Keywords

1 Introduction

Vanadium is the second most abundant metal in seawater at a concentration of about 35 nM [1] and universally [2] distributed in the soil. The metal oxide is the prosthetic group in the vanadium haloperoxidases from seaweeds and in a group of fungi, the dematiaceous hyphomycetes. Vilter [3] was the first to demonstrate evidence for the involvement of vanadium in the vanadium-dependent haloperoxidase. He showed that the bromoperoxidase isolated from Ascophyllum nodosum was inactivated by dialysis at pH 3.8 in phosphate buffer containing EDTA and that the enzyme was reactivated by vanadate in suitable buffers. Subsequently it was demonstrated [4, 5] that vanadate was the prosthetic group in the bromoperoxidase and that since phosphate is a structural and electronic analogue of vanadate it replaces vanadate in the enzyme [5, 6]. The vanadium chloroperoxidases with similar enzymatic and kinetic properties as the bromoperoxidases were detected a few years later in a family of terrestrial fungi [7, 8]. The oxidation state of the metal in the native form of the haloperoxidases is vanadium V that upon reduction is converted into the catalytically inactive IV state. As a d1 metal ion, vanadium IV has a single electron that is strongly coupled to the 51V nucleus (I = 7/2). This redox state of the metal can easily be detected by EPR since it gives rise to an EPR signal of either 8 or 2 sets of 8 overlapping lines. This technique allows the detection of relatively low concentrations of these enzymes [9]. Unfortunately at this redox state the enzyme is inactive. Unlike heme peroxidases the UV-VIS spectra of these vanadium enzymes show only a modest absorption in the optical spectra around 315 nm due to the bound cofactor.

Haloperoxidases catalyze the two-electron oxidation of halides (Cl−, Br−, I−) by H2O2 to hypohalous acids:

$$ {{\text{H}}_2}{{\text{O}}_2} + {{\text{H}}^{ + }} + {{\text{X}}^{ - }} \to {\text{HOX}} + {{\text{H}}_2}{\text{O}} $$
(5.1)

In fact this reaction can be regarded as an oxygen-transfer reaction from the peroxide to the halide ion. These hypohalous acids or related halogenating intermediates, such as OX−, X −3 and X+ are released from the active site during turnover and they may act nonspecifically on a variety of organic compounds (RH) that are susceptible for electrophilic attack resulting in the production of a diversity of halogenated compounds (RX).

$$ {\text{HOX}} + {\text{R}} - {\text{H}} \to {\text{RX}} + {{\text{H}}_2}{\text{O}} $$
(5.2)

In the absence of a nucleophilic acceptor a reaction may also occur between HOX and H2O2 resulting in the formation of singlet oxygen.

$$ {\text{HOX}} + {{\text{H}}_2}{{\text{O}}_2}{ \to^1}{{\text{O}}_2} + {{\text{H}}_2}{\text{O}} + {\text{HX}} $$
(5.3)

The historical nomenclature convention of the vanadium haloperoxidases is based on the most electronegative halide oxidized by these enzymes. Chloroperoxidases catalyze the oxidation of Cl− and Br− and I−, bromoperoxidases catalyze only the oxidation of Br− and I− and iodoperoxidases [10] are specific for iodide oxidation. However the distinction between the haloperoxidases is somewhat arbitrary since a bromoperoxidase may also oxidize chloride albeit with a low specificity constant.

2 Occurrence, Biological Function of Vanadium Iodo- and Bromoperoxidases and Impact of the Halogenated Products on the Environment

A striking array of iodinated and brominated products produced by natural sources is found in the biosphere and in particular the marine environment is a rich source. A majority of these products are probably formed by the enzymatic activity of the enzyme family of vanadium bromo- and iodoperoxidases. Iodo- and bromoperoxidases have been found in red and brown macro-algae from all over the world [11–13] and also in marine diatom cultures [14]. In the green macro-alga Ulvella lens probably also a vanadium dependent enzyme is present [15].

2.1 Brominated Compounds

Macro-algae and phytoplankton produce huge amounts of bromoform and other volatile brominated compounds [16] and oxidative stress as well as UV light result in the release of these compounds by seaweeds. Estimates [16] give a global annual source strength of 2.2 1011 g CHBr3 by macro-algae. The brominated compounds are ventilated to the atmosphere and may reach the lower region of the stratosphere in which they may play a natural role in the regulation of the ozone concentration [17] and references therein]. Thus, macro-algae have a dominant role in producing bromoform and other volatile brominated compounds that plays a role in the breakdown of ozone in tropo- and lower stratosphere [18, 19]. However, there are other natural sources of these halogenated compounds as well as evidenced [19] by halogen mediated ozone destruction over the biological active tropical Atlantic Ocean. Halogenated compounds are produced as shown [14] by marine diatoms that occur in all oceans. Also bromoform in the Artic Ocean [20, 21] appears to originate from phytoplankton. In some seaweed species the iodo- and bromoperoxidases are extracellularly located [22–25]. When these and other seaweeds are triggered by changes in the environment such as stress and light they produce not only HOBr [22, 25, 26] but also a variety of volatile brominated compounds [27–30]. It is well documented [31] that plant when exposed to stress such as wounding, pathogenic attack and light generate reactive oxygen species and H2O2. The vanadium bromoperoxidase will convert this rapidly into HOBr.

HOBr has a direct antimicrobial and virucidal effect [32, 33] and as shown [26] the oxidized halogen species produced by the marine alga L. digitata deactivate homoserine lactones. These lactones are important communication signals between bacteria and as a result of deactivation biofilm formation and fouling by bacteria and fungi of the surface of the seaweeds is prevented. Thus, oxidized halogens produced by the extra cellular peroxidase not only control biofouling by a bactericidal mechanism but also interfere with bacterial signaling systems.

A question that has not yet clearly been answered is how formation of HOBr by bromoperoxidases is linked to the formation of these volatile halogenated compounds. At least two mechanisms are operational. Theiler et al. [34] have proposed that HOBr generated by seaweeds reacts with ketoacids in the seaweeds. These compounds are susceptible to halogenation; the halogenated ketoacids formed are unstable and decay via the haloform reaction to a mixture of mono-, di-, and tribrominated compounds. Indeed as shown [15] incubation of oxaloacetate with BPO/hydrogen peroxide/Br− leads to formation of CHBr3 and CH2Br2. Alternatively it has been proposed [22] that HOBr produced by these seaweeds and which diffuses into the seawater reacts in an abiotic process with dissolved organic matter naturally present in seawater. The brominated compounds are unstable and decay again leading to the formation of bromoform and other compounds. This process occurs also when drinking water containing organic matter is exposed to HOCl or HOBr [35].

