apparent NPSV redox hysteresis for models [1] and [2] ...................................................................................................60 Figure 3.6. The effect of the intrinsic E½ on NPSV redox transitions in model [1] ......................62 Figure 3.7. The effect of FCN concentration on the observed and profiles .....................64 Figure 3.8. Kinetically limited NPSV reduction profile in the absence of FCN ...........................66 Figure 3.9. Effect of thermodynamic properties of the analyte on the apparent NPSV redox hysteresis when FCN is not present ..............................................................................................68 Figure 3.10. Comparison of experimental NPSV hysteresis with model-depended predictions in the absence of FCN ........................................................................................................................68 Figure 3.11. TauD active site structure and possible structural rearrangement .............................73 Figure 4.1. Catalytic cycle of TauD comparing the hydroxyl radical rebound (grey) and alkoxide- forming (blue) mechanisms ...........................................................................................................82 Figure 4.2. Transient absorption changes upon reaction 2OG-Fe-TauD with redox mediators ....86 Figure 4.3. The observed E½ of Mb and TauD upon reduction by TA ..........................................87 Figure 4.4. Normalized population kinetic traces of the amount of 2OG-Fe(II)-TauD oxidized or 2OG-Fe(III)-TauD reduced after addition into solutions containing various mediators ...............88 Figure 4.5. Transient changes in the redox potential of 2OG-Fe-TauD and three of its variants .90 Figure 4.6. Redox-coupled vibrational changes in WT and variants of 2OG-Fe-TauD ................93 Figure 4.7. Redox-coupled reorganization of 2OG-Fe-EFE .........................................................95 Figure 4.8. Selected residues at the TauD active site ....................................................................96 Figure 4.9. Proposed redox-linked conformational changes in TauD ..........................................98 Figure 5.1. pD sensitivity of Fe-TauD .........................................................................................108 x RdOx Figure 5.2. pD sensitivity of 2OG-Fe-TauD ................................................................................110 Figure 5.3. pD sensitivity of succinate-Fe-TauD ........................................................................111 Figure 5.4. pD sensitivity of taurine-2OG-Fe-TauD ..................................................................112 Figure 6.1. Comparison of E½ values of HRP, CYP450, and TauD in the ferrous, ferric, and ferryl states ...................................................................................................................................120 xi KEY TO ABBREVIATIONS 2OG 2-oxoglutarate AF alkoxide forming BDD boron doped diamond Cmp I compound I Cmp II compound II CYP450 cytochrome P450 DMSO dimethyl sulfoxide E½ Ea redox potential applied potential EFE 2OG-dependent ethylene-forming enzyme EOx Er ERd F3 F4 oxidation potential reference potential reduction potential Fe(III)-(hydr)oxo intermediate in TauD Fe(IV)-oxo intermediate in TauD FCN ferricyanide FTIR Fourier transform infrared FX alkoxide intermediate in TauD GSR global spectral regression HAT hydrogen atom transfer IR infrared xii KCl potassium chloride Mb myoglobin MB methylene blue MG methylene green MV methyl viologen NPSV normal pulsed spectrovoltammetry O.D. optical density OTTLE optically transparent thin layer electrochemistry Ox oxidation PSSV potential step spectrovoltammetry Rd reduction SSV staircase spectrovoltammetry TA thionine acetate TauD 2OG-dependent taurine oxygenase TMPD N,N,Nʹ,Nʹ-tetramethyl-p-phenylenediamine UV-Vis ultraviolet-visible WT wild type xiii CHAPTER 1: INTRODUCTION 1 C-H ACTIVATION C-H activation, the process of cleaving a C-H bond, is a fundamental step in the synthesis of a wide range of valuable products including plastics and pharmaceuticals. Ideally, the process by which these products are made begins with cheap and plentiful precursors, involves as few steps as possible, and can be performed at low temperature and atmospheric pressure to minimize energy consumption.1–3 While aliphatic hydrocarbons are abundant and relatively inexpensive, their thermodynamic stability presents a significant challenge for their use in synthetic chemistry. Many catalysts have been designed that are able to activate C-H bonds and incorporate new functional groups; however, these catalysts are usually composed of rare, expensive, and even toxic metals such as gold4,5, platinum5,6, palladium5,7, and mercury.5 Furthermore, reactions using these catalysts typically require high temperatures and pressure to function at an optimal rate. Interestingly, C-H activating enzymes, such as cytochromes P450 (CYP450) and 2-oxoglutarate (2OG) dependent oxygenases, are found throughout all forms of life and use cheap and abundant metals, such as iron, copper, and manganese, while functioning at comparably low temperatures and atmospheric pressure. For this reason, many researchers have turned to studying the properties of enzymes that control their reactivity in hopes of using this information to design better catalysts for more sustainable reaction schemes.8–10 CYTOCHROMES P450 CYP450 is the most studied family of C-H activating enzymes and utilizes the hydroxyl radical rebinding mechanism. Its mechanism (Fig. 1.1) begins with the enyzme’s resting state, a ferric center with a distal aqua ligand. Substrate binding displaces the aqua ligand and slightly 2 Figure 1.1. CYP450 mechanism displaces the iron from the plane of the porphyrin ring increasing its electron affinity. In this state, the ferric iron is reduced to ferrous iron by CYP450 reductase. The ferrous iron then binds molecular oxygen which subsequently undergoes a two electron reduction to form a ferric- peroxo intermediate first, followed by a ferric-hydroperoxo complex. The ferric-hydroperoxo is then protonated once more to form a Fe(IV)=O radical cation complex known at Compound I (Cmp I). Cmp I activates the C-H bond of the substrate leaving a carbon radical on the substrate and transferring the hydrogen atom to itself to form Compound II (Cmp II), a Fe(IV)-OH 3 species. Finally, the hydroxo ligand is rebound to the carbon radical thereby hydroxylating the substrate and returning the enzyme back to its resting state. The highly oxidized states of Cmp I and Cmp II have not been directly observed during the normal catalytic cycle of CYP450 because of their fast turnover rate but have been detected transiently in stop flow reactions.11 Historically, the structures of Cmps I and II were based on intermediates of the same name observed in peroxidases12 though these enzymes are incapable of performing C-H activation, causing some to question the role and structure of Cmp I in CYP450.13,14 Finally, work by Rittle and Green confirmed the structure of Cmp I as a ferryl and cation radical porphyrin (Fig 1.1).15 2OG-DEPENDENT OXYGENASES 2OG-dependent oxygenases have historically been considered to use a hydroxyl radical rebinding mechanism similar to that of CYP450. In this mechanism (Fig. 1.1, black), the enzyme starts in its ferrous resting state with three ligated waters. The co-substrate, 2OG, binds displacing two water ligands and is followed by binding of the substrate displacing the final water and leaving an open site for dioxygen to bind to the ferrous iron. The oxygen then attacks 2OG forming a five coordinate Fe(IV)=O succinate bound species (F4 in Fig. 1.2). The C-H bond of the substrate is then activated and the hydrogen atom is transferred to the oxygen forming an Fe(III)-OH intermediate denoted here as F3. The hydroxide is then rebound to the carbon radical of the substrate thereby hydroxylating the substrate. The enzyme returns to the ferrous state and is rehydrated. 4 Figure 1.2. Catalytic cycle of TauD including the CYP450 inspired hydroxyl radical rebinding route (black) and the proposed alkoxide intermediate route (blue). Brackets indicate intermediates that have not been directly observed. 5 THERMODYNAMICS OF C-H ACTIVATION As shown in Fig. 1.3, the F4 and F3 intermediates are catalytically equivalent to Cmp I and Cmp II, respectively, in CYP450. The thermodynamic properties of these intermediates have been directly tied to the ability of this family of enzymes to activate C-H bonds. Using Hess’ Law, Bordwell developed a relationship to determine the bond dissociation energy of the homolytic cleavage of C-H bonds.16 Mayer applied Bordwell’s thermodynamic cycle to C-H activation as it is performed by catalysts and enzymes (Fig. 1.3) and showed that the redox potential of Cmp I or F4 and the pKa of Cmp II or F3 together provide the energy needed to cleave the C-H bond of the substrate.17 Figure 1.3. Bordwell thermodynamic cycle represented by CYP450 and 2OG-dependent oxygenase mechanisms. The dissociation energy of the newly formed O-H bond is dependent on the Gibbs free energy of the electron transfer and proton transfer and the solvation energy of the H atom (C = 57.6 kcal/mol). 6 As previously mentioned, peroxidases, such as that found in horseradish (HRP), form Cmps I and II in their catalytic cycle, but are not able to perform C-H activation. Interestingly, the active sites of HRP and CYP450 share a very similar structure with one key difference; the axial ligands of CYP450 and HRP are cysteine and histidine, respectively (Fig. 1.4). The sulfur atom from cysteine in CYP450 provides a large amount of electron density that is pushed into the Fe-O bond thereby increasing the proton affinity of the oxygen atom. This process works as a type of push-pull mechanism where the “push” of electron density into the Fe-O bond provides a “pull” on a nearby hydrogen atom.18 This explanation is reflected in the observed pKas of these enzymes’ respective Cmp II intermediates (Fig. 1.4). Likewise, computational studies and model compounds of non-heme enzymes have indicated that the ferric F3 intermediate should likely have a large pKa with estimations of >20.19 This pKa is estimated to be larger than that observed in Cmp II because the ferric iron in F3 has a lower charge than the ferryl iron in Cmp II, thus providing a higher affinity for a proton. Surprisingly, a resonance Raman investigation of the structure of the F4 and F3 intermediates showed no evidence of a proton present in the F3 intermediate.20 This study involved the archetypal member of the 2OG-dependent oxygenase family, taurine:2OG dioxygenase (TauD). The lack of evidence supporting the presence of a proton on the F3 intermediate of TauD brought into question the applicability of the hydroxyl radical mechanism as it pertains to this enzyme. Instead, it was suggested that the F3 intermediate may be followed by the formation of an alkoxide intermediate (FX) by binding the Fe-oxo species directly to the carbon radical of taurine (Figure 1.2, blue). This unstable alkoxide intermediate would then quickly hydrolyze into the observed products. The formation of an alkoxide intermediate has been corroborated by a 7 crystallographic study of another 2OG-dependent dioxygenase, again indicating that the F3 state is not protonated.21 THERMODYNAMIC SIGNIFICANCE OF THE DEPROTONATED F3 INTERMEDIATE The absence of a proton on the oxo group of the F3 intermediate indicates that the pKa of F3 must actually be less than 8, the pH at which the resonance Raman study was performed.20 If TauD’s pKa is much lower than expected, TauD must be using some other method to perform C- H activation. According to the Bordwell thermodynamic cycle, the only other source for energy is the reduction potential of F4. Consequently, TauD would need to achieve a higher redox potential in its ferryl F4 intermediate than CYP450 in its ferryl cation radical Cmp I intermediate. This is a significant task considering Cmp I exists at a higher oxidation state than F4 (Fig. 1.3). Because of the debate surrounding TauD’s mechanism and our desire to understand the properties of enzymes necessary for their catalytic activity, it is important to study the formation of the F3 intermediate in order to 1) determine its thermodynamic properties including its pKa and redox potential and 2) determine the important structural features that contribute to these properties. Figure 1.4. Structures and observed pKa of Cmp II in HRP and CYP450, and the F3 state of TauD. 8 DEVELOPING AN IN SITU MODEL OF THE F3 INTERMEDIATE Figure 1.5 shows the alkoxide forming mechanism of TauD (black) overlaid with structural and electrochemical manipulations of TauD (blue) of the resting state, 2OG bound, and 2OG and taurine bound intermediates. Unlike the F4 and F3 intermediates, these states are stable in anaerobic environments making them easy to study and maintain. More importantly, the 2OG bound form of ferric TauD resembles the F3 intermediate providing an in situ model that can be used to estimate the thermodynamic properties of the F3 intermediate. Figure 1.5. Catalytic cycle of TauD (black) with emphasis on hydrogen atom transfer (HAT). Vertical transitions indicate a change in the oxidation state of the Fe center. Diagonal transitions indicate a change in a protonation state. Blue structures show how TauD can be electrochemically manipulated to model the F3 state and provide an estimate of the F3 state’s pKa. Using electrochemistry, these three stable forms of TauD can be oxidized and reduced between the ferrous and ferric states, a process that is controlled by the enzyme’s redox potential, E½. When this process is performed in a solution within a certain pH range, the bound water ligand will remain fully protonated in both the ferrous and ferric states and the E½ will remain constant. However, because of the difference in charge density between the ferrous and ferric 9 states, these two forms will have distinct pKa values. As the pH of the solution is increased, the bound water molecule on the ferric enzyme will lose a proton forming an Fe(III)-OH species that resembles a protonated version of the F3 intermediate while the ferrous state will remain protonated. As a result, the redox process between the ferrous and ferric states becomes a proton coupled electron transfer process. Under these conditions the redox potential becomes sensitive to the pH of the solution and will shift by -59 mV/pH unit as the pH increases (Figure 1.6). Further adjustment of the solution pH past the pKa of the ferrous state will cause the ferrous state of TauD to deprotonate as well. At this point, the redox process no longer involves a proton and the E½ returns to a constant value with increasing solution pH values. Therefore, a plot of the measured redox potential versus varied solution pH (Figure 1.6) will show boundaries where the redox potential becomes sensitive to the solution pH. The pH value observed at each of these boundaries indicates the pKa of the species that is (de)protonated at the indicated pH value. This relationship is most important for the 2OG bound intermediate as deprotonation of the water ligand will form a 2OG bound Fe(III)-OH and Fe(III)-O- pair that can be used as an in situ model of the F3 intermediate to provide an estimate of the thermodynamic properties of the F3 intermediate. Figure 1.6. Plot of E½ versus solution pH showing boundaries for pKa transitions. 10 Unfortunately, while electrochemistry is able to provide information about protonation and deprotonation events, it is unable to indicate where such events are taking place. This problem can be solved by combining electrochemistry with spectroscopy. Electrochemistry allows control over the enzyme’s oxidation state, while spectroscopy can be used to provide information about the structure and the populations that exist at any point in time. Spectroelectrochemistry therefore holds promise as a method to reveal redox-coupled protonation events in Fe(II)/Fe(III) states of TauD both from the vibrational spectra and the pH sensitivity of its redox transitions.22–24 Such data could be used to reconstruct the protonation states of transient Fe(III)-(hydr)oxo species using the known crystallographic structures of the Fe(II)-enzyme as a reference. DISSERTATION OUTLINE The following chapters discuss the development, validation, and implementation of spectroelectrochemical and computational methods used to analyze the thermodynamic and structural properties of TauD. Chapter 2 describes the development of a spectroelectrochemical method that combines normal pulsed voltammetry with Fourier transform infrared spectroscopy (NPSV) and a complementary computational method used to simulate the outcomes of such measurements. These methods are validated using the well-studied metalloprotein, myoglobin. These methods are then applied to TauD in Chapter 3 where we discover an unexpected redox- linked structural rearrangement in the active site of TauD. Chapter 4 sees the application of NPSV to variants of TauD. From this information, we are able to propose a mechanism for the redox-linked structural rearrangement. The pD sensitivity of TauD is evaluated in Chapter 5 11 where we observe evidence for protonation events. Finally, Chapter 6 concludes the dissertation and outlines remaining unanswered questions and the future directions of this project. 12 REFERENCES 13 REFERENCES (1) Hendrickson, J. B. Systematic Synthesis Design. IV. Numerical Codification of Construction Reactions. J. Am. Chem. Soc. 1975, 97 (20), 5784–5800. https://doi.org/10.1021/ja00853a023. (2) Hoffmann, R. W. Protecting-Group-Free Synthesis. Synthesis. 2006, No. 21, 3531–3541. https://doi.org/10.1055/s-2006-950311. (3) Norskov, J. K.; Bligaard, T.; Rossmeisl, J.; Christensen, C. H. Towards the Computational Design of Solid Catalysts. Nat. Chem. 2009, 1, 37–46. https://doi.org/10.1038/nchem.121. (4) Hughes, M. D.; Xu, Y.; Jenkins, P.; Mcmorn, P.; Landon, P.; Enache, D. I.; Carley, A. F.; Attard, G. A.; Hutchings, G. J.; King, F.; et al. Tunable Gold Catalysts for Selective Hydrocarbon Oxidation under Mild Conditions. Nature 2005, 437, 1132–1135. https://doi.org/10.1038/nature04190. (5) Hashiguchi, B. G.; Bischof, S. M.; Konnick, M. M.; Periana, R. A. Designing Catalysts for Functionalization of Unactivated C-H Bonds Based on the CH Activation Reaction. Acc. Chem. Res. 2012, 45 (6), 885–898. https://doi.org/10.1021/ar200250r. (6) Sen, A.; Lin, M.; Kao, L.; Hutson, A. C. C-H Activation in Aqueous Medium. The Diverse Roles of Platinum(II) and Metallic Platinum in the Catalytic and Stoichiometric Oxidative Functionalization of Organic Substrates Including Alkanes. J. Am. Chem. Soc. 1992, 114, 6385–6392. https://doi.org/10.1021/ja00042a014. (7) Topczewski, J. J.; Sanford, M. S. Chemical Science Carbon – Hydrogen (C–H) Bond Activation at Pd(IV) : A Frontier in C–H Functionalization Catalysis. Chem. Sci. 2015, 6, 70–76. https://doi.org/10.1039/c4sc02591a. (8) Lewis, J. C.; Coelho, P. S.; Arnold, F. H. Enzymatic Functionalization of Carbon- Hydrogen Bonds. Chem. Soc. Rev. 2011, 40, 2003–2021. https://doi.org/10.1039/c0cs00067a. (9) Comba, P. Oxidation Catalysis with High-Valent Nonheme Iron Complexes. 2014, 123– 145. (10) Smith, E. D.; Robbins, M. O.; Daly, C.; Krim, J.; Volokitin, A. I.; Ogletree, D. F.; Thiel, P. A.; Salmeron, M.; Ishii, M.; Kamo, M.; et al. A Predictably Selective Aliphatic C-H Oxidation Reaction for Complex Molecule Synthesis. Science. 2007, 318 (November), 783–787. https://doi.org/10.1126/science.1148597. (11) Denisov, I. G.; Makris, T. M.; Sligar, S. G.; Schlichting, I. Structure and Chemistry of Cytochrome P450. Chem. Rev. 2005, 105, 2253–2277. https://doi.org/10.1021/cr0307143. 14 (12) Dunford, H. B.; Stillman, J. S. On the Function and Mechanism of Action of Peroxidases. Coord. Chem. Rev. 1976, 19, 187–251. (13) Chandrasena, R. E. P.; Vatsis, K. P.; Coon, M. J.; Hollenberg, P. F.; Newcomb, M. Hydroxylation by the Hydroperoxy-Iron Species in Cytochrome P450 Enzymes. J. Am. Chem. Soc. 2004, 126 (1), 115–126. https://doi.org/10.1021/ja038237t. (14) Jin, S.; Makris, T. M.; Bryson, T. A.; Sligar, S. G.; Dawson, J. H. Epoxidation of Olefins by Hydroperoxo-Ferric Cytochrome P450. J. Am. Chem. Soc. 2003, 125 (12), 3406–3407. https://doi.org/10.1021/ja029272n. (15) Rittle, J.; Green, M. T. Cytochrome P450 Compound I: Capture, Characterization, and C- H Bond Activation Kinetics. Science. 2010, 330, 933–938. (16) Bordwell, F. G.; Ji, G. Z.; Satish, A. V.; Zhang, X.; Cheng, J. P. Bond Dissociation Energies in DMSO Related to the Gas Phase. J. Am. Chem. Soc. 1991, 113, 9790–9795. https://doi.org/10.1021/ja00026a012. (17) Mayer, J. M. Hydrogen Atom Abstraction by Metal-Oxo Complexes: Understanding the Analogy with Organic Radical Reactions. Acc. Chem. Res. 1998, 31, 441–450. https://doi.org/10.1021/ar970171h. (18) Groves, J. T. Enzymatic C-H Bond Activation: Using Push to Get Pull. Nat. Chem. 2014, 6 (2), 89–91. https://doi.org/10.1038/nchem.1855. (19) Gupta, R.; Borovik, A. S. Monomeric MnIII/II and FeIII/II Complexes with Terminal Hydroxo and Oxo Ligands : Probing Reactivity via O - H Bond Dissociation Energies. J. Am. Chem. Soc. 2003, 125, 13234–13242. (20) Grzyska, P. K.; Appelman, E. H.; Hausinger, R. P.; Proshlyakov, D. A. Insight into the Mechanism of an Iron Dioxygenase by Resolution of Steps Following the FeIV=O Species. Proc. Natl. Acad. Sci. U. S. A. 2010, 107, 3982–3987. https://doi.org/10.1073/pnas.0911565107. (21) Mitchell, A. J.; Dunham, N. P.; Martinie, R. J.; Bergman, J. A.; Pollock, C. J.; Hu, K.; Allen, B. D.; Chang, W.; Silakov, A.; Bollinger, J. M.; et al. Visualizing the Reaction Cycle in an Iron(II)- and 2-(Oxo)-Glutarate-Dependent Hydroxylase. J. Am. Chem. Soc. 2017, 139, 13830–13836. https://doi.org/10.1021/jacs.7b07374. (22) Proshlyakov, D. A.; Hausinger, R. P. Transient Iron Species in the Catalytic Mechanism of the Archetypal α-Ketoglutarate-Dependent Dioxygenase, TauD. In Iron-Containing Enzymes: Versatile Catalysts of Hydroxylation Reactions in Nature; Visser, S. P. De, Kumar, D., Eds.; Royal Society of Chemistry, 2012; pp 67–87. (23) Brischwein, M.; Scharf, B.; Engelhard, M.; Mäntele, W. Analysis of the Redox Reaction of an Archaebacterial Copper Protein, Halocyanin, by Electrochemistry and FTIR 15 Difference Spectroscopy. Biochemistry 1993, 32, 13710–13717. https://doi.org/10.1021/bi00212a041. (24) Hirst, J. Elucidating the Mechanisms of Coupled Electron Transfer and Catalytic Reactions by Protein Film Voltammetry. Biochim. Biophys. Acta - Bioenerg. 2006, 1757, 225–239. https://doi.org/10.1016/j.bbabio.2006.04.002. 