Analysis of three 2,3-dihydroxybiphenyl 1,2-dioxygenases found in Rhodococcus globerulus P6. Identification of a new family of extradiol dioxygenases

Novel genes encoding 2-aminophenol 1,6-dioxygenase from Pseudomonas species AP-3 growing on 2-aminophenol and catalytic properties of the purified enzyme

2-Aminophenol 1,6-dioxygenase was purified from the cell extracts of Pseudomonas sp. AP-3 grown on 2-aminophenol. The product from 2-aminophenol by catalysis of the purified enzyme was identified as 2-aminomuconic 6-semialdehyde by gas chromatographic and mass spectrometric analyses. The molecular mass of the native enzyme was 140 kDa based on gel filtration. It was dissociated into molecular mass subunits of 32 (alpha-subunit) and 40 kDa (beta-subunit) by sodium dodecyl sulfate-polyacrylamide gel electrophoresis, indicating that the dioxygenase was a heterotetramer of alpha2beta2. The genes coding for the alpha- and beta-subunits of the enzyme were cloned and sequenced. Open reading frames of the genes (amnA and amnB) were 816 and 918 base pairs in length, respectively. The amino acid sequences predicted from the open reading frames of amnA and amnB corresponded to the NH2-terminal amino acid sequences of the alpha-subunit (AmnA) and beta-subunit (AmnB), respectively. The deduced amino acid sequences of AmnB showed identities to some extent with HpaD (25.4%) and HpcB (24.4%) that are homoprotocatechuate 2,3-dioxygenases from Escherichia coli W and C, respectively, belonging to class III in the extradiol dioxygenases. On the other hand, AmnA had identity (23.3%) with only AmnB among the enzymes examined.

Analysis of a new dimeric extradiol dioxygenase from a naphthalenesulfonate-degrading sphingomonad

A new extradiol dioxygenase was cloned by screening a gene bank from the naphthalenesulfonate-degrading bacterial strain BN6 for colonies with 2,3-dihydroxybiphenyl dioxygenase (DHBPDO) activity. A 1.6 kb DNA fragment was sequenced and an ORF of 954 bp identified. Comparison of the deduced amino acid sequence of DHBPDO II from strain BN6 with previously published sequences showed the closest relationship to a metapyrocatechase (MpcII) from Alcaligenes eutrophus JMP 222. Thus, the enzyme was only distantly related to the main groups of catechol 2,3-dioxygenases or DHBPDOs. The dioxygenase was expressed using a T7 expression vector and the enzymic characteristics of the protein were examined. The enzyme oxidized 2,3-dihydroxybiphenyl, 3-isopropylcatechol, 3-methylcatechol, 4-fluorocatechol and 1,2-dihydroxynaphthalene. Comparison of the UV/visible spectrum of the product formed from 3,5-dichlorocatechol with previous reports suggested that this substrate is oxidized by different extradiol dioxygenases either by proximal or distal ring cleavage. The enzyme required Fe2+ for maximal activity. In contrast to most other extradiol dioxygenases, the enzyme consisted of only two identical subunits.

Heterologous expression of biphenyl dioxygenase-encoding genes from a gram-positive broad-spectrum polychlorinated biphenyl degrader and characterization of chlorobiphenyl oxidation by the gene products

The bphA1A2A3A4 gene cluster, encoding a biphenyl dioxygenase from Rhodococcus globerulus P6, a gram-positive microorganism able to degrade a wide spectrum of polychlorobiphenyls (PCBs), was expressed in Pseudomonas putida, thereby allowing characterization of chlorobiphenyl oxidation by this enzyme. While P6 biphenyl dioxygenase activity was observed in P. putida containing bphA1A2A3A4, no activity was detected in Escherichia coli cells containing the same gene cluster. In E. coli, transcription of genes bphB and bphCl, located downstream of bphA1A2A3A4, was shown to be driven solely by a vector promoter, which indicated that the lack of biphenyl dioxygenase activity was not due to a lack of mRNA synthesis. Radioactive labelling of bph gene products in E. coli implied inefficient translation of the bphA gene cluster or rapid degradation of the gene products. The biosynthesis of functional P6 biphenyl dioxygenase in P. putida cells containing the same plasmid construct that yielded no activity in E. coli emphasizes the importance of the host strain for heterologous expression and shows that synthesis, correct folding, and assembly of a Rhodococcus biphenyl dioxygenase can be achieved in a gram-negative organism. Dioxygenation of six mono- and dichlorinated PCB congeners by P. putida containing the P6 bphA gene cluster indicates the following ring substitution preference for this reaction (from most to least preferred): un-, meta-, para-, and ortho-substitution. No indications were found for dioxygenation of meta/para carbon pairs, or for hydroxylation of chlorinated carbons at any position of a monochlorinated ring, suggesting a strict specificity of this biphenyl dioxygenase for attack at nonhalogenated ortho/meta vicinal carbons. This contrasts the properties of an analogous enzyme from Pseudomonas sp. strain LB400, which can both dioxygenate at meta and para positions and dehalogenate substituted ortho carbons during ortho and meta dioxygenation.

