Keywords

1 Aromatic-Ring Hydroxylating Dioxygenases

Aromatic compounds are widely distributed and abundant organic compounds in nature. They are derived from aromatic amino acids, lignin components, flavonoids, quinones, crude oils, etc. Some of the toxic xenobiotics, such as dioxins, polychlorinated biphenyls, and nitroaromatics, also contain aromatic ring(s). Aromatic compounds are utilized by microorganisms as the carbon source and/or electron donor for respiration. Bacterial aerobic metabolism of aromatic compounds has been extensively studied for their applications of bioremediation and chemical and pharmaceutical industries. Initial reaction of aromatic-ring degradation is generally dihydroxylation by Rieske nonheme iron dioxygenase (RO) to yield cis-dihydrodiols and derivatives with strict stereo- and regio-specificity (Boyd and Sheldrake 1998; Gibson and Parales 2000; Ullrich and Hofrichter 2007). ROs catalyze not only dioxygenation but also monooxygenation, sulfoxidation, denitrification, and demethylation (Fig. 9.1) (Rosche et al. 1995; Herman et al. 2005; Nojiri 2012). In contrast to bacterial ROs, less numbers of these family enzymes in higher organisms are found, but known ROs in plants and insects play essential roles in biological processes, such as chlorophyll metabolism, cell death, development, and cholesterol metabolism (Gray et al. 1997; Tanaka et al. 1998; Yoshiyama et al. 2006; Yoshiyama-Yanagawa et al. 2011). Among the ROs, relatively simple compounds, such as naphthalene and biphenyl, are the substrates for the best-studied RO members. In these ROs, enzymatic mechanisms based three-dimensional structures are understood in detail.

Fig. 9.1
figure 1

Oxidation of aromatic compounds catalyzed by Rieske oxygenases. Compounds in parenthesis are unstable and spontaneous reaction (dotted arrow) to form the diol compounds is shown, except for the dicamba monooxygenation reaction (in the monooxygenated compound of dicamba in parenthesis is proposed intermediate of the monooxygenation reaction and the 3,6-dichlorosalicylic acid as the final product is shown)

2 RO Structures

2.1 Classification

ROs are composed of terminal oxygenase and electron transfer components. Oxygenation reaction is catalyzed by the terminal oxygenase component and two electrons are required for the reaction (Wolfe et al. 2001). Electron transfer components transfer the electrons from NAD(P)H to terminal oxygenase component. Electron transfer components are only reductase or ferredoxin and ferredoxin reductase (Table 9.1). All the components contain redox center(s) for storage of electrons. According to Batie’s classification system, ROs are divided into five groups by their component numbers and prosthetic groups (Table 9.1) (Batie et al. 1991). In class I ROs, the electron transfer component is only reductase which contains FMN (class IA) or FAD (class IB) and plant-type [2Fe-2S] cluster as the cofactors. The reductase component possesses NAD(P)H binding site for transferring electrons from NAD(P)H to terminal oxygenase via FMN or FAD and the plant-type [2Fe-2S] cluster. ROs in class II have two electron transfer components, ferredoxin and ferredoxin reductase. The ferredoxin reductase in class II has only FAD as a cofactor. Ferredoxin components in class IIA and IIB have putidaredoxin-type [2Fe-2S] cluster and Rieske-type [2Fe-2S] cluster, respectively. The [2Fe-2S] cluster of putidaredoxin-type ferredoxin is coordinated by four cysteine residues, whereas that of Rieske-type ferredoxin is coordinated by two cysteine and two histidine residues. There are few examples of RO systems with [3Fe-4S]-type ferredoxin as the electron transfer component (Martin and Mohn 1999; Kim et al. 2006; Takagi et al. 2005). However, so far, they have not been classified into any class in Batie’s classification. Therefore, here we propose “class IIC” into which ROs with [3Fe-4S]-type ferredoxin as the examples are classified (Table 9.1). Class III ROs have both ferredoxin reductase and ferredoxin component as well as class II ROs, but the ferredoxin reductase component contains plant-type [2Fe-2S] cluster and the FAD cofactor similarly to the reductase component of class IB ROs.