16 CHAPTER 2: FOURIER TRANSFORM INFRARED SPECTROVOLTAMMETRY AND QUANTITATIVE MODELING OF ANALYTES IN KINETICALLY CONSTRAINED REDOX MIXTURES 17 INTRODUCTION Spectroelectrochemistry is a powerful technique that has been widely used to examine structural determinants and thermodynamic properties of a variety of inorganic, organic and biological samples, including metalloenzymes. Staircase voltammetry with spectroscopic (UV- visible) detection, or staircase spectrovoltammetry (SSV), is the most common spectro- electrochemical method1–6 used with such strong chromophores as hemes because absorption by their cofactors is intense and the background drifts are insignificant compared to the redox- induced spectral changes. In this technique, the sample is subjected to square steps of changing potential in an optically transparent thin layer electrochemical (OTTLE) cell.1,7,8 Spectra are recorded after equilibration delay at the applied potential, Ea,1,2,8–12 and are reported as a difference vs. an initial single reference spectrum, Sr. The small pathlength of an OTTLE cell aids in exhaustive electrolysis, but the relatively low solubility of proteins limits optical SSV to analytes with strong absorption. It is natural that pioneering work by Mäntele et al.1–5,13,14 focused primarily on chlorophyll and heme proteins with their intense, oxidation state-sensitive UV-Vis absorption bands. This technique was rapidly adapted to multi-heme cytochromes,5,6 photosynthetic reaction centers,15–17 terminal oxidases,18–20 bc1 complex,21,22 copper enzymes,23,24 and Ni-Fe hydrogenases25,26 where optical detection of redox transitions was the primary technique or preceded IR studies.27 The information-rich mid-IR region can be used for the detection of redox-coupled vibrational changes in proteins, including peptide backbone and amino acid side chain conformations, protonation events, and changes in hydrogen bonding.28,29 Potentiometric titrations are typically not undertaken in this region as spectra are affected by large, temperature- sensitive water absorption and are susceptible to drifts over time. Typically, IR spectroscopy is 18 combined with potential step spectrovoltammetry (PSSV) where a square potential wave is repeated between a reference potential (Er) and Ea. Several redox-difference IR spectra of proteins have been reported,1,4,15,16,22,23,30,31 where the redox behavior of the cofactor was first established optically. Here we demonstrate that mid-IR normal pulse spectrovoltammetry (NPSV) with non- linear deconvolution analysis can resolve mixtures of analytes, including those with unknown properties. IR-NPSV can report redox properties and coupled structural events in a wider range of redox-active analytes than is possible with optical detection. We show that IR-NPSV can be used to investigate analytes that are electrochemically slow or inactive in the absence of mediators and that computational modeling can yield quantitative interpretation of reactions in the mixture. Furthermore, such modeling enables rational design of the optimal experimental conditions for mediated electrochemical analysis, which is currently lacking. The sensitivity of IR-NPSV in resolving both vibrational and electrochemical properties of individual components of a redox active mixture is demonstrated here using myoglobin (Mb), a metalloprotein with slow direct electrochemical kinetics and facile mediated electrochemistry, and two redox mediators. We revisit a general “1:100” rule for mediator/analyte pairs32,33, examine its limits, and greatly expand its practical utility by taking into account solution and electrode electron transfer kinetics. EXPERIMENTAL PROCEDURES SAMPLE PREPARATION Horse heart myoglobin, mediators, and general laboratory chemicals were of reagent or better grade from Sigma-Aldrich (St. Louis, MO) and were used as acquired. All samples were 19 prepared and measured under anaerobic conditions. Samples of Mb were prepared in 25 mM Bis-Tris in D2O, pD 7.0, containing 0.5 M KCl. The pD of the buffer was measured at 10 °C with 0.5 M KCl and corrected for the activity of deuterium ions.34 D2O (99.9%, Sigma-Aldrich) is used in place of H2O to allow investigation of the 1,500-1,700 cm-1 frequency region. Methylene green (MG) and thionine acetate (TA) were added to their indicated concentrations. Stock solutions were stored at 4 oC. NPSV MEASUREMENTS NPSV was performed using a previously reported OTTLE cell35 (Fig. 2.1) with a 12.5 μm pathlength over a nano-crystalline boron-doped diamond film on a silicon substrate as a working electrode35–37 and BaF2 back window at 10 °C. Coiled platinum wire was used as the counter electrode, and Ag/AgCl in saturated KCl as a reference electrode. The cell was placed in the glove box without the window, inner o-ring and cap in place. After the sample was prepared, 70 µL of the sample was placed directly onto the BDD electrode. The cell was sealed by putting the window, inner o-ring, and cap in place. The cell was removed from the glove box and transferred to a sample compartment of the Equinox 55/S spectrometer (Bruker) that was purged with nitrogen gas. A nitrogen stream was chilled in an ethanol/dry ice bath and re-heated by a resistive wire heater controlled by a Model 340 temperature controller (Lake Shore Cryotronics) to maintain the temperature at 10.0(1) °C throughout the measurement. The cell was left for two hours to allow the temperature of the sample to equilibrate. Reduction was performed using a Er of + 0.5 V and potential step of -0.05 V. Oxidation was performed against Er of -0.5 V with a +0.05 V potential step. Potentials were applied using a computer-controlled potentiostat (Model CHI1202b, CH Instruments). The applied potential 20 range varied between samples as indicated. Potential step durations were 150 s, 300 s, or 600 s as noted. Spectra were integrated during the final 90 s (for 150 s steps) or 120 s (for 300 and 600 s steps). Figure 2.1. Optically transparent thin layer electrochemical (OTTLE) cell. a) Deconstructed OTTLE cell with labeled components. b) Constructed OTTLE cell ready for spectroelectrochemical measurements. FTIR spectroscopy was performed using an Equinox 55/S spectrometer (Bruker). DETERMINATION OF SOLUTION KINETICS: Individual samples of 20 µM MG and 20 µM TA were prepared in 25 mM Tris, pH 8 in an N2-purged anaerobic chamber. Samples were placed in an airtight, stirred, 1-cm cuvette and transferred to a UV-Vis spectrophotometer (Hewlett Packard 8453). Sodium dithionite was injected to reduce 95% of the mediator population, as observed spectrophotometrically. The reaction was initiated by injection of oxidized Mb to achieve an equimolar ratio with the mediator. UV-Vis spectra were collected every 2 s for 20 s before and 10 min after the start of the reaction. Rate constants were obtained by exponential fit of temporal changes at the absorption maxima of mediators as described in the results. 21 DATA ANALYSIS AND SIMULATION All data analysis and simulation procedures were developed and performed using Igor Pro as described below. Source code of the KinESim simulation package and demonstration Igor Pro experiment can be found on GitHub.38 RESULTS NPSV requires twice as many measurements as SSV, but is substantially more robust against background drifts since the FTIR difference spectrum (ΔSi) at each i-th step is calculated against re-acquired reference. All spectra were acquired following an equilibration period at the present potential (Fig. 2.2). The raw NPSV spectra were collected as an absorbance, A, vs. frequency, ν, vs. the sampling step dataset, where each step i = [1, ns] included measurements at variable Ea,i and constant Er. This dataset was reduced to a ΔAi,f  νf  Ea,i matrix as follows: (2.1) where Sa,i is the equilibrium spectrum at Ea while Sr,i and Sr,(i+1) are flanking equilibrium spectra at Er (Fig. 2.3). Figure 2.2. Potential profile of NPSV. Ea alternates between a constant reference, Er, and variable applied, Ea,i, potentials over successive cycles. 22 ,,,112iaiririΔSSSS In multicomponent samples, such as analytes/mediators mixture, separation and identification of individual components becomes the main analytical challenge. Each experimentally observed difference spectrum of the mixture is a 1  nν vector, where nν is the number of sampled frequencies, f = [1, nν]. is the sum of nc vectors representing the spectra of individual components, , where k = [1, nc]. In addition, includes experimental error due to baseline fluctuations, which can be described by a polynomial of np-th order with parameters : (2.2) The complete experimental nν  ni dataset is also a sum of nc spectral matrices describing the complete spectral contribution of each component k, , plus a single nν  ns baseline matrix P: (2.3) The step-dependent spectrum of k is , where is a 1  n vector of extinction coefficients at the sampled frequencies, vf, and is the population fraction of the reduced (or oxidized) form of k as determined by the Nernst equation for the current Ea,i and thermodynamic properties of k: redox potential, , and the stoichiometry of the electron transfer, nk. Deconvolution of the experimental matrix into individual contributions and further into corresponding values of , , and was accomplished here by a custom global non- linear spectral regression (GSR) procedure for Igor Pro.39 Such deconvolution can be computationally demanding for complex mixtures due to the number of fitted variables, which 23 obsiSobsiSkiSobsiSmip10pcnnobskmmiiikmpfSSkS11scobsobskinnikSSSPkkkiiSεkε½,(,,)kkkiaifEEn½kEobsSkSkε½kEkn equals to . A typical dataset described here requires up to 2500 independent variables to fit. Deconvolution can be further hindered by similarity in values and/or spectral uncertainties at limiting S/N ratios, leading to singularity errors and collapse of the regression. To remediate this problem, GSR was first performed on NPSV data of all mediators individually to determine their , , and , which typically converged quickly. Providing approximate values for and as initial guesses facilitated convergence, but was not required for a one- component sample. An initial guess for the spectral vector was set to a uniform value (0 or 1) without bias for the expected spectral features. The polynomial term was suppressed over the initial iterations to resolve an inherent singularity problem between the baseline in (and, hence, ) and , providing bias against the inclusion of the polynomial term in until suppression of was lifted in subsequent iterations. A linear baseline (1st order) was sufficient in most cases described here. 24 ½kkfcpvEnnni½kEkε½kEkn½kEknkεkεkiSiPkεiP Figure 2.3. Deconvolution of NPSV FTIR data. Equation (2.1) was used to process a raw FTIR data (top) into a redox-difference spectral matrix (bottom), which was contributed by all components (k), including redox mediators. These data were further deconvoluted into ΔStot spectra and profiles of individual components, as shown for TauD by the superimposed blue and red traces, respectively. Fig. 2.4 shows the full occupancy redox difference FTIR spectra ( , top) and Nernstian profiles ( , bottom) of two mediators. The fitted Nernstian profiles are described by continuous functions while the experimental profiles represent discrete absorbance values, , at characteristic frequencies, vf, of 1603 cm-1 and 1605 cm-1 for MG and TA, respectively. The experimental values were normalized in Fig. 2.4 to the corresponding intensity in of the fully electrolyzed sample for direct comparison with . 25 kΔSk½,(,,,)kkkkiaivfEEnNΔε,ifA,ifAkΔSk Individual of mediators obtained by GSR are in good agreement with values reported earlier using traditional techniques.40,41 A similar approach can be used to describe potential-dependent vibrational changes of the electrode, although this was not necessary for BDD in this study. Figure 2.4. Resolution of spectra and Nernstian population profiles of mediators. Experimental NPSV spectra ( , panel a) and GSR population profiles ( , panel b) of MG and TA. Experimental ( , ) and fitted ( profiles at vf of 1603 cm-1 and 1605 cm-1 for MG and TA, respectively, were normalized to the absorbance of the complete redox transition. The calculated E½ values of the mediators are shown. ) Nernstian profiles are shown. Experimental , The pre-determined , , and values of each known mediator were held constant during the subsequent unbiased analysis of the Fe(II)/Fe(III) redox couple of Mb to test the ability of NPSV with GSR to resolve redox transitions of an unknown analyte. In the absence of mediators, (Fig. 2.5b) there was a large hysteresis between the observed reduction and oxidation profiles, and their width indicated an n << 1. Moreover, the extent of the transition was sensitive to the timing of the potential pulse with a larger observed for a longer pulse. These observations show that the direct electron transfer between Mb and the electrode is slower than 26 ½kEkiSkkε½kEkn,1657iA necessary to maintain a redox equilibrium. GSR could not be reliably conducted in the absence of a complete transition of Mb under these conditions. , ) and the combined spectrum of the mediator cocktail ( Figure 2.5. NPSV/GSR resolution of redox transition in Mb. a) isolated redox difference spectrum of Mb ( Experimental ΔAi,f for mediated ( , ) and unmediated ( absorbance at 1657 cm-1. A 150 s NPSV pulse width was used for the unmediated ( mediated sample; a 600 s pulse was used only for the unmediated sample ( profiles ) represent the relative intensity of , ) samples were obtained from the at any given Ea. and ( , ). b) ) and ). GSR population In the presence of mediators, reversible and exhaustive reduction and oxidation of Mb were observed (Fig. 2.4b). In this case, the experimental profiles were accurately described by GSR, yielding with E½ = -157 mV vs. Ag/AgCl, n = 1, in good agreement with the reported value of -153 mV.42 Control GSR analysis with unrestricted nMG yielded coefficients of 0.9 – 1.1 without significant impact on . The 1:100 rule requires at least 10 µM of reduced mediator in equilibrium with Ea = to support kinetically effective oxidation of 1 mM analyte. Exhaustive electrolysis of Mb shows that the 100 µM total concentration of TA and MG satisfies this threshold at Ea = . The full occupancy redox-difference GSR spectrum (left) is in remarkable agreement with that reported earlier by Mäntele and co-workers using unmediated SSV at up to 9x higher analyte concentration.4 Notice that in Fig. 2.4 does not contain contribution of the redox mediators since they are described by separate vectors. The 27 MbSMbRdMbOxMbSMb½MbE½AE½MbEMbSMbSkε combined redox difference spectrum of the mediator cocktail at the concentration used here is shown by the dotted trace. Figure 2.6. Sensitivity of NPSV to mediator concentration and pulse width. Sample contained 1 mM Mb and either 25 or 15 µM of each mediator was sampled with a 300 or 150 s potential pulse width. The ( ) profiles were obtained by normalizing the ΔAi,1657 ( ) and to the GSR population profiles. Simulated ( ) and ( ) profiles were calculated as described in the text using experimental parameters shown in Table 2.1. The sensitivity of the NSPV response of Mb to the mediator concentration and potential pulse width (Fig. 2.6) can be used to test the limits of the 1:100 rule. The and profiles of Mb completely converge and exhibit a Nernstian response at pulse widths ≥300 s for a 25 µM mixture (each mediator). When either the concentration of mediators was too low or the pulse width was too short, the observed and profiles deviated from Nernstian behavior. This distortion is observed as a combination of one or more of the following: (i) loss in the amplitude, (ii) shift of the apparent transition away from , and (iii) broadening of the transition, i.e. n < 1. The concentration of 25 µM mediator mixture at 300 s pulse is lower than the 60 µM each mediator minimum predicted by the 1:100 rule for 1 mM Mb and all known values. A larger discrepancy is expected for longer pulses. Since the original rule does not account for equilibration timing or rationalize kinetic artifacts (Fig. 2.6), we developed a quantitative pre- equilibrium kinetics model of mediated electrochemistry to predict and interpret the characteristic response of an analyte under specific experimental conditions. 28 MbRdMbOxMbRdMbOxMbRdMbOxMbRdMbOx½AE½E Figure 2.7. Kinetic model for the heterogeneous mediated electrolysis of an analyte in solution. a) Complete model, including adsorption equilibria (kon, koff, Kbind) and intrinsic electron transfer (k°el) steps under Ea on the electrode (hatched). b) Reduced model condenses electron and mass transfer steps to pseudo-homogeneous, Ea-independent (klim) and Ea-dependent (k*el) rate limiting steps. These two steps are combined into a single reduced rate constant . The homogenous reaction between the mediator and the analyte (kf,sol, kr,sol) controlled by their E½ in a layer of thickness d is identical for both models. See text for details. Fig. 2.7a shows the minimal chemical model for the electrolysis of one mediator and one analyte in the solution, consisting of two processes: a heterogeneous reaction in the two- dimensional space of the electrode surface and a homogeneous reaction in the three-dimensional space of the bulk solution. These spaces constitute two separate sub-systems and freely exchange matter as determined by the binding affinity. The adsorption/desorption process results in mass transfer of dnM moles of the reduced or oxidized mediator M over time, dt. The net mass transfer can be expressed as changes in concentrations of mediators in the bulk, CM, and surface, M, sub- systems based on their dimensions as follows: (2.4) 29 elkMsolelMMdnVdCSAd where Vsol is the volume of the bulk solution with layer thickness hsol over an electrode area SAel so that . Resulting changes in the concentrations in the two sub-systems are also related via , which is the only extensive property of the model: (2.5) The net rate of the mass transfer with binding affinity is: (2.6) In this study, we assumed that binding is not affected by the redox state, so that and , but corresponding rates and are distinct. The complete system of ordinary differential equations (ODE) describing MOx and MRd in each of the two sub-systems includes three rates: RMT (above), the rate of the redox conversion on the electrode, RET, and the rate of redox reactions in the bulk solution, RBL. RET is described by and Butler-Volmer electrode rates kOx,el and kRd,el: (2.7) (2.8) (2.9) where is the intrinsic electron transfer rate constant and α is the symmetry of the energy barrier. This expression does not account for possible binding competition between components of the mixture, which is not expected in dilute solutions. 30 solsolelVhSAsolh1MMsoldCdhbindonoffKkk1MMMMMTonoffsoldCdRkCkdthdt,,onOxonRdkk,,offOxoffRdkk,MTOxR,MTRdRM,,MMMMOxRdETOxelRdRdelOxddRkkdtdt½/0,aFRTEEOxelelkke½1/0,aFRTEERdelelkke0elk The bulk rate, RBL, depends on the reactivity between the mediator and the analyte. In the simplest case, it is described as a homogeneous, bimolecular, stoichiometric reaction: (2.10) where kOx,sol and kRd,sol are dependent on the E½ of mediator and analyte: (2.11) The contribution of additional reactions between pairs of mediators in the solution has been tested and they were found to have no effect on the resulting profiles, but required longer simulation times. Therefore, cross-mediator reactions were not included in further studies. We also assumed that all solution concentrations are uniform across the 12.5 µm thin layer since we focus on investigating time scales that are long relative to diffusion rates. Lastly, homogeneous reactions were constrained to n = 1 based on the known redox stoichiometry of Mb. The entire chemical system is described by a set of ODEs defined for each mediator/analyte pair: 31 ,,MMMAMAOxRdBLOxsolOxRdRdsolRdOxdCdCRkCCkCCdtdt½½/,,AMnFRTEEOxsolsolRdsolkKek,,onoffMMAMAMMOxfsolOxRdrsolRdOxOxOxdCkkdtkCCkCCC,,onoffMMMMMOxOxOxfelRdrelOxdkkdtCkk,,onoffMMAMAMMRdfsolOxRdrsolRdOxRdRddCkkdtkCCkCCC,,onoffMMMMMRdRdRdfelRdrelOxdkkdtCkk,,onoffAMAMAAAOxfsolOxRdrsolRdOxOxOxdCkkdtkCCkCCC,,onoffAAAAAOxOxOxfelRdrelOxdkkdtCkk The total differential rates of and are determined by integrating RBL over all solution reactions. Direct redox reactions of an analyte on the electrode can be accounted for using Eq. 2.6 and 2.7, as was done here for the case of Mb. The numerical integration of the complete system of ODEs over time can predict temporal changes in the concentrations of individual components in response to Ea waveform that mimic experimental spectroelectrochemistry. While the Euler integrator43 is often adequate for homogenous reactions, the heterogeneous electrolysis described here imposes computational challenges. First, unequal partitioning of mediators between the electrode surface and bulk solution results in a large even for small and , which is exacerbated with increasing hsol. To meet the requirement, Δt must be reduced by orders of magnitude relative to that needed for . Second, the exponential dependence of kel on requires further reduction of Δt when exceeds 0.3-0.4V. These two factors are multiplicative, leading to a stiff ODE problem: rapid shifts of equilibrium in small populations of and , interference between , and and the resulting fast and alternating shifts in the quotients of reactions over successive integration steps. Lastly, large and repetitive changes in Ea over NPSV profile and the resulting wide variation of rates made it impractical to use a uniform Δt across the experiment. We implemented two approaches to decrease the computational load when simulating the long (> 5 h) real reaction times. First, we used Runge-Kutta integration43 with adaptive Δt, which 32 ,,onoffAMAMAAARdfsolOxRdrsolRdOxRdRddCkkdtkCCkCCC,,onoffAAAAARdRdRdfelRdrelOxdkkdtCkkARdCAOxCMdMdnMdCkelRconstksolRconst½kakEEEkE,OxelM,RdelMETR,MTOxR,MTRdR varied based on the boundary conditions of the preceding step. This balanced a small time granularity under strongly off-equilibrium conditions immediately following Ea changes and with large Δt under near-equilibrium conditions at a stable Ea. The second approach involved simplification of the chemical model (Fig. 2.7b). The development of a reduced reaction scheme alleviated the stiffness problem by eliminating the disparity between the surface and solution fractions without sacrificing the accuracy of the simulation. The virtual analyte collectively represents either or , depending on the overall direction of the process under the given Ea. is engaged in two sequential reactions. The first, Ea-independent reaction with a rate-limiting step of klim represents the overall RMT. Even though for any particular process, it is an adequate approximation to use the same klim for both directions since only one of them, or , can be limiting at a time. The second, Ea-dependent reaction converts into one of the states of the mediator in the bulk solution and represents RET. Since is defined in the bulk phase, it is not affected by unequal partitioning between the electrode surface and the solution, reducing stiffness of the model. The reduced model is described by an overall apparent limiting rate (Eq. 2.12) and reduced rate constants and . First, the overall forward reaction direction at a given moment is determined by the larger value between kOx,el and kRd,el, i.e. when kRd,el > kOx,el, reduction is the forward reaction and = kRd,el and = kOx,el. Reduced rate constants are then found as = , and = , where: (2.12) 33 *M,elOxM,elRdM*M1K,MTOxR,MTRdR*M*METROxkRdk,felk,relkOxkrkRdkfkMMMMOxRdETOxRdRdOxdCdCRkCkCdtdt (2.13) (2.14) Figure 2.8. Modeling pre-equilibrium NPSV changes in the mediated reaction. a) Simulated concentration profiles of the oxidation of 1 mM analyte ( ) in the presence of 0.5 mM mediator ( spectral acquisition windows ( ) to obtain average populations ). The concentration of the analyte is integrated over the ) in response to the changing Ea ( and at applied (Ea) and reference (Er) NPSV potentials, respectively. b) Potential dependent population profile obtained from and per Eq. 2.1. E½ of the analyte and the mediator are both 50 mV. Conversely, when kOx,el > kRd,el, oxidation is the forward reaction so that = kOx,el and = kRd,el. Eq. 2.13 determines the overall rate in the forward (faster) direction and Eq. 2.14 corrects the