Manganese(II) active site mutants of 3,4-dihydroxyphenylacetate 2,3-dioxygenase from Arthrobacter globiformis strain CM-2

Whereas all other members of the extradiol-cleaving catechol dioxygenase family are iron-dependent, the 3,4-dihydroxyphenylacetate 2,3-dioxygenase (MndD) from Arthrobacter globiformis CM-2 is dependent on manganese for catalytic activity. Recently, the endogenous iron ligands of one family member, the 2,3-dihydroxybiphenyl 1,2-dioxygenase (BphC), were identified crystallographically as two histidines and a glutamic acid [Sugiyama, K., et al. (1995) Proc. Jpn. Acad., Ser. B 71, 32-35; Han, et al. (1995) Science 270, 976-980; Senda, T., et al. (1996) J. Mol. Biol. 255, 735-752]. Though BphC and MndD have low overall sequence identity (23%), the three BphC metal ligands are all conserved in MndD (H155, H214, and E266). In order to determine whether these residues also act as ligands to manganese in MndD, site-directed mutants of each were constructed, purified, and analyzed for activity and metal content. Mutations H155A, H214A, and E266Q yielded purified enzymes with specific activities of <0.1% of that of the wild-type dioxygenase and bound 0.4, 1.8, and 33% of the wild-type level of manganese, respectively. The relatively high level of manganese [with a Mn(II) EPR signal distinctly different from that of the wild-type enzyme] observed for E266Q suggests that the glutamine may act as a weak ligand to the metal. Mutant E266D, which retains the potential metal binding capability of a carboxylate group, exhibited 12% of the wild-type activity in crude extracts, suggesting that Mn remains bound; however, this mutant protein was too unstable to be purified and analyzed for metal content. On the basis of the low activity and metal content of mutant proteins, we propose that the conserved residues H155, H214, and E266 ligate manganese in MndD. As is the case with the superoxide dismutases, the extradiol-cleaving catechol dioxygenases appear to utilize identical coordinating residues for their iron- and manganese-dependent enzymes.

Evolutionary relationships among extradiol dioxygenases

A structure-validated alignment of 35 extradiol dioxygenase sequences including two-domain and one-domain enzymes was derived. Strictly conserved residues include the metal ion ligands and several catalytically essential active site residues, as well as a number of structurally important residues that are remote from the active site. Phylogenetic analyses based on this alignment indicate that the ancestral extradiol dioxygenase was a one-domain enzyme and that the two-domain enzymes arose from a single genetic duplication event. Subsequent divergence among the two-domain dioxygenases has resulted in several families, two of which are based on substrate preference. In several cases, the two domains of a given enzyme express different phylogenies, suggesting the possibility that such enzymes arose from the recombination of genes encoding different dioxygenases. A phylogeny-based classification system for extradiol dioxygenases is proposed.

Catechol dioxygenases from Escherichia coli (MhpB) and Alcaligenes eutrophus (MpcI): sequence analysis and biochemical properties of a third family of extradiol dioxygenases

The nucleotide sequence of the Escherichia coli mhpB gene, encoding 2,3-dihydroxyphenylpropionate 1,2-dioxygenase, was determined by sequencing of a 3.1-kb fragment of DNA from Kohara phage 139. The inferred amino acid sequence showed 58% sequence identity with the sequence of an extradiol dioxygenase, MpcI, from Alcaligenes eutrophus and 10 to 20% sequence identity with protocatechuate 4,5-dioxygenase from Pseudomonas paucimobilis, with 3,4-dihydroxyphenylacetate 2,3-dioxygenase from E. coli, and with human 3-hydroxyanthranilate dioxygenase. Sequence similarity between the N- and C-terminal halves of this new family of dioxygenases was detected, with conserved histidine residues in the N-terminal domain. A model is proposed to account for the relationship between this family of enzymes and other extradiol dioxygenases. The A. eutrophus MpcI enzyme was expressed in E. coli, purified, and characterized as a protein with a subunit size of 33.8 kDa. Purified MhpB and MpcI showed similar substrate specificities for a range of 3-substituted catechols, and evidence for essential histidine and cysteine residues in both enzymes was obtained.

