The pyranopterin dithiolene ligand is remarkable in terms of its geometric and electronic structure and is uniquely found in mononuclear molybdenum and tungsten enzymes. The pyranopterin dithiolene is found coordinated to the metal ion, deeply buried within the protein, and non-covalently attached to the protein via an extensive hydrogen bonding network that is enzyme-specific. However, the function of pyranopterin dithiolene in enzymatic catalysis has been difficult to determine. This focused account aims to provide an overview of what has been learned from the study of pyranopterin dithiolene model complexes of molybdenum and how these results relate to the enzyme systems.
1. Introduction
The pyranopterin molybdenum (Mo) enzymes factor prominently in global biogeochemical cycles and are critical to the life processes of most organisms on Earth
[1,2,3,4,5,6,7][1][2][3][4][5][6][7]. In humans, these enzymes catalyze reactions that contribute to the production of reactive oxygen species associated with postischemic reperfusion injury
[8,9][8][9] and oxidative stress
[10], xenobiotic detoxification
[10[10][11][12][13][14][15][16],
11,12,13,14,15,16], drug metabolism
[11,17,18,19,20,21,22,23,24][11][17][18][19][20][21][22][23][24] and prodrug activation
[21[21][25],
25], nitrite to NO conversion
[24,26,27,28,29,30,31[24][26][27][28][29][30][31][32][33],
32,33], sulfite oxidation
[1,34,35,36][1][34][35][36], molybdenum cofactor (Moco) sulfuration
[26[26][33][37][38][39][40][41][42],
33,37,38,39,40,41,42], and amino acid catabolism
[43]. The importance of these enzymes in humans is underscored by the fact that Moco deficiency can result in early infant mortality
[24]. More recently, pyranopterin Mo enzymes have been found to play key roles in the gut microbiome
[17,44,45,46][17][44][45][46] and as methionine sulfoxide reductases in respiratory pathogens (e.g.,
Haemophilus influenzae)
[1,47,48,49,50][1][47][48][49][50]. These are unusual metalloenzymes since they employ second- (Mo) and third-row (W) transition metal ions to perform a myriad of two-electron redox transformations
[4]. Furthermore, Moco in these molybdoenzymes is unique in possessing a pyranopterin dithiolene ligand (PDT; also known as molybdopterin)
[37[37][51][52][53][54][55][56],
51,52,53,54,55,56], and Moco is biosynthesized in a complex series of reactions by nine different gene products in bacteria and seven in plants and humans
[37,39,54,55,56,57,58,59,60][37][39][54][55][56][57][58][59][60]. The related pyranopterin W enzymes possess a closely analogous cofactor, the tungsten cofactor, or Tuco
[3,61,62][3][61][62].
2. Currently Known about Moco in the Enzymes
2.1. Protein X-ray Crystallography Gives Atomic Level Views of Moco
The W-containing aldehyde ferredoxin oxidoreductase
[76][63] and Mo
dimethylsulfoxide reductase DMSOR(DMSOR) [77][64] were structurally characterized in 1995 and 1996, respectively, and the structures represent the first for any pyranopterin tungsten or molybdenum enzyme. Now there are crystal structures for a large number of molybdoenzymes, and this has led to a dramatic increase in our understanding of these enzymes.
Figure 21 shows Moco as found in each of the three canonical molybdoenzyme families: sulfite oxidase (SUOX), xanthine oxidase (XDH), and dimethylsulfoxide reductase (DMSOR). These structures are depicted both as a three-dimensional image and in a bond line drawing representation. The 3D views emphasize the non-planar, bent nature of the pyranopterin component of Moco. As more examples of Moco structures located in different protein environments became available through protein crystallography, the dramatic range of pyranopterin conformations within the PDT ligand became apparent. This flexibility in pyranopterin conformation was noted as early as 1997
[78][65], and it is depicted as an overlay of the pyranopterin portions of Moco from different protein crystal structures (
Figure 3a2). A more recent analysis of the metrical differences in the folding of 319 pyranopterins in 102 molybdenum protein structures led to the identification of two main pyranopterin conformations observed in the protein structures and to the proposal that the pterin might have different oxidation states among the three families (
Figure 3b2)
[79][66]. From the work emerged the proposal that the highly bent PDT ligands displayed in the XDH family enzymes corresponded to fully reduced pyranopterin structures, whereas the less bent pyranopterins in the PDTs from SUOX family proteins better fit a dihydropyranopterin structure (
Figure 3b2). Intriguingly, the two PDT pyranopterins in the DMSOR family of proteins exhibited different conformations, where the highly bent proximal pyranopterin fits a reduced pyranopterin description while the distal PDT ligand is less bent and consistent with a dihydropterin assignment.