Table 9.1 ROs classification by electron transfer components [modified classification by Batie et al. (1991)]

2.2 Terminal Oxygenase Component

Terminal oxygenase in RO has Rieske-type [2Fe-2S] cluster and mononuclear nonheme iron. Crystal structures of naphthalene dioxygenase, biphenyl dioxygenase, carbazole dioxygenase, nitrobenzene dioxygenase, dicamba monooxygenase, 2-oxoquinoline 8-monooxygenase, toluene dioxygenase, cumene dioxygenase, and polycyclic hydrocarbon dioxygenase have been determined (Table 9.2). Terminal oxygenase component is usually an α3 or α3β3 subunit configuration (Fig. 9.2). The terminal oxygenase component of naphthalene dioxygenase (NDO) is the first example of terminal oxygenase structure of ROs (Kauppi et al. 1998). The terminal oxygenase of NDO is α3β3 configuration and resembles shape of mushroom. The height of the mushroom is 75 Å and the diameters of the cap and stem are 102 Å and 50 Å, respectively. The terminal oxygenase component of carbazole dioxygenase (CARDO) and 2-oxoquinoline 8-monooxygenase are the first examples of RO which have α3 configuration (Nojiri et al. 2005; Martins et al. 2005). The terminal oxygenase of CARDO is a doughnut shape with 100 Å in diameter, a central hole in 30 Å, and 45 Å thickness. Both of the redox centers (Rieske [2Fe-2S] cluster and mononuclear nonheme iron) and substrate-binding site are in the α-subunit. The β-subunit is thought to have a structural role in the holoenzyme but, in biphenyl dioxygenase (BDO) from Pandoraea pnomenusa (formerly Comamonas testosteroni) strain B-356, the β-subunit influences substrate specificity (Hurtubise et al. 1998). The α-subunit is consist of two domains, Rieske and catalytic domains. One iron (Fe1) of the Rieske cluster is coordinated by two cysteine residues and the other iron (Fe2) is coordinated by two histidine residues. Two inorganic sulfide ions bridge the two iron ions, forming a flat rhombic arrangement. This domain of ROs is similar to the Rieske iron-sulfur protein of the cytochrome bc 1 complex in the electron transport chain for ATP generation (Iwata et al. 1996) and chloroplast b 6 f complex (Carrell et al. 1997). The Rieske domain of terminal oxygenase component is likely to be derived from a common ancestral Rieske protein with these Rieske proteins because of their similarities. The distance between the Rieske center and the mononuclear iron within a single α subunit is further from the distance between the Rieske center and the iron in a neighboring α subunit. Therefore, electrons are thought to transfer from a Rieske center to a mononuclear iron in the neighboring α subunit (Fig. 9.2). The catalytic domain of RO is dominated by a stranded antiparallel β-sheet which extends into the middle of the domain. One side of the large β-sheet is covered with α-helices that contain ligands, an aspartate, and two histidines, to the mononuclear iron. A channel from molecular surface, which enables substrates to access to the active site, is formed by the sheet and the helices. The region close to the active site iron is primarily composed of hydrophobic residues so that the terminal oxygenase of ROs can accommodate hydrophobic substrates.

Table 9.2 A list of PDB entry codes of terminal oxygenase and electron components of ROs
Fig. 9.2
figure 2

Overall structures of Rieske oxygenase components. The iron atom, sulfur atom, FAD, and carbazole are shown in brown sphere, yellow sphere, yellow sticks, and yellow and blue sticks (imino nitrogen), respectively. (a) The terminal oxygenase component with α3 configuration (blue; PDB entry code: 2DE7), ferredoxin (red; PDB entry code: 1VCK), and ferredoxin reductase (orange; unpublished) of carbazole 1,9a-dioxygenase. Schematic electron transfer is also presented. Magnified view of the substrate-binding pocket is shown in the in set. (b) The terminal oxygenase of naphthalene dioxygenase with α3β3 configuration (PDB entry code: 1NDO). The α and β subunits of the terminal oxygenase component are colored in blue and green, respectively

2.3 Electron Transfer Components

The crystal structures of the electron transfer components of ROs have not been determined as many as the terminal oxygenase components (Table 9.2), but the structures have offered important insights into structure-function relationships of electron carrier proteins.