Cloning, sequencing, and characterization of the Bacillus subtilis biotin biosynthetic operon

A 10-kb region of the Bacillus subtilis genome that contains genes involved in biotin-biosynthesis was cloned and sequenced. DNA sequence analysis indicated that B. subtilis contains homologs of the Escherichia coli and Bacillus sphaericus bioA, bioB, bioD, and bioF genes. These four genes and a homolog of the B. sphaericus bioW gene are arranged in a single operon in the order bioWAFDR and are followed by two additional genes, bioI and orf2. bioI and orf2 show no similarity to any other known biotin biosynthetic genes. The bioI gene encodes a protein with similarity to cytochrome P-450s and was able to complement mutations in either bioC or bioH of E. coli. Mutations in bioI caused B. subtilis to grow poorly in the absence of biotin. The bradytroph phenotype of bioI mutants was overcome by pimelic acid, suggesting that the product of bioI functions at a step prior to pimelic acid synthesis. The B. subtilis bio operon is preceded by a putative vegetative promoter sequence and contains just downstream a region of dyad symmetry with homology to the bio regulatory region of B. sphaericus. Analysis of a bioW-lacZ translational fusion indicated that expression of the biotin operon is regulated by biotin and the B. subtilis birA gene.

Catechol 2,3-dioxygenases functional in oxygen-limited (hypoxic) environments

We studied the degradation of toluene for bacteria isolated from hypoxic (i.e., oxygen-limited) petroleum-contaminated aquifers and compared such strains with other toluene degraders. Three Pseudomonas isolates, P. pickettii PKO1, Pseudomonas sp. strain W31, and P. fluorescens CFS215, grew on toluene when nitrate was present as an alternate electron acceptor in hypoxic environments. We examined kinetic parameters (K(m) and Vmax) for catechol 2,3-dioxygenase (C230), a key shared enzyme of the toluene-degradative pathway for these strains, and compared these parameters with those for the analogous enzymes from archetypal toluene-degrading pseudomonads which did not show enhanced, nitrate-dependent toluene degradation. C230 purified from strains W31, PKO1, and CFS215 had a significantly greater affinity for oxygen as well as a significantly greater rate of substrate turnover than found for the analogous enzymes from the TOL plasmid (pWW0) of Pseudomonas putida PaW1, from Pseudomonas cepacia G4, or from P. putida F1. Analysis of the nucleotide and deduced amino acid sequences of C23O from strain PKO1 suggests that this extradiol dioxygenase belongs to a new cluster within the subfamily of C23Os that preferentially cleave monocyclic substrates. Moreover, deletion analysis of the nucleotide sequence upstream of the translational start of the meta-pathway operon that contains tbuE, the gene that encodes the C230 of strain PKO1, allowed identification of sequences critical for regulated expression of tbuE, including a sequence homologous to the ANR-binding site of Pseudomonas aeruginosa PAO. When present in cis, this site enhanced expression of tbuE under oxygen-limited conditions. Taken together, these results suggest the occurrence of a novel group of microorganisms capable of oxygen-requiring but nitrate-enhanced degradation of benzene, toluene, ethylbenzene, and xylenes in hypoxic environments. Strain PKO1, which exemplifies this novel group of microorganisms, compensates for a low-oxygen environment by the development of an oxygen-requiring enzyme with kinetic parameters favorable to function in hypoxic environments, as well as by elevating synthesis of such an enzyme in response to oxygen limitation.

p-Cumate catabolic pathway in Pseudomonas putida Fl: cloning and characterization of DNA carrying the cmt operon