2.2 Iodine and Iodinated Compounds

The brown seaweeds of the genus Laminaria (kelp) are the most efficient iodine accumulators among all living species and accumulate up to 1% of their dry weight of this halogen. In coastal regions Laminariales are the major contributor to the iodine flux from the oceans through the production of volatile iodocarbons. The iodoperoxidases that are found in these seaweed species play not only a prime role in the production of iodine and iodocarbons [36] but also in the accumulation of iodide. Phylogenetic analyses [37] indicate that these vanadium IPO’s shares a close common ancestor with brown algal vanadium-dependent bromoperoxidases. As discussed in more detail later the most striking difference in the active site of vanadium IPO is the presence of an alanine, instead of an invariant serine that is present in both in vanadium CPO and vanadium BPO.

Exposure to UV light and stress of L. digitata induces an efflux of volatile iodocarbons and bursts of molecular iodine I2 [38–40] and there is a strong relation to the H2O2 released by this alga as well. It is likely that the observed bursts are a consequence of the known stress-induced production of H2O2 in L. digitata. These results strongly support the model [41] that direct biogenic emission of I2 by algae is a very important natural process in the marine boundary layer, that impacts the tropospheric photochemistry both on regional and global scales. Iodine mediated coastal particle formation also occurs which is driven by emission of molecular iodine I2 by macro algal species under stress conditions at low tide. Particle formation strongly correlates with IO produced at high concentrations following photolysis of I2 and subsequent rapid reaction with O3. These particles grow to larger size and these particles should be able to act as cloud condensation nuclei at reasonable atmospheric super saturations. Indeed, iodocarbons and iodine containing particles produced in coastal areas by seaweeds provide condensation nuclei for coastal cloud formation [42–44]. A recent study [45] shows that other seaweed species also emit I2 when stressed for example by low tide. Further this study confirms that condensable particle forming gases derive from the photochemical oxidation of I2 emitted by these plants. Figure 5.1 gives a simplified picture of the processes that may occur and are triggered by the activity of the seaweeds.

Fig. 5.1
figure 1_5

Events occurring when seaweed is exposed to stress or light. An oxidase is triggered by stress signals e.g. low tide, rain or damage of the seaweed and produces hydrogen peroxide. Apparently during photosynthesis also hydrogen peroxide is generated. The iodoperoxidase converts this into reactive HOI, which is either converted into I2 or it reacts directly with organic matter in seawater. The formed iodinated compounds ventilate to the atmosphere, are photolyzed by UV light and iodine radicals I− are generated. The iodine radicals react further in the atmosphere and lead to particle formation that acts as condensation nuclei for coastal cloud formation. A major emission flux of I2 from the algae to the atmosphere also, which are directly photolysed leading iodine radicals

Full size image

The physiological function of the production of these iodinated compounds by these seaweeds is not clear and here is considerable debate as to the role of the iodoperoxidases in the Laminaria species and the involvement in iodide accumulation. Kupper et al. proposed [46] that iodide accumulation in kelp constitutes an extra cellular protection against oxidative stress such as O3 in air or H2O2 [47] generated under stress conditions by these plants. However, the O3 concentrations in air are likely to be low. It is also possible that by using the hydrogen peroxide generated the vanadium iodoperoxidases act as a chemical defense system against invading (micro) organisms.

The accumulation and distribution of iodine in these brown algae has been the subject of many studies [46] and references therein] and has led to various mechanisms and considerable debate. Micro chemical imaging studies of iodine distribution in L. digitata and careful fixation procedures have recently shown [48] that the distribution of iodine shows a huge, decreasing gradient from the meristoderm to the apoplasm with very high levels in the external cell layers. The meristoderm contains about 80% of the total iodine content and the peripheral tissue is therefore the main storage compartment of iodine and means that iodine is easily accessible and can be mobilized for potential antioxidant activities and chemical defense. Interestingly also the highest contents of bromine are also detected in the peripheral tissue. The results also suggested that iodine is mainly chelated as labile inorganic species by apoplastic molecules e.g. sulfated polysaccharides or protein like molecules. On the basis of these results a modified hypothesis (Fig. 5.2) for the mechanism of iodine uptake was proposed.

Fig. 5.2
figure 2_5

Mechanism of iodine uptake according to [48] and adapted. Iodide in seawater (0.3 μM) diffuses into the peripheral tissue were it is oxidized by the vanadium iodoperoxidases to I2. This inorganic species becomes bound to apoplastic macromolecules and accumulates. These macromolecules are conceivably charged polysaccharides, which are abundant in these seaweeds

Full size image

An apolastic oxidase which is activated by light or stress produces hydrogen peroxide which oxidizes iodide from seawater to a yet undefined oxidation state and after which the iodine species become bound to apolastic macromolecules or alternatively diffuse into seawater and the environment. It would have been nice to have data as well about the distribution of the vanadium iodoperoxidase. This could have led to a better understanding of the role of the iodoperoxidases and maybe also of the systems that are involved in the generation of hydrogen peroxide.

3 Presence and Biological Function of Vanadium Chloroperoxidases

Fungi belonging to the dematiaceous hyphomycetes that were isolated from plant material and soils collected in Death Valley were shown to contain haloperoxidase activity that could function at more elevated pH and temperature [49, 50]. Some years later it was shown [7] that the non-heme chloroperoxidase produced by the fungus Curvularia inaequealis also contains vanadate as the prosthetic group in the enzyme. The vanadium enzyme is produced in the secondary growth phase, when nutrients become limiting. In the group of the dematiaceous hyphomycetes several species produce vanadium haloperoxidases (e.g. Drechslera biseptata, D. subpapendorfii, Embellisia didymospora, Ulocladium chartarum [8, 51] and Botrytis cinera [52]). Many of these hyphomycetes are phytopathogenic and/or grow on decaying plant material. To date no halogenated metabolites have been reported to be formed by these organisms directly and it has been proposed that physiological role of the vanadium chloroperoxidase is an oxidative attack mechanism on the lignocellulose in cell walls of plants or the cuticle layer on the leaves of plant [53]. The oxidative degradation will facilitate penetration of the fungal hyphen into the host to reach nutrients in the cell. It is very likely that HOCl generated by fungal species in soils reacts with organic matter resulting in the formation of organohalogens. Since chloroperoxidase-producing fungi are ubiquitous in decaying lignocellulose and plant debris and occur widely in nature these enzymes are responsible in the natural production of high-molecular-weight chloro-aromatics and in lignin breakdown [54]. Because lignin is the earth’s most abundant aromatic substance chlorolignin produced by this route and humus derived from it are probably significant components of the global chlorine cycle [55].

Recently it has become clear that haloperoxidases are not the only class of enzymes that are able to incorporate halogens in complex organic molecules. The iron (II)- and a-oxoglutarate -dependent halogenases [56] are able to specifically chlorinate specific compounds by a Cl-Fe(IV)-oxo species through ‘rebound’ of a chloride radical, analogous to the hydroxyl radical rebound mechanism postulated for heme and non-heme hydroxylases. The FADH2 halogenases chlorinate specific compounds by formation of an enzyme-bound chloronium intermediate that is targeted via a protein tunnel to the region of the bound substrate. These classes of enzymes will not be discussed here but see [57] for details.

