CHAPTER 1: Spectroscopic and Electronic Structure Studies Probing Mechanism: Introduction and Overview

The first Gordon Research Conference on “Molybdenum and Tungsten Enzymes” was held July 4–9 1999 at Plymouth State College (now Plymouth State University) in New Hampshire and was chaired by Ed Stiefel and Russ Hille. This meeting proved to be a transformative one for our field in that it provided an intellectual forum for molecular biologists, synthetic chemists, enzymologists, theorists, crystallographers and spectroscopists to converge, discuss their latest results and develop long-standing relationships that would foster new collaborations that have allowed us to understand the structure and function of Mo- and W-containing enzymes as well as the intricate details of their catalytic mechanisms of activity. These enzymes continue to be the subject of intense research efforts, and this is a direct result of their unusual geometric and electronic structures, their key roles in the global C, N and S cycles, their pharmacological importance, and their importance in human health. This volume will detail how spectroscopy, structure, electrochemistry and theory have been used to develop a comprehensive description of the active site electronic structure contributions to reactivity in pyranopterin Mo enzymes and the Mo-dependent nitrogenase. A particular emphasis is placed on how these important studies have been used to reveal critical components of enzyme mechanisms.

With the sole exception of nitrogenase, the pyranopterin Mo (and W) enzymes are the only metalloenzymes that utilize second-^1,2  (and third-) row transition metal ions to catalyze a myriad of redox transformations involving enzyme substrates.^3,4  The uniqueness of these enzymes is further underscored by the fact that they possess the molybdenum cofactor (Moco, Figure 1.1),^5,6  which is comprised of a high-valent Mo^IV,V,VI (W^IV,V,VI) ion coordinated by a unique dithiolene ligand (the pyranopterin dithiolene). This dithiolene ligand is connected to a pterin ring system by a pyran ring that may be in either a closed (typical) or an open ring configuration. The pyranopterin dithiolene (PDT) is a highly complex ligand (vide infra) that is unique to the mononuclear Mo (W) enzymes. The remarkable nature of this ligand is exemplified by its potential for electronic flexibility, including changing its redox and/or tautomeric state to exert additional control of the Mo redox potential.⁶  The pyranopterin molybdenum enzymes have been historically divided into three broad families: the sulfite oxidase (SO) family, the xanthine oxidase (XO) family and the dimethylsulfoxide reductase (DMSOR) family of enzymes (Figure 1.2).⁵  This broad classification has been based on the coordination geometry of their active sites, the nature of their respective protein folds and the type and breadth of reactions that are catalyzed by the enzymes.

The vast majority of pyranopterin Mo (W) enzymes catalyze the formal transfer of an oxygen atom between the substrate and the Mo ion. The generalized oxygen atom transfer reactions possess the following stoichiometry:^5,7–9 

Regarding substrate oxidations, reductive oxygen atom transfer is given by the reverse of this reaction. In marked contrast to the oxotransferases, substrate hydroxylations (formal insertion of an O atom into a substrate C–H bond) follow the reaction stoichiometry given below:

Excellent reviews^3,5,7–17  have highlighted recent advances in our understanding of these enzymes, in addition to reviews that have covered the Mo-dependent nitrogenase,^18–20  which will be introduced later in this chapter. More specific reviews have focused on the sulfite oxidase,^21,22  DMSO reductase²³  and xanthine oxidase^7,24  enzyme families, with additional reviews covering Moco biosynthesis,^25–27  spectroscopic studies^17,28  and computational probes of the enzyme reaction coordinates.^16,28 