2.3.1 Reductase

The first three-dimensional structure of class I reductase component is phthalate dioxygenase from Burkholderia cepacia strain DB01 (Table 9.2) (Correll et al. 1992). This protein has three individual domains: FMN, NAD, and [2Fe-2S] domains (N-terminal to C-terminal). The FMN-binding, [2Fe-2S] cluster, and NADH-binding domains are brought together near a central cleft in the molecule with only 4.9 Å separating the 8-methyl group of FMN and a cysteine sulfur ligated to iron. The phthalate dioxygenase belongs to class IA but another example of reductase structure of class I, benzoate dioxygenase, belongs to class IB. The reductase component of benzoate dioxygenase from Acinetobacter baylyi strain ADP1 also has three domains, [2Fe-2S] cluster, FAD-binding, and NADH-binding domains (N-terminal to C-terminal), with different order from that of phthalate dioxygenase (Karlsson et al. 2002).

2.3.2 Ferredoxin

ROs in class II have a ferredoxin component which transfers electrons from ferredoxin reductase component to terminal oxygenase. Crystal structure of ferredoxin component of BDO (BphF) from Burkholderia xenovorans strain LB400 was determined as the first example of RO ferredoxin (Colbert et al. 2000). This small protein composed of 109 amino acid residues has a Rieske [2Fe-2S] cluster which is located at the apex of this molecule with its two histidine ligands exposed to solvent. Redox potential of BphF was approximately −150 mV, whereas Rieske proteins from the bc 1 or b 6 f complexes have redox potentials near +300 mV. Some other ferredoxin structures of ROs; two different CARDOs, NDO, toluene dioxygenase; and other two different BDOs have been determined (Table 9.2). They are primarily similar to BphF protein at structural level except for ferredoxin of CARDO from Novosphingobium sp. strain KA1, which is not Rieske-type ferredoxin but putidaredoxin type. This protein structure is round shape rather than wedge shape as the Rieske-type ferredoxins (Umeda et al. unpublished). This difference probably influences interaction with terminal oxygenase and ferredoxin reductase component.

2.3.3 Ferredoxin Reductase

Crystal structure of ferredoxin reductase (BphA4) in class IIB from Acidovorax sp. strain KKS102 with and without NADH has been determined as the first example of ferredoxin reductase in this class (Senda et al. 2000). This protein has significant similarity to not only ferredoxin reductases of ROs in other bacteria but also putidaredoxin reductase of P450cam system of Pseudomonas putida, and it has essentially the same fold as the enzymes of the glutathione reductase family. This protein has three domains, FAD-binding, NADH-binding, and C-terminal domain. The FAD-binding and C-terminal domains have interacting surface areas for its counterpart, ferredoxin component (BphA3) (Senda et al. 2007). Crystal structures of ferredoxin reductase components of CARDOs from Janthinobacterium sp. strain J3 and Novosphingobium sp. strain KA1 and toluene dioxygenase from Pseudomonas putida strain F1 have been also determined (Table 9.2). They are essentially similar to BphA4 structure.

3 Electron Transfer in Rieske Oxygenase

Both intra- and intermolecular electron transfers in ROs are essential for catalytic activity. Redox potential of each redox center is an important factor for electron transfer but limited numbers of reports which demonstrated experimentally measured redox potentials (Table 9.3). The Rieske proteins have reduction potentials ranging from −150 to +400 mV which is proposed to be due to electrostatic environment of the cluster and/or structural differences, whereas the redox centers in RO are relatively lower (Colbert et al. 2000; Brown et al. 2008). The redox potential of Rieske proteins is known to be dependent on pH probably because of the two iron-coordinating histidines. The wide range redox potential of Rieske proteins is first proposed to be due to solvent accessibility or protein structure. However, determined structures of Rieske proteins are significantly similar and do not have large variations in solvent accessibility. Thus, the redox potential is likely to be determined by structural difference rather than solvent accessibility. This is supported by biochemical studies using mutated enzymes (Klingen and Ullmann 2004).