rate in the reverse (slower) direction to maintain an overall equilibrium constant at its thermodynamic value. Despite such a seemingly gross oversimplification, this reaction scheme yielded mediator concentration profiles that were indistinguishable from the full simulation with over two orders of magnitude reduction in CPU time across all conditions tested here (Fig. 2.9). The apparent reduced electron transfer rate is related to the full model as follows: 34 ,lim111ffelkkk½,lim111arnFEERTrelkekk,ai,ri,Aai,Aai,Ari,felk,relk*elk (2.15) Figure 2.9. Comparison of NPV concentration profiles of a single analyte calculated by the complete and reduced models (Fig. 2.7). Complete model simulations (solid lines) at various Kbind, as indicated, and constant k° el, of 0.1 M-1s-1. The simplified model (open circles) used same parameters in the empirical relationship: k*el = k° el, = 0.1 M-1s-1, the layer el x (Kbind) x h where k° thickness (h) was 12 µm, and klim = kon. The reduced model was used to predict changes in CM and CA under experimental NPSV conditions (Fig. 2.8a). Concentration profiles were integrated over the spectral acquisition window, , followed by reversible changes (Eq. 2.1) and the population profiles (right), which are directly equivalent to the experimental NPSV vectors , but offer insight into kinetic limitations. The NPV simulation in Fig. 2.8 was carried out with the assumption that the Figure 2.10. Sensitivity of the unmediated redox transition in Mb to NPSV pulse width. Experimental (markers) and simulated (lines) reduction ( , panel a) and oxidation ( , panel b) profiles were obtained in the absence of mediators at 300 s ( , , , ) and 600 s ( , , , ) pulses. is shown by the vertical line. 35 *0elelbindkkKh,Aai,AaiOxkMbRdMbOx½MbE analyte is fully dependent on the mediator for redox transitions, as evident from the lag in the analyte response behind the mediator in real time (Fig. 2.8a). The reduced heterogeneous electrochemical model was validated against experimental NPSV profiles of Mb under a variety of conditions. First, unmediated NPSV of Mb with incomplete electrolysis was simulated for two different pulse durations (Fig. 2.10). The unknown parameters klim and were determined to be 1.25×10-3 M-1s-1 and 1.0×10-4 M-1s-1, respectively, using a series of simulations of both and with a 300 s pulse, and then verified by extrapolation to a 600 s pulse without further adjustment. Simulations adequately predict the pulse-dependent increases in the amplitude of the early pre-equilibrium NPSV response (Fig. 2.10), demonstrating that the reduced model is an adequate description of heterogenous reactions in a thin layer and can be used for determination of the reduced kinetic parameters from experimental NPSV profiles. Figure 2.11. Determination of klim and of MG at high concentrations. MG samples were prepared in 25 mM Tris, pD 7.0, 0.5 M KCl and transferred to the OTTLE cell for FTIR measurements. MG samples were exposed to 50 mV, 100 mV, and 200 mV overpotentials over 15 s (blue) and 30 s (red) steps. IR spectra were integrated for 15 s. The intensity at 1603 cm-1 was normalized to the intensity of the completely oxidized sample at the same frequency and then multiplied by the concentration of the sample. Experimental data ( , ) were simulated ( , ) by adjusting the values of klim and to find the best fit. 36 *elkMbRdMbOx*elk*elk Table 2.1. Empirical kinetic properties of the redox mediators. Mediator k*el (s-1) 10 µM klim (s-1) 15 µM 25 µM MG1 0.025 0.124 0.186 0.31 TA 0.5 0.467 0.7 1.17 kf,sol (M-1s-1) 1.4x104 1.7x104 1Applies to both redox transitions (-111 mV and -243 mV). See text for interpretation. NPSV profiles of mediators were used to estimate their klim and , except that 1 mM Mb was used as a reporter when concentrations of mediators were too low for spectral detection (<100 µM, Fig. 2.13). NPSV of MG were measured directly at higher concentrations (≥200 µM, Fig. 2.11), while TA was not sufficiently soluble. Both TA and MG required shorter pulses and/or increased concentrations of Mb to detect pre-equilibrium populations. At low concentrations, klim of both mediators show linear concentration dependences (Table 2.1), effectively making the single 2e- transition in TA and the two resolved 1e- transitions in MG second order rate constants in the mediator. This dependence was no longer observed for MG at higher concentrations with Mb as a reporter, suggesting saturation of klim=0.5 s-1 at approximately 40 µM MG (Fig. 2.12). A slight decrease in of MG to 0.015 s-1 was observed when CMG ≥ 0.5 mM (not shown), but not enough to establish a clear concentration dependence. Low solubility of TA under our experimental conditions (<50 µM) precluded investigation of its klim saturation. 37 *elk*elk Figure 2.12. klim values for MG determined at various concentrations. klim follows a linear profile at low mediator concentrations and reaches a constant value of 0.5 at 40.3 µM MG. Figure 2.13. Concentration dependence of mediator-limited NPSV response. Experimental ( profiles in , ) and simulated ( and ) , , , , , , , , , the presence of no mediators ( , ) and MG and TA at 15 µM ( , ) and 25 µM ( , ). The determination of empirical parameters was completed by estimating ksol of mediator/analyte pairs upon oxidation of Mb by either MG or TA (Fig. 2.14) and they were used to examine the cumulative effect of mediator mixture on the NPSV profiles. Simulations performed using a single set of parameters (Table 2.1) can adequately model observed NPSV profiles of Mb across multiple experimental conditions (Fig. 2.6, lines). All three key characteristics – amplitude, apparent potential, and the width of transition – are reproduced computationally, including the asymmetry originating from the difference in potentials between the analyte and mediators. Cooperativity between TA and MG in supporting the redox transition 38 MbRdMbOx of Mb was examined by modeling their mixtures (Fig. 2.15), which showed that the mediator capacity of the mixture equals to the sum of the individual contributions of mediator/analyte pairs when measured relative to the residual activity of unmediated Mb. This observation is limited to mediator concentrations in the linear dependence range, since at the saturating concentrations electrolysis of Mb became exhaustive, precluding similar analysis. The agreement between experimental and computational results validates our assumption that no substantial competition or cooperativity should occur in dilute solutions of mediators. This includes homogeneous mediator/mediator reactions, presumably because MG and TA exhibit comparable limiting kinetics and provide no favorable reduction pathway. The latter conclusion may not hold if, for example, a homogeneous reaction between mediators is fast and electrode kinetics of one mediator is much faster than another. Figure 2.14. Determination of homogenous bimolecular rates of reactions between mediators and Mb. Optical changes at characteristic wavelengths were recorded in 2 s intervals as a 0.5 molar equivalent of Mb was added to the reduced MG or TA. Absorbance profiles were taken from 611 nm and 598 nm for MG and TA, respectively, and converted to concentrations using their corresponding ε. The quantitative model was used to find the best ksol value to simulate ( ) the experimental ( ) concentration profiles. 39 Figure 2.15. Cumulative effect of mediator mixtures. Experimental ( , , , ) and simulated ( , , , ) (left) and (right) profiles in the presence of individual and combined mediators. All samples were measured using a 150 s pulse width and contained 1 mM Mb and 15 µM of the indicated mediator(s), MG ( , ), ), MG and TA ( , ), TA ( , or no mediators at all ( , ). DISCUSSION Spectroelectrochemistry with GSR analysis is a powerful tool for resolving components in a multicomponent system that can be used in optical (SSV) and vibrational (NPSV) domains. Its major advantage is the ability to report spectra and electrochemical properties of individual analytes simultaneously, especially when analytes do not exhibit fast, direct electrochemistry. In most cases it is not susceptible to spectral overlaps or regions of substantial background absorption and can further benefit from optimization of mediator concentrations using quantitative modelling. It also permits detection of redox-active insoluble analytes, such as particulate suspensions or immobilized samples. Spectroelectrochemistry is not as sensitive to the rate of the electron transfer as electric detection methods and can identify sites undergoing spectral transition. IR-NPSV eliminates the dependence of OTTLE techniques on the strong visible absorption by redox cofactors, opening the majority of analytes beyond porphyrins for redox-linked vibrational analysis. Here, we illustrate the ability of GSR/NPSV to isolate the contribution of redox mediators and extract an unbiased redox difference spectrum of an unknown analyte using a well-characterized model, Mb, whose NPSV spectrum (Fig. 2.4) is 40 MbRdMbOx nearly identical to that previously reported using SSV.4 The NPSV-measured E½ of Mb is also in remarkable agreement with the reported value of -46.0 mV vs. NHE42 obtained by chemical titrations. The overall resolution of a multi-dimensional technique, such as NPSV, is inherently higher than either spectroscopy or voltammetry separately. Several strategies can facilitate GSR convergence and further improve the resolution, especially when the analyte and the mediators have common spectral features and/or similar . Particularly beneficial is a priori knowledge of , , and for all redox mediators. GSR uses the entire of every mediator at the sampled frequencies instead of characteristic minima/maxima, improving the overall signal to noise ratio. Therefore, it is best to either obtain in situ or to perform interpolation of the known reference spectra. If redox properties of a mediator are sensitive to conditions, they can be obtained at the same time by performing NPSV/GSR on an isolated component. An appropriate integer value can be assigned to nk, if known from the chemical nature and the reaction stoichiometry, and constrained to further reduce the computational load and improve the resolution of . This assumption provided an excellent fit for the NPSV profiles of isolated mediators in this study, but non-unity values can be used in other cases. Such preparatory steps significantly reduce the number of unknown variables in GSR on samples containing unknown component(s) and facilitates convergence as long as there are no tight binary interactions between the components of the mixture with the formation of new complexes. The latter assumption requires validation in preliminary tests. The accuracy of redox properties reported by NPSV/GSR is highly dependent on the properties of the mediators. Slow kinetics of the mediator/analyte reaction and a low population of the effective redox state of the mediator at are the key factors that lead to low amplitudes and an apparent hysteresis (Fig. 2.5). The original 1:100 rule does not consider the time or rate of 41 ½kEkε½kEknkεkε½kE½AE a process. Its literal application to our Mb model inflates the required concentrations threshold several fold for the E½ =157 mV of Mb and potentials of mediators (Fig. 2.5). As the minimal concentration is approximately inversely proportional to the potential pulse in multi-turnover mediation, a larger difference can be predicted for longer potential pulses. While the 1:100 rule may provide a reasonable initial estimate for a general reaction, specific kinetic constraints are particularly critical for mediated reactions with slow kinetics, low stability, mediators with low solubilities, or electrodes with a small relative surface area, which lead us to the development of a quantitative pre-equilibrium model. The ability of the kinetically constrained, pre-equilibrium model to report observables can be used to make experimentally testable predictions well beyond the limits of the original 1:100 rule. For example, by selecting 95% of the redox transition of the analyte as the minimal pre-equilibrium threshold, it is possible to predict the potential step duration necessary to reach that extent of reaction (t95) across a wide range of E½, CM, and ksol values (Fig. 2.16). Such analysis shows that t95 for a single arbitrary mediator/analyte pair is inversely proportional to ksol for slow reactions, but reaches a minimum when ksol exceeds 1000 M-1s-1 using empirical parameters of TA or MG as an example, which arises from changing the relative contributions of ksol and in the overall rate limiting step. Fast solution reactions are limited by , which, in turn, is determined by either the electron transfer rate or the adsorption/desorption process klim. The relative contributions of and klim to the overall rate limiting step can be assessed by varying E½ and CM with or without a reporter analyte and subsequently used to prevent or interpret possible distortion of NPSV profiles (Fig. 2.11, 2.12). 42 elkelk*elk*elk Figure 2.16. The role of mediator in determining the minimum NPSV pulse duration. The minimal pulse duration (t95) is calculated for of an arbitrary 1 mM analyte with k* el of 0.1 and klim of 0.4 following a potential step of Ea = + 100 mV in the presence of a mediator with given bimolecular rate constant ksol. Left: the effect of mediator concentration at = . Right: the effect of relative to for a constant mediator concentration of 100 µM. It is important to emphasize that while NPSV can yield structural and thermodynamic signatures simultaneously, it may be difficult to draw qualitative distinctions between intrinsic analyte properties on one hand and kinetic mediator artifacts on another based solely on the experimental voltammetric data. Utilization of GSR for the resolution of redox active components followed by quantitative modeling of the redox response can alleviate such ambiguity and provide additional insight into underlying phenomena. A similar approach can be implemented across a variety of spectroelectrochemical techniques improving experimental design and interpretation. CONCLUSION We report the development of a set of complementary experimental and computational methods for the characterization of analytes in redox-active, multicomponent mixtures, including analytes that require mediators for an effective electron transfer. Utilization of FTIR spectroelectrochemistry in the NPSV mode opens this method to a broad range of redox active 43 ½AE½AE½ME½ME½AE samples irrespectively of their optical absorption. Complementary GSR and semi-empirical computational modeling of a kinetically limited redox system provided quantitative rationalization of experimental observations, revealed the characteristic effects of kinetic limitations on the NPSV profiles, and permitted accurate predictions of optimal experimental conditions. The optimized quantitative model reported here provides a flexible tool that can be used to determine the empirical properties of analytes on the electrode and interpret mediator/analyte interactions in the solution. Its generalization refines and expands the utility of the 1:100 rule over a wide range of conditions. 44 REFERENCES 45 REFERENCES (1) Moss, D.; Nabedryk, E.; Breton, J.; Mäntele, W. Redox‐linked Conformational Changes in Proteins Detected by a Combination of Infrared Spectroscopy and Protein Electrochemistry: Evaluation of the Technique with Cytochrome c. Eur. J. Biochem. 1990, 187, 565–572. https://doi.org/10.1111/j.1432-1033.1990.tb15338.x. (2) Moss, D. A.; Leonhard, M.; Bauscher, M.; Mantele, W. Electrochemical Redox Titration of Cofactors in the Reaction Center from Rhodobacter Sphaeroides. FEBS Lett. 1991, 283 (1), 33–36. (3) Fritz, F.; Moss, D. A.; Mäntele, W. Electrochemical Titration of the Cytochrome Hemes in the Rhodopseudomonas Viridis Reaction Center Cyclic Equilibrium Titrations Yield Midpoint Potentials without Evidence for Heme Cooperativity. FEBS Lett. 1992, 297, 167–170. https://doi.org/10.1016/0014-5793(92)80352-H. (4) Schlereth, D. D.; Mäntele, W. Redox-Induced Conformational Changes in Myoglobin and Hemoglobin: Electrochemistry and Ultraviolet-Visible and Fourier Transform Infrared Difference Spectroscopy at Surface-Modified Gold Electrodes in an Ultra-Thin-Layer Spectroelectrochemical Cell. Biochemistry 1992, 31, 7494–7502. https://doi.org/10.1021/bi00148a009. (5) Schlereth, D. D.; Mäntele, W.; Fernandez, V. M. Protein Conformational Changes in Tetraheme Cytochromes Detected by FTIR Spectroelectrochemistry: Desulfovibrio Desulfuricans Norway 4 and Desulfovibrio Gigas Cytochromes c3. Biochemistry 1993, 32, 9199–9208. https://doi.org/10.1021/bi00086a027. (6) Arciero, D. M.; Collins, M. J.; Hooper, A. B.; Collins, M. J.; Haladjian, J.; Bianco, P. Resolution of the Four Hemes of Cytochrome C554 from Nitrosomonas Europaea by Redox Potentiometry and Optical Spectroscopy. Biochemistry 1991, 30 (48), 11459– 11465. https://doi.org/10.1021/bi00112a013. (7) Yao, C.; Capdevielle, F. J.; Kadish, K. M.; Bear, J. L. Thin-Layer Microcell for Transmittance Fourier Transform Infrared Spectroelectrochemistry. Anal. Chem. 1989, 61, 2805–2809. (8) Arciero, D. M.; Hooper, A. B.; Collins, M. J. Preliminary Note Low Volume Spectroelectrochemistry : An Anaerobic Optically Transparent Thin Layer Electrode Cell and a Potentiostat 1 UV-Vis Spectrophotometer Interface for Computer-Controlled Data Collection. J. Electroanal. Chem. 1994, 371, 277–281. (9) Schlereth, D. D.; Mäntele, W. Electrochemically Induced Conformational Changes in Cytochrome c Monitored by Fourier Transform Infrared Difference Spectroscopy: Influence of Temperature, PH, and Electrode Surfaces. Biochemistry 1993, 32, 1118– 46 1126. https://doi.org/10.1021/bi00055a018. (10) Graham, P. B.; Curran, D. J. Characterization of a Gold Minigrid Cell for Fourier Transform Infrared Spectroelectrochemistry: Experimental vs Digitally Simulated Response. Anal. Chem. 1992, 64 (22), 2688–2692. https://doi.org/10.1021/ac00046a007. (11) Tsujimura, S.; Kuriyama, A.; Fujieda, N.; Kano, K.; Ikeda, T. Mediated Spectroelectrochemical Titration of Proteins for Redox Potential Measurements by a Separator-Less One-Compartment Bulk Electrolysis Method. Anal. Biochem. 2005, 337 (2), 325–331. https://doi.org/10.1016/j.ab.2004.11.017. (12) Bernad, S.; Mäntele, W. An Innovative Spectroelectrochemical Reflection Cell for Rapid Protein Electrochemistry and Ultraviolet/Visible/Infrared Spectroscopy. Anal. Biochem. 2006, 351, 214–218. https://doi.org/10.1016/j.ab.2005.12.024. (13) Berthomieu, C.; Boussac, A.; Breton, J.; Nabedryk, E.; Mäntele, W. Molecular Changes Following Oxidoreduction of Cytochrome B559 Characterized by Fourier Transform Infrared Difference Spectroscopy and Electron Paramagnetic Resonance: Photooxidation in Photosystem II and Electrochemistry of Isolated Cytochrome B559 and Ir. Biochemistry 1992, 31, 11460–11471. https://doi.org/10.1021/bi00161a026. (14) Mäntele, W.; Wollenweber, A. M.; Rashwan, F.; Heinze, J.; Nabedryk, E.; Berger, G.; Breton, J. Fourier Transform Infrared Spectro- Electrochemistry of the Bacteriochlorophyll a Anion Radical. Photochem Photobiol 1988, 47 (3), 451–456. (15) Leonhard, M.; Mäntele, W. Fourier Transform Infrared Spectroscopy and Electrochemistry of the Primary Electron Donor in Rhodobacter Sphaeroides and Rhodopseudomonas Viridis Reaction Centers: Vibrational Modes of the Pigments in Situ and Evidence for Protein and Water Modes Affected by P+ Formation. Biochemistry 1993, 32, 4532–4538. https://doi.org/10.1021/bi00068a007. (16) Bauscher, M.; Leonhard, M.; Moss, D.; Mäntele, W. Binding and Interaction of the Primary and the Secondary Electron Acceptor Quinones in Bacterial Photosynthesis. An Infrared Spectroelectrochemical Study of Rhodobacter Sphaeroides Reaction Centers. Biochim. Biophys. Acta 1993, 1183, 59–71. https://doi.org/10.1016/0005-2728(93)90005- Z. (17) Kuglstatter, A.; Hellwig, P.; Fritzsch, G.; Wachtveitl, J.; Oesterhelt, D.; Mäntele, W.; Michel, H. Identification of a Hydrogen Bond in the Phe M197-->Tyr Mutant Reaction Center of the Photosynthetic Purple Bacterium Rhodobacter Sphaeroides by X-Ray Crystallography and FTIR Spectroscopy. FEBS Lett. 1999, 463, 169–174. https://doi.org/10.1016/S0014-5793(99)01614-2. (18) Hellwig, P.; Rost, B.; Kaiser, U.; Ostermeier, C.; Michel, H.; Mäntele, W. Carboxyl Group Protonation upon Reduction of the Paracoccus Denitrificans Cytochrome c Oxidase: Direct Evidence by FTIR Spectroscopy. FEBS Lett. 1996, 385, 53–57. 47 https://doi.org/10.1016/0014-5793(96)00342-0. (19) Hellwig, P.; Behr, J.; Ostermeier, C.; Richter, O. M. H.; Pfitzner, U.; Odenwald, A.; Ludwig, B.; Michel, H.; Mäntele, W. Involvement of Glutamic Acid 278 in the Redox Reaction of the Cytochrome c Oxidase from Paracoccus Denitrificans Investigated by FTIR Spectroscopy. Biochemistry 1998, 37, 7390–7399. https://doi.org/10.1021/bi9725576. (20) Hellwig, P.; Ostermeier, C.; Michel, H.; Ludwig, B.; Mäntele, W. Electrochemically Induced FT-IR Difference Spectra of the Two- and Four-Subunit Cytochrome c Oxidase from P. Denitrificans Reveal Identical Conformational Changes upon Redox Transitions. Biochim. Biophys. Acta - Bioenerg. 1998, 1409, 107–112. https://doi.org/10.1016/S0005- 2728(98)00151-0. (21) Ritter, M.; Anderka, O.; Ludwig, B.; Mäntele, W.; Hellwig, P. Electrochemical and FTIR Spectroscopic Characterization of the Cytochrome Bc1 Complex from Paracoccus Denitrificans: Evidence for Protonation Reactions Coupled to Quinone Binding. Biochemistry 2003, 42 (42), 12391–12399. https://doi.org/10.1021/bi035103j. (22) Baymann, F.; Robertson, D. E.; Dutton, P. L.; Mäntele, W. Electrochemical and Spectroscopic Investigations of the Cytochrome Bc1 Complex from Rhodobacter Capsulatus. Biochemistry 1999, 38, 13188–13199. https://doi.org/10.1021/bi990565b. (23) Brischwein, M.; Scharf, B.; Engelhard, M.; Mäntele, W. Analysis of the Redox Reaction of an Archaebacterial Copper Protein, Halocyanin, by Electrochemistry and FTIR Difference Spectroscopy. Biochemistry 1993, 32, 13710–13717. https://doi.org/10.1021/bi00212a041. (24) Clair, C. S. S.; Gray, H. B.; Valentine, J. S. Spectroelectrochemistry of Copper-Zinc Superoxide Dismutase. Inorg. Chem. 1992, 31, 925–927. https://doi.org/10.1021/ic00031a041. (25) DeLacey, A. L.; Stadler, C.; Fernandez, V. M.; Hatchikian, E. C.; Fan, H. J.; Li, S.; Hall, M. B. IR Spectroelectrochemical Study of the Binding of Carbon Monoxide to the Active Site of Desulfovibrio Fructosovorans Ni-Fe Hydrogenase. J. Biol. Inorg. Chem. 2002, 7, 318–326. https://doi.org/10.1007/s00775-001-0301-7. (26) DeLacey, A. L.; Fernandez, V. M.; Rousset, M.; Cavazza, C.; Hatchikian, E. C. Spectroscopic and Kinetic Characterization of Active Site Mutants of Desulfovibrio Fructosovorans Ni-Fe Hydrogenase. J. Biol. Inorg. Chem. 2003, 8, 129–134. https://doi.org/10.1007/s00775-002-0397-4. (27) Ash, P. A.; Vincent, K. A. Spectroscopic Analysis of Immobilised Redox Enzymes under Direct Electrochemical Control. Chem. Commun. 