Mechanistic aspects of molybdenum-containing enzymes

In the past several years, the number of enzymes known to possess mononuclear molybdenum centers in their active sites has increased significantly, and well over 50 such enzymes have been identified that catalyze a variety of hydroxylation, oxygen atom transfer and other oxidation-reduction reactions. Many of these enzymes have been isolated from either obligate anaerobes or facultative anaerobes grown under anaerobic or microaerobic conditions, and are involved in a variety of anabolic, catabolic and energy-conserving metabolic pathways. There are now several enzymes having known crystal structure, and this new structural information has provided the basis for an increasingly detailed understanding of the mechanisms of action of these enzymes. In the present review, an overview of our present understanding of the mechanism of action of these enzymes will be presented, and those of a few selected enzymes will be considered in greater detail. Several alternative classification schemes have been suggested for these molybdenum-containing enzymes, but for the purposes of the present discussion they will be considered to fall into three families, based on the structures of their molybdenum centers in their oxidized Mo(VI) state (Fig. 1) [1]. These families include: (i) the molybdenum hydroxylases, a large and broadly dispersed family of enzymes that possess an MoOS unit and catalyze the hydroxylation of a broad range of aldehydes and aromatic heterocycles; (ii) the eukaryotic oxo transferases, a family that at present includes only sulfite oxidase and the assimilatory nitrate reductases, enzymes which possess an MoO2 unit in their active sites and which catalyze oxygen atom transfer to or from a substrate; and finally (iii) a diverse group of prokaryotic enzymes that catalyze either oxo atom transfer or other oxidation-reduction reactions. As will be seen below, however, some may catalyze significantly more complex (and interesting) reactions. Enzymes of this last group have a common molybdenum center structure in which the metal is coordinated by a pair of dithiolene ligands contributed by an unusual pterin cofactor [2]. This cofactor is common to all the mononuclear molybdenum (and tungsten) enzymes, but whereas the first two families possess only a single equivalent bound to the metal, in members of the third family two equivalents of the cofactor coordinate to the metal. It should be emphasized that this final group is structurally more diverse than the first two, and can be subdivided into enzymes that possess an Mo-O(Ser), Mo-S(Cys) or Mo-Se(Se-Cys) group contributed by the polypeptide; in addition, some of these enzymes appear to possess an Mo=S group rather than the more commonly encountered Mo=O (see [1] for a review). Active site structures of the mononuclear molybdenum enzymes. From left to right, the molybdenum hydroxylases, the eukaryotic oxotransferases and the prokaryotic oxotransferases (and related enzymes). Active site structures of the mononuclear molybdenum enzymes. From left to right, the molybdenum hydroxylases, the eukaryotic oxotransferases and the prokaryotic oxotransferases (and related enzymes). The pterin cofactor that is common to all these enzymes has the structure shown in Fig. 2. In addition to the pterin itself, there is an additional pyran ring that possesses an exocyclic dithiolene moiety by which the cofactor coordinates the metal, plus a short phosphorylated side chain. In eukaryotic systems the cofactor possesses the structure shown, but in prokaryotic systems the cofactor is usually found elaborated as the guanine, adenine, cytidine or hypoxanthine dinucleotide [2]. The cofactor has been called ‘molybdopterin’ or ‘moco’ (for ‘molybdenum cofactor’), although this nomenclature is misleading for two reasons: (i) the term refers to the organic cofactor itself without molybdenum bound to it; and (ii) the identical cofactor is found in tungsten-containing enzymes (where it has occasionally been referred to as ‘tungstopterin’). The term pyranopterin has also been suggested, and this has the advantage of lacking the confusion associated with either of the above terms. Structure of the pterin cofactor associated with molybdenum- and tungsten-containing enzymes. In prokaryotic enzymes, the cofactor exists as the dinucleotide of guanine, adenine, cytidine or hypoxanthine. Structure of the pterin cofactor associated with molybdenum- and tungsten-containing enzymes. In prokaryotic enzymes, the cofactor exists as the dinucleotide of guanine, adenine, cytidine or hypoxanthine. With this general background concerning the active site structures of these enzymes, we will now turn to a consideration of