Figure 21. Representative examples of Moco structures for each of the SO, XO, and DMSO families, and one example of Moco possessing one PDT having an open, uncyclized pyran ligand. (a) Sulfite Oxidase (PDB 1SOX). (b) Xanthine Oxidase (PDB 3NRZ). (c) Dimethylsulfoxide Reductase (PDB 1EU1). (d) Perchlorate Reductase (PDB 5CH7) where the red arrow points to the open pyran ring position.
Figure 32. (
a) Two views of the range of pyranopterin conformations observed in 1997. (
b) Two distinct conformations suggest that pyranopterins in Moco have different oxidation states in different families. P* denotes a phosphate or a dinucleotide terminus. Adapted from Ref.
[79][66].
A second type of structural anomaly is observed within the pyranopterin portion of Moco. Among the large number of molybdoenzyme structures, there are three proteins—all members of the DMSOR family—whose structures clearly show the distal PDT ligand with no pyran ring
[80][67]. The first such example identified was dissimilatory nitrate reductase (NarGHI) from
E. coli [81][68], followed by ethyl benzene dehydrogenase (EBDH)
[82][69]. The most recent example is perchlorate reductase (PcrAB)
[83][70] from
Azospira suillum, which is shown as a representative example for this structural type in
Figure 21d.
Lastly, the protein environment that encapsulates and protects Moco from degradation is recognized to play a role in Moco function. The abundance of H-bonds tethering pyranopterin to the protein is recognized to enforce the proper orientation and conformation of the cofactor. However, H-bonding analysis shows several other ways that H-bonds—or indeed, their absence—might be involved in catalysis. A study of all known PDT-containing protein structures that analyzed patterns of hydrogen bonding interactions between protein residues and Moco revealed multiple conserved features within each protein family
[80][67]. These are summarized pictorially in
Figure 43.
Figure 43. H-bonding interactions between the pyranopterin of Moco and adjacent protein residues. (
a) Conserved H-bonding interactions identified for all 7 members of the XDH family. (
b) Conserved H-bonding interactions identified for three members of the SUOX family. (
c) H-bonding interactions within the DMSOR protein structure that are representative of 15 other members of the DMSOR family. (
d) H-bonding interaction in
E. coli nitrate reductase, whose Moco exhibits one non-cyclized pterin structure.
2.2. Information about the PDT of Moco Obtained from Spectroscopy
Direct spectroscopic studies that inform on the PDT component of Moco are sparse
[89[71][72],
90], and this is due to the fact that the vast majority of pyranopterin Mo enzymes possess additional highly absorbing chromophores, including flavin, [2Fe-2S], [4Fe-4S], and heme. However, spectroscopic probes of the PDT in molybdoproteins that lack these chromophores and in relevant model systems that possess a coordinated PDT ligand are important, and these studies will assist in defining both the proposed and still unknown role(s) of the PDT in catalysis. As detailed in the prior section, the PDT has no covalent interactions with the protein, and X-ray crystallography provides strong evidence that the PDT is extensively hydrogen bonded to the protein
[80][67].