Table 9.3 Redox potentials of Rieske oxygenase componentsa

One redox center must be in close proximity to the other so that electron transfer occurs. In class II or III ROs, the ferredoxin component shuttles between the ferredoxin reductase and terminal oxygenase components and thus the ferredoxin noncovalently binds to the other components. To date, two electron transfer complexes have been reported: ferredoxin reductase (BphA4) and ferredoxin (BphA3) of BDO from Acidovorax sp. strain KKS102 and ferredoxin and terminal oxygenase of CARDO (from Pseudomonas resinovorans strain CA10 and Janthinobacterium sp. strain J3, respectively). These crystal structures have revealed interacting site, binding force, and conformational changes by complex formation. The crystal structure of BphA4 and BphA3 complex demonstrated BphA3 binds to middle of BphA4 by hydrophobic interaction (four tryptophan residues in BphA4 and two proline residues in BphA3), electrostatic-like interaction (negatively charged surface of BphA4 interacts with positively charged BphA3 surface), and hydrogen bonds (Fig. 9.3a) (Senda et al. 2007). Electrons are likely to be transferred from FAD in BphA4 to the [2Fe-2S] cluster in BphA3 via Trp320 of BphA4 and His66 of BphA3. Structural comparison between the free and complex forms of BphA3 and BphA4 revealed conformational changes, clockwise rotation of NADH and C-terminal domain of BphA4, side chain rotation of His66 (a ligand in [2Fe-2S] cluster) of BphA3, and the peptide bond between Gly46 and Glu47 of BphA3. This report also presents the butterfly-like motion of the isoalloxazine ring of FAD along with redox status. The other electron transfer complex of ferredoxin and terminal oxygenase of CARDO revealed interacting amino acid residues between the components (Fig. 9.3b, c) (Ashikawa et al. 2006). Ferredoxin binds to boundary of the terminal oxygenase subunits by electrostatic and hydrophobic interactions. Free forms of ferredoxin and terminal oxygenase component of CARDO have also been determined. Detailed comparison revealed two significant conformational changes of main chain, Lys12-Trp15 and Asp347-Asn352 of the terminal oxygenase component. These amino acid residues are located at the component boundary to form hydrophobic interactions and the hydrophobic residues: Trp15 and Val351 of Oxy interact with Phe67 and Pro83 of Fd, respectively. On the interacting molecular surfaces, electrostatic interactions (Lys13, Arg118, Arg210, and Glu353 of the terminal oxygenase interact with Glu64, Glu43, Glu55, and His68 of the ferredoxin component, respectively) were also observed. According to the complex structure, electron transfer pathway from the terminal oxygenase and the ferredoxin was proposed (Fig. 9.3c). The distance between the Rieske cluster of ferredoxin and mononuclear iron of the terminal oxygenase is approximately 22 Å. This is too far for the direct electron transfer without Rieske cluster of the terminal oxygenase. However, the average distance between the ligand atoms of the Rieske clusters of terminal oxygenase and ferredoxin was approximately 12–13 Å which is shorter than 14 Å threshold, the limit of electron tunneling in a protein medium (Page et al. 1999).