4 Haloperoxidase Activity Assays

Classical heme peroxidase substrates such as guaiacol, o-dianisidine and benzidine are not oxidized by haloperoxidases and other non-standard assays should be used. The activity of the bromo- and chloroperoxidases is routinely determined using scavengers reacting with the very reactive intermediates HOCl or HOBr [58] and that result in marked changes in absorbance. A quantitative assay to determine the halogenating activity of these enzymes is the bromination or chlorination of monochlorodimedone [59]. This compound is a 1,3 diketone with an activated carbon atom that is brominated or chlorinated by HOX or X2 to the dihalogenated compound. This reaction is monitored spectrophotometrically at 290 nm and the loss of absorbance of monochlorodimedone (ε = 20,000 M−1 cm−1) is recorded. This assay can not be used to measure iodoperoxidase activity due to the poor reactivity of oxidized iodine species towards monochlorodimedone, high instability of the iodinated product and formation of I3 − that also strongly absorbs in the UV region. In general iodoperoxidase activity is measured by spectrophotometric detection of triiodide.

The MCD formation is quantitative with respect to H2O2 consumption but only at neutral or slightly alkaline pH values. At higher pH values a competing reaction with hydrogen peroxide occurs resulting in the formation of dioxygen Eq. 5.3. This competing reaction results in singlet oxygen formation and a decrease in the rate of bromination or chlorination of monochlorodimedone [60]. This decrease in rate should be taken into account when the activity measurements are carried at higher pH values. In fact the vanadium chloroperoxidase has been used as a highly efficient catalyst for the production of singlet oxygen and the enzyme is 103–104 times faster than the molybdate- peroxide system [61]. The halogenation of phenol red is an alternative assay. This test may only been used for qualitative purpose due to polyhalogenation of the dye [62]. However the bromination of the dye results in a marked color change, which also allows visual screening of haloperoxidase activity of large numbers of samples. This assay has been used in high-throughput screening of mutant libraries of vanadium chloroperoxidase [63]. Also the dye thymol blue may acts as a scavenger for HOBr. The kinetic parameters obtained [58] are essentially the same as that reported [59] using monochlorodimedone. Thymol blue has also a high reactivity towards HOI formed and it has been used in the steady-state kinetic analysis [58] of the oxidation of I− by vanadium bromoperoxidase from A. nodosum.

Figure 5.3 illustrates the reaction and the results show that this method is quantitative and superior to the standard I −3 assay.

Fig. 5.3
figure 3_5

Bromination of iodination of thymol blue (1). At pH 8 there is marked color change due a shift in the pKa upon halogenation of the dye, the dihalogenated compound (2) has a deep blue color with an absorbance maximum at 620 nm

Full size image

The vanadium iodoperoxidases oxidize iodide to HOI according to Eq. 5.4 and the HOI formed reacts further to I2 (Eq. 5.5).

$$ {{\text{H}}_2}{{\text{O}}_2} + { }{{\text{I}}^{ - }} \to {\text{HOI }} + {{\text{H}}_2}{\text{O}} $$
(5.4)
$$ {\text{HOI}} + {{\text{I}}^{ - }} + {{\text{H}}^{ + }} \to {{\text{I}}_2} + {{\text{H}}_2}{\text{O}} $$
(5.5)
$$ {{\text{I}}_2} + {{\text{I}}^{ - }} \leftrightarrow {{\text{I}}_3}^{ - } $$
(5.6)

The triiode formation is the result of reversible complexation of I2 with I− according to Eq. 5.6. This method when used [36] in an assay is not free from intrinsic problems and due the value of the equilibrium constant of Eq. 5.6 iodoperoxidase activity can only be measured quantitatively at iodide concentrations higher than 20 mM. Also a reaction occurs between HOI and H2O2 at pH values > 6.5 [58] and thus reported values for K m and V max should be treated with caution. To date the synthesis of iodinated organic compounds by vanadium iodoperoxidases has not been reported.

4.1 Steady-State Kinetics of Bromoperoxidases

Detailed steady-state kinetic studies on the oxidation of bromide by vanadium bromoperoxidases have been carried out, which suggested a bi-bi ping-pong two substrate mechanism in which hydrogen peroxide first binds to the active site followed by halide oxidation. The kinetic mechanisms for the bromoperoxidases from various sources appear to be very similar [59, 60]. Only slight inhibition of these enzymes by high concentrations of H2O2 (120–400 mM) was observed [60]. Thus vanadium bromoperoxidases are very resistant towards oxidative inactivation [61]. By contrast most heme peroxidases would be completely destroyed by these concentrations of H2O2.

Evidence that hydrogen peroxide reacts only with a deprotonated group in the enzyme with a pKa of 5.7–6.7 was obtained from measurement of the log K m value for H2O2 as a function of pH. This log K m decreased linearly with increasing pH but levels off above pH 6. The second-order rate constant that was derived for the binding of hydrogen peroxide to the bromoperoxidase from the specificity constant [62] is 2.5 × 106 M−1 s−1 at pH > 6. In contrast the Km for bromide is hardly affected by pH [59]. Table 5.1 gives values of the kinetic constants of the various haloperoxidases.

Table 5.1 Kinetic parameters of the oxidation of I−, Br− and Cl− catalyzed by vanadium iodo bromo- and chloroperoxidases
Full size table

4.2 Steady State Kinetics of the Iodide Oxidation by Iodo- and Bromoperoxidases

A careful analysis carried out by [58] showed that the oxidation of I− by the vanadium bromoperoxidase from A. nodosum occurs also by an ordered two-substrate mechanism. The K m value for I− of 0.18 mM is about 100-fold smaller than that found for bromide oxidation by the same enzyme and conditions. The k cat value of 75 s−1 for iodide oxidation is similar to that found for bromide oxidation at the same pH.

Table 5.1 gives a summary of some of the kinetic constants of the iodo-, bromo- and of the chloroperoxidase that will be discussed later.

These early kinetic studies suggested the presence of a peroxo-intermediate. The primary reaction products of the enzyme-mediated peroxidation of bromide appear to be Br3, Br2 or HOBr. Although HOBr is probably the first oxidized identity this is difficult to prove since different bromine species may be present that are in rapid equilibrium [59]. It was not possible to demonstrate specificity with regard to bromination of various organic nucleophilic acceptors for the enzyme from the brown seaweed A. nodosum. This suggests a mechanism in which the enzyme releases a diffusible halogen intermediate into the seawater. This brown seaweed produces simple halohydrocarbons such as CH2Br2 and CHBr3 that are probably the result of the enzymatically generated HOBr with organic matter in seawater or ketoacids in the seaweed. A brominated surface tryptophane is found in the enzyme from A. nodosum that is the result from free diffusion of the HOBr species generated by the enzyme during turnover [67]. Single molecule fluorescence microscopy also gave evidence for a specific bromination. By this technique it was possible to study the diffusion behavior of bromonium species produced by a vanadium peroxidase from the fungus Curvularia verruculosa [68]. Under steady-state conditions hypobromite was able to diffuse over 800 nm in the bulk solution before it reacted with organic substrates. Scheme 5.1 gives a simplified model for the reaction mechanism of the enzyme.