Early resonance Raman spectroscopic studies were performed on
R. sphaeroides and
R. capsulatus DMSORs
[93,94,95][73][74][75] and biotin sulfoxide reductase
[96][76], but more direct probes of hydrogen bonding between the protein and the pyranopterin derive from Raman studies on bovine and bacterial XO/XDH
[89,90,97,98][71][72][77][78]. These latter studies have been particularly revealing from the standpoint of observing low-frequency modes assignable to the PDT. Using lumazine as the reducing substrate, it has been shown that XO/XDH catalyzes the two-electron conversion of this substrate to violopterin
[89,90,97,98,99,100[71][72][77][78][79][80][81][82],
101,102], which subsequently binds strongly to the Mo(IV) center, allowing for spectral probing of an enzyme-product complex by optical spectroscopies. This Mo(IV)-violopterin state possesses an intense charge transfer band that absorbs light in the red/NIR region of the optical spectrum, providing an opportunity to probe a catalytically relevant product-bound species formed by enzymatic turnover by optically pumping into this band and probing the nature of resonantly enhanced protein and product vibrations. The importance of this band being in the red/NIR region of the spectrum is underscored by the fact that spectral contributions from both the [2Fe-2S] clusters and FAD are effectively eliminated, and this includes any notable background fluorescence from the FAD. The early resonance Raman studies by Hille and coworkers
[98][78], which indicated that the low-energy charge transfer band was Mo → violapterin in nature, showed that numerous vibrational modes associated with the violopterin product were observed. The lower frequency vibrations in the 250–1100 cm
−1 region were postulated to arise from the Mo coordination sphere. These studies suggested that the product was bound end-on to Mo(IV) in an Mo-O-R fashion
[98][78].
Subsequently, Kirk and coworkers used a combination of electronic absorption and resonance Raman spectroscopies to spectroscopically interrogate the nature of the Mo(IV)-product species in XO/XDH through the use of two different heavy atom congeners of the lumazine substrate
[89,90,97][71][72][77]. The two-electron oxidized 4-thioviolapterin (4-TV) and 2,4-thioviolapterin (2,4-TV) bind tightly to the Mo(IV) centers of wt-XDH and the Q102G and Q197A variants. These important studies provided deep insight into specific Moco-protein interactions. The electronic absorption and rR spectroscopies were evaluated in the context of vibrational and spectroscopic computations, and this enabled an unambiguous assignment of the intense Mo → violapterin charge transfer transition as being a Mo(xy) → violapterin (π*) metal-to-ligand charge transfer (MLCT) excitation
[97][77]. The intensity of this low-energy MLCT band derives from the Mo(xy) redox orbital being oriented orthogonal to the product ring plane, since this allows for strong overlap between the Mo(xy) orbital and the π* orbitals of the thioviolapterin product molecules
[89,90,97][71][72][77]. Thus, this MLCT transition can be described as a one-electron promotion from the doubly occupied Mo(xy) orbital to the LUMO of the product, and this effectively produces a hole on the Mo center (e.g., a formal Mo(V) center with the transfer of a full electron).
2.3. What Is Known about Pterin Oxidation State and Pterin Redox Reactivity in Moco
2.3.1. Earliest Redox Studies on PDT in Molybdenum Enzymes
Rajagopalan was the first to probe the redox state of Moco in several studies initiated shortly after his proposal of its tetrahydropterin structure using detailed absorption spectral analyses
[106,107][83][84]. The absorption spectrum of XO includes a 300 nm absorption consistent with either a tetrahydropterin or an unstable quinonoid dihydropterin, but it eliminated the possibility of a 7,8-dihydropterin structure. Oxidation of sulfite oxidase and xanthine oxidase by the redox dye dichlorophenylindophenol (DCIP) showed a 2e
−/2H
+ reaction occurred at the PDT to produce a fully oxidized pterin that was identified by electronic absorption spectroscopy, and this result indicated that Moco in both XO and SO possesses a pterin at the dihydro-level of reduction
[106,107][83][84]. On the basis of extensive reactivity studies in the Rajagopalan labs
[106[83][84],
107], the native state of the pterin of Moco in XO was proposed to be a quinonoid dihydropterin, whereas SO was argued to have a different tautomeric dihydropterin structur.
2.3.2. Redox Studies on Pyranopterin
Following the discovery of the pyranopterin structure of PDT
[76,77][63][64], Burgmayer et al. investigated the redox behavior of a synthetic reduced pyranopterin in reactions with either DCIP or ferricyanide to corroborate the Rajagopalan studies
[92,111,112,113][85][86][87][88]. The work demonstrated that a reduced tetrahydropyranopterin (i.e., the pyrano-dihydroneopterin in
Figure 74) reacted as a dihydropterin; that is, it required 1 eq DCIP or 2 eq ferricyanide to generate the oxidized pterin product neopterin (
Figure 74).