Fig. 9.3
figure 3

The crystal structures of the electron transfer complexes. (a) The ferredoxin and ferredoxin reductase complex of biphenyl dioxygenase from Acidovorax sp. strain KKS102 (PDB entry code: 2YVJ). The ferredoxin and ferredoxin reductase are colored in red and orange, respectively. The [2Fe-2S] cluster, FAD, and NAD+ molecules are shown as the brown and yellow balls, yellow sticks, and green sticks, respectively. The Rieske [2Fe-2S] cluster of the ferredoxin binds closely to the FAD and NAD+ molecules in the middle of ferredoxin reductase component. (b) The terminal oxygenase and ferredoxin complex of carbazole dioxygenase from Janthinobacterium sp. strain J3 and Pseudomonas resinovorans strain CA10 (PDB entry code: 2DE7). Chains A, B, C (terminal oxygenase) and chains D, E, F (ferredoxin) are colored in light blue, marine, deep blue, light pink, red, and dark pink, respectively. The iron atom, sulfur atom, and carbazole are shown in brown sphere, yellow sphere, and yellow and blue (imino nitrogen) sticks, respectively. The ferredoxin components bind to the terminal oxygenase one for one in close proximity between each Rieske [2Fe-2S] cluster. (c) Stereoview of proposed electron transfer pathways between the ferredoxin and terminal oxygenase of carbazole dioxygenase. Two possible routes are presented as orange and red arrows, according to a report by Ashikawa et al. (2006)

The distance from Rieske [2Fe-2S] cluster of one terminal oxygenase α subunit and mononuclear iron of the neighboring α subunit is shorter than 14 Å as well. A conserved aspartate residue (D180 in CARDO; Fig. 9.3c) binds to a histidine ligand to the Rieske cluster and a histidine ligand to the mononuclear iron. A mutational study of the corresponding residue to the aspartate residue in toluene dioxygenase from Pseudomonas putida strain F1 (Asp219) demonstrated this residue is essential for enzyme activity (Jiang et al. 1996; Friemann et al. 2009), indicating the aspartate residue construct a main path for electron transfer from the Rieske cluster to mononuclear iron.

4 Catalytic Mechanism

In ROs, dioxygen is reductively activated as well as other oxidase and oxygenase enzymes. Biochemical studies of the ROs, especially for single turnover studies by regulating the enzymes to be oxidized and reduced, have provided important insights into reaction mechanisms. A scheme of proposed catalytic mechanism of ROs based on the previous studies (Ashikawa et al. 2012; Karlsson et al. 2003; Neibergall et al. 2007; Wolfe and Lipscomb 2003; Wolfe et al. 2001, 2002; Kovaleva and Lipscomb 2008) is presented in Fig. 9.4. Single turnover studies of NDO have revealed two electrons (reduction of one Fe of Rieske [2Fe-2S] cluster and the mononuclear iron), and substrate binding is required for oxygen reactivity (Wolfe et al. 2001). After substrate binding and oxygen reactivation, an Fe(III)-OOH intermediate is likely to undergo O–O bond homolytic fission to yield an Fe(V)-oxo-hydroxo intermediate. The cis-hydroxylated product is released when one or two electron(s) come from the electron transfer component. ROs with the Rieske cluster oxidized and the mononuclear iron reduced are under resting state. Addition of H2O2 to NDO under resting state allows reaction with the substrate to yield the product (Fig. 9.4) (Wolfe and Lipscomb 2003). This is so-called “peroxide shunt” reaction that does not require a reduced Rieske cluster. In benzoate 1,2-dioxygenase, even fully oxidized form with the Rieske cluster oxidized and mononuclear iron in the Fe(III) state can utilized H2O2 as a source of reduced oxygen atom to form the appropriate product from benzoate (Neibergall et al. 2007). Several crystal structures with substrate and oxygen and reaction intermediates suggested further detailed reaction mechanisms. Karlsson and coworkers revealed that a molecular oxygen species bound to the mononuclear iron in a side-on fashion by determining crystal structures of NDO with substrate and dioxygen (Karlsson et al. 2003). The side-on binding of dioxygen enables each oxygen to attach neighboring carbons from the same face of the aromatic ring to produce the cis-dihydrodiols. However, when NO is introduced instead of O2, NO is bound end-on to the mononuclear iron, whereas NO is commonly used as an analogue for dioxygen (Karlsson et al. 2005). Crystal structures of CARDO with substrate, dioxygen, and both carbazole and dioxygen revealed that a room for oxygen binding in the substrate-binding pocket is created by substrate binding, suggesting the substrate binds to the active site before dioxygen binding (Ashikawa et al. 2012). On the other hand, CARDO with the substrate revealed conformational changes with closure of the entranceway of the substrate by substrate binding (Fig. 9.5b, c) (Ashikawa et al. 2006). This conformational change seems to trap the substrate at the substrate-binding site and to exclude water from the active site during turnover in order to minimize the risk of leakage of partially processed substrates.