Pseudomonas putida F1 utilizes p-cumate (p-isopropylbenzoate) as a growth substrate by means of an eight-step catabolic pathway. A 35.75-kb DNA segment, within which the cmt operon encoding the catabolism of p-cumate is located, was cloned as four separate overlapping restriction fragments and mapped with restriction endonucleases. By examining enzyme activities in recombinant bacteria carrying these fragments and sub-cloned fragments, genes encoding most of the enzymes of the p-cumate pathway were located. Subsequent sequence analysis of 11,260 bp gave precise locations of the 12 genes of the cmt operon. The first three genes, cmtAaAbAc, and the sixth gene, cmtAd, encode the components of p-cumate 2,3-dioxygenase (ferredoxin reductase, large subunit of the terminal dioxygenase, small subunit of the terminal dioxygenase, and ferredoxin, respectively); these genes are separated by cmtC, which encodes 2,3-dihydroxy-p-cumate 3,4-dioxygenase, and cmtB, coding for 2,3-dihydroxy-2,3-dihydro-p-cumate dehydrogenase. The ring cleavage product, 2-hydroxy-3-carboxy-6-oxo-7-methylocta-2,4-dienoate, is acted on by a decarboxylase encoded by the seventh gene, cmtD, which is followed by a large open reading frame, cmtI, of unknown function. The next four genes, cmtEFHG, encode 2-hydroxy-6-oxo-7-methylocta-2,4-dienoate hydrolase, 2-hydroxypenta-2,4-dienoate hydratase, 4-hydroxy-2-oxovalerate aldolase, and acetaldehyde dehydrogenase, respectively, which transform the decarboxylation product to amphibolic intermediates. The deduced amino acid sequences of all the cmt gene products except CmtD and CmtI have a recognizable but low level of identity with amino acid sequences of enzymes catalyzing analogous reactions in other catabolic pathways. This identity is highest for the last two enzymes of the pathway (4-hydroxy-2-oxovalerate aldolase and acetaldehyde dehydrogenase [acylating]), which have identities of 66 to 77% with the corresponding enzymes from other aromatic meta-cleavage pathways. Recombinant bacteria carrying certain restriction fragments bordering the cmt operon were found to transform indole to indigo. This reaction, known to be catalyzed by toluene 2,3-dioxygenase, led to the discovery that the tod operon, encoding the catabolism of toluene, is located 2.8 kb downstream from and in the same orientation as the cmt operon in P. putida F1.

Crystal structure of the biphenyl-cleaving extradiol dioxygenase from a PCB-degrading pseudomonad

Polychlorinated biphenyls (PCBs) typify a class of stable aromatic pollutants that are targeted by bioremediation strategies. In the aerobic degradation of biphenyl by bacteria, the key step of ring cleavage is catalyzed by an Fe(II)-dependent extradiol dioxygenase. The crystal structure of 2,3-dihydroxybiphenyl 1,2-dioxygenase from a PCB-degrading strain of Pseudomonas cepacia has been determined at 1.9 angstrom resolution. The monomer comprises amino- and carboxyl-terminal domains. Structural homology between and within the domains reveals evolutionary relationships within the extradiol dioxygenase family. The iron atom has five ligands in square pyramidal geometry: one glutamate and two histidine side chains, and two water molecules.

Sequence and expression of the bpdC1C2BADE genes involved in the initial steps of biphenyl/chlorobiphenyl degradation by Rhodococcus sp. M5

The nucleotide (nt) sequence of the bpdC1C2BADE genes which encode the first three enzymes in the biphenyl (BP) degradation pathway of Gram+ Rhodococcus sp. M5 (formerly Arthrobacter M5) was determined. Except for the ferredoxin component (BpdB) of the initial BP dioxygenase, the predicted amino acid (aa) sequences of the remaining proteins are found to be more closely related to the counterpart proteins (TodC1C2BADE) present in the toluene-degrader, Pseudomonas putida F1, than those of three BP-degrading pseudomonads. The cloned bpd genes were verified by their expression in the Escherichia coli T7 RNA polymerase/promoter system. In E. coli, BpdA was able to complement TodC1C2B in indigo biosynthesis, although the M5 native or cloned BP dioxygenase does not carry out this reaction.