Scheme 5.1
scheme 1_5

Simplified reaction mechanism of the vanadium peroxidases based on enzyme kinetics. The enzyme species in boxes represent inhibited forms of the enzymes. For Cl− oxidation a proton is needed a since the K m for Cl− is strongly pH dependent whereas protonation is not needed for Br− oxidation by either bromoperoxidase or chloroperoxidase

Full size image

In addition to the simple volatile brominated compounds produced mostly by brown seaweed other seaweed species contain more elaborate and complex brominated compounds such as halogenated sequiterpenes. In this case there may be a more specific interaction between organic substrate and an enzyme bound halogenating intermediate. Vanadium bromoperoxidases isolated from marine red algae (species of Laurencia, Plocamium, Corallina) convincingly catalyze [69, 70] the asymmetric bromination and cyclization of a terpenoid precursor (E)-(+) – nerolidol, producing single diastereoisomers that are also found in red algae. These vanadium bromoperoxidases produce bromonium ions or equivalents in the active site that brominates one face of the terminal olefin that docks within the active site cavity in a specific orientation. Surprisingly cyclic terpenes are only formed in mixtures of water and organic solvents. When the organic solvent is deleted only halohydrines are formed due to a competing reaction with water. This raises the question as to the formation of the cyclic terpenes under physiological conditions.

As will be discussed bromoperoxidases from A. nodosum and Corallina officinalis catalyze sulfoxidation of aromatic sulfides enantioselectively. This suggests specific binding of organic molecules possibly close to or at the active site of this enzyme as well.

4.3 Properties of the Prosthetic Group in Bromoperoxidase

When the vanadium bromoperoxidase from A. nodosum is exposed to phosphate buffers containing EDTA at low pH the enzymatic activity is rapidly lost. In phosphate buffers at neutral pH the enzyme also inactivates but more slowly, however, in Tris buffers the enzyme remains active for months. Complete recovery of the enzymatic activity is possible by incubation of the apo-enzyme at neutral pH with orthovanadate (HVO 2−4 ) [3]. This reconstitution by vanadate was taken as early evidence that vanadate is present in the active site. Indeed the X-ray structures of vanadium bromoperoxidases confirmed that vanadate is indeed the prosthetic group in these enzymes [71, 72]. Vanadate binds strongly to apo-bromoperoxidase with K d values of 35–55 nM. These values are close to the concentration of vanadate in seawater of 35 nM [1]. Thus, no additional enzyme system is required to incorporate vanadate in the enzyme. Also molybdate may be incorporated in the active site of the bromoperoxidase, however, the enzyme reconstituted with molybdate is inactive [9]. The purified enzyme only displays weak bands at 300–330 nm in the optical absorption spectrum [73]. These bands diminish in intensity when H2O2 is added and the original intensity is restored by addition of bromide. This was taken as the first direct indication for the existence of an enzyme-peroxide intermediate in line with the steady state kinetic studies.

When substrates are added to the bromoperoxidase or during turnover the vanadium (IV) or (III) states are not observed by EPR [9] or by K-edge X-ray absorption studies [74]. Thus and since also the reduced enzyme is inactive, the redox state of the metal during turnover remains vanadium (V). Hence it has been proposed that vanadium binds and activated hydrogen peroxide by acting as a Lewis acid [9]. In this metal-assisted mechanism electron density from the bound peroxide is withdrawn in such a way that the nucleophilic halide reacts with this activated peroxide intermediate to yield hypohalous acid. These spectroscopic methods have not only yielded insight into the mechanism but also valuable structural data were obtained. EXAFS studies [74, 75] showed that the geometry of the metal oxide bound in the active site was five-coordinate. The multiple scattering effects from outer atoms of a group pointed to a histidine ligated to the vanadate. An ESEEM study [76] of the reduced enzyme showed the presence of nitrogen atoms in the direct coordination sphere of the metal oxide. The X-ray structure of the vanadium bromoperoxidase from A. nodosum [71] confirmed these findings.

4.4 Kinetic and Optical Properties of Vanadium Chloroperoxidases

Although expression of active recombinant bromoperoxidases has been reported it is not straightforward due to the formation of inactive inclusion bodies that have to be solubilized and subsequent activation of the enzyme [77, 78]. In contrast cloning and expression of vanadium chloroperoxidase is fairly easy. The enzyme has both been expressed in the yeast Saccharomyces cerevisiae [79] and in E.coli [63]. Activation of the recombinant apo-form of the enzyme is easily achieved by addition of vanadate to either the growth medium or the isolated apo-enzymes. Most soils contain a high concentration (100 ppm) of vanadium is and since the affinity of the apochloroperoxidase for vanadate is high the amount of vanadium is not a limiting factor in the conversion of the apo-enzyme into the holo-enzyme. Further low concentrations of H2O2 increase the affinity of the apo-enzyme for vanadate at least 200-fold. This is due to the formation of pervanadate that binds the active site with a dissociation constant of less than 5 nM [80].

The substrate inhibited Bi Bi ping-pong mechanism of the chloroperoxidase [7, 65, 66] resembles those of the vanadium bromoperoxidases. The chloroperoxidase exhibits a pH profile similar to vanadium bromoperoxidases although the optimal pH of 4.5–5.0 is at a lower pH value. The enzyme is inhibited at lower pH values by chloride in a competitive way whereas at higher pH values normal Michaelis-Menten kinetics is observed. As in the bromoperoxidases the log K m for hydrogen peroxide decreases with increasing pH and very low values of less than 5 μM are observed. In contrast as shown in Fig. 5.4 the log K m for chloride increases linearly with pH.

Fig. 5.4
figure 4_5

Log K m for chloride and hydrogen peroxide as a function of pH. Figure was constructed from data published in [7, 65]

Full size image

This demonstrates that protons are involved in binding and activation of substrates [66, 81]. The enzyme reacts first with peroxide to form a peroxo-intermediate after which a chloride ion and a proton react resulting in the formation of an enzyme-HOCl intermediate. This intermediate decays in a rate-determining step to enzyme and free HOCl. The linear dependency of the log K m for hydrogen peroxide on pH suggests that like the bromoperoxidase an ionizable group is involved in the binding of hydrogen peroxide. Hydrogen peroxide is unable to bind when this group is protonated. The linear dependence of the log K m for chloride on pH shows that for binding of chloride to occur, protonation of a group is essential. However protonation of this group is clearly not essential for bromide oxidation. Both the K m for bromide of the native bromoperoxidase from A. nodosum and the vanadium chloroperoxidase oxidizing bromide are hardly dependent upon pH. Thus it has been proposed [66, 81] that chloride oxidation requires protonation of the peroxo-complex in the active site as confirmed by model studies [82] and DFT calculations [83] that suggested the necessity of protonation of the peroxo- state for chloride oxidation on each turnover. Scheme 5.2 illustrates the proposal.