Fig. 9.4
figure 4

Scheme of proposed catalytic cycles of Rieske oxygenases. Oxidized Rieske [2Fe-2S] cluster, reduced Rieske [2Fe-2S] cluster, oxidized nonheme iron, and reduced nonheme iron are shown as [2Fe-2S]ox, [2Fe-2S]red, Fe3+, and Fe2+, respectively. Oxidized, reduced, and resting state of the redox centers in the terminal oxygenase are shown in red, blue, and purple background. Solid arrows indicate the proposed main pathway taking account into the biochemically and structurally well-characterized ROs. Broken arrow shows the possible pathway based on biochemical studies and the complex structures determined

Fig. 9.5
figure 5

Substrate-binding site of RO with a substrate (crystal structure of carbazole 1,9 a-dioxygenase and carbazole complex by Ashikawa et al. 2006). (a) Stereoview of carbazole bound to the substrate-binding pocket of CARDO. The amino acid residues which contact the substrate are shown in light pink sticks and carbazole is shown in a white and blue (imino nitrogen) sticks. (b) Superposition of the entrance of the substrate-binding pocket of CARDO. The carbazole-free and carbazole-binding structures are shown in blue green and green and orange and pink, respectively. (c and d) The surface plots of the carbazole-free (c) and carbazole-binding structures (d). White arrows indicate the entrance of the substrate-binding pocket. The insets show the molecular surfaces viewed from the orientation directed by the white arrows (Reprinted from Ashikawa et al. (2006), with permission of Elsevier)

5 Substrate Specificity

ROs are known for their broad range of substrate (not only aromatic compounds but cholesterols and chlorophylls as described above). ROs also catalyze monooxygenation, sulfoxidation, O- and N-dealkylation. Substrate specificity analyses and recent structural studies revealed what determines the substrate specificity and the relationship between structure and substrate specificity, which provides critical clues to the modification of the substrate-binding pocket to obtain an enhanced catalyst for bioremediation (degradation) of toxic contaminants and enzymatic synthesis of chiral precursors for the production of specialty compounds. Several examples of studies on mutant enzymes demonstrated that regio- and enantio-selectivity can be altered by the mutation as follows (Boyd and Sheldrake 1998).

BDO is one of the best-characterized RO by its mutant and chimeric enzymes (Furukawa et al. 2004). Randomly recombined terminal oxygenase components of BDO from Pseudomonas pseudoalcaligenes strain KF707 and Burkholderia xenovorans strain LB400 by DNA shuffling showed expanded substrate specificities not only for biphenyl derivatives and polychlorinated biphenyl (PCB) but for single aromatic compounds (Kumamaru et al. 1998). Subsequent random priming and site-directed mutants of BDO from strain KF707 showed expanded substrate range and different regio-specificities for various polychlorinated biphenyls from the wild-type enzyme (Suenaga et al. 2001, 2002). Many of the mutants which showed different substrate specificity from the wild-type enzyme had mutation at hydrophobic amino acid residues, such as Phe227, Ile335, and Phe377. According to structural studies of ROs, the substrate-binding pocket is commonly composed of hydrophobic residues consistent with RO’s preference for hydrophobic substrates (Fig. 9.5a). These mutational studies indicate that hydrophobic residues also play an important role in determining substrate-binding orientation. BDO from Burkholderia xenovorans strain LB400 is also well-studied enzyme by mutations. Mutated terminal oxygenase BphAEp4 (Thr335Ala/Phe336Met) and its derivative BphAERR41 metabolize dibenzofuran two and three times faster than wild-type enzyme, respectively (Mohammadi et al. 2011). The crystal structures of BphAEp4 and BphAERR41 revealed that replacement altered the constraints (hydrogen bonds) or plasticity of the substrate-binding pocket and increased available space for bulkier compounds than biphenyl, such as dibenzofuran and the doubly ortho-chlorinated congener 2,6-dichlorobiphenyl (Kumar et al. 2011; Mohammadi et al. 2011).