their mechanisms of action. We focus here on work done since 1990, referring the reader to recent reviews for work done prior to this time [1,3]. As indicated above, the molybdenum hydroxylases constitute a large and widely dispersed group of enzymes. These enzymes are unique among the many biological systems that catalyze hydroxylation of either aromatic or aliphatic substrates in that water rather than dioxygen is the source of the oxygen incorporated into the product, and reducing equivalents are generated rather than consumed in the course of the reaction. In humans, there is a xanthine oxidoreductase (or dehydrogenase) and several aldehyde oxidases that have specific tissue distributions. In bacteria there are a large number of enzymes that are expressed when organisms are grown on various heterocyclic compounds as sole carbon source, and other enzymes that act on aldehyde substrates (taking these to the corresponding carboxylic acid). The sole member of this family that has been examined crystallographically is the aldehyde oxidoreductase from Desulfovibrio gigas[4,5], an enzyme that is involved in the generation of H2 using reducing equivalents obtained from aldehyde substrates. The structure of the molybdenum center of this enzyme (in the oxidized Mo(VI) form) is that shown in Fig. 1, which represents a composite of the X-ray crystal work in conjunction with the results of X-ray absorption spectroscopy (XAS) studies of other members of this family (most notably the xanthine oxidase from cow’s milk). The coordination geometry inferred from the crystallographic work is that of a distorted square pyramid, with an apical Mo=S and the two sulfurs of the pterin cofactor, an Mo=O and a fifth oxygen ligand defining the basal plane [4]. This last ligand has been formulated as a water molecule on the basis of an Mo-O distance of ∼2.2 Å, which is in the range for a metal-bound water (as opposed to hydroxide). As has been pointed out recently, however, this distance must be regarded as an upper limit for the true Mo-O distance, owing to series truncation artifacts around the heavy metal ion that occur in generation of the electron density map for the protein – these artifacts may in some cases lead to an overestimation of the true Mo-O distance by as much as 0.3 Å[6]. It remains to be seen whether this is a concern in the case of the molybdenum center of the aldehyde oxidoreductase. A hydroxide ligand is preferred here on the following additional grounds: (i) the XAS work with xanthine oxidase indicates a (weak) Mo-O (or N) scattering at a shorter distance of 1.98 Å rather than the crystallographically observed 2.2 Å; (ii) in small inorganic complexes of Mo(VI) oxygen ligands are most frequently found as the fully deprotonated Mo=O, occasionally as the singly protonated Mo-OH, but only extremely rarely as the fully protonated Mo-OH2; and (iii) computational studies have shown that an Mo-OH is expected to be considerably more stable from a thermodynamic standpoint than an Mo-H2O in the specific coordination geometry found in the active site of the aldehyde oxidoreductase. The mechanistic significance of the degree of protonation of this oxygen ligand lies in recent work indicating quite strongly that it is this oxygen which is incorporated into the hydroxyl group of the product [7–9], with the site subsequently regenerated by hydroxide/water from the solvent for the next catalytic cycle. It has been known for some time that these enzymes (as represented by xanthine oxidase) possess such a catalytically labile oxygen, but in the absence of specific knowledge of an Mo-OH(2) ligand to the metal, the Mo=O has been considered the likely candidate for this oxygen. In light of this more recent work, however, a mechanism can be formulated in which the substrate becomes hydroxylated by nucleophilic attack of a metal-bound hydroxide on the substrate, followed by hydride transfer from the carbon to become hydroxylated to the Mo=S group (known from earlier XAS work to become protonated upon reduction of the molybdenum center) (Fig. 3). Proposed mechanisms for the molybdenum hydroxylases. The reaction with aldehydes (top) and xanthine (bottom). Proposed mechanisms for the molybdenum hydroxylases. The reaction with aldehydes (top) and xanthine (bottom). An active site glutamate (Glu-869 in the aldehyde oxidoreductase, conserved among molybdenum hydroxylases), has been proposed to act as a base catalyst in the formation of the C-O bond of product, and possibly also in the displacement of the molybdenum-bound product that is generated by this step of the reaction [5]. Indeed, the pH dependence of the reaction of the bovine xanthine oxidase provides clear evidence for the participation of an active