Although it has long been suggested that the PDT may play key roles in facilitating vectorial electron transfer, modulating enzyme reduction potentials and providing an anchor for the Mo ion in the catalytic active site, there is still much to be learned about this complex ligand. Some of the latest work on the PDT has detailed a relationship between pyranopterin dithiolene geometric structure and enzyme function.⁶  A bioinformatics study of 309 pyranopterins from 109 separate enzyme structures showed that these enzymes possessed geometries that can be described by a well-defined distortion coordinate that is related to the nature of PDT out-of-plane distortions that were revealed by X-ray crystallography.⁶  This distortion coordinate was analyzed in the context of DFT calculations that were performed on geometry-optimized PDT structures. The results suggest that differences in the nature of the PDT out-of-plane distortions are related to PDTs that adopt different oxidation states. Specifically, the analysis suggests that biological PDTs may not all exist in the fully reduced “tetrahydro” oxidation state, as has been previously thought.⁵  These researchers hypothesized that the PDT may also be present in two electron oxidized “dihydro” forms.⁶  Remarkably, the observed PDT distortions can be associated with specific enzyme families. For example, the PDT distortions observed for XO family enzymes are consistent with the PDT being in the fully reduced “tetrahydro” oxidation state, while SO family PDTs adopt “dihydro” structures. Interestingly, DMSOR family enzymes, which contain two PDTs coordinated to Mo, appear to possess one PDT that displays an SO type distortion and one PDT possessing an XO type distortion. The results are of interest in that they suggest a link between enzyme function and the oxidation state(s) of the PDT.⁶  In support of this idea, the first structurally characterized oxomolybdenum complex to incorporate a pyranopterin dithiolene ligand was found to possess the ligand in the “dihydro” oxidation state.²⁹  A combination of X-ray crystallography and ¹H NMR spectroscopy was used to show that this complex possessed a complete pyranopterin dithiolene ligand, and that reversible pyran ring opening and closing may represent a dynamic process in pyranopterin molybdenum enzymes. Although these addressed apparent relationships between PDT geometric structure, oxidation state and enzyme function, our understanding of how PDT electronic structure contributes to enzyme catalysis remains to be determined.

From a spectroscopic and electronic structure viewpoint, the pyranopterin molybdenum (and tungsten) containing enzymes are unique among metalloenzymes. This primarily derives from the terminal Mo-oxo ligation coupled with the high oxidation states accessible to the metal ion during the course of catalysis. This results in a large splitting of the t_2g orbital set with the Mo(xy) redox active orbital being well separated energetically from the Mo–O_oxo dπ* antibonding orbitals.³⁰  Thus, the Mo(v) ion possesses an (xy)¹ configuration with the redox orbital oriented perpendicular to a Mo–O_oxo vector³⁰  and the Mo(iv) ion possesses a low-spin (xy)² configuration with a diamagnetic ground state. Spectroscopic studies probing the Mo ion in pyranopterin Mo enzymes have not been trivial. This results from the fact that the majority of pyranopterin Mo enzymes possess strongly absorbing flavin, iron–sulfur and/or heme chromophores⁵  that mask the electronic absorption spectra associated with the Mo active sites. Furthermore, this problem is exacerbated by the fact that the only relevant paramagnetic state is the Mo(v) oxidation state, which represents an obligatory catalytic intermediate in the electron transfer regeneration half-reaction that progresses via sequential one-electron transfers that interconvert the diamagnetic Mo(iv) and Mo(vi) oxidation states. Thus, in order to use high resolution paramagnetic probes of pyranopterin Mo enzyme active site electronic structure, it is necessary to either trap reaction intermediates³¹  or study inhibited enzyme forms.^32,33  This critical issue cannot be overstated, as the general inability to probe resting, fully oxidized or fully reduced enzyme forms using paramagnetic spectroscopies has provided a significant challenge to understanding crucial interrelationships between spectroscopic features of the active sites and their corresponding geometric and electronic structures.