2012, 48, 1400–1409. https://doi.org/10.1039/C1CC15871F. 48 (28) Melin, F.; Hellwig, P. Recent Advances in the Electrochemistry and Spectroelectrochemistry of Membrane Proteins. Biol. Chem. 2013, 394, 593–609. https://doi.org/10.1515/hsz-2012-0344. (29) Murgida, D. H.; Hildebrandt, P. Redox and Redox-Coupled Processes of Heme Proteins and Enzymes at Electrochemical Interfaces. Phys. Chem. Chem. Phys. 2005, 7, 3773– 3784. https://doi.org/10.1039/b507989f. (30) Berthomieu, C.; Boussac, A.; Breton, J.; Nabedryk, E.; Mäntele, W. Molecular Changes Following Oxidoreduction of Cytochrome B559 Characterized by Fourier Transform Infrared Difference Spectroscopy and Electron Paramagnetic Resonance: Photooxidation in Photosystem II and Electrochemistry of Isolated Cytochrome B559 and Ir. Biochemistry 1992, 31 (46), 11460–11471. https://doi.org/10.1021/bi00161a026. (31) Hellwig, P.; Rost, B.; Kaiser, U.; Ostermeier, C.; Michel, H.; Mäntele, W. Carboxyl Group Protonation upon Reduction of the Paracoccus Denitrificans Cytochrome c Oxidase: Direct Evidence by FTIR Spectroscopy. FEBS Lett. 1996, 385 (1–2), 53–57. https://doi.org/10.1016/0014-5793(96)00342-0. (32) Meckstroth, M. L.; Norris, B. J.; Heineman, W. R. Mediator-Titrants for Thin-Layer Spectroelectrochemical Measurement of Biocomponent U°’ and n Values. Bioelectrochemistry Bioenerg. 1981, 8, 63–70. (33) Fultz, M. Lou; Durst, R. A. Mediator Compounds for the Electrochemical Study of Biological Redox Systems: A Compilation. Anal. Chim. Acta 1982, 140 (1), 1–18. https://doi.org/10.1016/S0003-2670(01)95447-9. (34) Covington, A. K.; Paabo, M.; Robinson, R. A.; Bates, R. G. Use of the Glass Electrode in Deuterium Oxide and the Relation between the Standardized PD (PaD) Scale and the Operational PH in Heavy Water. Anal. Chem. 1968, 40 (4), 700–706. https://doi.org/10.1021/ac60260a013. (35) Dai, Y.; Zheng, Y.; Swain, G. M.; Proshlyakov, D. A. Equilibrium and Kinetic Behavior 3-/4- and Cytochrome c in Direct Electrochemistry Using a Film Electrode Thin- of Fe(CN)6 Layer Transmission Cell. Anal. Chem. 2011, 83, 542–548. (36) Swain, G. M.; Ramesham, R. The Electrochemical Activity of Boron-Doped Polycrystalline Diamond Thin Film Electrodes. Anal. Chem. 1993, 65 (4), 345–351. https://doi.org/10.1021/ac00052a007. (37) Show, Y.; Witek, M. A.; Sonthalia, P.; Swain, G. M. Characterization and Electrochemical Responsiveness of Electrodes. Chem. Mater. 2003, 15, 879–888. https://doi.org/10.1021/cm020927t. (38) John, C. W.; Proshlyakov, D. A. KinESim. https://github.com/Dap-Biospec/KinESim, 2019. (Accessed June 14, 2019). 49 (39) Grzyska, P. K.; Appelman, E. H.; Hausinger, R. P.; Proshlyakov, D. A. Insight into the Mechanism of an Iron Dioxygenase by Resolution of Steps Following the FeIV=O Species. Proc. Natl. Acad. Sci. U. S. A. 2010, 107, 3982–3987. https://doi.org/10.1073/pnas.0911565107. (40) Zhou, D. M.; Fang, H. Q.; Chen, H. Y.; Ju, H. X.; Wang, Y. The Electrochemical Polymerization of Methylene Green and Its Electrocatalysis for the Oxidation of NADH. Anal. Chim. Acta 1996, 329 (1–2), 41–48. https://doi.org/10.1016/0003-2670(96)00117-1. (41) Chen, H. Y.; Zhou, D. M.; Xu, J. J.; Fang, H. Q. Electrocatalytic Oxidation of NADH at a Gold Electrode Modified by Thionine Covalently Bound to Self-Assembled Cysteamine Monolayers. J. Electroanal. Chem. 1997, 422 (1–2), 21–25. https://doi.org/10.1016/S0022-0728(96)04887-5. (42) Taylor, J. F.; Morgan, V. E. Oxidation-Reduction Potentials of the Metmyoglobin- Myoglobin System. J. Biol. Chem. 1942, 144, 15–20. (43) Butcher, J. C. Numerical Methods for Ordinary Differential Equations; John Wiley & Sons, 2016. 50 CHAPTER 3: STRONGLY COUPLED REDOX-LINKED CONFORMATIONAL SWITCHING AT THE ACTIVE SITE OF THE NON-HEME IRON-DEPENDENT DIOXYGENASE, TAUD 51 INTRODUCTION The 2-oxoglutarate (2OG)-dependent dioxygenases activate C-H bonds while catalyzing a variety of biologically relevant reactions,1 including synthesis of a wide range of commercial products.2 2OG:taurine dioxygenase (TauD), the archetypical member of this enzyme family,3,4 is found in Escherichia coli where it metabolizes taurine (2-aminoethane sulfonate) as a sulfur source for sulfur-starved cells.5 TauD activates taurine via hydrogen atom transfer (HAT) to an Fe(IV)=O intermediate J (shown as F4 in Fig. 1.5),6 followed by hydroxyl rebound and product degradation.7 A time-resolved resonance Raman study of TauD revealed the existence of two transient intermediates that trailed F4 in time and assigned them to the νFe-O and νas modes of the ferric (hydr)oxo (F3) and alkoxo (FX) intermediates, respectively.8 HAT by F4 in TauD is analogous to HAT by Compound I in cytochromes P450 (CYP450)9 except that TauD is expected to yield a transient non-heme Fe(III)-OH complex (Fig. 1.5) instead of a heme Fe(IV)-OH in Compound II of CYP450. However, the Raman study found no vibrational evidence for protonation of the oxo group in F3, attributing its absence to a proton transfer to a nearby base that allows for the formation of an ensuing FX species. While the alkoxo structure of FX received additional support in a recent crystallographic study of another 2OG-dependent hydroxylase, VioC,10 the lack of proton sensitivity of F3 is unexpected for two reasons. First, the Raman study8 was conducted at pH 8 and well below the pKa  25 (in DMSO) reported for Fe(III)-OH model complexes.11 The low dielectric environment of the active site12 is also expected to increase the pKa over that in an aqueous medium. Second, the pKa of F3 contributes to the HAT capacity of F4 per Bordwell’s thermodynamic cycle (Fig. 1.3),13,14 as has been demonstrated for the analogous pKa of Compound II in CYP450 (pKa = 12)15 in contrast to peroxidases (pKa < 4).16,17 52 Since the geometry of the Fe(III)-OH ligand may reduce the sensitivity of the νFe-OH Raman mode to 1H/2H substitution, it is important to examine the protonation state of F3 using other techniques. Spectroelectrochemistry can reveal protonation events either directly from spectroscopic changes or indirectly from the pH sensitivity of the Fe(II)/Fe(III) redox potential.3,18,19 This information can then be used to reconstruct the protonation states of Fe(III)-TauD as an in situ model of the transient F3 species from the known crystallographic structures of Fe(II)-TauD. The weak UV-Vis absorption of the non-heme iron center in TauD20 makes direct detection of the Fe(II)/Fe(III) redox transition by optical spectroscopy impractical. Much of the pioneering work by Mäntele et al.21–23 focused on proteins with porphyrin cofactors, with their strong, oxidation state-sensitive optical absorption spectra. Similar methods were applied to the relatively intense chromophores in copper enzymes18,24 and Ni-Fe hydrogenases.25,26 As an alternative to optical spectroscopy, one can exploit reaction-induced infrared (IR) difference spectrocopy27,28 for the detection of reversible vibrational changes of ligands in response to the redox state changes of the metal.29,30 Here, we investigated several complexes of TauD (Fig. 1.5) using normal pulse spectrovoltametry (NPSV) with IR detection to identify structural and redox properties simultaneously.31 This study revealed a redox-linked conformational change at the active site of TauD that modulates its redox potential at an unprecedented magnitude. EXPERIMENTAL PROCEDURES SAMPLE PREPARATION TauD apoprotein was purified as previously described32 with the following modifications: cell cultures were grown in six 2 L flasks, each containing 1 L of Terrific Broth medium, at 37 53 °C and with shaking at 200 rpm. A final concentration of 1 mM isopropyl β-D-1- thiogalactopyranoside was added to each flask after reaching an O.D. of 0.6 at 600 nm. The temperature was reduced to 30 °C and the cultures were grown overnight. Pelleted cells were stored at -80 °C until needed. The cells were thawed and resuspended in a lysis buffer containing 20 mM Tris, pH 8, 1 mM ethylenediaminetetraacetic acid, and 1 mM phenylmethanesulfonyl fluoride prior to being lysed by sonication. Fe(III)-TauD was prepared by adding a 1:1 molar ratio of ferrous ammonium sulfate to TauD apoprotein under anaerobic conditions, followed by oxidation using a 5-fold excess of ferricyanide (FCN) for 1 h and the removal of ferri/ferrocyanide using a GE Healthcare PD-10 desalting column. Fe(III)-TauD samples were stored on ice in 25 mM Tris buffer, pH 8. The final sample buffer was exchanged using a 10 kDa centrifugation membrane unit (Amicon). 2OG-Fe(III)-TauD and taurine-2OG-Fe(III)-TauD were prepared by adding 2-fold excess of 2OG and/or 3-fold excess of taurine to approximately 1 mM Fe(III)-TauD. FCN, methylene green (MG), and thionine acetate (TA) mediators were added to achieve 100 μM of each in the final solution, unless noted otherwise. All samples were prepared in 25 mM Tris buffer at pD 8.5 in D2O with 0.5 M KCl. SPECTROSCOPIC MEASUREMENTS NPSV measurements were performed using a 12.5 μm pathlength optically transparent thin layer electrochemical (OTTLE) cell31 over a boron-doped diamond working electrode and a Ag/AgCl (0.5 M KCl) reference electrode at 10 °C.33,34 Electrochemical measurements were performed using a computer-controlled potentiostat (Model CHI1202b, CH Instruments). Reference potentials (Er) in reduction and oxidation modes were +0.5 V and -0.6 V, respectively, 54 with a potential increment of 0.04 V. The potential pulse duration was 300 s, unless otherwise noted, with FTIR spectra acquisition (Equinox 55/S, Bruker) during the final 120 s (Fig. 2.1). NPSV data were analyzed by a non-linear global spectral regression (GSR)35 and kinetic simulations were performed using KinESim36 packages for Igor Pro.31 The solution rate constants ksol for the reaction of MG and TA with Fe(III)-TauD were determined as previously described for myoglobin (Mb).31 The ksol for the oxidation of Fe(II)- TauD by FCN was determined anaerobically while stirring against 100 µM FCN in 25 mM Tris buffer, pH 8. The reaction was initiated by the injection of Fe(II)-TauD to achieve an equimolar ratio with the mediator. UV-Vis spectra (Hewlett Packard 8453) were collected every 2 s for 20 s before and 10 min after the start of the reaction. The bimolecular rate constant was obtained by fitting the temporal changes at 420 nm. RESULTS NPSV (Fig. 2.1) utilizes a series of applied potentials, Ea,i, each incremented by a defined potential step and preceded by a reference potential, Er, i. The absolute Fourier transform infrared (FTIR) spectra were acquired at the end of each potential step over a spectral integration period when the reaction had reached equilibrium. Spectra acquired sequentially under alternating potentials Ea,i (Sa,i) and Er,i (Sr,i) formed a raw absolute data matrix (Fig. 2.3, top). This matrix was reduced into a redox difference matrix of spectra (ΔSi, Fig. 2.3, bottom): (1) Experimental values of ΔSi at 1681 cm-1 are shown by markers in Fig. 3.1, right. For the oxidation NPSV segment (blue) was measured vs. Er,i = -0.6 V as Ea,i was increased from - 55 ,,,112iaiririΔSSSS1681A 0.14 to 0.34 V. For the reduction sweep (red), Ea,i was decreased from 0.14 to -0.38 V and was measured vs. Er,i = +0.5 V. The redox-difference dataset (ΔS) was deconvoluted into a full occupancy spectra, ΔStot, and Nernstian population profiles, , so that each redox difference spectrum ΔSi at the i-th NPSV step is described as: (2) where is a property of the analyte and the applied potential (Ea).31 The resulting profiles (lines in Fig. 3.1, right) represent Ea-dependent intensities of the entire spectra ΔStot (Fig. 3.1, left) as opposed to experimental at a selected frequency (markers). Fe-TauD is redox inactive on the electrode without mediators (MG, TA, and FCN), each of which is also described by its own ΔStot , E½, and n.31 Fig. 3.1 shows ΔStot (left) and (right) of the Fe(II)/Fe(III) transition in Fe-TauD, 2OG-Fe-TauD, and taurine-2OG-Fe-TauD. Notably, mediators are described by their own matrices and do not contribute to TauD spectra. The close similarities between the reduction and oxidation FTIR spectra in all three complexes show that the observed redox processes are fully reversible. The spectra were dominated by vibrational changes in the 1550-1700 cm-1 region, outside of which only a vibration at ~1400 cm-1 was consistently observed in all forms. Redox changes in the 1650-1700 cm-1 region were specific to the 2OG-containing complexes and were altered upon taurine binding (Fig. 3.1). All complexes exhibited changes in the amide I stretching region at 1632/1624 cm-1 (Fe-TauD) or 1638/1630 cm-1 (2OG-Fe-TauD and taurine-2OG-Fe-TauD).37 A prominent trough was observed at 1580 cm-1 in Fe-TauD with a likely corresponding peak at 56 1681AiitotΔSΔS,½(,,)aiiEEnfAtotΔSΔS 1614 cm-1; these features showed a 20 cm-1 downshift upon binding of 2OG (2OG-Fe-TauD and taurine-2OG-TauD). Figure 3.1. NPSV transitions in Fe-TauD, 2OG-Fe-TauD, and taurine-2OG-Fe-TauD. Redox mediators: 100 µM MG, 100 µM TA, and 100 µM FCN Left: ΔStot of the reduction (—) and oxidation steps (—). Positive modes represent ferric TauD and negative modes represent ferrous TauD. Right: Experimental profiles of reduction steps ( ) and oxidation steps ( ),normalized using , and the fitted profiles (—, —). , , and In contrast to the redox spectra, the reduction and oxidation population profiles, and , were sharply different, exhibiting a large separation of apparent potentials (Fig. 3.1, right) and the appearance of minor oxidation phases in Fe-TauD and taurine-2OG-Fe-TauD (Fig. 3.1). Results of fitting to Nernstian profiles with n = 1 are shown in Table 3.1. The appearance of a large redox hysteresis in all three forms of TauD contrasts with the behavior of Mb using the same OTTLE cell, electrode, and mediators (except FCN),31 raising the possibility that the shift in potentials originates from the difference in the specific interactions of reduced and oxidized 57 Fe-TauD1632A2OG-Fe-TauD1681Ataurine-2OG-Fe-TauD1692ARdOx forms of the mediator with the protein moiety rather than of the metal. This possibility was excluded using Zn2+-substituted TauD,20 which was found to be redox inactive (Fig. 3.2), attributing the observed changes to the redox transitions of the iron center. Table 3.1. Reduction and oxidation potentials of TauD calculated from experimental NPSV profiles. The maximal magnitude of the observed hysteresis (ΔE½) is shown. TauD forms ERd (mV) EOx (mV) Fe-TauD 2OG-Fe-TauD Taurine-2OG-Fe- -154 ± 22 -127 ± 21 -124 ± 16 TauD Phase 2 Phase 1 -153 ± 20 187 ± 19 171 ± 17 -134 ± 48 174 ± 48 - ΔE½ (mV) 341 298 298 Figure 3.2. NPSV of redox transitions in 2OG-TauD with Zn or Fe bound. Redox mediators: 100 µM MG, 100 µM TA, and 100 µM FCN The second possibility is that the apparent shift in potentials arises from the kinetic or thermodynamic limitations of the TauD/mediator reactions, particularly due to the redox gap between the E½ of FCN and the other two mediators (Table 3.2), which is comparable in magnitude to the hysteresis observed in TauD. Since this potential gap could not be bridged experimentally due to low solubility, instability, or irreversible interactions of other mediators with TauD (Table 3.3), the role of kinetic limitations of a discontinuous ladder of mediators was investigated using a semi-empirical model of heterogeneous mediated electrochemistry (Fig. 2.7).31 Since this model was developed using the same mediators and electrochemical system as 58 ½AE reported here, it required only parametrization of FCN, including the mediator-specific heterogeneous potential-dependent (k* el) and potential-independent (klim) effective rate-limiting constants on the electrode and the bimolecular rate constant (ksol) for the homogeneous FCN/TauD reaction (Table 3.2). These kinetic parameters were used to simulate concentration profiles of TauD, which were integrated over specified time periods to obtain and compared with experimental profiles obtained under identical conditions. Figure 3.3. Titration of 70 µM 2OG-Fe(II)-TauD into 100 µM FCN. The initial electron transfer rate between TauD and FCN was calculated to be >> 15,000 M-1s-1. The value of k2, the conversion rate in model [2], is estimated to be 9 x 10-4 s-1 based on the rate of the continued reduction of FCN beginning at about 5 s. Two alternative reaction schemes were investigated to assess the origin of the redox hysteresis in TauD (Fig. 3.4). In the single state model [1], identical to what was used for Mb,31 TauD exists in a simple redox equilibrium with a single redox potential, . The redox hysteresis using [1] could originate only from the inefficiency of the mediators. The redox-linked switching model [2] includes at least two distinct conformations of TauD, conformers A and Aʹ, each with a distinct redox potential (Fig. 3.4, right). Since the experimental NPSV data report the combined contribution of all isomers of TauD, simulated populations of conformers A and Aʹ were combined as well. The combined populations yielded only one or two apparent NPSV 59 ½AE Figure 3.4. Chemical models for the mediated electrochemistry of TauD. The single state model [1] describes a homogeneous population of A, with a single redox potential, . The redox switching model [2] allows for redox-linked isomerization of the analyte between conformations A and Aʹ with distinct redox potentials, and , respectively, and preferential stability in the oxidized and reduced forms. Direct electron transfer is allowed only for the mediator M. See text for the estimates and interpretation of forward and reverse isomerization rates. Figure 3.5. Effect of thermodynamic properties of the analyte on the apparent NPSV redox (blue) profiles of an analyte (A) hysteresis for models [1] and [2]. Simulated (red) and that is completely dependent on mediators (Table 3.2). The intrinsic E½ of A (left) or A, Aʹ (right) used in simulations are indicated to the right of each plot. The observed E½ values are indicated next to each transition, including minor transitions, where present. 60 ½AE½AE'½AERdOx transitions even if the responses of individual conformers were more complex. Conformational changes of TauD using [2] are described by separate sets of rates and equilibria constants in the reduced and oxidized states. The values of k1 and k2 were estimated from the slow phases of oxidation by FCN and were found to be approximately 9 × 10-4 s-1 (Fig. 3.3). The extent of the oxidation of Fe(II)-TauD by FCN over the first 30 s allowed for the estimation of K1 ≥ 102 and the reverse isomerization rate k-1 10-5 s-1. Since ΔG = 0 for a cyclic process, one can calculate K2 = 9.6×10-6 from experimental estimates of , , and K1. Table 3.2. List of mediators used during NPSV measurements with their thermodynamic and kinetic properties. E½ is measured versus a Ag/AgCl reference, n is the electron transfer el and klim are defined separately,a and ksol is a homogeneous bimolecular rate coefficient, k* constant for the reaction with TauD. All values were determined empirically. Mediator E½ (mV) FCN MG TA 215 b -111, -243 -197 c n 1 1, 1 2 k* el (s-1) 0.1 0.025 0.5 klim (s-1) 0.2 0.5 1.88 ksol (M-1s-1) 2.0 × 104 1.4 × 104 1.7 × 104 a John, C. W.; Proshlyakov, D. A. Anal. Chem. 2019, xx, xx–xx. b Kolthoff, I. M.; Tomsicek, W. J. J. Phys. Chem. 1934, 39 (7), 945 c Chen, H. Y.; Zhou, D. M.; Xu, J. J.; Fang, H. Q. J. Electroanal. Chem. 1997, 422 (1–2), 21 The NPSV response was predicted for the two alternative models for 2OG-Fe-TauD while varying (for [1]) or and (for [2]). The simulated and profiles shown in Fig. 3.5 are directly comparable to those obtained from experimental NPSV data in Fig. 3.1, right. Simulations using [1] with an within 50 mV of that of any mediator show nearly complete reversibility, as was observed for Mb.31 As approached the middle of the potential gap between MG and FCN, exhibited an increasing degree of distortion with the development of an apparent hysteresis between -60 and +140 mV in the reduction and oxidation modes, respectively, and the appearance of minor redox transitions (Fig. 3.6 and Table 3.4). At = 0 mV the hysteresis reaches a maximum amplitude of 213 mV, which is much smaller than the 298 61 ½AE'½AE½AE½AE½AEAOxARd½AE½AEA½AE mV hysteresis observed in TauD. The largest hysteresis of 199 mV was observed using [1] with symmetrical and and = 60 mV (Fig. 3.5). Table 3.3. List of other mediators tested for NPSV of TauD. Mediator Ru(NH3)6 a 2,3,5,6-TMPD b,c N,N,Nʹ,Nʹ-TMPD b,c indophenol b 1,2-naphthoquinone b Fe-EDTA b methylene blue b methyl viologen b E½ (mV) 100 52 25 -9 -42 -82 -236 -687 n 1 1 1 2 2 1 2 2 a Meyer, T. J.; Taube, H. Inorg. Chem. 1968, 7 (11), 2369 b Fultz, M. Lou; Durst, R. A. Anal. Chim. Acta 1982, 140 (1), 1 c Rawson, F. J.; Downard, A. J.; Baronian, K. H. Sci. Rep. 2015, 4 (1), 5216 Figure 3.6. The effect of the intrinsic E½ on NPSV redox transitions in model [1]. Simulations and NPSV integrations were performed under realistic conditions. Marker sizes represent relative amplitudes of the major and minor fitted phases (EObs) of corresponding profiles in the reduction ( ) and oxidation ( ) NPSV modes. The intrinsic illustrated in Fig. 3.5, left, are indicated by vertical dashed lines and their intercepts with the plot represent the apparent transitions. The diagonal dashed line represents ideal Nernstian behavior. Arrows indicate the E½ of individual mediators. Analogous simulations using model [2], with the addition of , confirmed that the profile of and the magnitude of the observed hysteresis depend primarily on the intrinsic 62 AOxARd½AEAA½E½AETauD potentials and (Fig. 3.5) when their difference is large. When ≈ , [2] reduces to [1] and the observed was determined by the gap in the mediator ladder. Therefore, there were multiple conditions where the shape of alone was not sufficient to unambiguously distinguish between [1] and [2] based on the hysteresis alone. However, the two models predict different sensitivities to mediator concentrations, particularly that of FCN due to the absence other mediators in its effective potential range. Table 3.4. Calculated oxidation and reduction potentials of profiles generated using model [1]. Phases representing >50% of the observed oxidation or reduction are bolded. The largest observed hysteresis is italicized. EOx1 (mV) ΔE½,max (mV) ERd2 (mV) EOx2 (mV) ERd1 (mV) E½ actual (mV) 200 180 160 140 120 100 80 60 40 20 0 -20 -40 -60 -80 -100 -120 -140 -160 -180 -200 ΔE½,major (mV) 126 101 59.5 38.2 18.9 3.55 -10.1 -22.6 -35 -47.6 -61.3 -78.7 -99.0 -120 -140 -160 -180 -200 200 180 160 169 164 160 158 156 155 153 151 146 130 113 113 201 187 174 162 151 139 127 114 88.7 41.9 -10.4 -17.3 -17.7 -2.02 16.7 -15 -19.5 -27.8 -33.2 -37.4 -40.5 -42.6 -45.1 -49.1 -62 -65.6 -70.3 -74.6 -83.6 -102 -120 -140 -160 -180 -200 183 195 180 197 197 197 199 199 200 202 213 212 200 188 197 3 0 0 0 0 0 1 7 14 7 197 197 199 199 200 202 213 30.6 22.7 13.3 4.9 3 0 0 0 0 0 The effects of FCN concentration on and using models [1] vs. [2] are illustrated in Fig. 3.7. For [1], we selected E½ = 60 mV for which the hysteresis is the most sensitive to FCN 63 ½AE½AE½AE½AETauDTauDOxRd Figure 3.7. The effect of FCN concentration on the observed and profiles. Experimental ( ) and ( ) for 1 mM TauD (center) are compared with corresponding profiles predicted by models [1] (left) and [2] (right) at the indicated concentrations of FCN. The using [1] was 60 mV. and using [2] were -130 and 100 mV, respectively. The EObs (mV) of major phases are shown. concentration. Simulations using [2] were conducted with = -140 mV and = 100 mV, as these values yielded an apparent hysteresis comparable to that observed experimentally. An increase in the concentration of FCN from 0.1 mM to 2.5 mM decreased the oxidation EObs by 66 mV using [1] (131 mV – 65.2 mV) and 78 mV using [2] (158 mV – 80 mV) vs. the experimentally observed decrease of 121 mV (171 mV – 50 mV). Oxidation EObs using [2] decreased below in agreement with the continuous changes in experimental value (Fig. 3.7, middle) and contrary to the exponential saturation predicted for [1]. However, the largest discrepancy was observed in the apparent reduction potential, which increased using [1] and remained essentially unchanged using [2]. As a result, [1] predicts the loss of the hysteresis at 2.5 mM FCN, in contrast to the hysteresis of 198 mV observed experimentally and 184 mV 64 RdOxTauDRdTauDOx½AE½AE'½AE½AE'½AE'½AE predicted using [2]. The contribution of the minor phase also diminishes at high FCN concentration. The complete removal of FCN is expected to hinder the oxidation process and exacerbate the non-ideal behavior of TauD differently for models [1] and [2], allowing for further discrimination between the models. As the applied potentials and approach E½ (Fig. 3.8, left; NPV cycles 4 and 5), facile reduction of the analyte leads to large changes in its population during the and steps. This effect increases the reversible NPSV response using [1] ( and ), which is defined as .31 However, this reduction is followed by only a minimal re-oxidation during and steps due to the limited oxidizing capacity of low potential mediators. Subsequent steps show exhaustive reduction with smaller amplitudes ( ) and the NPSV response of the mostly reduced sample falls ( ; Fig 7, right). This process results in a pseudo-equilibrium region where changes in become potential-independent due to inefficient oxidation of TauD by MG and TA even at a high Er (shaded area in Fig. 3.8). The magnitude of in this region represents only the extent of the change in redox state ( <20%, right) and not the major redox state of TauD (>80% reduced and , left). 