Scheme 5.2
scheme 2_5

Protonation state of the side-on bound peroxide that is attacked by the incoming halide. (a) less oxidizing non-protonated form of the side-on bound hydrogen peroxide. (b) strongly oxidizing protonated form

Full size image

A similar band in the near UV (316 nm) as observed in the optical spectrum of vanadium bromoperoxidase [73] is present in the chloroperoxidase [80], which is not found in the apo-enzyme. The intensity of the band at 316 nm decreases upon addition of hydrogen peroxide but a weak band is formed at 480 nm. Chloride restores the original spectrum. This has allowed monitoring the pre-steady state binding of peroxide to the enzyme and made it also possible to study the reaction of the peroxo-intermediate with the second substrate chloride.

The first order rate constant for the binding of hydrogen peroxide to the enzyme at pH 5 was 0.8 × 105 M−1 s−1. At higher pH values the reaction between enzyme and peroxide was too fast to be measured. This increase in rate constant may relate to the decrease in K m for peroxide at higher pH values (see Fig. 5.5). It was also possible to study the rate of chloride oxidation by the peroxo-intermediate using a sequential stopped flow [80]. The first-order rate constant of 7.0 s−1 is similar in magnitude to the k cat obtained from steady-state kinetics.

Fig. 5.5
figure 5_5

Active sites of VCPO from C. inaequalis (grey) and VBPO from A. nodosum (pink) have been superimposed to demonstrate the similar geometries. Active site residue labels correspond to VCPO (top) and VBPO (bottom) (From [84]. Reproduced with permission of the American Chemical Society)

Full size image

The UV-VIS spectra of several mutants have been obtained [80, 81] and the UV spectra together with the known X-ray data [85] confirm that amino acid residue in the active site contribute to the optical spectrum in the near UV. The crystal structure of the His496Ala mutant shows that tetrahedral vanadate is present in the active site but not bound to this histidine [85]. The optical spectrum [81] of this mutant lacks the feature at 315 nm indicating that the V-N bond is crucial for the observed spectrum.

The optical absorption spectra including the effect of peroxide addition, change of pH and mutation of active site residues are in line with TD-DFT and quantum mechanical calculations [86]. In the native enzyme a His496 p → V 3d transition dominates, whereas in the peroxide form a peroxide p* → V 3d CT transition found at 480 nm is most intense. On the basis of analogy to well-studied inorganic imidazole-peroxovanadate complexes this band was already attributed [80] to a peroxo- to vanadium charge-transfer band. Thus by using reliable models of active sites in chloroperoxidase the observed spectral features can be calculated theoretically.

The bromoperoxidase activity of chloroperoxidase has also been studied [64]. The chloroperoxidase has much larger very oxidation efficiency for bromide (K m about 10 μM) than the bromoperoxidase, which has a K m for bromide at pH 5.2 of 5.1 mM. The k cat at pH 5 is about 250 s−1. This value is similar to that of the bromoperoxidase from A. nodosum. The chloroperoxidase is strongly inhibited by excess bromide and hardly any activity is observed at 100 mM bromide. Interestingly the vanadium peroxidase that was isolated [87] from the lichen Xanthoria parietina that grows on rocks and stones also has a very small value (28 μM) for the K m for bromide. This enzyme was also inhibited strongly by excess bromide. It is conceivable therefore that this enzyme is also a vanadium chloroperoxidase. However, chloroperoxidase activity has not been tested.

Table 5.1 shows that the specificity constants for the reaction of hydrogen peroxide with the peroxidases and mutants are nearly the same in all cases and not affected by pH. These specificity constants correspond [59] to the lower limits of the bimolecular rate constant of the binding of peroxide to the vanadate and this means that the protein environment in the active site does not control the reaction of peroxide with vanadate. In contrast the specificity constants for the reaction with the halides vary 1000-fold. This suggests that in line with the oxidation-reduction potentials of the halides the halide reactivity is driven by chemical reactivity of the peroxo-vanadium intermediate rather than molecular recognition in a binding pocket [58].

5 Sulfoxidation and Other Oxidation Reactions

In the absence of halide the vanadium haloperoxidases are capable of catalyzing the enantioselective sulfoxidation of organic sulfides [88, 89]. The bromoperoxidase from the red seaweed C. pilulifera produces the S-enantiomer of methyl phenyl sulfoxide (55% e.e.) whereas the bromoperoxidase from the brown seaweed A. nodosum produces the R-enantiomer of this sulfide (96% e.e.) [89, 90]. Organic sulfides that structurally resemble indenes and also small sulfides, possessing a cis- positioned carboxyl group with respect to the sulfur atom, are rapidly converted to the corresponding sulfoxides with selectivities exceeding 95% e.e. The oxygen of the peroxide is directly transferred to the sulfide in a selective manner, strongly suggesting that the aromatic sulfide binds near or at the active site with a relatively high K m value > 1.5 mM [89]. Compared to the brominating activity of the bromoperoxidase from A. nodosum (166 s−1). The k cat value for the sulfoxidation reactions (1 min−1) is low. This enzyme also catalyzes the sulfoxidation of racemic non-aromatic cyclic thioethers with kinetic resolution [91]. In contrast to the vanadium bromoperoxidases the chloroperoxidase was observed to mediate the formation of only racemic sulfoxides [90]. The highly reactive chloroperoxidase peroxo intermediate probably directly abstracts an electron from the substrate. A positively charged sulfur radical is subsequently formed which migrates from the enzyme and is subsequently non-enzymatically converted into a sulfoxide.

Surprisingly, the recombinant vanadium chloroperoxidase from the fungus Curvularia inaequalis catalyzes the oxidation of 2,2′-azino-bis(3-ethylbenzthiazoline-6-sulfonic acid) (ABTS), a classical chromogenic heme peroxidase substrate. The enzyme mediates the oxidation of ABTS in the presence of hydrogen peroxide with a turnover frequency of 11 s−1 at pH 4.0 and has a K m for ABTS of approximately 35 μM [92]. Also the industrial sulfonated azo dye Chicago Sky Blue 6B was bleached by recombinant vanadium chloroperoxidase in the presence of hydrogen peroxide. No further studies on this bleaching reaction have been reported.