Figure 74. Redox reactivity of a model pyranopterin.
2.3.3. Role of PDT Oxidation State in Reductive Activation of DMSO Family Enzymes
It is well known that enzymes within the DMSOR family are isolated as heterogeneous samples that can be activated by a prereduction step that reduces unknown species. A recent investigation of such a reductive activation in nitrate reductase concludes with the proposal that it is the pterin of PDT that requires reduction [115,116,117]. Dissimilatory Redox reactivity of a model pyranopterin.
2.3.3. Role of PDT Oxidation State in Reductive Activation of DMSO Family Enzymes
It is well known that enzymes within the DMSOR family are isolated as heterogeneous samples that can be activated by a prereduction step that reduces unknown species. A recent investigation of such a reductive activation in nitrate reductase concludes with the proposal that it is the pterin of PDT that requires reduction [89][90][91]. Dissimilatory E. coli nitrate reductase (Ec Nar) was studied using protein film voltammetry to obtain kinetic parameters for the reductive activation [115]. Based on the kinetic analysis, there are two inactive species in equilibrium in the Nar enzyme, and only one of these is reductively activated by sodium dithionite. Furthermore, it is proposed that the equilibrium involves the cyclization of an open pterin form of PDT to a cyclized pyranopterin form of PDT prior to the reduction step that produces the active Nar enzyme.
nitrate reductase (Ec Nar) was studied using protein film voltammetry to obtain kinetic parameters for the reductive activation [89]. Based on the kinetic analysis, there are two inactive species in equilibrium in the Nar enzyme, and only one of these is reductively activated by sodium dithionite. Furthermore, it is proposed that the equilibrium involves the cyclization of an open pterin form of PDT to a cyclized pyranopterin form of PDT prior to the reduction step that produces the active Nar enzyme.
2.3.4. Pterin Protein Environment in DMSOR Family Enzymes Correlates with Mo Reduction Potential
Nitrate reductase NarGHI from
E.
coli was the first enzyme identified by protein crystallography to possess a Moco structure with one pyranopterin (proximal) and one bicyclic, pyran-opened pterin (distal) (
Figure 21d)
[81][68]. Variants were made at amino acid residues having H-bonding interactions at the O atom of the open, distal PDT pterin (
Figure 21d) to assess the effect on the Mo redox potential E
m [118][92]. It was found that when Ser719 was replaced by alanine, there was very little effect on Mo E
m, whereas the H1163A and H1184A variants caused large effects (ΔE
m values of −88 and −36 mV, respectively). On this basis, it was proposed that a charge transfer relay involving both His residues and three water molecules regulates the protonation state of the pyran-OH and thereby the Mo reduction potential. This charge relay was also proposed as initiating the pyranopterin ring opening reaction of the distal PDT via proton abstraction. A second mutation investigated the amino acid bridging the proximal and distal pterins at their N5 atom positions within each pterin. For NarGHI nitrate reductase and most members of the DMSOR family, this bridging residue is a histidine (His1092 in
Figure 43d), whose H-bonding to the proximal PDT at pterin N5 is believed to maintain the reduced pyranopterin structure. Alanine variants of His1092 and His1098 also caused large ΔE
m values of −143 and −101 mV, respectively. The results of the work support the hypothesis that changes in the pterin component of the PDT, both in terms of its oxidation state and its structure (or tautomeric form), can affect the Mo reduction potential. This modulation of the reduction potential may be used to tune an enzyme to function with a variety of substrates, th