65 ,4aE,5aE,4,4rar,5,5a45,,,10.5iairiai,4,5ar,5,6ar,6a6i7,7a,7r Figure 3.8. Kinetically limited NPSV reduction profile in the absence of FCN. Left: The concentration of oxidized analyte (solid line, E½ = 80 mV) and oxidized MG (dashed line) predicted by [1] over successive NPSV cycles (dotted line). Simulation conditions for a realistic mediator cocktail are identical to the 1000 s pulse width data in Fig. 3.10 (left). TA is present, but not shown for clarity. Integrated that determine and are shown in blue and red, respectively. Right: The integrated NPSV profile for decreasing from +0.1 V to - 0.2 V over eight steps against = 0.1 V. The maximal amplitude is observed for due to facile reduction during the → step. The amplitudes of are decreased due to the slow reduction, i.e. → step. An expanded view of this profile is shown in Fig. 3.10 (left top, open markers) in comparison with shorter pulse duration and the oxidation mode profiles. The pseudo-equilibrium region is highlighted in grey. A different response can be expected using [2] in the absence of FCN, because protein isomerization provides an alternative pathway for oxidation of Aʹ via A. Furthermore, both models predict that non-equilibrium profiles would exhibit substantial dependence on pulse duration and analyte potentials. Predictions of the two models (Fig. 3.10 left and right) differ in three characteristic parameters. i) The onsets of the and transitions using [1] are narrow (up to in Fig. 3.8) and follow the profiles of an n = 1 redox process. Using [1] this is followed by a sharp transition to the pseudo-equilibrium phase ( in Fig. 3.8). In contrast, the onset is much broader using [2] (n = 0.2 – 0.5, for the best fit lines in Fig. 3.10, right) and no distinct pseudo-equilibrium region is observed. ii) The alternative oxidation pathway using [2] results in a larger NPSV amplitude following the onset of the redox transition. A second NPSV peak in the reduction mode is predicted using [2] under some conditions (Fig. 3.9), which 66 i57aErE5,5r,5a6,7a,8rOxRd56 reflects a direct contribution of the low potential redox transition. The pseudo-equilibrium region using [1] is always featureless and its amplitude diminishes rapidly at high E½. iii) The apparent potentials of maximal and responses are close to the intrinsic using [1] ( = 110 mV) in contrast to a much larger apparent difference for [2] ( = 330 mV) (Fig. 3.10, black lines). The experimental , obtained at 300 s and 1000 s NPSV pulse widths, are compared to the matching simulations using models [1] and [2] in Fig. 3.9 and 3.10. As argued above for Fig. 3.5, experimental observations for the full mediator cocktail limit using [1] to the range of 40 - 80 mV, but these values result in a substantially larger amplitude in the absence of FCN than observed experimentally (Fig. 3.9). The maximum experimental NPSV amplitude of 2OG-Fe-TauD with 300 s and 1000 s pulses in the absence of FCN (Fig. 3.10) was > 0.7 mM and < 0.18 mM for the reduction and oxidation modes, respectively. Such an amplitude was observed using [1] only for of +100 to +130 mV, where the hysteresis in the full mediator cocktail is already smaller than observed for TauD. The experimental NPSV onset was much broader than predicted for [1] with E½ < +130 mV and showed a better correlation using simulation [2] with > +140 mV. The relative amplitudes of and with 300 s or 1000 s NPV pulse widths further support model [2] over [1]. 67 OxRd½AEObsEObsETauD½AE½AE½AE'½AETauDOxTauDRd Figure 3.9. Effect of thermodynamic properties of the analyte on the apparent NPSV redox hysteresis when FCN is not present. Simulations were carried out for 1 mM analyte in the presence of 0.5 mM MG and TA. The E½ of the analyte is indicated above each simulation set for model [1] and both values are shown for model [2]. Population profiles using model [2] are the sum of both conformations A and Aʹ. Simulations were carried out using a 300 s pulse width (closed circles) and 1000 s pulse width (open circles). Figure 3.10. Comparison of experimental NPSV hysteresis with model-depended predictions in the absence of FCN. Experimental ( , ) and ( , ) of 1 mM TauD with 0.5 mM MG and saturated TA (<150 µM) but no FCN (center) are compared with corresponding profiles predicted by models [1] (left) and [2] (right). Experimental and simulated profiles were acquired for 300 s ( , ) and 1000 s ( , ) NPSV pulse widths. Best fit Nernstian ; [1]: n = 1, ΔEobs = 0.11 V; [2]: nox = 0.53, nrd = 0.23, ΔEobs profiles are shown for reference ( = 0.33 V). 68 TauDRdTauDOx DISCUSSION The redox cycle of anaerobic Fe-TauD is vibrationally fully reversible (Fig. 3.1). The observation of an electrochemical hysteresis in all forms of TauD (Fig. 3.1) was surprising and sharply contrasts with the simple redox transition of myoglobin.31 The lack of suitable mediators with E½ of 0 – 100 mV required us to conduct an investigation into the role of kinetic and thermodynamic limitations of mediated electrochemistry using two alternative models (Fig. 3.4) before being able to state unequivocally that the hysteresis is attributed to intrinsic properties of Fe-TauD. The application of a semi-empirical model of heterogeneous thin layer electrochemistry showed that both models can yield comparable NPSV results using specific Fe- TauD potentials at the given mediator concentrations. However, no single value in model [1] could yield and consistent with experimental observations across all the conditions tested here, giving strong support for intrinsic redox-linked reorganization in Fe-TauD, as follows. The largest hysteresis predicted for model [1] was 213 mV ( = 0.0 mV), which is 85 mV smaller than that experimentally observed in 2OG-Fe-TauD and taurine-2OG-Fe-TauD and 128 mV smaller than that observed in Fe-TauD (Table 3.1). The magnitude of the hysteresis and apparent values in 2OG-Fe-TauD and taurine-2OG-Fe-TauD could be reproduced by model [2] using = -130 mV and = 180 mV (Fig. 3.5). The observed reduction are within the effective potential range of MG and TA, as is evident for Mb,31 and a negligible hysteresis is predicted for [1] with E½ = -130 mV (Fig. 3.5). This observation is important for two reasons. First, if TauD has a single E½ that falls within the effective range of TA and MG, both the reduction and oxidation should occur at this potential unless there is a major difference in the interaction of TauD with the reduced or oxidized forms of both MG and TA, which falls 69 ½TauDEOxRd½TauDE½TauDE½AE½AE½TauDE into the definition of [2]. Second, the E½ of Mb (-157 mV by NPSV, no hysteresis) is similar to the observed reduction potential of Fe-TauD (-154 mV, large hysteresis).31 The differences in the bimolecular rates of electron transfer between the mediators and Mb or TauD are negligible for the NPSV pulse duration and an equilibrium is expected to be reached before each spectral acquisition. The observed oxidation may be biased positive if it falls below the effective range for the current FCN concentration and well above the E½ of MG. Such bias can be reduced or eliminated at higher FCN concentrations in both models in agreement with the observed (Fig. 3.7); however, the two models predicted an increasingly divergent response of the apparent reduction potential with increasing FCN. The observed showed distinctly better agreement with model [2] than with [1]. The complete removal of FCN as a mediator resulted in several characteristic features in the calculated and profiles that arise from the imbalance between the reduction and oxidation rates31 and could be used to further discriminate between models [1] and [2] (Fig. 3.10, Fig. 3.9). The redox transition using [1] starts with an escalating amplitude that follows an n = 1 profile until most of the analyte is reduced and cannot return to the oxidized state due to the kinetic limitations. The ability of mediators to support the reverse transition is limited by mass transfer and/or the bimolecular reaction and, therefore, is not accelerated at increasing Ea. In contrast, model [2] always provides a pathway to return the sample to the oxidized state. In the absence of FCN this oxidation process is accomplished via isomerization of the protein into the low potential form, which can be effectively oxidized by MG and TA. The direct and isomerization pathways are mixed using [2], yielding a broad NPSV profile with an apparent n > 1. The relative contributions of the two pathways depend on Ea so that the overall NPSV 70 D½TauETauDOxTauDRdTauDRdTauDOx amplitude continuously increases with increasing Ea in contrast to a distinctive pseudo- equilibrium region using [1]. The low overall amplitudes of the and profiles in the absence of FCN provide the final argument in support of model [2]. Model [1] predicts a contradiction between > 80 mV, required to reproduce the amplitudes without FCN (Fig. 3.9), and the ≈ 60 mV, required to reproduce an artificial hysteresis with FCN (Figs. 4 and S6). Considering all the experimental conditions examined here, we conclude that the observed redox hysteresis arises mostly from an intrinsic redox-linked isomerization of TauD per model [2], although kinetic limitations may contribute to the observed NPSV response. Some redox-linked reorganization of the active site is expected due to electrostatic and electronic effects on the coordinated ligands. Distinct structural conformations of metal ligands in reduced and oxidized states lead to changes in E½.38–43 The magnitude of such changes in TauD greatly exceeds those of other currently known examples with values of < 100 mV.44 The hysteresis of ≈300 mV (Table 3.1) is equivalent to a reversible reorganization with ΔG > 7 kcal/mol ( ) or pKa > 5, if coupled to a single protonation event ( ). This observation raises an intriguing question of the extent to which ligand conformation contributes to the unusually low pKa of the ferric (hydr)oxo F3 species of TauD, detected by transient Raman spectroscopy.3,8 An uncompensated change in the E½ of 7 kcal/mole would represent a substantial stabilization of TauD following the redox transition. Considering the magnitude this electrochemical relaxation and the likelihood of redox-linked protonations, it is possible that the observed decrease in is partially compensated by an increase in , resulting in a small in the context of the Bordwell relationship (Fig. 1.5). This possibility can be examined experimentally from the pH-dependence of the E½ and associated vibrational changes, which is currently under investigation. The E½ of the ferrous isomer of TauD reported here is over 0.5 V 71 TauDRdTauDOxD½TauETauDRdD½TauE1nHennoeGoHGototG higher than the Fe2+/3+ transition in CYP450.45 If this isomer conformation is transiently retained during the catalytic cycle, the F3/F4 transition in TauD may also have a much higher E½ than that of Cmp I/II of CYP450 while the pKa of F3 is lower than that of Cmp II. In this case, the ferrous isomer conformation would promote deprotonation of F3 that favors the alkoxide pathway in TauD over the hydroxyl radical rebinding pathway, as found in CYP450. The structural rearrangements may lead to a reduction in the oxidation potential of F4 as a protection mechanism against long-range oxidation if the reaction occurs in the absence of taurine. The magnitude of the redox hysteresis in TauD suggests that the redox-linked reorganization involves at least one of the metal ligands: His99, Asp101, and His255 (Fig. 3.11a). The redox-difference vibrational changes of the amide mode (Fig. 3.1) suggest that structural changes propagate beyond the first coordination shell. The redox-linked changes in protonatable residues, particularly deprotonation of histidine(s) upon oxidation, can maintain charge neutrality in the active site (a redox Bohr effect) and alter the hydrogen bonding network.46 Initial de-protonations in TauD are likely to occur much faster than can be probed by NPSV due to accessibility of the active site for water molecules.47 A 2.6 Å distance between N1 of His99 and the peptide carbonyl oxygen of Asn97 suggest fairly strong interactions in Fe(II)- TauD. 72 Figure 3.11. TauD active site structure and possible structural rearrangement. a) Selected residues at the TauD active site. The carbon atoms of 2OG and taurine are shown in orange. The peptide segment proposed to be linked to structural rearrangement is highlighted by use of stick mode. Selected hydrogen bonding interactions (yellow) and water molecules (red) are shown.47 b) Proposed reversible redox-linked structural rearrangement. Deprotonation of His99 upon oxidation would disrupt this interaction, allowing the Asn97- Asp101 peptide backbone to undergo further changes. This perturbation could disrupt weak interactions between Asp101 and Trp248 (3.6 Å) in Fe(II)-TauD, allowing bidentate carboxylate binding of Asp101 upon oxidation (Fig. 3.11b), which has been observed in other non-heme iron enzymes.48 Such a change involving Asp101 is likely to cause a larger backbone reorganization than the initial deprotonation of histidine residue(s), but it may not be the primary trigger since Asp101 is already deprotonated in Fe(II)-TauD. The interaction of His255 with water 521 also may be involved. Changes in both the carboxyl and imidazole moieties are consistent with redox-difference NPSV spectra of TauD. Vibrational effects of cosubstrate binding suggest that 73 neither metal-bound water molecule(s) nor 2OG directly control isomerization, although both are affected by it. A detailed analysis of the structural origin for the redox-linked switching and its role in catalysis is currently under investigation. CONCLUSION Our results reveal extensive, fully reversible redox-linked conformational changes in three forms of TauD. The hysteresis between the oxidation and reduction Nernstian NPSV profiles arises primarily from isomerization between two separate ferric/ferrous redox couples of the protein. Quantitative kinetic modeling shows that a redox-linked conformational switching process substantially improves the fitness of simulations compared to the experimental data across various conditions over a simple model with a single redox potential. Changes in the redox potential of up to 0.3 V are attributed to the reversible reorganization of the metal center primary ligands, leading to further reorganization of the protein backbone. 74 REFERENCES 75 REFERENCES (1) Herr, C. Q.; Hausinger, R. P. Amazing Diversity in Biochemical Roles of Fe(II)/2- Oxoglutarate Oxygenases. Trends Biochem. Sci. 2018, 43, 517–532. https://doi.org/10.1016/j.tibs.2018.04.002. (2) Lewis, J. C.; Coelho, P. S.; Arnold, F. H. Enzymatic Functionalization of Carbon- Hydrogen Bonds. Chem. Soc. Rev. 2011, 40, 2003–2021. https://doi.org/10.1039/c0cs00067a. (3) Proshlyakov, D. A.; Hausinger, R. P. Transient Iron Species in the Catalytic Mechanism of the Archetypal α-Ketoglutarate-Dependent Dioxygenase, TauD. In Iron-Containing Enzymes: Versatile Catalysts of Hydroxylation Reactions in Nature; Visser, S. P. De, Kumar, D., Eds.; Royal Society of Chemistry, 2012; pp 67–87. (4) Proshlyakov, D. A.; McCracken, J.; Hausinger, R. P. Spectroscopic Analyses of 2- Oxoglutarate-Dependent Oxygenases: TauD as a Case Study. J. Biol. Inorg. Chem. 2017, 22, 367–379. https://doi.org/10.1007/s00775-016-1406-3. (5) Eichhorn, E.; Van Der Ploeg, J. R.; Kertesz, M. A.; Leisinger, T. Characterization of α- Ketoglutarate-Dependent Taurine Dioxygenase from Escherichia coli. J. Biol. Chem. 1997, 272, 23031–23036. https://doi.org/10.1074/jbc.272.37.23031. (6) Price, J. C.; Barr, E. W.; Glass, T. E.; Krebs, C.; Bollinger, J. M. Evidence for Hydrogen Abstraction from C1 of Taurine by the High-Spin Fe(IV) Intermediate Detected during Oxygen Activation by Taurine:α-Ketoglutarate Dioxygenase (TauD). J. Am. Chem. Soc. 2003, 125, 13008–13009. https://doi.org/10.1021/ja037400h. (7) Martinez, S.; Hausinger, R. P. Catalytic Mechanisms of Fe(II)- and 2-Oxoglutarate- Dependent Oxygenases. J. Biol. Chem. 2015, 290, 20702–20711. https://doi.org/10.1074/jbc.R115.648691. (8) Grzyska, P. K.; Appelman, E. H.; Hausinger, R. P.; Proshlyakov, D. A. Insight into the Mechanism of an Iron Dioxygenase by Resolution of Steps Following the FeIV=O Species. Proc. Natl. Acad. Sci. U. S. A. 2010, 107, 3982–3987. https://doi.org/10.1073/pnas.0911565107. (9) Poulos, T. L. Heme Enzyme Structure and Function. Chem. Rev. 2014, 114, 3919–3962. https://doi.org/10.1021/cr400415k. (10) Mitchell, A. J.; Dunham, N. P.; Martinie, R. J.; Bergman, J. A.; Pollock, C. J.; Hu, K.; Allen, B. D.; Chang, W.; Silakov, A.; Bollinger, J. M.; et al. Visualizing the Reaction Cycle in an Iron(II)- and 2-(Oxo)-Glutarate-Dependent Hydroxylase. J. Am. Chem. Soc. 2017, 139, 13830–13836. https://doi.org/10.1021/jacs.7b07374. 76 (11) Gupta, R.; Borovik, A. S. Monomeric MnIII/II and FeIII/II Complexes with Terminal Hydroxo and Oxo Ligands : Probing Reactivity via O - H Bond Dissociation Energies. J. Am. Chem. Soc. 2003, 125, 13234–13242. (12) Li, L.; Li, C.; Zhang, Z.; Alexov, E. On the Dielectric “Constant” of Proteins: Smooth Dielectric Function for Macromolecular Modeling and Its Implementation in DelPhi. J. Chem. Theory Comput. 2013, 9 (4), 2126–2136. https://doi.org/10.1021/ct400065j. (13) Bordwell, F. G.; Ji, G. Z.; Satish, A. V.; Zhang, X.; Cheng, J. P. Bond Dissociation Energies in DMSO Related to the Gas Phase. J. Am. Chem. Soc. 1991, 113, 9790–9795. https://doi.org/10.1021/ja00026a012. (14) Mayer, J. M. Hydrogen Atom Abstraction by Metal-Oxo Complexes: Understanding the Analogy with Organic Radical Reactions. Acc. Chem. Res. 1998, 31, 441–450. https://doi.org/10.1021/ar970171h. (15) Yosca, T. H.; Rittle, J.; Krest, C. M.; Onderko, E. L.; Silakov, A.; Calixto, J. C.; Behan, R. K.; Green, M. T. Iron(IV)Hydroxide pKa and the Role of Thiolate Ligation in C-H Bond Activation by Cytochrome P450. Science 2013, 342, 825–829. https://doi.org/10.1126/science.1244373. (16) Green, M. T.; Park, U. V; Pennsyl, V. Application of Badger’s Rule to Heme and Non- Heme Iron - Oxygen Bonds : An Examination of Ferryl Protonation States Depends on the One-Electron Reduction Potential of Compound I. J. Am. Chem. Soc. 2006, 128, 1902– 1906. (17) Sitter, A. J.; Reczek, C. M.; Terners, J. Heme-Linked Ionization of Horseradish Peroxidase Compound II Monitored by the Resonance Raman Fe(IV)=O Stretching Vibration. J. Biol. Chem. 1985, 260, 7515–7522. (18) Brischwein, M.; Scharf, B.; Engelhard, M.; Mäntele, W. Analysis of the Redox Reaction of an Archaebacterial Copper Protein, Halocyanin, by Electrochemistry and FTIR Difference Spectroscopy. Biochemistry 1993, 32, 13710–13717. https://doi.org/10.1021/bi00212a041. (19) Hirst, J. Elucidating the Mechanisms of Coupled Electron Transfer and Catalytic Reactions by Protein Film Voltammetry. Biochim. Biophys. Acta - Bioenerg. 2006, 1757, 225–239. https://doi.org/10.1016/j.bbabio.2006.04.002. (20) Grzyska, P. K.; Hausinger, R. P.; Proshlyakov, D. A. Metal and Substrate Binding to an Fe(II) Dioxygenase Resolved by UV Spectroscopy with Global Regression Analysis. Anal. Biochem. 2010, 399, 64–71. https://doi.org/10.1016/j.ab.2009.11.022. (21) Moss, D.; Nabedryk, E.; Breton, J.; Mäntele, W. Redox‐linked Conformational Changes in Proteins Detected by a Combination of Infrared Spectroscopy and Protein Electrochemistry: Evaluation of the Technique with Cytochrome C. Eur. J. Biochem. 77 1990, 187, 565–572. https://doi.org/10.1111/j.1432-1033.1990.tb15338.x. (22) Moss, D. A.; Leonhard, M.; Bauscher, M.; Mäntele, W. Electrochemical Redox Titration of Cofactors in the Reaction Center from Rhodobacter sphaeroides. FEBS Lett. 1991, 283, 33–36. (23) Fritz, F.; Moss, D. A.; Mäntele, W. Electrochemical Titration of the Cytochrome Hemes in the Rhodopseudomonas viridis Reaction Center Cyclic Equilibrium Titrations Yield Midpoint Potentials without Evidence for Heme Cooperativity. FEBS Lett. 1992, 297, 167–170. https://doi.org/10.1016/0014-5793(92)80352-H. (24) Clair, C. S. S.; Gray, H. B.; Valentine, J. S. Spectroelectrochemistry of Copper-Zinc Superoxide Dismutase. Inorg. Chem. 1992, 31, 925–927. https://doi.org/10.1021/ic00031a041. (25) DeLacey, A. L.; Stadler, C.; Fernandez, V. M.; Hatchikian, E. C.; Fan, H. J.; Li, S.; Hall, M. B. IR Spectroelectrochemical Study of the Binding of Carbon Monoxide to the Active Site of Desulfovibrio fructosovorans Ni-Fe Hydrogenase. J. Biol. Inorg. Chem. 2002, 7, 318–326. https://doi.org/10.1007/s00775-001-0301-7. (26) DeLacey, A. L.; Fernandez, V. M.; Rousset, M.; Cavazza, C.; Hatchikian, E. C. Spectroscopic and Kinetic Characterization of Active Site Mutants of Desulfovibrio fructosovorans Ni-Fe Hydrogenase. J. Biol. Inorg. Chem. 2003, 8, 129–134. https://doi.org/10.1007/s00775-002-0397-4. (27) Mantele, W. Reaction-Induced Infrared Difference Spectroscopy for the Study of Protein Function and Reaction Mechanisms. Trends Biochem. Sci. 1993, 18, 197–202. (28) Vincent, K. A. Triggered Infrared Spectroscopy for Investigating Metalloprotein Chemistry. Philos. Trans. A. Math. Phys. Eng. Sci. 2010, 368, 3713–3731. https://doi.org/10.1098/rsta.2010.0055. (29) Melin, F.; Hellwig, P. Recent Advances in the Electrochemistry and Spectroelectrochemistry of Membrane Proteins. Biol. Chem. 2013, 394, 593–609. https://doi.org/10.1515/hsz-2012-0344. (30) Ash, P. A.; Vincent, K. A. Spectroscopic Analysis of Immobilised Redox Enzymes under Direct Electrochemical Control. Chem. Commun. 