These caveats aside, paramagnetic spectroscopies, including EPR, ESEEM, ENDOR, MCD and Mössbauer, have played critical roles in determining the electronic structure of the Mo(v) state and have contributed to a greater understanding of pyranopterin molybdenum enzyme reaction coordinates. The first spectroscopic studies of pyranopterin Mo enzyme geometric and electronic structure employed EPR spectroscopy. Notable among these are the early work of Bray and coworkers, who used EPR to investigate rapidly appearing Mo(v) EPR signals in reduced forms of XO.³⁴  These early EPR studies were often calibrated against data collected on small molecule models,^35,36  as there were not yet any high-resolution protein crystal structures. The xanthine oxidase very rapid species,^37,38  which can be trapped under turnover conditions with specific substrates,³⁹  represents a clear example of how multiple spectroscopic probes of an enzyme intermediate have played a significant role in enhancing our understanding of the enzyme's mechanism. The very rapid intermediate is a Mo(v)-product species that has been extensively probed by a combination of EPR^37–39  (Figure 1.3), ENDOR⁴⁰  and MCD³¹  spectroscopies. Collectively, these data have been used to show that the very rapid intermediate (1) possesses an apical oxo and an equatorial sulfido ligand that are oriented cis relative to one another,³¹  (2) possesses a highly covalent Mo=S d–p π bonding interaction, (3) does not involve the formation of an organometallic Mo–C bond in the catalytic cycle of the enzyme⁴⁰  and (4) possesses product bound to the Mo ion as the enolate tautomer.⁴⁰  The combination of computational, spectroscopic and reactivity studies on xanthine oxidase very rapid has contributed greatly to our understanding of the reductive half-reaction in xanthine oxidase and related Mo hydroxylase family enzymes.³  Of course, the xanthine oxidase very rapid story is only one of many studies where paramagnetic spectroscopies have proven to be crucial for developing a greater understanding of pyranopterin molybdenum enzyme electronic structure contributions to reactivity.^3,24  Diamagnetic probes of enzyme electronic structure have primarily involved electronic absorption and resonance Raman spectroscopies. It is important to note that when spectroscopic studies on the enzymes can be directly compared to analogous studies on small molecule analogs of their active sites, tremendous insight into the relationships between electronic and geometric structure contributions begins to emerge. This is due, in part, to the fact that the small molecule analogs do not possess the competing chromophores that are present in many of the enzymes, and they allow for specific structural components found in the enzymes to be evaluated individually (Figure 1.4). This has resulted in the detailed spectral probing of numerous elegantly designed small molecules using electronic absorption spectroscopy, S K-edge XAS, resonance Raman and MCD spectroscopies.^30,41–59  Spectroscopic studies on diamagnetic model compounds in the catalytically relevant Mo(iv) and Mo(vi) oxidation states have also contributed greatly to our understanding of pyranopterin Mo enzymes in these “spectroscopically challenged” oxidation states.

The initial contribution to this volume provides a detailed overview of how spectroscopy and computations have been used in concert to probe the canonical members of each pyranopterin Mo enzyme family, as well as the pyranopterin dithiolene ligand itself. The discussion focuses on how a combination of enzyme geometric structure, spectroscopy and biochemical data have been used to arrive at an understanding of electronic structure contributions to reactivity in all of the major pyranopterin Mo enzyme families. A unique aspect of this discussion is that spectroscopic studies on relevant small molecule model compounds have been melded with analogous studies on the enzyme systems to arrive at a sophisticated description of active site electronic structure. As the field moves forward, it will become increasingly important to understand the structure, function and reaction mechanisms for the numerous non-canonical (i.e. beyond sulfite oxidase, xanthine oxidase, DMSO reductase) pyranopterin Mo enzymes.

Guigliarelli and coworkers then present a detailed overview of how EPR spectroscopy has been applied to the study of pyranopterin Mo enzymes. A specific emphasis has been placed on the challenging nature of studying reaction intermediates and how analysis of the magnitude and anisotropy of the g-tensor have contributed to a greater understanding of the enzymes. They discuss advances in pulsed EPR techniques, including ENDOR, ESEEM and HYSCORE, and how these have been used with isotope perturbation and computational studies to understand the nature of the hyperfine tensor and how this relates to structure and mechanism. Finally, they detail how one may obtain long-range structural details from spin–spin coupling between distinct paramagnetic centers in the enzymes.

The first pyranopterin molybdenum enzyme to be fully characterized by X-ray crystallography was the aldehyde oxidoreductase from Desulfovibrio gigas.⁶⁰  The subsequent determination of numerous molybdenum protein structures, including nitrogenase, by X-ray crystallography have revolutionized the way we think about geometric structure contributions to catalysis. These data have provided electronic structure researchers, synthetic chemists and spectroscopists with a structural reference for the starting point in the catalytic cycles of these enzymes. When combined with EXAFS analysis, a clear picture of the active site metal–ligand first coordination sphere begins to emerge for both the oxidized and reduced states of the enzyme. George has provided a thorough tutorial-type presentation of X-ray absorption spectroscopy (XAS), including the analysis of EXAFS data, before entering into a discussion of how XAS has enhanced our understanding of pyranopterin Mo and W enzymes. A thoughtful component of the contribution details the relative strengths and weaknesses of XAS as a structural probe by using a series of illustrative examples.

As mentioned previously, the vast majority of pyranopterin Mo enzymes catalyze two-electron redox reactions that are coupled to the formal transfer of an oxygen atom between the Mo center and the substrate. As such, electrochemical studies of the enzymes provide insight into these two-electron redox processes, in addition to the one-electron transfer events that are associated with the completion of the catalytic cycle. Kalimuthu and Bernhardt have provided a wonderful and timely discussion of direct and mediated protein electrochemistry of pyranopterin Mo enzymes that begins with a tutorial on enzyme electrochemistry (DC voltammetry). DC voltammetry is of particular utility as it is well suited for probing metalloenzymes that are undergoing catalysis (i.e. oxidizing or reducing substrates) since the observed current (catalytic current) is amplified. A description of how XOR-, sulfite oxidase-, nitrate reductase- and DMSO reductase-based electrochemistry can be used for the development of amperometric biosensors is included in their contribution.