3. What Has Been Learned about Moco from Model Studies Directly Probing PDT-Mo Interactions?
3.1. Studies That Define “Simple” Mo-Ditholene Interactions
3.1.1. Tp*MoO(bdt)
Some of the first comprehensive spectroscopic studies on oxo-molybdenum dithiolene model complexes were performed by Kirk and Enemark on Tp*MoO(dithiolene) complexes (Tp* =
tris-(3,5-dimethylpyrazolyl)hydroborate; dithiolene = bdt, tdt, qdt) (
Figure 85)
[2,84,91,122][2][93][94][95]. The work showed evidence for low-energy dithiolene → Mo LMCT transitions that indicated a three-center, pseudo-σ, Mo(xy)—S(dithiolene) bonding interaction is present in this system. From an electron transfer viewpoint, these results supported the hypothesis that in-plane Mo-S covalency could be important in modulating active site reduction potentials by destabilizing the Mo(xy) redox orbital in mono-oxo sites. For mono-oxo enzyme active sites
[84[93][96][97],
123,124], the strong ligand field produced by the Mo≡O bond orients the Mo(xy) orbital to be orthogonal to this bond, with implications for both atom and electron transfer reactivity
[123][96]. Thus, if the pyranopterin component of the PDT is involved in electron transfer regeneration of catalytically competent active sites, there must be a long-range superexchange pathway that couples the Mo(xy) redox orbital into the PDT
[84,110,125][93][98][99]. The low-frequency rR spectra of these key molecules show two important totally symmetric vibrational modes that can be described: S-Mo-S stretching and bending (
Figure 85, bottom left)
[2,84,91,122][2][93][94][95]. Since these vibrations have the same symmetry, they can mix to yield the vibrational modes given at the bottom right of
Figure 85.
Figure 85. Top: Tp*MoO(dithiolene) first-generation model complexes that have been extensively probed spectroscopically using a combination of MCD, electronic absorption, photoelectron, electron paramagnetic resonance, and resonance Raman spectroscopies. Bottom: Symmetry coordinates for two totally symmetric low-frequency normal modes and their respective linear combinations.
3.1.2. Remote Charge Effects on Oxygen Atom Transfer Reactivity
Differences in the electron-donating ability of the PDT, which could result from S-fold distortions, contributing thiol-thione resonance forms, PDT protonation, etc., possess the potential to affect the rates of oxygen atom transfer reactivity in pyranopterin Mo enzymes. Recent model compound studies have shown that changing the molecular charge by a single unit at a position remote from the Mo ion can have dramatic effects on thermodynamic parameters and reaction kinetics related to oxygen atom transfer reactivity
[126][100]. In comparing Tp*Mo
VIO
2Cl with [Tpm*Mo
VIO
2Cl]
1+, differences in their respective molecular charges arise from a single atom substitution (N → C). The change in charge at the virtual parity of their geometric structures leads to a dramatic +350 mV change in the Mo
VI/Mo
V reduction potential. A comparative analysis of the frontier molecular orbitals and electrostatic potential energy surfaces between Tp*Mo
IVO
2Cl and [Tpm*Mo
IVO
2Cl]
1+ showed that the remarkable shift in the reduction potential can be explained by a stabilization of the [Tpm*Mo
IVO
2Cl]
1+ LUMO. This LUMO stabilization results in an increase in the oxygen atom transfer reaction rate by several orders of magnitude, and the observed rate acceleration was accompanied by a larger thermodynamic driving force in accordance with the Bell-Evans-Polanyi principle. Thus, the Mo reduction potential in the enzymes can be modified by a few hundred mVs with changes in charge that are remote from the Mo center. This charge effect study conclusively showed that the structural changes that accompany charge changes are likely to be difficult or even impossible to observe in the enzymes using protein X-ray crystallography.