Characterization of a 2,3-dihydroxybiphenyl dioxygenase from the naphthalenesulfonate-degrading bacterium strain BN6

An extradiol dioxygenase was cloned from the naphthalenesulfonate-degrading bacterial strain BN6 by screening a gene bank for colonies with 2,3-dihydroxybiphenyl dioxygenase activity. DNA sequence analysis of a 1,358-bp fragment revealed an open reading frame of only 486 bp. This is the smallest gene encoding an extradiol dioxygenase found until now. Expression of the gene in a T7 expression vector enabled purification of the enzyme. Gel filtration and sodium dodecyl sulfate-polyacrylamide gel electrophoresis analysis showed that the protein was a dimer with a subunit size of 21.7 kDa. The enzyme oxidized 2,3-dihydroxybiphenyl, 3-isopropylcatechol, 3- and 4-chlorocatechol, and 3- and 4-methylcatechol. Since the ability to convert 3-chlorocatechol is an unusual characteristic for an extradiol-cleaving dioxygenase, this reaction was analyzed in more detail. The deduced amino-terminal amino acid sequence differed from the corresponding sequence of the 1,2-dihydroxynaphthalene dioxygenase, which had been determined earlier from the enzyme purified from this strain. This indicates that strain BN6 carries at least two different extradiol dioxygenases.

Molecular and biochemical characterization of two meta-cleavage dioxygenases involved in biphenyl and m-xylene degradation by Beijerinckia sp. strain B1

Beijerinckia sp. strain B1 is able to grow on either biphenyl or m-xylene as the sole source of carbon and is capable of cooxidizing many polycyclic aromatic hydrocarbons. The catabolic pathways for biphenyl and m-xylene degradation are coinduced and share common downstream enzymatic reactions. The catabolic pathway for biphenyl degradation involves two meta-cleavage steps, one for 2,3-dihydroxybiphenyl and a second for catechol. The catabolic pathway for m-xylene involves one m-cleavage step for 3-methylcatechol. The genes for two meta-cleavage dioxygenases were cloned from Beijerinckia sp. strain B1 on a single fragment of genomic DNA. The two genes are located approximately 5.5 kb away from one another. Expression of each gene separately in Escherichia coli and analysis of the meta-cleavage dioxygenase produced showed that one enzyme was more specific for 2,3-dihydroxybiphenyl while the second was more specific for catechol. The genes for the two meta-cleavage enzymes were thus labeled bphC and xylE for 2,3-dihydroxybiphenyl 1,2-dioxygenase and catechol 2,3-dioxygenase, respectively. Nondenaturing polyacrylamide gel electrophoresis followed by enzyme activity staining showed that the two meta-cleavage dioxygenases could be easily separated from each other. Similar analyses of Beijerinckia sp. strain B1 grown on succinate, biphenyl, or m-xylene indicate that both meta-cleavage enzymes are induced when cells are grown on either biphenyl or m-xylene. The nucleotide sequence was determined for both bphC and xylE. The two genes are transcribed in opposite directions, demonstrating that at least two operons must be involved in biphenyl degradation by Beijerinckia sp. strain B1. Analysis of the deduced amino acid sequence indicates that 2,3-dihydroxybiphenyl 1,2-dioxygenase (BphC) falls into the class of meta-cleavage dioxygenases acting on dihydroxylated polycyclic aromatic hydrocarbons and is somewhat distinct from the main group of meta-cleavage dioxygenases acting on 2,3-dihydroxybiphenyl. Catechol 2,3-dioxygenase (XyIE) falls into the class of meta-cleavage enzymes acting on dihydroxylated monocyclic aromatic hydrocarbons but shows little similarity to the canonical TOL plasmid-encoded catechol 2,3-dioxygenase.

Characterization of biphenyl catabolic genes of gram-positive polychlorinated biphenyl degrader Rhodococcus sp. strain RHA1