Theory has played a central role in our understanding of molybdenum enzyme active site structure, dynamics, mechanism, spectroscopy and electronic structure contributions to reactivity. Ryde and coworkers have reviewed how computations have enhanced our understanding of enzymes that belong to the SO, DMSOR and XOR enzyme families. Particular emphasis is given to understanding the effects of basis set size, nature of the functional, solvation and dispersion effects, etc. on the determination of computed reaction coordinate energies, including transition states. The quantum mechanical approaches that are highlighted have been diverse. Modern DFT and configuration interaction methodologies will continue to be utilized to understand these remarkable biological catalysts that employ a second-row transition metal ion. Coupled with the ever-increasing power of modern computers, we can expect continued and accelerated applications of theory to provide additional insight into the interpretation of experimental results, and to greatly assist in solving some of the most difficult problems related to these enzymes.

The Mo dependent nitrogenase is the sole exception among all of the Mo containing enzymes in that it does not possess a pyranopterin dithiolene ligand bound to the Mo ion. Molybdenum-dependent nitrogenase consists of an Fe protein and a MoFe protein, the latter of which possesses an 8Fe7S P-cluster and an Fe₇S₉Mo cluster possessing a homocitrate bound to Mo and an unusual carbide ligand embedded in the center of the core. The FeMo-cofactor (FeMo-co) represents the locus of dinitrogen reduction to ammonia according to:

The catalytic mechanism of nitrogenase is discussed in terms of the Lowe–Thorneley scheme (Figure 1.5), which details eight sequential proton coupled electron transfer steps in the catalytic cycle. The importance of nitrogenase is underscored by its dominant role in biogeochemical nitrogen fixation. Spectroscopic studies probing the Mo ion in nitrogenase have also not been trivial due to the large protein iron content and the complex nature of the exchange coupled paramagnetic spin centers. For nitrogenase, Mössbauer spectroscopy has been used to probe iron oxidation states and spin states of FeMoco, and has proven to be complementary to the high information content of paramagnetic resonance spectroscopies. The goal of these spectroscopic studies is to provide a geometric and electronic structure description of the various coupled electron–proton transfer events that make up the various components of the Lowe–Thorneley scheme in order to develop a detailed understanding of the nitrogenase catalyzed six-electron reduction of dinitrogen to ammonia. Nitrogenase provides a prime example of how spectroscopic studies can be used to gain insight into enzyme structure and mechanism. Most recently, pulsed EPR and ENDOR methods have been coupled with approaches that include freeze-quench and the study of enzyme variants in order to identify several catalytic intermediates and their relationship to the nitrogenase catalytic cycle.

As such, the final chapter of the volume effectively ends as it begins, with a detailed discussion of how model and enzyme structural, computational and spectroscopic studies have led to an increased understanding of enzyme (nitrogenase) reactivity. Here, Tuczek provides an excellent state-of-the-art review of how these approaches have been utilized in order to unravel the intricacies of how Nature fixes nitrogen. He begins with a description of how model studies have provided detailed information regarding how dinitrogen binds to metal centers and is subsequently activated for cleavage of the strong N≡N bond by a series of protonation steps and the introduction of reducing equivalents. This work is evaluated in the context of computational studies on the enzyme system, with a specific emphasis on correlating the computational work to available experimental data, including spectroscopic studies on nitrogenase.

Clearly, there have been numerous advances in our understanding of how pyranopterin Mo enzymes and nitrogenase function, including how the underpinning electronic structure of their respective catalytic active sites contributes to their unique reactivities. The individual contributions to this volume highlight how we, as a community, have greatly expanded the knowledge base by using a combined structural, spectroscopic, voltammetric and computational approach to ultimately understand the nature of their catalytic cycles. Of course, this would not be possible without the numerous contributions from researchers engaged in the biological, structural and small-molecule studies that have been detailed in the two other volumes.

M. L. K. would like to thank all of his graduate students, postdoctoral associates and collaborators who have contributed to the various works that have been described in this chapter. M. L. K. also acknowledges the National Institutes of Health (GM-057378) for continued support of the author’s work, part of which is included in this chapter.