2012, 48, 1400–1409. https://doi.org/10.1039/C1CC15871F. (31) John, C. W.; Proshlyakov, D. A. FTIR Spectro-Voltammetry and Quantitative Modeling of Analytes in Kinetically Constrained Redox Mixtures. Anal. Chem. 2019, xx, xx–xx. https://doi.org/10.1021/acs.analchem.9b00859 (32) Ryle, M. J.; Padmakumar, R.; Hausinger, R. P. Stopped-Flow Kinetic Analysis of Escherichia coli Taurine/α-Ketoglutarate Dioxygenase: Interactions with α-Ketoglutarate, 78 Taurine, and Oxygen. Biochemistry 1999, 38, 15278–15286. https://doi.org/bi9912746 [pii]. (33) Yao, C.; Capdevielle, F. J.; Kadish, K. M.; Bear, J. L. Thin-Layer Microcell for Transmittance Fourier Transform Infrared Spectroelectrochemistry. Anal. Chem. 1989, 61, 2805–2809. (34) Arciero, D. M.; Hooper, A. B.; Collins, M. J. Preliminary Note Low Volume Spectroelectrochemistry : An Anaerobic Optically Transparent Thin Layer Electrode Cell and a Potentiostat 1 UV-Vis Spectrophotometer Interface for Computer-Controlled Data Collection. J. Electroanal. Chem. 1994, 371, 277–281. (35) https://github.com/dap-biospec/G3F. (36) https://github.com/dap-biospec/KinESim. (37) Susi, H. Infrared Spectra and Protein in Aqueous Solutions. J. Biol. Chem. 1967, 242, 5460–5467. (38) Costentin, C. Electrochemical Approach to the Mechanistic Study of Proton-Coupled Electron Transfer. Chem. Rev. 2008, 108, 2145–2179. https://doi.org/10.1021/cr068065t. (39) Murgida, D. H.; Hildebrandt, P. Redox and Redox-Coupled Processes of Heme Proteins and Enzymes at Electrochemical Interfaces. Phys. Chem. Chem. Phys. 2005, 7, 3773– 3784. https://doi.org/10.1039/b507989f. (40) Armstrong, F. A.; Camba, R.; Heering, H. A.; Hirst, J.; Jeuken, L. J.; Jones, A. K.; Léger, C.; McEvoy, J. P. Fast Voltammetric Studies of the Kinetics and Energetics of Coupled Electron-Transfer Reactions in Proteins. Faraday Discuss. 2000, 116, 191–203; discussion 257-268. https://doi.org/10.1039/b002290j. (41) Léger, C.; Bertrand, P. Direct Electrochemistry of Redox Enzymes as a Tool for Mechanistic Studies Direct Electrochemistry of Redox Enzymes as a Tool for Mechanistic Studies. Chem. Rev. 2008, 108, 2379–2438. https://doi.org/10.1021/cr0680742. (42) Jeuken, L. J. C.; Armstrong, F. A. Electrochemical Origin of Hysteresis in the Electron- Transfer Reactions of Adsorbed Proteins: Contrasting Behavior of the “Blue” Copper Protein, Azurin, Adsorbed on Pyrolytic Graphite and Modified Gold Electrodes. J. Phys. Chem. B 2001, 105, 5271–5282. https://doi.org/10.1021/jp004283t. (43) Pulcu, G. S.; Elmore, B. L.; Arciero, D. M.; Hooper, A. B.; Elliott, S. J. Direct Electrochemistry of Tetraheme Cytochrome C554 from Nitrosomonas europaea: Redox Cooperativity and Gating. J. Am. Chem. Soc. 2007, 129, 1838–1839. https://doi.org/10.1021/ja065657k. (44) Hosseinzadeh, P.; Lu, Y. Design and Fine-Tuning Redox Potentials of Metalloproteins 79 Involved in Electron Transfer in Bioenergetics. Biochim. Biophys. Acta - Bioenerg. 2016, 1857, 557–581. https://doi.org/10.1016/j.bbabio.2015.08.006. (45) Mittra, K.; Green, M. T. Reduction Potentials of P450 Compounds I and II: Insight into the Thermodynamics of C-H Bond Activation. J. Am. Chem. Soc. 2019, 141 (13), 5504– 5510. https://doi.org/10.1021/jacs.9b00242. (46) Papa, S. Proton Translocation Reactions in the Respiratory Chains. Biochim. Biophys. Acta 1976, 456, 39–84. https://doi.org/10.1016/0304-4173(76)90008-2. (47) Elkins, J. M.; Ryle, M. J.; Clifton, I. J.; Hotopp, J. C. D.; Lloyd, J. S.; Burzlaff, N. I.; Baldwin, J. E.; Hausinger, R. P.; Roach, P. L. X-Ray Crystal Structure of Escherichia coli Taurine/α-Ketoglutarate Dioxygenase Complexed to Ferrous Iron and Substrates. Biochemistry 2002, 41, 5185–5192. (48) Karlsson, A.; Parales, J. V; Parales, R. E.; Gibson, D. T.; Eklund, H.; Ramaswamy, S. Crystal Structure of Naphthalene Dioxygenase : Side-on Binding of Dioxygen to Iron. Science (80-. ). 2003, 299, 1039–1043. 80 CHAPTER 4: STRUCTURAL ORIGIN OF THE LARGE REDOX-LINKED REORGANIZATION IN THE 2-OXOGLUTARATE DEPENDENT OXYGENASE, TAUD 81 INTRODUCTION Taurine:2-oxoglutarate (2OG) dioxygenase (also termed taurine hydroxylase or TauD) is a non-heme mononuclear iron enzyme and the archetypal member of the 2OG-dependent oxygenase family.1 These enzymes typically share a 2-His-1-carboxylate iron binding motif and utilize 2OG to activate O2, generating an Fe(IV)=O (F4) intermediate that performs substrate C- H bond activation (Fig. 4.1).2 This reactivity allows the 2OG-dependent oxygenases to catalyze a wide range of biochemical reactions.3 TauD is found in Escherichia coli where it catalyzes the production of sulfite from taurine in sulfur-starved cells.4 Historically, TauD was proposed to utilize a hydroxyl radical rebound mechanism similar to that observed in cytochromes P450;1,5 however, recent studies have suggested that TauD and its fellow family members instead use an alkoxide-forming (AF) mechanism (Fig. 4.1).6,7 Figure 4.1. Catalytic cycle of TauD comparing the hydroxyl radical rebound (grey) and alkoxide-forming (blue) mechanisms. Vertical transitions indicate a change in the oxidation state of the Fe center. Diagonal transitions indicate a change in a protonation state. Red structures show the artificial manipulation of the enzyme used in this study to generate an in-situ model of the F3 intermediate. 82 In the AF mechanism, the F4 intermediate activates the C-H bond at C1 of taurine by hydrogen atom transfer (HAT)8 yielding a substrate radical and the Fe(III)-OH species, which rapidly transfers its proton to a nearby base and forms a vibrationally detectable Fe(III)-O– intermediate (F3). The deprotonated oxygen ligand forms the Fe-O-C bond of the alkoxo intermediate (FX).6,7 Subsequent cleavage of the Fe-O bond leads to the formation of sulfite and aminoacetaldehyde products.1 This mechanism, initially proposed based on the transient Raman data for TauD6 and more recently supported by X-ray crystallography of the arginine hydroxylase VioC,7 leaves two important questions to be answered: i) the feasibility of the Fe(III)-O– intermediate under physiological conditions and ii) the identity of the base that is responsible for accepting the proton from the Fe(III)-OH group formed immediately after HAT to the ferryl state. The ability of the 2OG-dependent oxygenases to activate C-H bonds arises from the combined energy of the reduction potential (ERd) of the F4 state and the pKa of the resulting F3 state.1 The Raman evidence suggests the transient F3 intermediate is deprotonated, a particularly surprising finding that suggests that the pKa of this intermediate is low. Computational studies and synthetic models have suggested that this pKa should be greater than 20,9 making it difficult to rationalize the formation of an alkoxy species in the AF mechanism. The free energy for HAT that leads to the formation of an alkoxy species via the Fe(III)-O– intermediate could be gained from either a highly positive ERd of the F4 state or a dynamic modulation of pKa of the proton donor/acceptor pair associated with the subsequent proton transfer. The lack of experimental data in support of either one of these possibilities can be addressed by characterization of the redox, vibrational, and pH-dependent properties of the redox transitions between the in situ Fe(III) and Fe(IV) models of the transient F3 and F4 intermediates of TauD (Fig. 4.1). 83 To better understand the F3 state, we have focused on characterizing the redox and pH- dependent properties of the off-pathway Fe(III)-OH states for Fe-TauD, 2OG-Fe-TauD, and taurine-2OG-Fe-TauD. Unexpectedly, our initial investigation of the Fe(II)/(III) redox couples for these species revealed reversible, redox-linked conformational changes that modulate their redox potentials by at least 300 mV.10 Based on the observed vibrational changes and the magnitudes of the redox hysteresis, we proposed an isomerization that involves more than one primary metal ligand and the reorganization of the protein backbone. In this study, we use site- directed mutagenesis to test this hypothesis and to characterize the contributions of individual ligands to the overall redox response of the active site. We demonstrate that isomerization is a reversible, multi-step process and show that the total magnitude of the electrochemical hysteresis is substantially larger than suggested by the initial spectro-electrochemical study. In addition, we demonstrate that redox hysteresis is not unique to TauD, but also is observed in the 2OG- dependent ethylene-forming enzyme (EFE)11 and may be a more general feature of this enzyme family. EXPERIMENTAL PROCEDURES PROTEIN PURIFICATION AND SPECTROELECTROCHEMICAL MEASUREMENTS Mutagenesis of the gene encoding TauD was performed as previously described.12,13 EFE apoprotein was purified by established procedures.11 Isolation of the TauD variant apoproteins, preparation of Fe(II) and Fe(III) holoenzymes of EFE and the versions of TauD, and spectroelectrochemical measurements were performed as previously described for WT TauD in Chapter 3. 84 CHEMICAL REDOX TITRATIONS Chemical redox titrations were performed under anaerobic conditions at room temperature. UV-Vis spectra were acquired using a Hewlett Packard 8453 spectrophotometer. A long-pass filter with a cutoff of 400 nm was used to suppress photochemical reduction of mediators. All samples were prepared in 25 mM Tris, pH 8, buffer. The reactions were initiated by addition of a 70% equivalent of anaerobic TauD or EFE to an anaerobic solution of reducing or oxidizing mediator in a stirred, anaerobic optical cuvette with 1 cm pathlength. Spectra were acquired for 1 h with progressively increasing time points of the sampling. RESULTS Shifts in the E½ of TauD associated with redox-linked conformational changes can be probed by observable time-dependent changes during the equilibration of the enzyme with other redox active analytes, provided that the initial bimolecular reaction rate is faster than the ensuing protein reorganization. Whereas the visible optical absorption of TauD is too weak to detect these changes,14 the progress of the redox reactions can be followed by monitoring the redox- dependent optical changes in the spectra of mediators. Typical time-dependent absorption changes upon reduction of ferricyanide (FCN) by 2OG-Fe(II)-TauD (Fig. 4.2) show a multiphasic response. The initial fast kinetic phase (Table 4.1) arises from the redox equilibration controlled by the slower of two steps - the mediator binding and the electron transfer in the protein-mediator complex. Following that phase are multiple slower kinetic phases, which cannot be limited by the bimolecular rate and, therefore, are attributed to the changes in the properties of the protein. This behavior contrasts with a simple, monophasic oxidation of ferrous Mb by FCN,10 or the reduction of ferric myoglobin (Mb) by thionine acetate 85 (TA, Fig. 4.3), as expected for a sample with reversible electrochemical behavior. Reactions with Mb quickly reach a redox equilibrium that corresponds to a E½ of -155 mV, in good agreement with the reported value of -153 mV15 and show little change after the first 5 s. The lack of change in the extent of oxidation of TA by Mb confirms that the sample remains anaerobic over a 1 h measurement and that gradual changes observed in TauD are not due to a slow ingress of atmospheric O2. Figure 4.2. Transient absorption changes upon reaction of 2OG-Fe-TauD with redox mediators. Left: Reduction of 100 µM ferricyanide was initiated by the addition of 70 µM 2OG- Fe(II)-TauD, for both the wild type enzyme and selected variants. The ferricyanide absorption at 420 nm was used to monitor the progress of the reaction. Right: Oxidation of 20 µM thionine acetate was initiated by the addition of 28 µM 2OG-Fe(III)-TauD for the same set of proteins. The oxidized thionine acetate absorption at 598 nm was used to monitor the progress of the reaction. Similar multiphasic changes were observed upon the reaction of 2OG-Fe-TauD in the ferrous and ferric states with several oxidized and reduced mediators (Fig. 4.2 right and Fig. 4.4), including methylene blue (MB), methylene green (MG), and N,N,Nʹ,Nʹ-tetramethyl-p- phenylenediamine (TMPD), in addition to FCN and TA. It can be seen in Fig. 4.4 that the extents of reduction of 2OG-Fe(III)-TauD or oxidation of 2OG-Fe(II)-TauD are inversely related for any single mediator. The extent of the reaction correlates with changes in the redox potentials between mediators, especially for FCN, TA, and MB, which exhibit single accessible redox transitions. FCN (E½ = 215 mV, n = 1) showed a fast initial oxidation of TauD to approximately 86 50%, followed by a slow oxidation to near completion. We observed no evidence that 2OG- Fe(III)-TauD can be reduced by ferrocyanide to any detectable extent. MB (E½ = -236 mV, n = 2) showed the opposite behavior; it slowly reduced 2OG-Fe(III)-TauD (Fig. 4.4B) and showed little oxidation of 2OG-Fe(II)-TauD (Fig. 4.4A). TA (E½ = -197 mV, n = 2) was effective at partial reduction and oxidation of 2OG-Fe-TauD to levels intermediate between those of FCN and MB. Although the two-electron redox transitions of TA and MB theoretically could slow the kinetics by requiring simultaneous reduction of two TauD molecules, the kinetics of the observed changes were distinctly multiphasic after the reduction or oxidation commenced. Figure 4.3. The observed E½ of Mb and TauD upon reduction by TA. 28 µM Mb or 2OG- Fe(III)-TauD was injected into 20 µM reduced TA. An equilibrium E½ of -155 mV vs. Ag/AgCl is observed for Mb (horizontal line). During the same time period, the E½ of TauD continues to change. similar trends. The E½ = +25 mV of TMPD was too high to be an effective reductant of 2OG-Fe(III)-TauD, whereas TMPD2+ caused nearly complete oxidation of 2OG-Fe(II)-TauD. The initial, large-amplitude oxidation was likely due to the TMPD/TMPD●+ redox couple (E½ = +25 mV), while the subsequent small-amplitude oxidation is due to the TMPD●+/ TMPD2+ redox couple (E½ = +425 mV). Similarly, the E½ = -111 mV redox transition of MG made it a more 87 efficient oxidant of 2OG-Fe(II)-TauD than the structurally similar and more reducing MB and TA. Comparison of the reduction/oxidation efficiency of TA and MG also suggests that the n = 2 Figure 4.4. Normalized population kinetic traces of the amount of 2OG-Fe(II)-TauD oxidized or 2OG-Fe(III)-TauD reduced after addition into solutions containing various mediators. A) 2OG-Fe(II)TauD was titrated into solutions containing 1.4-fold excess electron equivalents of the listed mediators, all in their oxidized forms. B) 2OG-Fe(III)TauD was titrated into solutions containing 1.4-fold excess electron equivalents of the listed mediators, all in their reduced forms. Time zero indicates the point at which TauD was added to the sample. The reported E½ values of the mediators (vs. Ag/AgCl) are shown in parentheses. MG and TMPD, both of which have two accessible n = 1 redox transitions, showed redox couple of TA was not a determining factor in its reactions with TauD. The smaller extent of reduction of 2OG-Fe(III)-TauD by MG than by TA, however, suggests that the E½ = -243 mV redox couple of MG is ineffective in reducing TauD. Finally, the exhaustive and monotonic reduction of 2OG-Fe(III)-TauD by methyl viologen (MV) (Fig. 4.4B) confirms that the entire population of TauD is redox active. This demonstration, along with the facile oxidation of TauD by FCN and TMPD allows one to attribute the transient kinetics observed with the structurally similar TA, MG, and MB to thermodynamic changes, rather than kinetic limitations. Table 4.4. Apparent initial rate constants of the oxidation (kOx) or reduction (kRd) of 2OG- Fe-TauD by various mediators. Mediator kOx (µM-1s-1) kRd (µM-1s-1) FCN TMPD MG TA MB MV 0.69 0.43 0.11 0.28 0 ‒ 88 ‒ 0.11 0.35 0.013 0.14 0.91 Since the bimolecular reactions of TauD with mediators (Table 1) are much faster than the ensuing slow changes in TauD and because the E½ values of the mediators are known (Fig. 4.4), the time-dependent E½ of TauD can be calculated using the Nernst equation. This analysis can be accomplished using a single chromogenic component of the mixture as long as initial concentrations of the mediator, , and TauD, , are known. In the case of the oxidation of 2OG-Fe(II)-TauD (n = 1) by FCN (n = 1), for which only the Fe(II) and Fe(III) redox states are accessible: (4.1) where E½,FCN is the standard redox potential of FCN, and E½,TauD(t) is the time-dependent redox potential of TauD. In this example, both and can be determined from the optical absorption at 420 nm. Such analysis for the reduction of FCN by the WT protein and several variants of 2OG-Fe(II)-TauD is shown in Fig. 4.5 (top), along with the analogous oxidation of TA (n = 2) by 2OG-Fe(III)-TauD (bottom). 89 0M0TauD½00FCN,20½,FCNTauDFCNFCNln1FCNFCNOxtttTauDRTEtEF0FCNFCNt Figure 4.5. Transient changes in the redox potential of 2OG-Fe-TauD and three of its variants. EOx and ERd values (markers) were calculated using Equation 4.1 from the absorbance data. Top row: oxidation of 2OG-Fe(II)-TauD species by FCN. Bottom row: reduction of 2OG- Fe(III)-TauD species by TA. The uncertainty derived from error propagation is shown by the grey area (see SI). *Absorbance data for the oxidation of the H99A variant indicated that the entire population of H99A TauD was immediately oxidized, allowing for the estimation of only the upper limit of EOx assuming 99.8% oxidation of H99A TauD. Estimates of the time-dependent redox potentials (Fig. 4.5) must consider unavoidable errors that arise from i) pipetting precision and ii) the temporal stability of optical measurements. Equations 4.2 and 4.3 were derived using propagation of uncertainty and assuming a 1% error in the total concentration of TauD, d[TauD], error of the initial concentrations of the oxidized mediators, d[FCN]0 and d[TA]0, and variances reported by the spectrophotometer for each absorbance measurement, d[FCN]t and d[TAOx]t. In the case of WT TauD, changes in E½ were substantially larger than the estimated errors (shaded areas in Fig. 4.5). In contrast, the oxidation of H99A protein by FCN was nearly complete immediately after the start of the reaction, allowing only for an estimate of the upper limit of its E½ at <25 mV, and the estimated errors were substantially larger than any subsequent changes. The metal ligand variants all showed only minimal reduction by TA at the start the reaction, which indicated decrease in their E½ relative to 90 that of the WT protein and increased the uncertainty in their calculated values. As the reduction progressed, the uncertainties were reduced. (4.2) (4.3) Overall, substitutions of the metal ligands led to a decrease in the ERd and hindered reorganization upon oxidation. The H99A variant exhibited a decrease in the EOx by at least -175 mV relative to the WT protein at 10 s with only a minimal decrease in the corresponding ERd. In contrast, the initial reduction E½ in D101Q TauD decreased by 50 mV relative to the WT enzyme, reaching comparable values at the end of the reduction. While the initial exponential phase of the oxidation of D101Q was very similar to that of the WT protein, oxidation did not continue beyond 100 s, suggesting that reorganization of its carboxylate group is required for the second step of the oxidative structural response. Both the reduction and oxidation E½ were lower in H255Q TauD, with no noticeable oxidative reorganization. The reductive reorganization for this protein reached saturation after 1200 s. Noticeably, EOx and ERd of the D101Q and H255Q variants do not equilibrate at the same E½, indicating that two distinct conformations likely exist for the ferrous and ferric forms of these proteins 1 h following redox transition. The structural origins of the thermodynamic changes in TauD were further examined using normal pulsed spectrovoltammetry (NPSV) with FTIR detection (Fig. 4.6), which yields vibrational information about redox-coupled groups.16 Global spectral regression of the 91 22220222200½0Ox,TauD24200dFCNdTauDdFCNdFCNFCN2FCNFCNdFCNdFCNTauDFCNFCNdFCNTauDFCNFCNFCNFCNtttttttttEO½OR22222Oxx0000Ox4224Ox0OxOx0Ox00462OxxOx,0dT0u0aD2TauD-TA+TAdTauD+dTA+dTAdTATA3TA-TAdTA-dTATauD-TA+TAdTATauD-TA+TATA-TAttttttttttE reversible NPSV response allows one to deconvolute a 3-D experimental data set into a full occupancy frequency spectrum, ΔStot (Fig. 4.6, left), and its Nernstian population profile, (Fig. 4.6, right), as a function of applied potential Ea.16 Therefore, in addition to structural information, NPSV reports thermodynamic properties of the reversible redox transitions. To balance the time necessary for TauD samples to undergo substantial redox-coupled reorganization (Fig. 4.5) against the practical limit of the FTIR measurement (up to 6-8 h), 300 s NPSV time pulses were used in this study with spectral collection between 120 s and 300 s at the given Ea. This timeframe corresponds to approximately half of the overall reorganization amplitude observed optically. In agreement with the transient EOx and ERd values at 300 s (Fig. 4.5), all variants of TauD with bound 2OG show a hysteresis between the and profiles indicating that some structural rearrangement occurs in the altered proteins. The magnitudes of the redox hysteresis and redox-difference IR spectra were similar between WT TauD and the Y73I variant, which is expected since Y73 is not an Fe ligand. However, Y73 is involved in taurine binding at the active site and can be oxidized to a tyrosyl radical during the catalytic cycle if taurine is not present.13 The formation of a new vibrational mode at 1702 cm-1 in Y73I indicates that this residue is involved in the redox-linked response (possibly via a hydrogen bonding network involving the metal ligands), but not enough to significantly alter the redox-linked conformational change. The NPSV-detected redox hysteresis in all other variants examined here was smaller than that found in WT TauD. This behavior was in general agreement with the transient E½ measurements, except for the initial E½ of D101Q. Several observations from the redox- difference FTIR spectra corroborate the conclusion that structural rearrangement is hindered in all variants involving metal-binding residues. In particular, the variant proteins exhibited i) 92 OxRd substantial vibrational changes, ii) overall lower intensities, and iii) increased discrepancies between reduction and oxidation spectra. Particularly noticeable is the loss of the amide I stretching modes at 1630 and 1638 cm-1 in the H99A, D101Q, and H255Q variants. These features were previously proposed to arise from redox-linked peptide backbone reorganization. Figure 4.6. Redox-coupled vibrational changes in WT and variants of 2OG-Fe-TauD. Left: Redox difference IR-NPSV spectra. Right: Experimental (markers) and fitted (lines) (red) and (blue) NPSV profiles for the indicated TauD proteins were normalized using ΔA1682 (WT), ΔA1683 (Y73I), ΔA1669 (H99A), ΔA1659 (D101Q), and ΔA1692 (H255Q). 