6 Stability of Bromo and Chloroperoxidases

The bromo and chloroperoxidases are remarkable stable [23, 73, 87, 93–95]. In particular the chloroperoxidase is highly resistant towards elevated temperatures; the midpoint temperatures determined from thermal denaturation curves of the chloroperoxidase range from 82°C to 90°C [65]. Detergents such as SDS or water miscible organic solvents e.g. ethanol or dioxane have little effect on activity. The vanadium chloroperoxidase also remains fully active and is stable in micro-emulsions containing 10% non-ionic surfactant [96]. Oxidative agents such as H2O2 at concentrations in the M range, HOCl and singlet oxygen have little effect on the activity of the chloroperoxidase [61]. In comparison to heme peroxidases the vanadium haloperoxidases are unique and the ability of the enzymes to handle the oxidative substrate and product are related to the nature of the active site. This stability is a requirement for the potential application of the vanadium peroxidase as (industrial) biocatalyst in organic synthesis, their use in disinfection formulations [33, 97] and as bleaching enzymes [63].

7 X-Ray Structures of Vanadium Bromoperoxidases

Three X-ray structures of bromoperoxidases have been determined, the homodimeric structure of the enzyme from the brown seaweed A. nodosum [71] and those of the red seaweeds Corallina pilulifera and C. officinalis [72]. The enzymes from the red seaweeds crystallize as dodecamers and the structures are made up of 6 homo-dimers. The subunit in the bromoperoxidases from Cor. officinalis [98] measures approximately 85 × 56 × 55 Å in size. Twelve subunits are arranged within a 23 cubic point group symmetry. The vanadate-binding site is located at the bottom of the active site cleft, which is about 20 Å deep and 14 Å wide. The active site cleft is formed predominantly from residues of two different subunits in the dimer. These subunits are intertwined and this suggests that the subunit by itself will show no activity and that reconstitution of activity of a recombinant enzyme is a difficult process. It is possible to superimpose the secondary structure of the vanadium chloroperoxidase [99] from the fungus Curvularia inaequalis on the Corallina bromoperoxidase dimer. Many of the a-helices of each chloroperoxidase domain are structurally equivalent [98] to the a-helices in the Corallina bromoperoxidase dimer.

8 Active Site of Vanadium Bromoperoxidase from A. Nodosum and Chloroperoxidase from Curvularia inaqualis

Most of the amino acid residues in the active site of the haloperoxidases are conserved. Figure 5.5 illustrates the active site in the bromoperoxidase from A. nodosum [71] in pink and also that of the vanadium chloroperoxidase from Curvularia inaequalis [99] in grey.

In the bromoperoxidase from A. nodosum vanadium is bound as orthovanadate on the N-terminal side of a four-helix bundle and the vanadium metal is the center of a trigonal bipyramid with three equatorial oxo ligands, one oxygen in the axial position and a covalent linkage to the \( {{\text{N}}^{{{{{\varepsilon }}_{{2}}}}}} \) of His486. The positively charged amino acids, including Lys341, Arg349 and Arg480 form a kind of a positively charged cage, which compensates the negative charge of the oxygen atoms in the equatorial plane. There are hydrogen bonds of the Ser416 and Gly417 to the bound vanadate. The axial oxygen atom forms a hydrogen bond with the \( {{\text{N}}^{{{{{\varepsilon }}_1}}}} \) of the distal His418, which is a conserved catalytic residue in vanadium peroxidases. As in the enzyme from Cor. officinalis amino acid residues from both monomers contribute to the substrate access channel of each active center.

Both a 2.1 Å crystal structure of native chloroperoxidase and of recombinant chloroperoxidase at 1.66 Å resolution has been reported [85, 99, 100]. The enzyme molecule crystallizes as a monomer and has an overall cylindrical shape (not shown) with a length of about 80 Å and a diameter of 55 Å. The protein fold is mainly α-helical with two four-helix bundles as the main structural motifs. In contrast to the vanadium bromoperoxidase from A. nodosum there are no disulfide bridges in the chloroperoxidase. The vanadium-binding center is located on top of the second four-helix bundle and orthovanadate is bound in the same way as in the bromoperoxidases. Three oxygen atoms form a plane and the fourth oxygen is found at the apex. It is hydrogen bonded to the nitrogen \( {{\text{N}}^{{{{{\varepsilon }}_1}}}} \) of His404. The \( {{\text{N}}^{{{{{\varepsilon }}_2}}}} \) atom from His496 ligates directly to the metal. Again the negative charge of the vanadate group is compensated by hydrogen bonds to three positively charged residues (Arg360, Arg490, Lys353). Further, the vanadate forms hydrogen bonds with a glycine (Gly403) and a serine residue (Ser402). There is also a histidine residue (His404) close to and hydrogen bonded to the apical hydroxyl moiety of the vanadate that may act as an acid/base group in catalysis. Two water molecule hydrogen bonded to the His 404 are also present in the chloroperoxidase [99]. As Fig. 5.6 shows the active sites of the two enzymes are structurally nearly super-imposable.

Fig. 5.6
figure 6_5

Suggested location of the V–O and V–N distances obtained from the vanadium K-edge EXAFS recorded for native VCPO (left) after the addition of H2O2 (peroxo- intermediate) (right), on the basis of the coordination numbers determined by protein crystallography for crystals obtained at pH 8.0 [99]. PDB VCPO (Wildtype): 1idq, resolution 2.03 Å at pH 8.0. PDB PeroxoVCPO (wildtype): 1idu, resolution 2.24 Å and pH 8.0. The vanadium centre (V) is shown in grey and the oxygens (red, labeled 1, 2, 3 and 4) and nitrogens (blue) are labeled based on [99] (Reproduced with permission from [101])

Full size image

X-ray structures are not yet available for a vanadium iodoperoxidase. However, the full-length cDNA has been obtained [37] for the iodoperoxidase from L. digitata and phylogenetic analyses indicated that this iodoperoxidase shares a close common ancestor with brown algal vanadium-dependent bromoperoxidases. A three-dimensional structure model of the iodoperoxidase active site and comparisons with those of other vanadium- dependent haloperoxidases revealed [37] a striking difference in the catalytic region of iodoperoxidase. Instead of an invariant serine in vanadium bromo- and chloroperoxidase an alanine residue (Ala481) is present at the same position (Fig. 5.6). This substitution results in the loss of one of the hydrogen bonds involving the negatively charged oxygen atoms of the vanadate cofactor in vanadium chloro and bromoperoxidase.

Recently [101] vanadium K-edge X-ray Absorption Spectra were reported for the native and peroxo-forms of vanadium chloroperoxidase from Curvularia inaequalis at pH 6.0. The Extended X-ray Absorption Fine Structure (EXAFS) regions provided a refinement of crystallographic data. As shown in Fig. 5.6 vanadium is coordinated in the native enzyme to two oxygen atoms at 1.69 Å, another oxygen atom at 1.93 Å and the nitrogen of an imidazole group at 2.02 Å. Further one short V-O bond (1.54 Å) is present. In addition to H bonds oxygen atom 1 and 2 have electrostatic interaction with the positively charged residues Arg360 and Arg489 and Lys353 (O2). The EXAFS data [74] on the vanadium bromoperoxidase from A. nodosum and the X ray structure of this enzyme [71] also show one short V=O bond. Reduction of the chloroperoxidase with dithionite has a clear influence on the spectrum, showing a change from vanadium (V) to vanadium (IV).