site base in this portion of the catalytic sequence with both heterocyclic substrates [10] and simple aldehydes [9]. It thus appears that these enzymes catalyze hydroxylation of both general classes of substrate, aldehydes and aromatic heterocycles, by this same basic mechanism. Another aspect of catalysis that has been much informed by the crystal structure of the aldehyde oxidoreductase has to do with the manner in which reducing equivalents, once introduced at the molybdenum center, are removed from the metal to complete the catalytic sequence. In the structure of the aldehyde oxidoreductase, the amino group of the pterin cofactor that lies distal to its dithiolene group is found to be hydrogen-bonded to a cysteine residue that is a ligand to one of the two 2Fe/2S iron-sulfur clusters found in the enzyme [4]. On the basis of the physical disposition of the cofactor, it seems very likely that it is involved in facilitating electron transfer from the molybdenum to the iron-sulfur center, the physiological direction of electron flow in the enzyme. A loosely similar arrangement is found in other molybdenum enzymes, but the situation is not a general one: thus the cofactor is found to intervene between the molybdenum and 4Fe/4S center of the Escherichia coli formate dehydrogenase H (a bis-dithiolene enzyme) [11] but not between the molybdenum and heme of chicken sulfite oxidase [12]. The molybdenum-containing oxo transferases from eukaryotic sources, sulfite oxidase from vertebrate liver or the assimilatory nitrate reductase from higher plants, algae, yeast and fungi, catalyze a chemistry that is fundamentally more straightforward than that of the molybdenum hydroxylases. The enzyme from this group that has been structurally characterized by X-ray crystallography is the sulfite oxidase from chicken liver [12]. As indicated in Fig. 1, the coordination geometry of the molybdenum center of this enzyme is distorted square pyramidal, with an apical Mo=O and a basal plane defined by the two dithiolene sulfurs of the pterin cofactor, the remaining Mo=O and a cysteine thiolate contributed by the polypeptide. The present evidence suggests quite strongly that these enzymes alternate between Mo(VI)O2 and Mo(IV)O centers in the course of their catalytic sequences, directly analogous to the now well-characterized chemistry of small molybdenum complexes thought to represent functional models of these enzymes [13,14] (Fig. 4). Catalytic cycles for molybdenum model complexes (A–C) and sulfite oxidase (D). In all cases, the system cycles between dioxo Mo(VI) and mono-oxo Mo(IV) states. Catalytic cycles for molybdenum model complexes (A–C) and sulfite oxidase (D). In all cases, the system cycles between dioxo Mo(VI) and mono-oxo Mo(IV) states. These model systems are found to be quite robust and able to turn over repeatedly in the presence of a suitable oxygen atom donor/acceptor pair. In light of the large negative entropy of activation associated with the reaction of the Mo(VI)O2 model with a phosphine acceptor, an associative reaction mechanism having a highly ordered transition state has been proposed, with a phosphine lone electron pair undertaking nucleophilic attack on one of the Mo=O groups to initiate the reaction. Indeed, this chemistry has been successfully examined computationally, and the reaction is understood in considerable detail [15]. Briefly, the phosphine is thought to approach the MoO2 unit at 90°, thereby maximizing orbital overlap between the phosphine lone pair and a π* orbital of the Mo=O group, followed by rotation about the Mo-O bond. This rotation results in formation of an O-P σ bond and reduction of the Mo-O bond order to 1; at the same time the stability of the remaining Mo=O bond becomes significantly stronger due to increased π-backbonding to the metal, with the bond order increasing to 3. It is this increase in bond order of the remaining Mo=O group – as the first is lost in the course of the reaction – that makes the as much of the of one of the quite stable Mo=O is in the stronger bond that This an in the of these MoO2 A mechanism is likely for the reaction of sulfite oxidase with sulfite (Fig. in which the is attack on an Mo=O group by a substrate lone pair. It is for that of the two of sulfite results in a increase in from the dependence of the of enzyme reduction upon reaction with but the of the reaction is from that observed with sulfite The results that the groups significantly to substrate but are not for of the the is that the substrate not coordinate directly to the molybdenum center in order for the reaction to This is with the known substrate site of the which of a of three around a of to sulfite (in the crystal structure this found to be by from