93 RdOx Several other vibrations were assigned to individual metal ligands. The complete loss of the 1558 cm-1 and 1659 cm-1 modes in the H99A variant and their retention in the D101Q variant allows us to attribute these modes to H99. Likewise, the 1695 cm-1 mode is attributable to D101 as it is retained in H99A (1690 cm-1) and H255Q (1692 cm-1) proteins, but not in D101Q TauD. Interestingly, the 1682 cm-1 mode found in WT TauD was not observed in the redox difference spectra of the variant proteins. This mode was previously assigned to the carboxylate vibration of 2OG based on the comparison of holo-, 2OG-, and 2OG-taurine complexes of WT TauD.10 The characteristic metal-to-ligand charge transfer band at 530 nm17 (data not shown) in 2OG complexes of the H99A and D101Q proteins confirms that 2OG was bound to the Fe center in these variants. This result suggests that while 2OG bound to the metal in the H99A and D101Q variants, it no longer affected the redox-linked reorganizations in the same way as in WT TauD. To assess whether the novel redox-linked reorganization behavior is unique to TauD among members of the 2OG-dependent oxygenase family, we examined EFE by NPSV and chemical titrations (Fig. 4.7). EFE shares the same His-X-Asp-Xn-His Fe binding motif as TauD and catalyzes similar hydroxylation chemistry in addition to its ethylene-generating activity.11 The NPSV spectrum of this protein exhibited the general pattern observed in TauD. Missing in the NPSV difference spectrum of EFE were the 2OG (1682 cm-1) and amide I (1638/1630 cm-1) vibrations. The lack of 2OG vibrations is consistent with the perturbed binding of 2OG by this protein compared to other family members; i.e., 2OG-Fe(II)-EFE favors monodentate binding over bidentate coordination by the substrate to the Fe.18 The lack of amide I vibrations suggested a lesser extent of redox-linked reorganization in EFE; consequently, the profile of 2OG-Fe- EFE indicated a shift towards lower potential with the development of redox heterogeneity. The of 2OG-Fe-EFE was similar to that of 2OG-Fe-TauD. The latter observation was 94 OxRd corroborated by the nearly identical oxidation profiles of the ferrous enzymes by FCN over the first 300 s. At later times, EFE exhibited faster oxidation that proceeded to a greater extent (Fig. 4.7C). Figure 4.7. Redox-coupled reorganization of 2OG-Fe-EFE. A) IR-NPSV redox difference (blue) NPSV profiles. spectrum. B) Experimental (markers) and fitted (lines) (red) and C) Transient changes in the EOx of EFE upon oxidation by FCN. DISCUSSION Redox chemistry often results in charge-induced reorganizations and redox-linked protonations that ultimately cause ERd and EOx values to differ; however, the magnitudes of these differences are usually small.19,20 Our results show that 2OG-Fe-TauD has an exceedingly large redox hysteresis between the stable ferric and ferrous forms of up to 468 mV in the WT protein and 497 mV in the D101Q variant (Fig. 4.5 and 4.6). We have shown that comparable redox hysteresis is observed in Fe-TauD and its complexes with substrates, but not in Zn-TauD, thus localizing the reorganization to the enzyme active site and attributing it primarily to metal- binding residues and the surrounding protein moiety.10 Our observation of similar 95 RdOx electrochemical properties by EFE (Fig. 4.7) suggests that the structural modulation of the active site and the resulting changes in the E½ may be common in this family of enzymes. Such structural flexibility may be one of the key properties that sets apart non-heme oxygen metabolizing enzymes from their heme counterparts. The rigid porphyrin structure severely limits the active involvement of the protein in tuning cofactor reactivity beyond controlling substrate approach and electronic effects of the proximal ligands.21,22 Figure 4.8. Selected residues at the TauD active site. The peptide segment proposed to be linked to structural rearrangement is highlighted in stick mode. Selected hydrogen bonding interactions (yellow) and water molecules (red) are shown. The carbon atoms of the substrates are shown in orange.23 Previously, we proposed that D101 and H99 are directly involved in the redox-linked conformational change of TauD based on the observed alterations in characteristic IR vibrations and because these residues would be most susceptible to electrostatic effects of the redox transitions.10 The present results provide strong evidence that all three metal-binding residues play a major role in the observed redox response. The involvement of amide vibrations in the redox-induced changes (Fig. 4.6) suggests that these initial rearrangements propagate outside of 96 the first ligand shell. TauD shares the same binding motif as many non-heme ferrous iron enzymes,23 with one Asp (D101) and two His (H99 and H255) ligands (Fig. 4.8). Considering the close proximity of D101 to H99 in an unstructured region of the protein’s secondary structure, it is highly likely that i) the H99-T100-D101 peptide segment undergoes redox- induced rearrangements and ii) the coupled changes of H99 and D101 are mediated through the peptide backbone (Fig. 4.9). Redox-linked (de)protonation is the most direct mechanism of charge compensation in redox transitions.24 Although direct crystallographic data on the protonation states of H99 and H255 are not available, their N atoms are within hydrogen bonding distances of the N97 keto oxygen and a crystallographic water 521, respectively.10 Thus, one can expect that H99 is protonated in ferrous TauD. The ability to form a hydrogen bond in this unstructured peptide region hints at a mechanism of dynamic linking between H99 and D101. We propose that oxidation of ferrous TauD leads to the deprotonation of H99 and the loss of its hydrogen bond, altering the conformation of the N97-D101 chain. Movement of this fragment (detected as the amide mode changes by FTIR) would then allow for bidentate ligation of D101 to the Fe ion.10 This coordination switch can explain the involvement of D101 in the redox hysteresis, whereas redox-linked protonation of a metal-bound carboxyl is highly unlikely. The peptide-triggered rearrangement of D101 following deprotonation of H99 is also consistent with the delayed effect of the D101Q substitution on the oxidation profile (Fig. 4.5, top) and further suggests that D101 reorganization in the second exponential phase is required for further redox tuning in the later phases. Substitutions of either H99 or D101 inhibit the redox-linked response to various extents (Figs. 4.3 and 4.4), which can be rationalized based on the available crystallographic data. 97 However, the disproportionally large effect of the H255Q variant is unexpected. H255 may play several roles in controlling the redox properties of the metal center. First, it may undergo redox- linked protonation directly. Second, sequential changes in the H99 and D101 residues may alter hydrogen bonding between H255 and water 521. Such interactions have been shown to play a significant role in modulating the redox potentials of enzymes,25,26 particularly those involving imidazole ligands for which hydrogen bonding may modulate the pKa of the ligand and translate into significant changes in the redox potential of the metal.27,28 Lastly, the H255Q variant may control positioning of the metal center relative to the H97-D101 peptide. Hydrogen bonding between H255 and water 521 may be maintained for both the imidazole and imidazolate states, unlike that between H99 and the keto group. The significant decrease in the oxidation potential of the H255Q variant (Fig. 4.5) suggests an excess of uncompensated negative charge in proximity to the metal while the observed reduction of hysteresis may indicate the preferential disruption of the H99 hydrogen bond during bidentate coordination of D101. Figure 4.9. Proposed redox-linked conformational changes in TauD. See text for details. Our results show that while all three metal-binding residues are directly involved in the redox-linked structural reorganization in TauD, no single variant resulted in the complete loss of 98 such reorganization as evidenced by the presence of continuous variability in the observed ERd (Fig. 4.5), a hysteresis in the NPSV profiles, and residual changes in amide vibrations in the redox difference spectra (Fig. 4.6). While the results of NPSV and stoichiometric redox reactions are consistent, it is important to note that the two methods may not probe the same reorganization events due to their differences in time scales. Stoichiometric reactions always start from the most stable configuration for a given redox state and probe the monotonic progression of the redox state for up to 1 h. NPSV probes the reversible redox transitions that occur within a pulse duration (5 min). On the one hand, NPSV may probe only a subset of states that are accessible for stoichiometric reactions. This possibility is supported by the pronounced differences between the reduction and oxidation NPSV spectra of H99A, D101Q, and H255Q variants (Fig. 4.6) in agreement with the optical data that suggest limited reorganization in these variants. On the other hand, redox cycling during NPSV may facilitate transitions over conformational energetic minima. The latter possibility is supported by the similarity of the reduction and oxidation NPSV spectra in the WT protein and the Y73I variant (Fig. 4.6).10 Whether similar phenomena take place during the native redox transitions between ferric, ferrous, and, especially, ferryl states in the catalytic cycle of TauD and related enzymes (Fig. 4.1) is an intriguing question that requires further investigation by experimental and computational methods. CONCLUSION The non-heme active site of TauD provides a flexible environment that allows the enzyme to modulate its redox potential reversibly by up to 0.5 V. All three amino acid ligands of the iron center (H99, D101, and H255) are intricately involved in the redox-linked structural 99 rearrangement that also affects the protein backbone, presumably the H99-D101 segment. Similar redox-linked reorganization occurs in another enzyme of the 2OG-dependent oxygenase family, EFE, which shares the Asp-X-His binding motif with TauD. This observation demonstrates that such thermodynamic flexibility is not unique to TauD. Our results strongly argue that the protein moiety may play an active role in redox catalysis by enzymes beyond positioning of the substrates and should not be viewed as a rigid spectator in a dynamic catalytic process. 100 REFERENCES 101 REFERENCES (1) Proshlyakov, D. A.; Hausinger, R. P. Transient Iron Species in the Catalytic Mechanism of the Archetypal α-Ketoglutarate-Dependent Dioxygenase, TauD. In Iron-Containing Enzymes: Versatile Catalysts of Hydroxylation Reactions in Nature; Visser, S. P. De, Kumar, D., Eds.; Royal Society of Chemistry, 2012; pp 67–87. (2) Martinez, S.; Hausinger, R. P. Catalytic Mechanisms of Fe(II)- and 2-Oxoglutarate- Dependent Oxygenases. J. Biol. Chem. 2015, 290, 20702–20711. https://doi.org/10.1074/jbc.R115.648691. (3) Herr, C. Q.; Hausinger, R. P. Amazing Diversity in Biochemical Roles of Fe(II)/2- Oxoglutarate Oxygenases. Trends Biochem. Sci. 2018, 43, 517–532. https://doi.org/10.1016/j.tibs.2018.04.002. (4) Eichhorn, E.; Van Der Ploeg, J. R.; Kertesz, M. A.; Leisinger, T. Characterization of α- Ketoglutarate-Dependent Taurine Dioxygenase from Escherichia coli. J. Biol. Chem. 1997, 272, 23031–23036. https://doi.org/10.1074/jbc.272.37.23031. (5) Meunier, B.; de Visser, S. P.; Shaik, S. Mechanism of Oxidation Reactions Catalyzed by Cytochrome P450 Enzymes. Chem. Rev. 2004, 104, 3947–3980. https://doi.org/10.1021/cr020443g. (6) Grzyska, P. K.; Appelman, E. H.; Hausinger, R. P.; Proshlyakov, D. A. Insight into the Mechanism of an Iron Dioxygenase by Resolution of Steps Following the FeIV=O Species. Proc. Natl. Acad. Sci. U. S. A. 2010, 107, 3982–3987. https://doi.org/10.1073/pnas.0911565107. (7) Mitchell, A. J.; Dunham, N. P.; Martinie, R. J.; Bergman, J. A.; Pollock, C. J.; Hu, K.; Allen, B. D.; Chang, W.; Silakov, A.; Bollinger, J. M.; et al. Visualizing the Reaction Cycle in an Iron(II)- and 2-(Oxo)-Glutarate-Dependent Hydroxylase. J. Am. Chem. Soc. 2017, 139, 13830–13836. https://doi.org/10.1021/jacs.7b07374. (8) Lansky, D. E.; Goldberg, D. P. Hydrogen Atom Abstraction by a High-Valent Manganese(V)-Oxo Corrolazine. Inorg. Chem. 2006, 45, 5119–5125. https://doi.org/10.1021/ic060491+. (9) Gupta, R.; Borovik, A. S. Monomeric MnIII/II and FeIII/II Complexes with Terminal Hydroxo and Oxo Ligands : Probing Reactivity via O-H Bond Dissociation Energies. J. Am. Chem. Soc. 2003, 125, 13234–13242. (10) Under review: John, C. W.; Swain, G. M.; Hausinger, R. P.; Proshlyakov, D. A. Strongly Coupled Redox-Linked Conformational Switching at the Active Site of the Non-Heme Iron-Dependent Dioxygenase , TauD. J. Phys. Chem. B xx, xx. 102 (11) Martinez, S.; Hausinger, R. P. Biochemical and Spectroscopic Characterization of the Non-Heme Fe(II)- and 2-Oxoglutarate-Dependent Ethylene-Forming Enzyme from Pseudomonas syringae pv. phaseolicola PK2. Biochemistry 2016, 55 (43), 5989–5999. https://doi.org/10.1021/acs.biochem.6b00890. (12) Grzyska, P. K.; Müller, T. A.; Campbell, M. G.; Hausinger, R. P. Metal Ligand Substitution and Evidence for Quinone Formation in Taurine/α-Ketoglutarate Dioxygenase. J. Inorg. Biochem. 2007, 101, 797–808. https://doi.org/10.1016/j.jinorgbio.2007.01.011. (13) Ryle, M. J.; Liu, A.; Muthukumaran, R. B.; Ho, R. Y. N.; Koehntop, K. D.; Mccracken, J.; Que, L.; Hausinger, R. P. O2- and α-Ketoglutarate-Dependent Tyrosyl Radical Formation in TauD , an α-Keto Acid-Dependent Non-Heme Iron Dioxygenase. Biochemistry 2003, 42, 1854–1862. (14) Grzyska, P. K.; Hausinger, R. P.; Proshlyakov, D. A. Metal and Substrate Binding to an Fe(II) Dioxygenase Resolved by UV Spectroscopy with Global Regression Analysis. Anal. Biochem. 2010, 399, 64–71. https://doi.org/10.1016/j.ab.2009.11.022. (15) Taylor, J. F.; Morgan, V. E. Oxidation-Reduction Potentials of the Metmyoglobin- Myoglobin System. J. Biol. Chem. 1942, 144, 15–20. (16) John, C. W.; Proshlyakov, D. A. FTIR Spectro-Voltammetry and Quantitative Modeling of Analytes in Kinetically Constrained Redox Mixtures. Anal. Chem. 2019, 91, 9536– 9570. (17) Ryle, M. J.; Padmakumar, R.; Hausinger, R. P. Stopped-Flow Kinetic Analysis of Escherichia coli Taurine/α-Ketoglutarate Dioxygenase: Interactions with α-Ketoglutarate, Taurine, and Oxygen. Biochemistry 1999, 38, 15278–15286. https://doi.org/bi9912746. (18) Martinez, S.; Fellner, M.; Herr, C. Q.; Ritchie, A.; Hu, J.; Hausinger, R. P. Structures and Mechanisms of the Non-Heme Fe(II)- and 2-Oxoglutarate-Dependent Ethylene-Forming Enzyme: Substrate Binding Creates a Twist. J. Am. Chem. Soc. 2017, 139 (34), 11980– 11988. https://doi.org/10.1021/jacs.7b06186. (19) Schlereth, D. D.; Mäntele, W. Redox-Induced Conformational Changes in Myoglobin and Hemoglobin: Electrochemistry and Ultraviolet-Visible and Fourier Transform Infrared Difference Spectroscopy at Surface-Modified Gold Electrodes in an Ultra-Thin-Layer Spectroelectrochemical Cell. Biochemistry 1992, 31, 7494–7502. https://doi.org/10.1021/bi00148a009. (20) Schlereth, D. D.; Mäntele, W.; Fernandez, V. M. Protein Conformational Changes in Tetraheme Cytochromes Detected by FTIR Spectroelectrochemistry: Desulfovibrio desulfuricans Norway 4 and Desulfovibrio gigas Cytochromes C3. Biochemistry 1993, 32, 9199–9208. https://doi.org/10.1021/bi00086a027. 103 (21) Dawson, J. H. Probing Structure-Function Relations in Heme-Containing Oxygenases and Peroxidases. Science 1988, 240, 433–439. https://doi.org/10.1126/science.3358128. (22) Eichhorn, G.; Marzilli, L.; Marzilli, P. Advances in Inorganic Chemistry, 7th ed.; Elsevier: New York, 1988. (23) Elkins, J. M.; Ryle, M. J.; Clifton, I. J.; Hotopp, J. C. D.; Lloyd, J. S.; Burzlaff, N. I.; Baldwin, J. E.; Hausinger, R. P.; Roach, P. L. X-Ray Crystal Structure of Escherichia coli Taurine/α-Ketoglutarate Dioxygenase Complexed to Ferrous Iron and Substrates. Biochemistry 2002, 41, 5185–5192. (24) Papa, S. Proton Translocation Reactions in the Respiratory Chains. Biochim. Biophys. Acta 1976, 456, 39–84. https://doi.org/10.1016/0304-4173(76)90008-2. (25) Hosseinzadeh, P.; Lu, Y. Design and Fine-Tuning Redox Potentials of Metalloproteins Involved in Electron Transfer in Bioenergetics. Biochim. Biophys. Acta - Bioenerg. 2016, 1857, 557–581. https://doi.org/10.1016/j.bbabio.2015.08.006. (26) Yanagisawa, S.; Banfield, M. J.; Dennison, C. The Role of Hydrogen Bonding at the Active Site of a Cupredoxin: The Phe114Pro Azurin Variant. Biochemistry 2006, 45 (29), 8812–8822. https://doi.org/10.1021/bi0606851. (27) Bowman, S. E. J.; Bren, K. L. Variation and Analysis of Second-Sphere Interactions and Axial Histidinate Character in c -Type Cytochromes. Inorg. Chem. 2010, 49 (17), 7890– 7897. https://doi.org/10.1021/ic100899k. (28) Sonnay, M.; Fox, T.; Blacque, O.; Zelder, F. Modulating the Cobalt Redox Potential through Imidazole Hydrogen Bonding Interactions in a Supramolecular Biomimetic Protein-Cofactor Model. Chem. Sci. 2016, 7, 3836–3842. https://doi.org/10.1039/C5SC04396D. 104 CHAPTER 5: EVIDENCE FOR PROTONATION EVENTS AT THE ACTIVE SITE OF TAUD 105 INTRODUCTION Analyzing the pH sensitivity of the 2OG-Fe-TauD in situ model of the F3 intermediate has become an increasingly important task for understanding the mechanism of TauD. Determining the pKa of the F3 intermediate would provide evidence to help us infer whether the AF mechanism or hydroxyl radical rebound mechanism is utilized by the enzyme. Furthermore, the pH sensitivity of residues within the active site of TauD will likely provide insight into the redox-linked structural rearrangements discussed in Chapters 3 and 4. We may be able to use this information to strengthen our proposed structural model (Fig. 4.9) by investigating the proton affinity of H99 and H255. NPSV is a suitable tool for evaluating pKas as it provides two separate ways of monitoring the pH sensitivity of an analyte. The first is by observing changes to the vibrational spectra. Protonation of a residue can be revealed by changes to vibrational modes that are observed in the FT-IR difference spectra. Plotting the absorption amplitude of these modes versus the solution pH will give a thermodynamic profile of the proton affinity. The pKa is found at the midpoint of the transition. This method comes with the added benefit of using the IR spectra to identify the source of the pH sensitivity. The second method relies on the observed redox potential of the analyte over many different solution pH values. The redox potential will shift by -59 mV/pH between the pKa values of the reduced and oxidized states (Fig. 1.6). The pKa is determined by finding the transition points on this plot. The benefit of this method is its selectivity to the site of electron transfer. This is important in our case as the aquo ligands of Fe in TauD are not detectable in the vibrational spectra since they appear outside the frequency range provided by our FTIR spectrometer. 106 This chapter provides preliminary data of the pD sensitivity of TauD in the Fe-TauD, 2OG-Fe-TauD, succinate-Fe-TauD, and taurine-2OG-Fe-TauD forms. We analyze the observed reduction and oxidation potentials of TauD as they relate to the solution pD and observe some pD sensitive vibrational modes. EXPERIMENTAL PROCEDURES SAMPLE PREPARATION The methods used to purify the apoprotein of TauD and prepare 2OG-Fe(III)-TauD can be found in the Experimental Procedures section of Chapter 3. The Bis-Tris and Tris buffers used in this study were prepared at 25 mM in D2O at 10 °C with 0.5 M KCl. Bis-Tris was used to prepare buffer at pD 7.0 and 7.5. Tris was used to prepare buffers at pD 8.5, 9.5, and 10. NPSV MEASUREMENTS The experimental procedures in Chapter 2 describe how NPSV measurements were performed. For this chapter, the potential step durations were always 300 s. RESULTS The pD sensitivity of TauD is shown in Figures 5.1 – 5.4 for each form of TauD between solution pD values of 7.0 – 10. All forms of TauD, including Fe-TauD (Fig. 5.1), 2OG-Fe-TauD (Fig. 5.2), succinate-Fe-TauD (Fig. 5.3), and taurine-2OG-Fe-TauD (Fig. 5.4), show some form of pD sensitivity both in their vibrational spectra and in their electrochemical behavior. The observed pD sensitivity of the EOx and ERd values is indicative of a protonation event at the 107 active site of TauD that is closely associated with the oxidation state of the Fe center. This sensitivity follows the theoretical -59 mV / pD trend indicating that this process is a single protonation event. Figure 5.1 pD sensitivity of Fe-TauD. Top: Redox difference FTIR spectra. pD sensitive modes are labeled. Bottom left: Experimental Nernst profiles of reduction (circles) and oxidation profiles (solid lines). Bottom right: E½ vs (diamonds) steps normalized using ΔA1632 with solution pD plots for the observed reduction (left) and oxidation (right) potentials. The dashed line shows the best fit -59 mV/pD slope for a single protonation process. The observed pD sensitive vibrational modes (Table 5.1) are mostly attributable to Asp and His ligands with the 1603 – 1605 cm-1 mode as the only common mode observed in all forms of TauD. Other modes observed at 1358 cm-1 in Fe-TauD and between 1644 – 1657 cm-1 in Fe- TauD, 2OG-Fe-TauD, and taurine-2OG-Fe-TauD could not be confidently assigned based on their frequency at this time. Since all pD sensitive vibrational modes observed for each form of 108  TauD individually follow a synchronous Nernstian response, they are attributed to the same protonation event. However, the pD sensitivity differs between various forms of TauD as demonstrated by a comparison of the E½ vs. pD plots. Table 5.1. pD sensitive modes (cm-1) and tentative assignments.1–3 taurine-2OG- Assignments Fe-TauD 1649 1603 1415 1358 1338 2OG- TauD 1657 1605 1456 succinate- TauD 1603 1515 TauD 1682 1644 1605 Asp (C=O) / 2OG? unassigned Asp, HisH(C=C / CO2 +) His (vRing/NH3 His (CH3, CN) Asp, His(CH2, CO2 -) -) unassigned Asp, His (CH2) 1333 1333 Using these plots, we are able to observe the second pKa transition, that of the ferrous form, in Fe-TauD, 2OG-Fe-TauD, and succinate-Fe-TauD. The reduction and oxidation processes of Fe-TauD and succinate-Fe-TauD have the same pKa transitions at about 9.5. The oxidation and reduction processes of taurine-2OG-Fe-TauD also follow the same trend but no pKa transition is observed over the measured pD range. Interestingly, two separate pKa transitions were observed in the reduction and oxidation processes of 2OG-Fe-TauD. The pKa transition for the reduction is about 9.5 and 8.0 for the oxidation. This shift in the pKa by about 1.5 pD units is likely caused by changes to the hydrogen bonding network associated with the de/protonating group.4 The pKa transition of the ferric state of TauD is never observed. 109 Figure 5.2 pD sensitivity of 2OG-Fe-TauD. Top: Redox difference FTIR spectra. pD sensitive modes are labeled. Bottom left: Experimental Nernst profiles of reduction (circles) and oxidation profiles (solid lines). Bottom right: E½ vs (diamonds) steps normalized using ΔA1632 with solution pD plots for the observed reduction (left) and oxidation (right) potentials. The dashed line shows the best fit -59 mV/pD slope for a single protonation process. Changes to the pD appear to have an effect on the isomerization of TauD. Fe-TauD, 2OG-Fe-TauD, and taurine-2OG-Fe-TauD each show multiple transitions in at least one solution pD measurement. In Fe-TauD, the minor phase has an EOx that is approximately equal to the ERd of the transition and follows the same pD sensitivity. The relative populations of the minor and major phase varies between pD measurements but without a clear trend. In contrast, the minor phase in taurine-2OG-Fe-TauD becomes the major phase as the pD is increased from pD 7.5 to 10. At pD 10, taurine-2OG-Fe-TauD shows only a single oxidation phase. The only minor phase observed in the profiles of 2OG-Fe-TauD is at pD 7.0. 110 OxRdOxOx Figure 5.3 pD sensitivity of succinate-Fe-TauD. Top: Redox difference FTIR spectra. pD sensitive modes are labeled. Bottom left: Experimental Nernst profiles of reduction (circles) and profiles (solid lines). Bottom right: oxidation (diamonds) steps normalized using ΔA1632 with E½ vs solution pD plots for the observed reduction (left) and oxidation (right) potentials. The dashed line shows the best fit -59 mV/pD slope for a single protonation process. 