9 Structure of the Peroxo-Intermediate of Vanadium Chloroperoxidase

By incubating native crystals in mother liquor containing peroxide and shock freezing these in cryobuffer it was possible to obtain the X-ray structure of the important catalytic peroxo-intermediate [99]. The peroxide is bound side-on with a distance of 1.47 Å between the two peroxide oxygen atoms O2 and O4. The resulting coordination geometry is that of a distorted tetragonal pyramid with only four oxygen atoms and one nitrogen atom. According to the X-ray data there are several changes in the hydrogen-bonding network when peroxide is bound. His404 is no longer bonded to any of the vanadate oxygens, but there are still two water molecules in the active site. Lys353 forms a hydrogen bond to one of the peroxide oxygens. The EXAFS study [102] gives a refinement of the X-ray structure, although it should be kept in mind that the X-rays structure was obtained at pH 8 and the EXAFS study was carried out at pH 6. As Fig. 5.6 shows a short V=O bond of 1.54 Å is still present, another V-O bond at 1.67 Å and the side-on bound peroxide involves V-O bonds of 1.67 and 1.88 Å to oxygen atoms 2 and 4, respectively. Oxygen atom 2 is more exposed to the solvent, whereas oxygen 4 is more buried in the protein cavity [63] and is hydrogen bonded to Lys 353. This Lys 353 polarizes the oxygen-oxygen bond and is very important in catalysis. Indeed mutation to an alanine [81] has a profound effect on the catalytic activity. Considering the location of oxygen 2 in the active site it is likely that at this atom nucleophilic attack by halide occurs. The shorter V-N bond observed in the peroxo-form with respect to the native enzyme is in line with the previously reported [80] stronger binding of peroxo vanadate to the enzyme.

10 Molecular Mechanism of Vanadium Haloperoxidases and Difference in Reactivity Between Iodo-, Bromo-, and Chloroperoxidase

Based on the effect of mutations of the active site residues and kinetic data the following mechanism arises for the chloroperoxidase. By the Lewis acid properties of the metal ion electron density is withdrawn from the bound peroxide assisted by the two positively charged arginine residues (Arg490 and Arg360). The conserved Lys353 that forms a hydrogen bond to one of the oxygen atoms further polarizes and activates the bound peroxide. One of the oxygen atoms of the peroxide is protonated to allow oxidation chloride ion (c.f. Fig. 5.4). After formation of the peroxide intermediate the next step in catalysis is the nucleophilic attack of the halide on the electrophilic hydrogen peroxide (oxygen 2 in Fig. 5.6) and breaking the peroxide bond, resulting in formation of HOX or XO−. All mutations of the active site residues result in kinetically crippled [64, 66, 80, 81] mutants. Most or all of the chloroperoxidase activity is lost, however, some mutants retain some bromoperoxidase activity.

Detailed structural data are available now and in principle it should be possible to explain the difference in reactivity between the vanadium iodo-, bromo- and chloroperoxidases and the effects of mutations on the activity. A prominent difference in the active site architecture of the chloroperoxidase and the iodo- and bromoperoxidases is the presence of a second histidine in the iodo- and bromoperoxidase, a residue that in chloroperoxidase is a phenylalanine. Since chloride is more easily oxidized than bromide and bromide more easily oxidized than iodide it has been suggested [63, 71, 81] that the histidine in bromoperoxidase accepts the proton from the bound peroxide and in this way reduces the oxidative strength of the bound peroxide and the reactivity of the peroxidase. This decrease in reactivity may also explain why vanadium chloroperoxidase is strongly inhibited [103] by the nucleophilic inhibitor azide whereas the bromoperoxidase is only weakly inhibited by azide. When by site-directed mutagenesis the Phe397 in chloroperoxidase was replaced by a histidine it was observed [66] that the specificity constant for the oxidation of chloride decreases 100-fold. However, the specificity constant for bromide oxidation decreased also10-fold and further the mutant rapidly inactivated during turnover. Thus the difference in oxidizing ability between the two enzymes is not simply due to a single residue but apparently more factors are involved. An attempt was also made to induce chlorinating activity in the bromoperoxidase from C. pulilifera by mutating the Arg397 in this bromoperoxidase to a tryptophane that is present in chloroperoxidase. Indeed chlorinating activity was observed [104], however, the reported K m value for chloride of about 700 mM is 100-fold higher than that of native chloroperoxidase. No satisfactory explanations have been offered for these observations and the factors that are important in tuning the reactivity of the haloperoxidases are not fully understood.

The active site of bromoperoxidase differs from the iodoperoxidase by the presence of serine instead of an alanine residue. The absence of the polar serine residue in the iodoperoxidase from L. digitata lowers the charge neutralization of the vanadate cofactor in its peroxo state and this may be the reason [37, 58] why the peroxo-vanadate in the iodoperoxidase is less capable of oxidizing highly electronegative halides. In this context the observation is important that a significant decrease in the chloride oxidation occurs when the Ser402 in vanadium chloroperoxidase is mutated to an alanine residue [66]. This mutation abolishes oxidation of the most electronegative halide, i.e. chloride, while maintaining still 20% oxidizing activity of bromide. Finally it should be noted the access cavity in vanadium iodoperoxidase is wide and open compared to the small deep channel to the active site in chloroperoxidase. This feature may affect the reactivity of the peroxo-intermediate to halides [105] as well. This exposed site may also explain why an acid phosphatase that has essentially the same active site as the chloroperoxidase [79] and in which vanadate is substituted has only a very low brominating activity [106]. This phosphatase has a very accessible active site [107, 108].

Driven by potential application of the chloroperoxidase in antifouling paints a directed evolution study [63] was carried out to improve the brominating activity of the vanadium chloroperoxidase from C. inaequalis at mildly alkaline pH. Two rounds of random mutagenesis were carried out, followed by saturation mutagenesis at a hot-spot position. A triple mutant was created by rational combination of positive mutants. This final triple mutant had an activity that was the highest ever measured for vanadium haloperoxidases. At pH 8 it showed an increase in the k cat from 1 to 100 s−1 and at pH 5 a fivefold higher brominating activity was found (k cat 575 s−1). Also the chlorinating activity at pH 5 was doubled to 36 s−1. The K m values for the substrates increased somewhat but overall the specificity constants k cat /K m for bromide and chloride were increased (Table 5.1). The mutants were in the first and second coordination sphere of the vanadate cofactor and the effect on the steady-state kinetic parameters confirm that fine-tuning of residues Lys353 and Phe397 and changes in the electrostatic potential in the active site are very important.