the [12]. It is to be however, that the substrate and that a complex also with this in addition to between the substrate and enzyme must occur between substrate and active site in the On the basis of the present it is likely that the reaction mechanism of nitrate reductase in the of that for sulfite The that the physiological direction for the is the enzyme an oxidase) that for the is the enzyme a has to do with the of the and which the of the as oxo and This has been on a and a thermodynamic by which the direction of reaction for a pair of can be there must be substrate between sulfite oxidase and nitrate reductase is able to as a oxidoreductase, although this reaction is it is known from XAS studies that the molybdenum centers of the two enzymes are very similar and the of the considerable amino sequence [1,3]. sulfite oxidase nitrate reductase has been examined from a mechanistic standpoint in the detail xanthine and the to which their reaction mechanisms are related remains to be Many of the from that constitute the third family of molybdenum-containing enzymes catalyze oxygen atom transfer that are similar to those by sulfite oxidase and the nitrate the nitrate and reductase all catalyze that oxygen atom that these enzymes likely possess only a single Mo=O group in their oxidized (see the is that they alternate between mono-oxo Mo(VI) and Mo(IV) and aspect to their reaction mechanisms (Fig. It has been that reductase is able to catalyze oxygen atom transfer between and a phosphine oxo a catalytically labile oxygen site and the found to be present in the Mo=O group on the basis of work XAS results with both oxidized and reductase are with the enzyme as indicated in Fig. A proposed reaction mechanism for The enzyme is proposed to between mono-oxo Mo(VI) and Mo(IV) states. A proposed reaction mechanism for The enzyme is proposed to between mono-oxo Mo(VI) and Mo(IV) states. The X-ray crystal structure of reductase has been by three with the enzyme of and the other two with the related enzyme) and the results a considerably more than that indicated above for the reaction mechanism of the enzyme. the structures of the in all three structures are the molybdenum centers are quite In the first structure by and the molybdenum center of the oxidized enzyme has a structure as shown in Fig. 1, with a distorted coordination reduction the Mo=O group is with the chemistry shown in Fig. of one of the two dithiolene groups is in the of the is quite at Å the of the other dithiolene are and the other has been lost with a distance of These have been to a in the dithiolene in which the at distance from the molybdenum possesses considerable In the crystal structure of oxidized enzyme by and with the the first dithiolene is present in the same as in the structure of the but the is found to have from the the for the dithiolene sulfurs are and Å, with the ligand coordination by a Mo=O group at In the final structure by and and of the enzyme the ligand coordination of the molybdenum in the oxidized enzyme is from to in a complex The additional ligand is as a Mo=O, that the oxidized enzyme is dioxo Mo(VI) rather than present it is to such structures to one in an reaction and much work remains to be done to the of these structures to the catalytic sequence of the enzyme. The other enzyme from this family that is of known crystal structure is the formate dehydrogenase H of This enzyme is of the and the possess only a molybdenum center) also has a 4Fe/4S The oxidized enzyme has the structure shown in Fig. 1, with a distorted coordination The oxygen found at 2.2 Å, however, and has been to an Mo-OH rather than Mo=O This distance has been by and it seems likely that one of the of for in the protein ligand to the metal is to the Mo=O bond. This is not as the increased of the bond in the series be expected to have the of other as the state of the metal is dehydrogenase H the of formate to the reducing equivalents thus generated into the of the enzyme catalyze the reaction by hydroxylation of formate to followed by to it appears to by of formate and is thus an of a molybdenum enzyme that not catalyze oxygen atom On the basis of the crystal structure of both the oxidized and the a mechanism has been proposed in which formate to the metal by the Mo-OH, with reduction of the metal with the participation of an active site residue protonation of the has been suggested to in the course of the prior to protonation of the This reaction not oxygen atom transfer from the standpoint of its and it is that formation of an bond in the course of the reaction. As above, in most prokaryotic enzymes the oxygen atom of the group from The in these is the that both water (or an and the substrate are which not with a of one