111  Figure 5.4 pD sensitivity of taurine-2OG-Fe-TauD. Top: Redox difference FTIR spectra. pD sensitive modes are labeled. Bottom left: Experimental Nernst profiles of reduction (circles) and profiles (solid lines). Bottom right: oxidation (diamonds) steps normalized using ΔA1632 with E½ vs solution pD plots for the observed reduction (left) and oxidation (right) potentials. The dashed line shows the best fit -59 mV/pD slope for a single protonation process. DISCUSSION The presence of pD sensitive vibrational modes that can, in most cases, be attributed to either Asp or His modes (Table 5.1) is not surprising considering the presence of the H99, D101, and H255 Fe ligands. The modes observed at 1603 – 1605 cm-1 and at 1333 cm-1 can also be attributed to Glu,1 however this possibility is not considered since the only Glu residues present in TauD are on the surface of the enzyme and would not be redox-coupled and detectable using the NPSV method used here (see Chapter 2). 112  The pKa values of Asp and His free in solution are 4 and 6.4 respectively.5 However, the protein environment can cause these values to vary significantly due to hydrophobicity and electrostatic interactions.5,6 In fact, the Fe ligation of D101, H99, an H255 is expected to decrease their pKa values. The E½ versus pD profiles indicate a pKa transition at pD 8.0 or greater that is associated with the ferrous pKa transition. The ferric pKa transition is not observed in any of the E½ versus pD profiles but will likely take place at pD 7 or less. These pKa values are more positive than that of the pKa values of both Asp and His. However, His has a second protonatable nitrogen atom with a pKa of 14.5.7 Due to the high pKa of this imidazolate ion, it might be assumed that His99 and His255 are both in the imidazole state as one nitrogen atom must be deprotonated to bind to the Fe atom. However, deprotonation of the second nitrogen atom in Fe ligated His residues has been observed in peroxidases.8,9 Therefore, H99 and H255 are reasonable sources of the pD sensitivity. This idea is consistent with our proposed structural rearrangement (Fig. 4.9) where H255 is protonated in the ferrous state and deprotonated in the ferric state. CONCLUSION The preliminary data presented here provide evidence for a protonation event at the active site of TauD within the solution pD range of 7.0 and 10. This protonation event is observed in all forms of TauD studied here and is most likely associated with H99 or H255. A protonation event on H99 is consistent with the redox linked structural rearrangement we proposed in Ch. 4. Unfortunately, the data presented here are not enough to make this assignment. Using NPSV to study the pD sensitivity of H99 and H255 variants may provide more insight into the origin of the pD sensitivity observed here. 113 REFERENCES 114 REFERENCES (1) Barth, a. The Infrared Absorption of Amino Acid Side Chains. Prog. Biophys. Mol. Biol. 2000, 74, 141–173. https://doi.org/10.1016/S0079-6107(00)00021-3. (2) Barth, A. Infrared Spectroscopy of Proteins. Biochim. Biophys. Acta - Bioenerg. 2007, 1767, 1073–1101. https://doi.org/10.1016/j.bbabio.2007.06.004. (3) Mesu, J. G.; Visser, T.; Soulimani, F.; Weckhuysen, B. M. Infrared and Raman Spectroscopic Study of pH-Induced Structural Changes of L-Histidine in Aqueous Environment. Vib. Spectrosc. 2005, 39 (1), 114–125. https://doi.org/10.1016/j.vibspec.2005.01.003. (4) Shokri, A.; Abedin, A.; Fattahi, A.; Kass, S. R. Effect of Hydrogen Bonds on pKa Values: Importance of Networking. J. Am. Chem. Soc. 2012, 134 (25), 10646–10650. https://doi.org/10.1021/ja3037349. (5) Harris, T. K.; Turner, G. J. Structural Basis of Perturbed pKa Values of Catalytic Groups in Enzyme Active Sites. IUBMB Life 2002, 53, 85–98. https://doi.org/10.1080/10399710290038972. (6) Pace, C. N.; Grimsley, G. R.; Scholtz, J. M. Protein Ionizable Groups: pKa Values and Their Contribution to Protein Stability and Solubility. J. Biol. Chem. 2009, 284 (20), 13285–13289. https://doi.org/10.1074/jbc.R800080200. (7) Walba, H.; Isensee, R. W. Acidity Constants of Some Arylimidazoles and Their Cations. J. Org. Chem. 1961, 26 (8), 2789–2791. https://doi.org/10.1021/jo01066a039. (8) Heimdal, J.; Rydberg, P.; Ryde, U. Protonation of the Proximal Histidine Ligand in Heme Peroxidases. J. Phys. Chem. B 2008, 112 (8), 2501–2510. https://doi.org/10.1021/jp710038s. (9) Capena, X.; Vidossich, P.; Schrottner, K.; Calisto, B. M.; Banerjee, S.; Stampler, J.; Soudi, M.; Furtmüller, P. G.; Rovira, C.; Fita, I.; et al. Essential Role of Proximal Histidine-Asparagine Interaction in Mammalian Peroxidases. J. Biol. Chem. 2009, 284 (38), 25929–25937. https://doi.org/10.1074/jbc.M109.002154. 115 CHAPTER 6: CONCLUSION AND FUTURE DIRECTIONS 116 INTRODUCTION Chemists are constantly seeking new solutions for the design and development of efficient and sustainable catalysts. While significant progress has been made over the past few decades, 1–3 more progress can be made by studying metalloenzymes that utilize cheap and abundant metals, such as Fe, Cu, and Mn, to carry out similar reactions at much lower temperatures and pressures than used in industrial synthesis. For this reason, much work has been focused on the study of mechanisms, structures, and thermodynamic properties of enzymes to understand how they are capable of performing powerful reactions such as C-H activation.4–6 Spectroelectrochemistry has been a useful tool for such studies7–9, but is limited to systems that have an accessible active site and/or can be probed optically. This makes it difficult to study metalloenzymes with no natural pathway for electron transfer to the exterior of the protein, and those that have very weak optical properties. One such enzyme that has remained elusive due to its weak optical properties and buried active site is TauD. This enzyme is the archetypical member of the 2OG-dependent oxygenase family, known for their ability to activate C-H bonds of their substrates.10,11 Many studies have been performed on TauD and other 2OG-dependent oxygenases and structural models in order to develop an understanding how these enzymes perform such a powerful reaction.12–15 However, the thermodynamic properties of these enzymes have remained elusive and this deficiency has caused much controversy over the mechanism performed by these enzymes.12,13 Therefore, we sought to develop a novel spectroelectrochemical method that would allow us to probe the thermodynamic properties of TauD and other systems that remain elusive with current methods. 117 NPSV AND COMPUTATIONAL METHODS The previous chapters describe the development, validation, and application of novel methods for analyzing the thermodynamic and structural properties of analytes like TauD. NPSV coupled with global spectral regression analysis is beneficial for the study of analytes that have low extinction coefficients in the visible and ultraviolet regions and require the use of electrochemical mediators that have a spectrum that overlaps with that of the analyte. NPV allows for the quantitative determination of redox potentials and proton affinities and FTIR provides information about the structure of the analyte. The computational program, KinESim,16 is both a tool that can be used to help improve our understanding of observations in experimental data and an environment for designing experiments and predicting their outcomes. KinESim is a flexible tool providing the ability to model a diverse range of chemical and electrochemical pathways. Both of these methods were first applied to Mb to validate their effectiveness. NPSV was able to reproduce the previously published redox difference spectrum and reported E½ of Mb.17 Simulations using KinESim were capable of reproducing our experimental observations and helped us to further understand the relationship between an electrochemical mediator and an analyte. INSIGHTS FROM TAUD EXPERIMENTS KinESim and NPSV were used to investigate the structural and thermodynamic properties of many forms of TauD including Fe-TauD, 2OG-Fe-TauD, succinate-Fe-TauD, and taurine-2OG-Fe-TauD. NPSV provided the surprising observation of an electrochemical hysteresis in TauD, but, on its own, was not enough to determine if the hysteresis was caused by 118 the experimental set up or was an actual property of the enzyme. KinESim provided the necessary insight to determine that this hysteresis was caused by structural rearrangements of TauD’s active site. Further investigation of TauD using NPSV and chemical redox titrations with WT TauD and the Y73I, H99A, D101Q, and H255Q demonstrated that the Fe ligating residues, H99A, D101Q, and H255, are all involved in the redox-linked structural rearrangement. Using this information, we were able to propose a mechanism (Fig. 4.9) for the structural rearrangement that, upon oxidation to the ferric state, involves deprotonation of H99 and subsequent loss of a hydrogen bond allowing for the flexibility of the H99-T100-D101 chain, which then allows D101 to form a bidentate ligand with the Fe center. NPSV of TauD at various solution pD values suggested that there is a protonation event on a His residue within the pD range of 7.0 – 10, providing support to our proposed mechanism. This residue is expected to be either H99 or H255 but cannot be assigned at this time. Ultimately, the use of the NPSV and KinESim methods along with chemical titrations revealed the unexpected ability of TauD to undergo redox-dependent structural rearrangements resulting in a 0.47 V difference in potential between the ferrous a ferric states of TauD. This magnitude is equivalent to an 11 kcal/mole change in energy of the state, or 8 pH unit shift in a pKa. However, the E½ of the ferrous state of the enzyme is observed to shift negative upon oxidation to the ferric state of the enzyme (Fig. 4.4 and 4.5). This means that the redox-linked conformational change results in a loss of 11 kcal/mole of energy thus making C-H activation less likely to take place. So if structural rearrangements within TauD that occur upon oxidation cause a significant loss of energy, how is TauD able to activate C-H bonds? Fig 6.1 compares the E½ values of the 119 ferrous, ferric, and ferryl forms (including Cmp I and Cmp II) of HRP and CYP450 with those of TauD in the same oxidation states. Notice that the ferric isomer of TauD has a similar E½ for the Fe(II/III) transition as HRP and the five coordinate form of CYP450. In contrast, the ferrous state of TauD is over 0.5 V more positive than CYP450 for the same transition. Also notice that potential gap between the E½ of the Fe(II/III) and E½ of the Fe(III/IV) transitions is about 1 V in HRP and CYP450. If we assume the same potential gap between the E½ of the Fe(II→III) and Fe(III) to F4 transitions in TauD, the E½ of the F4 intermediate is estimated to be about 1.2 V, providing about 4 kcal/mole of energy towards C-H activation than the E½ of Cmp I in CYP450. Figure 6.1. Comparison of E½ values of HRP, CYP450, and TauD in the ferrous, ferric, and ferryl states. All values are reported against a Ag/AgCl reference electrode. Values reported for HRP18,19 and CYP45020,21 can be found in the indicated references. The 1.196 V value reported for the F4 intermediate of TauD is only an estimate and has not been measured. 120 The BDEO-H of the F3 intermediate in TauD was determined to be 98.3 kcal/mole.22 If the pKa of the F3 intermediate is assumed to be 7, thereby providing 9.6 kcal/mole of energy towards C-H activation, the E½ of the F4 intermediate would have to be 1.35 V to provide the remaining 31 kcal/mole needed. This is about 0.15 V higher than the low estimation for the E½ of the F4 intermediate (Fig. 6.1). Notice that the Fe in TauD has six coordinating ligands in the ferrous and ferric states but only five coordinating ligands in the F4 intermediate. A similar change is observed in CYP450 after the substrate binds resulting in the loss of a water ligand. The loss of the oxygen ligand from that water results in a positive shift in the E½ of CYP450 by +0.12 V. Since the F4 intermediate is formed after becoming a five coordinate system after losing an oxygen ligand, it is reasonable to assume that the E½ of the F4 intermediate will be more positive than the estimated 1.196 V and likely closer to 1.35 V. Therefore, it appears that the ferrous state of TauD is finely tuned to have a significantly more positive E½ than is observed in similar heme systems and uses this high potential to achieve large oxidizing potentials for C-H activation while maintaining a low pKa to promote the alkoxide formation pathway (Fig 1.2). As for the redox-linked conformational change, the formation of the ferryl intermediate will likely result in much faster changes than observed in the ferric state. However, these changes will still likely take place much slower than the turnover rate of TauD. Therefore, we think that TauD may have a built in protection mechanism against self-oxidation. If TauD manages to form the F4 intermediate in the absence of taurine, the enzyme will slowly begin to change its structure to lower its oxidation potential by over 0.45 V and thereby reduce the risk of oxidizing nearby residues. 121 FUTURE DIRECTIONS New and exciting information about TauD has been discovered and described in this dissertation. However, these discoveries have led us to yet more questions. What is the source of the pD sensitivity revealed in Chapter 5? How accurate is our proposed mechanism of the structural rearrangement presented in Chapter 4? Will oxidation to the ferryl state result in faster structural rearrangements? We still haven’t been able to answer the question that led us to the discoveries described herein: What is the pKa of the F3 intermediate? Below is a list of future directions that can be pursued to answer these questions. CONTINUED STUDIES OF pD SENSITIVITY IN TAUD Using 2OG-TauD as an in-situ model of the F3 intermediate to estimate its pKa is one of the major goals of our lab’s studies of TauD as explained in Chapter 1. Furthermore, our proposed mechanism for the redox-linked structural rearrangement of TauD described in Chapter 4 depends heavily on the pKa of H99. Preliminary results on the pD sensitivity of TauD are shown in Chapter 5 but the source of the observed pD sensitivity has not been established. Due to the vibrational modes associated with this pD sensitivity, we expect that it may be caused by H99 or H255. This question can be answered by evaluating the pD sensitivity of the H99A and H255Q variants. If the pD sensitivity is lost in either of these variants, it can be assumed that the evaluated variant is the source of the pD sensitivity. RESONANACE RAMAN STUDIES OF AQUO LIGANDS An alternative to using the redox potential of the Fe center in TauD to determine the pKa of the aquo ligands is to monitor the Fe-oxo vibrational mode of an aquo ligand over various 122 solution pH values. The protonated and deprotonated ligands will have two distinct frequencies that are expected to appear near 600 cm-1 based on previous measurements of the Fe-oxo mode in TauD.12 As the solution pH changes the population of protonated and deprotonated ligand, the mode will appear to shift between the two frequencies. The halfway point between the two modes indicates that 50% of the ligand is protonated and deprotonated. The pH at this point is the pKa of the ligand. If successful, this method could be used to corroborate the electrochemical data. The Fe-oxo mode is expected to appear between 500 and 600 cm-1, placing it outside the range of our FTIR spectrometer. However, this mode could be directly detected by using resonance Raman spectroscopy which utilizes the electronic absorption of an analyte to enhance Raman scattering. Some experiments have been attempted by exciting the metal to ligand charge transfer band of 2OG, but they have not yet been successful in detecting water ligands. Another approach used catechol-bound TauD since catechol produces a significantly more intense optical band than 2OG that can be used to enhance the Raman signal. However, experiments in our lab performed by Allison Stettler, Yegor Proshlyakov, and myself showed that catechol is capable of extracting Fe from the active site of TauD, making the catechol-Fe-TauD complex a poor candidate for measuring the proton affinity of aquo ligands. Other aromatic cofactors such as bipyridine or phenanthroline may offer more success in detecting aquo ligands as long as they do not also extract Fe from TauD. OTHER STRUCTURAL STUDIES In Chapter 4, we proposed a mechanism for the redox-linked structural rearrangement in TauD based on the electrochemical behavior of the D101Q and H99A variants. While plausible, 123 this proposed mechanism should be corroborated with other structural studies of TauD. One such study would be to obtain the crystal structure of TauD in the ferric state. To date, the only available crystal structure of TauD is in the ferrous state. A crystal structure of the ferric state would affirm or disprove structural rearrangement associated with bidentate binding of D101 and movement of the H99, T100, and D101 region. However, this method would not be able to provide direct information about the protonation state of H99. Other methods such as X-ray absorption fine structure, magnetic circular dichroism, and nuclear resonance vibrational spectroscopy can be used to acquire more information about how the Fe atom is coordinated in the ferrous and ferric states. FERRYL IRON STUDIES The redox-linked conformational change in TauD has been observed between the ferrous and ferric states, but it is unknown what implications this has on the highly oxidized ferryl state of TauD. It is the ferryl intermediate (F4, Fig. 1.2) that oxidizes the C-H bond of taurine and has even been shown to oxidize Y73 resulting in the formation of a tyrosyl radical placing the redox potential at no less than 730 mV vs. Ag/AgCl, the oxidation potential of tyrosine.23 Therefore, it is of interest to know what the redox potential of ferryl TauD is and if structural rearrangements play a role in obtaining this potential. Unfortunately, it is expected that the redox potential of the ferryl state will be high enough to perform oxidation of water and has been observed to oxidize a nearby tyrosine residue.24 Therefore, maintaining stability of the ferryl state long enough to acquire information about its structure and thermodynamic properties is expected to be a significant challenge. One possible solution is to use electrochemical mediators and an electrode that will allow the ferryl 124 state to be continuously regenerated so that it remains detectable. Some highly oxidizing mediators are already available for initial experiments including hexachloroiridate, iron bipyridine, and iron phenanthroline. 125 REFERENCES 126 REFERENCES (1) Gandeepan, P.; Müller, T.; Zell, D.; Cera, G.; Warratz, S.; Ackermann, L. 3d Transition Metals for C-H Activation. Chem. Rev. 2019, 119 (4), 2192–2452. https://doi.org/10.1021/acs.chemrev.8b00507. (2) Melin, F.; Hellwig, P. Recent Advances in the Electrochemistry and Spectroelectrochemistry of Membrane Proteins. Biol. Chem. 2013, 394, 593–609. https://doi.org/10.1515/hsz-2012-0344. (3) Sauermann, N.; Meyer, T. H.; Qiu, Y.; Ackermann, L. Electrocatalytic C-H Activation. ACS Catal. 2018, 8 (8), 7086–7103. https://doi.org/10.1021/acscatal.8b01682. (4) Singh, R.; Kumar, M.; Mittal, A.; Mehta, P. K. Microbial Enzymes: Industrial Progress in 21st Century. 3 Biotech 2016, 6, 1–15. https://doi.org/10.1007/s13205-016-0485-8. (5) Mohamad, N. R.; Marzuki, N. H. C.; Buang, N. A.; Huyop, F.; Wahab, R. A. An Overview of Technologies for Immobilization of Enzymes and Surface Analysis Techniques for Immobilized Enzymes. Biotechnol. Biotechnol. Equip. 2015, 29, 205–220. https://doi.org/10.1080/13102818.2015.1008192. (6) Lewis, J. C.; Coelho, P. S.; Arnold, F. H. Enzymatic Functionalization of Carbon- Hydrogen Bonds. Chem. Soc. Rev. 2011, 40, 2003–2021. https://doi.org/10.1039/c0cs00067a. (7) Clair, C. S. S.; Gray, H. B.; Valentine, J. S. Spectroelectrochemistry of Copper-Zinc Superoxide Dismutase. Inorg. Chem. 1992, 31, 925–927. https://doi.org/10.1021/ic00031a041. (8) DeLacey, A. L.; Stadler, C.; Fernandez, V. M.; Hatchikian, E. C.; Fan, H. J.; Li, S.; Hall, M. B. IR Spectroelectrochemical Study of the Binding of Carbon Monoxide to the Active Site of Desulfovibrio Fructosovorans Ni-Fe Hydrogenase. J. Biol. Inorg. Chem. 2002, 7, 318–326. https://doi.org/10.1007/s00775-001-0301-7. (9) Tsujimura, S.; Kuriyama, A.; Fujieda, N.; Kano, K.; Ikeda, T. Mediated Spectroelectrochemical Titration of Proteins for Redox Potential Measurements by a Separator-Less One-Compartment Bulk Electrolysis Method. Anal. Biochem. 2005, 337 (2), 325–331. https://doi.org/10.1016/j.ab.2004.11.017. (10) Proshlyakov, D. A.; Hausinger, R. P. Transient Iron Species in the Catalytic Mechanism of the Archetypal α-Ketoglutarate-Dependent Dioxygenase, TauD. In Iron-Containing Enzymes: Versatile Catalysts of Hydroxylation Reactions in Nature; Visser, S. P. De, Kumar, D., Eds.; Royal Society of Chemistry, 2012; pp 67–87. 127 (11) Proshlyakov, D. A.; McCracken, J.; Hausinger, R. P. Spectroscopic Analyses of 2- Oxoglutarate-Dependent Oxygenases: TauD as a Case Study. J. Biol. Inorg. Chem. 2017, 22, 367–379. https://doi.org/10.1007/s00775-016-1406-3. (12) Grzyska, P. K.; Appelman, E. H.; Hausinger, R. P.; Proshlyakov, D. A. Insight into the Mechanism of an Iron Dioxygenase by Resolution of Steps Following the FeIV=O Species. Proc. Natl. Acad. Sci. U. S. A. 2010, 107, 3982–3987. https://doi.org/10.1073/pnas.0911565107. (13) Mitchell, A. J.; Dunham, N. P.; Martinie, R. J.; Bergman, J. A.; Pollock, C. J.; Hu, K.; Allen, B. D.; Chang, W.; Silakov, A.; Bollinger, J. M.; et al. Visualizing the Reaction Cycle in an Iron(II)- and 2-(Oxo)-Glutarate-Dependent Hydroxylase. J. Am. Chem. Soc. 2017, 139, 13830–13836. https://doi.org/10.1021/jacs.7b07374. (14) Ho, R. Y. N.; Mehn, M. P.; Hegg, E. L.; Liu, A.; Ryle, M. J.; Hausinger, R. P.; Que, L. Resonance Raman Studies of the Iron ( II ) -R -Keto Acid Chromophore in Model and Enzyme Complexes. J. Am. Chem. Soc. 2001, 123, 5022–5029. (15) Janardanan, D.; Wang, Y.; Schyman, P.; Que, L.; Shaik, S. The Fundamental Role of Exchange-Enhanced Reactivity in C-H Activation by S = 2 Oxo Iron(IV) Complexes. Angew. Chemie - Int. Ed. 2010, 49, 3342–3345. https://doi.org/10.1002/anie.201000004. (16) https://github.com/dap-biospec/KinESim. (17) John, C. W.; Proshlyakov, D. A. FTIR Spectro-Voltammetry and Quantitative Modeling of Analytes in Kinetically Constrained Redox Mixtures. Anal. Chem. 2019, 91, 9536– 9570. (18) Harbury, H. A. Oxidation-Reduction Potentials of Horseradish Peroxidase. J. Biol. Chem. 1956, 24, 1009–1024. (19) Hayashi, Y.; Yamazaki, I. The Oxidation-Reduction Potentials of Compound I/Compound II and Compound II/Ferric Couples of Horseradish Peroxidases A2 and C. J. Biol. Chem. 1979, 254, 9101–9106. (20) Kazlauskaite, J.; Westlake, A. C. G.; Wong, L.; Hill, H. A. Direct Electrochemistry of Cytochrome P450cam. Chem. Commun. 1996, 2189–2190. (21) Mittra, K.; Green, M. T. Reduction Potentials of P450 Compounds I and II: Insight into the Thermodynamics of C-H Bond Activation. J. Am. Chem. Soc. 2019, 141 (13), 5504– 5510. https://doi.org/10.1021/jacs.9b00242. (22) Visser, S. P. De. Propene Activation by the Oxo-Iron Active Species of Taurine / α- Ketoglutarate Dioxygenase (TauD) Enzyme . How Does the Catalysis Compare to Heme- Enzymes ? J. Am. Chem. Soc. 2006, 128, 9813–9824. 128 (23) Harriman, A. Further Comments on the Redox Potentials of Tryptophan and Tyrosine. J. Phys. Chem. 1987, 91 (24), 6102–6104. https://doi.org/10.1021/j100308a011. (24) Ryle, M. J.; Liu, A.; Muthukumaran, R. B.; Ho, R. Y. N.; Koehntop, K. D.; Mccracken, J.; Que, L.; Hausinger, R. P. O2- and α-Ketoglutarate-Dependent Tyrosyl Radical Formation in TauD , an α-Keto Acid-Dependent Non-Heme Iron Dioxygenase. Biochemistry 2003, 42, 1854–1862. 129