10.1 Theoretical Calculations on the Nature of the Vanadate Cofactor

A detailed molecular picture of the catalytic mechanism of the haloperoxidases requires understanding of the structure of the vanadate cofactor, its protonation state, protonation state of the peroxo intermediate and the relevant interactions between vanadate and surrounding amino acids. The present accuracy of the X-ray data does not allow an assignment of these protonation states and hydrogen bonding networks. Several theoretical investigations have therefore been performed on vanadium complexes that model the native VCPO and on the enzyme itself. However, it should be kept in mind that the theoretical calculations up to now are based upon the X-ray data and the more accurate distances obtained by EXAFS have not been taken into account yet.

DFT calculations [109] resulted in a doubly protonated vanadate with an axial hydroxo ligand and one hydroxo group in the equatorial plane. The same conclusion was reached in a time-dependent DFT study [110] as well as in a QM/MM study [111]. A solid-state 51V-NMR study on VCPO in combination with DFT calculations of the NMR observables using a series of small models [84] concluded this picture. The importance of including the protein environment into the QM models was also shown [107] which again led to a minimum with two protons, but now in the form of an axially bound water ligand. The 51V-NMR chemical shifts [101] were remodeled [112] using QM/MM-optimized models of the complete protein and systematically increasing the size of the QM region. It was concluded that the resting state of the cofactor is either the proposed doubly protonated vanadate or a triply protonated state in which there is an axially bound water and an equatorial hydroxide. These studies converge on the scenario in which the resting state of the enzyme can be described as doubly protonated trigonal bipyramidal vanadate with at least one hydroxyl group in the apical position. The second proton can be assigned to any of the equatorial oxygen atoms or to the apical oxygen forming a bound water molecule. According to [113] and based on quantum mechanics/molecular mechanics evaluations of ground state properties, UV-VIS spectra and NMR chemical shifts the most likely ground state configuration of vanadate in vanadium chloroperoxidase is an axially bound water molecule that is stabilized by a hydrogen bond to His404. Mutation experiments already showed that the mutant His404Ala was strongly impaired in its activity and ability to bind vanadate. Further water molecules that are present in the crystallographic structure are very important in explaining the spectral and NMR data. These computational studies and the studies of the effect of mutations on the enzymatic activity and structure show clearly that the enzyme has a very complex hydrogen-bonding network around the cofactor. Two water molecules also participate in this but whether they have a possible role in tuning the activity is not clear. Thus the factors that are important in tuning the reactivity are not fully understood but it is unlikely that halide specificity is due to a selective halide binding effect. In this respect it should be noted that the existence of halide binding pocket is doubtful. Haloperoxidases are inhibited by excess halide and if a halide is found in the active site of these enzymes [114] it may present binding to an inhibitory site. The ability to oxidize halides most likely relates to the action of amino acid residues on the peroxo-intermediate in or in close vicinity of the active site and the resulting hydrogen bonding network.

11 Bacterial Vanadium Enzymes

Thus far vanadium haloperoxidases have only been found in eukaryotes but they may also be present in prokaryotes. Three putative homologous genes that are present in the 43-kb napyradiomycin biosynthetic cluster found in marine sediment derived actinomycetes [115] show sequences that correspond to the active site residues of the vanadium haloperoxidases but also to the acid phosphatases. Heterologous expression of this biosynthetic cluster in bacterial host resulted in the synthesis of chlorinated meroterpenoids. This observation was taken as evidence that bacteria also contain vanadium enzymes. However, an alignment (Fig. 5.7) of the active site region of the vanadium chloroperoxidase with corresponding parts in the NapH1, NapH3 en NapH4 region show clearly that the active site residue His404 is lacking and replaced by either a serine residue (NapH1 and NapH4) or a phenylalanine at this position.

Fig. 5.7
figure 7_5

Alignment of the sequences of NapH1, NapH4, NapH3 and vanadium chloroperoxidase. Conserved residues in bold. *, Active site residues involved in hydrogen bonds and binding the vanadate cofactor. Clustal W was used for the phylogenic analysis

Full size image

Considering the importance of this residue in catalysis [81] it is unlikely that all these genes code for an active haloperoxidase and may have another function e.g. as discussed below as phosphatases.

12 Active Site Similarity of Haloperoxidases and Acid Phosphatases

The three regions providing the metal-anion binding site in the vanadium enzymes show [116–118] a high similarity to stretches that are present in a large group of acid phosphatases that were previously considered unrelated. Based on this sequence similarity and the structural and electronic homology with phosphate, it was concluded that the architecture of the active sites in the two classes of enzymes is very similar. Indeed the X-ray structure of the acid phosphatase [107] from Escherichia blattae and that [108] of Salmonella enterica confirmed the remarkable similarity of the residues binding vanadate in the vanadium enzymes and phosphate in the phosphatases. Not only the phosphatases from bacteria contain this motif for the active site but it is wide spread in nature and include for example the membrane-bound glucose-6-phosphatase, an enzyme involved in regulation of glucose levels in the blood of mammals and the large family of lipid phosphohydrolases [118].

Since the same supra-molecular environment is present in the two classes of enzymes, it is likely that they exhibit dual enzymatic activities. Indeed when vanadate was removed from the active site of the vanadium chloroperoxidase from C. inaequalis it exhibited phosphatase activity [79]. However, the phosphatase activity with p-nitrophenyl phosphate (pNPP) is low probably due to the deeply buried active site in the protein that is not easily accessible for substrates. The low turnover allowed pre-steady state-studies on the phosphatase activity of both on both native chloroperoxidase and mutants [81] and to obtain details of the mechanism. A rather stable phospho-intermediate is formed in which His496 acting as a nucleophile reacts with the phosphate group. By incubating apo-chloroperoxidase crystals with p-nitrophenylphosphate and subsequent flash cooling of the crystals it was possible to trap the phospho-intermediate intermediate and obtain a high resolution X-ray structure [119]. It consists of a metaphosphate anion PO −3 covalently bound via its phosphorous atom to the \( {{\text{N}}^{{{{{\varepsilon }}_{{2}}}}}} \) atom of the His 496 with a water molecule in the position for a nucleophilic attack on the phosphorus.

When vanadate is substituted in the active site of acid phosphates they exhibit bromoperoxidase activity and enantioselective sulfoxidation activity. Thus, the analogy between these classes of enzymes includes both structural and catalytic aspect. However, the turnover values of 3–30 min−1 are much lower than of native bromoperoxidase from A. nodosum. When vanadate is substituted in the phosphatase phytase sulfoxidation and bromoperoxidase activity is observed also [120]. However, phytase does not belong to the class of acid phosphatases and active site of phytase has an architecture that differs from the acid phosphatases. The phosphatases have very different physiological functions and are found in species varying from E. coli to humans. Since phosphate and phosphate metabolizing enzymes entered evolution at an early stage and in which oxygen and hydrogen peroxide were not yet present, it seems likely that vanadate was coined by nature to become the prosthetic group in the vanadium enzymes in a more recent period and thus these enzymes have evolved from the phosphatases.