3.1.3. Mo-Dithione Interactions Relevant to Molybdoenzymes
Although the Mo ion redox cycles between the Mo(IV) and Mo(VI) states in most molybdoenzymes, with one-electron nitrite to NO• and the tungstoenzyme-catalyzed non-redox hydration of acetylene being notable examples
[28[28][29][30][31][32][101][102][103][104],
29,30,31,32,127,128,129,130], the PDT has not been shown to be redox active in catalysis
[48], although it is capable of storing up to six redox equivalents. Two of these equivalents are localized on the dithiolene, and four are localized on the pterin. Spectroscopic and electronic structure studies on [Mo
4+O(
iPr
2Pipdt)
2Cl][PF
6] (Pipdt =
N,
N-piperazine-2,3-dithione) have been used to explore the potential non-innocence of the dithiolene in PDT
[131][105]. The electronic absorption spectrum of this complex is unusual for a Mo(IV) complex in that it possesses a relatively intense (ε ~ 1400 M
−1cm
−1) low-energy (E ~ 13,500 cm
−1) metal-to-ligand charge-transfer (LMCT) band. Typically, low-energy LMCT transitions in mono-oxo Mo sites are not observed due to the large terminal oxo-derived ligand field splitting of the t
2g orbitals and the double occupancy of the lowest energy Mo(xy) orbital. However, if the dithiolene is oxidized to a dithione and ligand acceptor orbitals are available, low-energy MLCT may be observed. This is the case for [Mo
4+O(
iPr
2Pipdt)
2Cl]
1+, where the MLCT has been assigned as Mo(xy) → dithione(π*) HOMO → LUMO transition based on spectral computations and resonance Raman enhancement of bands with C–C and C–S stretching characters. The
iPr
2Pipdt ligand was described in valence bond terms using a natural bond orbital approach to be comprised of a hybrid of contributing dithione (63%) and di-zwitterionic dithiolene (37%) resonance structures. The π-acceptor character of this type of dithione was also shown in studies on MoO(SPh)
2(
iPr
2Dt
0) (
iPr
2Dt
0 = N,N′-isopropyl-piperazine-2,3-dithione), where an intense thiolate → dithione ligand-to-ligand CT band was assigned at ~18,000 cm
−1. This assignment was based on spectroscopic computations and resonance Raman enhancement of a 378 cm
−1 vibration that was shown to possess dithione ligand S−Mo−S + C−N stretch character. The π-acceptor character of the ligand is also exemplified in a dramatic dithione ligand fold angle distortion of 70°, which derives from the pseudo-Jahn–Teller effect
[91][94].
3.1.4. Donor-Acceptor Quinoxaline Dithiolene Ligands
Non-innocent metal-ligand redox behavior in molybdenum dithiolene complexes that possess ligands comprised of nitrogen heterocycles was initially reported by Pilato in a series of pyridinyl- and quinoxalinyl- dithiolene complexes of molybdenum of the type Cp
2Mo(S
2C
2(heterocycle)H)
[132][106], foreshadowing the results scholars would obtain using pyranopterin dithiolene ligands. The potential for such non-innocent behavior in the molybdenum cofactor was originally demonstrated using a ligand (pyrrolo-S
2BMOQO) comprised of an N-heterocycle (quinoxaline) that is appended to a dithiolene fragment that was covalently bound to a Mo(IV) ion
[133][107]. These quinoxalyldithiolene ligands effectively served as first-generation models for how the PDT may function in Moco. Tp*MoO(pyrrolo-S
2BMOQO) is formed from the dehydration of TEA[Tp*MoO(S
2BMOQO)] (TEA = tetraethylammonium; Tp* = hydrotris(3,5-dimethylpyrazolyl)-borate), where an intramolecular cyclization within the S
2BMOQO ligand occurs. A combination of DFT computations, which were interpreted in the context of resonance Raman and electronic absorption spectroscopies and complemented by X-ray crystallographic studies, revealed that an asymmetric dithiolene chelate was present in Tp*MoO(pyrrolo-S
2BMOQO). Additionally, it was shown that this five-membered MoS
2C
2 chelate ring possessed considerable thione-thiolate character. A valence bond description was used to describe the observed Mo-ligand chelate ring thione-thiolate bonding character, and this analysis showed that there were two dominant resonance structures that contribute to the electronic structure description (
Figure9 6).
Figure 96. (left) Bond lengths determined from X-ray crystallography for Tp*MoO(pyrrolo-S2BMOQO). (right) Contributing resonance structures for the ligand showing dominant dithiolene (
A) and thione-thiolate (
B) structures.
3.2. Model Systems That Incorporate Both Dithiolene and Pyranopterin Structures on Molybdenum
The first Moco model compound to successfully incorporate a pterin dithiolene ligand on a Mo(4+) ion was reported from Pilato’s labs in 1991
[134,135][108][109]. These Cp
2Mo(IV)(S
2C
2(pterin)(COMe)) systems were constructed on a
bis-cyclopentadienyl-Mo(IV) structure that lacked a terminal oxo ligand. Limited studies of this molecule demonstrated one electron oxidation to Mo(V) and reactivity towards acids. Subsequently, Garner and coworkers
[136][110] reported a pterin dithiolene ligand in a related complex, CpCo(S
2C
2(pterin)(H)).