of the is for such an is that the molybdenum coordinates a water molecule or a hydroxyl ion and it to the substrate as a of the substrate occur by which with water either directly or by of the molybdenum In the case of xanthine oxidase which is one of the most molybdenum-containing enzymes (see and Fig. about a mechanisms have been proposed In some of these the molybdenum-bound (or is as a in the oxygen is from an Mo=O as an in first a bond is with the substrate, followed by addition of in of or these mechanisms have been in recent reviews the of molybdenum in is a of there are some molybdenum-containing enzymes that catalyze without hydroxyl or oxo and a oxidoreductase and clusters but other The enzyme a large number of to the corresponding recently, that a formate dehydrogenase H from coli formate to carbon without hydroxyl transfer (see 4). The of the molybdenum cofactor in these is but it is that molybdenum-containing enzymes can similar to those by or enzymes. in detail with This anaerobic and carboxylic to three of The are first and is into which and to which is oxidized and to three The reaction to represents a It shown that in at to the reaction The enzyme (and equivalent to physiological reaction in the presence of by higher to and specific of various that this enzyme transfer hydroxyl groups also from other and transfer to other from these a reaction mechanism suggested which the reaction as a hydroxyl transfer between two and to and a new molecule of which as a with a new to Fig. This is an to the but is not unique in enzyme also the enzymes and advantage of their and It shown by an with that there is of oxygen from water in the and that the hydroxyl groups are only between the substrates and Proposed of as in the reaction. Proposed of as in the reaction. From a at the reaction it is that a hydroxyl transfer from the to occur of suggested a catalytic mechanism as in Fig. as a first step an of to the corresponding is the now to a nucleophilic attack by an of and the between substrate and can be to the product and the of of the with of a molecule the catalytic cycle. mechanism for the reaction a mechanism for the reaction a this mechanism we the two and (Fig. these in a the compounds are to oxygen and not be obtained in with the in those In our both successfully by and the compounds characterized by and and spectroscopy and to be of the of the in the absence of the in the presence of 50 into three new compounds of the one of which identified by its time and as This and with time only in the presence of the third one is and is an product of the The may have to and the of (Fig. can be for the (i) the is not an of the catalytic (ii) there is a for its transfer to the active or (iii) the enzyme is in the oxidized In the reaction in the presence of as a reducing In the first of into very but the a small but of with of the of the once A similar obtained with the in the presence of as reducing These results that the and do with but their a as it found that a dinucleotide cofactor and clusters and to the reductase family The reaction mechanism above (Fig. is in with the widely of molybdenum-containing enzymes to which the molybdenum center is the and of hydroxyl or oxo an alternative mechanism such a for the molybdenum cofactor is in Fig. to this an by the group at of The following are which not that they occur on the enzyme. is an it is not it is to the reaction cycle. mechanism for the reaction of the hydroxyl group by the molybdenum mechanism for the reaction of the hydroxyl group by the molybdenum also a mechanism in which the substrate by to the and thus may be by the nucleophilic group coordinated to the molybdenum center (Fig. This to as which may turn at the active site and with the molybdenum center in a as in the first of the cycle. for the of a of the may be that its is only by a from the enzyme it may with for the active a of the molybdenum center in the is analogous to the xanthine dehydrogenase are to either one of the reaction mechanisms suggested mechanism for the reaction of the substrate by to an and nucleophilic hydroxylation by a mechanism for the reaction of the substrate by to an and nucleophilic hydroxylation by a is unique it not catalyze a hydroxylation of as do the hydroxylases above the is as to the of that is as to the reaction to (Fig. is generated in the with other and also oxygen have provided evidence that they all can act as hydroxyl to initiate the of to by formation of a catalytic of and since one of these oxidized is to the under these do not a to this have been that are involved in anaerobic of aromatic compounds by the hydroxylation of or of to These may be by enzymes, but since these enzymes have not been a of molybdenum remains a of