Rhodococcus sp. strain RHA1 is a gram-positive polychlorinated biphenyl (PCB) degrader which can degrade 10 ppm of PCB48 (equivalent to Aroclor1248), including tri-, tetra-, and pentachlorobiphenyls, in a few days. We isolated the 7.6-kb EcoRI-BamHI fragment carrying the biphenyl catabolic genes of RHA1 and determined their nucleotide sequence. On the basis of deduced amino acid sequence homology, we identified six bph genes, bphA1A2A3A4, bphB, and bphC, that are responsible for the initial three steps of biphenyl degradation. The order of bph genes in RHA1 is bphA1A2A3A4-bphC-bphB. This gene order differs from that of other PCB degraders reported previously. The amino acid sequences deduced from the RHA1 bph genes have a higher degree of homology with the tod genes from Pseudomonas putida F1 (49 to 79%) than with the bph genes of Pseudomonas sp. strains KF707 and KKS102 (30 to 65%). In Escherichia coli, bphA gene activity was not observed even when expression vectors were used. The activities of bphB and bphC, however, were confirmed by observing the transformation of biphenyl to a meta-cleavage compound with the aid of benzene dioxygenase activity that complemented the bphA gene activity (S. Irie, S. Doi, T. Yorifuji, M. Takagi, and K. Yano, J. Bacteriol. 169:5174-5179, 1987). The expected products of the cloned bph genes, except bphA3, were observed in E. coli in an in vitro transcription-translation system. Insertion mutations of bphA1 and bphC of Rhodococcus sp. strain RHA1 were constructed by gene replacement with cloned gene fragments.(ABSTRACT TRUNCATED AT 250 WORDS)

The evolutionary relationship of biphenyl dioxygenase from gram-positive Rhodococcus globerulus P6 to multicomponent dioxygenases from gram-negative bacteria

The Gram+ bacterium Rhodococcus globerulus P6 (RgP6) catabolizes a range of polychlorinated biphenyl (PCB) congeners, thus being of interest in bioelimination processes for PCB. The first step in the pathway, a dioxygenase attack of one of the biphenyl (BP) rings, is catalyzed by biphenyl dioxygenase (BDO). In this study, the nucleotide (nt) sequences of the four clustered cistrons, bphA1A2A3A4, encoding the subunits of BDO and forming part of the bph operon of RgP6 for BP degradation, were determined. A conserved motif proposed to bind a Rieske-type [2Fe-2S] cluster was identified in the deduced amino acid (aa) sequence of both the a subunit of the terminal oxygenase (BphA1) and ferredoxin (BphA3). The ferredoxin reductase subunit (BphA4) contains conserved sites for FAD and NADH binding. Deduced aa sequences of the BDO subunits shared homologies to multicomponent aromatic ring-hydroxylating dioxygenases from Gram- microorganisms. Stronger identity was found to toluene dioxygenase (TDO) of Pseudomonas putida F1 than to other BDO. Aa sequence comparisons suggest that BP degradation genes of RgP6 may have originated in Gram- microorganisms, probably Pseudomonas, and subsequently transferred to this Gram+ bacterium.

A manganese-dependent dioxygenase from Arthrobacter globiformis CM-2 belongs to the major extradiol dioxygenase family

Almost all bacterial ring cleavage dioxygenases contain iron as the catalytic metal center. We report here the first available sequence for a manganese-dependent 3,4-dihydroxyphenylacetate (3,4-DHPA) 2,3-dioxygenase and its further characterization. This manganese-dependent extradiol dioxygenase from Arthrobacter globiformis CM-2, unlike iron-dependent extradiol dioxygenases, is not inactivated by hydrogen peroxide. Also, ferrous ions, which activate iron extradiol dioxygenases, inhibit 3,4-DHPA 2,3-dioxygenase. The gene encoding 3,4-DHPA 2,3-dioxygenase, mndD, was identified from an A. globiformis CM-2 cosmid library. mndD was subcloned as a 2.0-kb SmaI fragment in pUC18, from which manganese-dependent extradiol dioxygenase activity was expressed at high levels in Escherichia coli. The mndD open reading frame was identified by comparison with the known N-terminal amino acid sequence of purified manganese-dependent 3,4-DHPA 2,3-dioxygenase. Fourteen of 18 amino acids conserved in members of the iron-dependent extradiol dioxygenase family are also conserved in the manganese-dependent 3,4-DHPA 2,3-dioxygenase (MndD). Thus, MndD belongs to the extradiol family of dioxygenases and may share a common ancestry with the iron-dependent extradiol dioxygenases. We propose the revised consensus primary sequence (G,T,N,R)X(H,A)XXXXXXX(L,I,V,M,F)YXX(D,E,T,N,A)PX(G,P) X(2,3)E for this family. (Numbers in brackets indicate a gap of two or three residues at this point in the sequence.) The suggested common ancestry is also supported by sequence obtained from genes flanking mndD, which share significant sequence identity with xylJ and xylG from Pseudomonas putida.