More recently, a number of pterin- and quinoxaline-dithiolene Mo compounds have been designed in the Burgmayer labs. Each is synthesized from the reaction of a molybdenum tetrasulfide precursor [Tp*MoS(S
4)]
− with a suitably substituted pterinyl- or quinoxalyl-alkyne, as depicted in
Figure 107. The pivaloyl group added to the exocyclic amine group of the pterin overcomes the notorious insolubility of pterins. Those complexes shown in
Figure 107 have been studied in detail to provide considerable insight about the (pterin-dithiolene)-Mo system.
Figure 107. Pterin-dithiolene model compounds for Moco.
4. What We Have Learned from Model Systems That Pertain to Moco in Enzymes? An Update
4.1. Previous Roles of the Pterin Defined
The PDT is the least understood critical component of Moco in all pyranopterin Mo and W enzymes. This is remarkable, given the complex biosynthetic pathway and the ubiquitous nature of PDT in all the enzymes. At present, there is good evidence for three key roles of the PDT in catalysis, and these include functioning as an anchor for the Mo/W ion, serving as a through-bond electron transfer conduit for obligatory one-electron transfers in the electron transfer half reaction of the enzymes, and, with respect to the dithiolene component of the PDT, enabling redox potential modulation of the active site. Early protein crystallography studies indicated that the dithiolene component of the PDT could be either completely or partially dissociated from the metal ion, suggesting a role for this behavior in the catalytic cycles of some pyranopterin-containing enzymes. However, it is now understood that these structures represent active sites that have been damaged by the high flux of the X-ray beam during data collection. Thus, bidentate coordination of the dithiolene moiety is necessary for catalysis. Studies by Hille and coworkers were among the first to suggest a role of the PDT in electron transfer regeneration of the active site in xanthine oxidase (
Figure 23) [148,149,150]. An extensive amount of model compound studies, including those detailed here, strongly suggest that the nature of the dithiolene ligand, remote charge effects, and the degree of the sulfur-fold angle can all affect the effective nuclear charge of the metal ion to drive large changes in the Mo redox potential. However, new roles for the PDT have been suggested that involve different oxidation states of the pterin component of the PDT, hydrogen bonding and proton transfer involving the pterin, and the role of thione-thiol resonance structure contributions to the electronic structure of the dithiolene chelate.8) [111][112][113]. An extensive amount of model compound studies, including those detailed here, strongly suggest that the nature of the dithiolene ligand, remote charge effects, and the degree of the sulfur-fold angle can all affect the effective nuclear charge of the metal ion to drive large changes in the Mo redox potential. However, new roles for the PDT have been suggested that involve different oxidation states of the pterin component of the PDT, hydrogen bonding and proton transfer involving the pterin, and the role of thione-thiol resonance structure contributions to the electronic structure of the dithiolene chelate.
Figure 238. The electron transfer chain in XO, indicating a vectorial pathway for electron egress involving the Mo ion, the PDT, two spinach ferredoxin type 2Fe2S clusters, and a flavin. Electrons exit the enzyme at FAD.
4.2. Recent Results Define New Roles for the PDT in Catalysis
An early hypothesis that the PDT might serve to modulate the Mo redox potential has now been demonstrated and quantified for pterin-dithiolene model complexes. The partially reduced pyranopterin structure is electron withdrawing with respect to the Mo-dithiolene unit, and this results in the Mo(V) ion being a stronger oxidant in these systems. The redox-flexible dithiolene responds by accessing a partially oxidized thione/thiolate resonance structure. Pyran ring cleavage severs the PDT electron conduit, as the dithiolene is now electronically isolated from the pterin, which can now rotate out of the Mo-dithiolene plane. Reduction of pyranopterin is expected to also decrease this electronic relay from Mo-dithiolene to pterin as the sp
3-hybridized bridgehead carbon between pterin and dithiolene interrupts extended π-conjugation in the PDT.
It is worth emphasizing that the degree of thione-thiolate character in the chelate ring is Mo oxidation state-specific, with thione-thiolate character being observed only for Mo(IV)
[86,92,139][85][114][115]. Higher oxidation states of Mo typically result in the dithiol resonance form of the ligand dominating and an increase in the chelate ring S-S fold angle to reduce the effective nuclear charge on the Mo ion
[79,88,91,122][66][94][95][116].