NDO is also one of the extensively studied ROs by mutational studies. Mutant enzymes of NDO from Pseudomonas putida strain NCIB9816-4 have demonstrated alteration of a series of amino acid residues composing the substrate-binding pocket can affect the regio-selectivity of product formation (Parales et al. 2000a, b; Yu et al. 2001). Detailed comparison between wild-type NDO and its amino acid replacement variants in which Phe352 is substituted by valine demonstrated region- and stereo-selectivity was principally dependent on the orientation of the substrate binding at the active site (Ferraro et al. 2006).

Another example of mutational study based on the crystal structure is CARDO from Janthinobacterium sp. strain J3. A series of site-directed mutagenesis variants in which the mutated amino acid residues are located at substrate-binding pocket wall showed different oxygenation site to hetero-aromatic compounds (Uchimura et al. 2008). Among the mutant enzymes, Ile262Val and Gln282Tyr converted carbazole to 1-hydroxycarbazole which is probably a dehydrated compound of lateral dioxygenation product (cis-1,2-dihydroxy-1,2-dihydrocarbazole) with higher yield than the wild-type enzyme. Another mutated variant Phe275Trp converted fluorene to 4-hydroxyfluorene (probable dehydration product of cis-3,4-dihydroxy-3,4-dihydrofluorene), whereas wild-type enzyme cannot catalyze this substrate. This variant also could oxidize fluoranthene to yield unidentified cis-dihydrodiol. Subsequent structural studies of mutated CARDO variants revealed difference in substrate-binding orientation between the variants and wild-type CARDO (Inoue et al. unpublished). In Ile262Val and Gln282Tyr variant structures with carbazole, the substrate bound to the active site rotated approximately 15° and 25°, respectively, from the substrate position of the wild type. In CARDO, imino nitrogen of carbazole forms hydrogen bonds with carbonyl oxygen of Gly178 at the active site (Fig. 9.5a). This hydrogen bond is likely to stabilize the substrate binding [in 2-oxoquinoline-8-monooxygenase from Pseudomonas putida, strain 86 also has similar hydrogen bond between NH group of the substrate, 2-oxoquinoline, and carbonyl oxygen of Gly216 (Martins et al. 2005)]. In the mutant enzymes, there is no “direct” hydrogen bonds between carbazole and the carbonyl oxygen of Gly178 but, instead, two water molecules were introduced into the space which created by the substrate rotation, and they formed different hydrogen bond networks with substrate and carbonyl oxygen of Gly178. The structural analysis of the mutant enzymes of CARDO suggested that different substrate orientation stabilized by the hydrogen bond networks increased formation of the lateral dioxygenation product.

Protein engineering technology combined with three-dimensional structure information has enabled alteration of substrate specificity of ROs. However, limited numbers of modification of the substrate specificity of ROs were reported so far. Further structural and biochemical studies of ROs will develop novel catalytic reactions and applicable enzymes.

6 Outlook

In this chapter structural and functional properties of aromatic-ring hydroxylating dioxygenases which is known as Rieske oxygenases were presented. In addition to the genetic and biochemical studies, recent structural studies of ROs have demonstrated interactions between the components, the substrate recognition, and catalytic mechanisms at the atomic level. Considering that ROs have been found to be extensively diverse and widespread in nature (Capyk and Eltis 2012; Iwai et al. 2011), basic structural and functional studies on each enzyme are becoming more and more important. The structural information and mutational studies with genetic and biochemical approach help each other to generate enzyme which is attractive to application for bioremediation and industrial biocatalysis.