INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

From the relatively low frequency of its appearance in the literature during the last 2 decades, the gentisate (2,5-dihydroxybenzoate) pathway in which a p-dihydroxylated aromatic ring is oxidatively cleaved might appear to be a less common route for bacterial aromatic catabolism than either of the more extensively studied pathways through catechols (o-dihydroxybenzenes). However, gentisate and substituted gentisates serve as key intermediates in the aerobic pathways for the metabolism of a large number of aromatic compounds, including 3-hydroxybenzoate (15, 24), substituted phenols (8, 23, 39), salicylate (37, 40), 3,6-dichloro-2-methoxybenzoate (53), and naphthalene (14, 17, 34). Ring cleavage of gentisate is catalyzed by gentisate 1,2-dioxygenase (GDO; EC1.13.11.4) to form maleylpyruvate (26). The further conversion of maleylpyruvate to central metabolites has been reported to proceed either by direct hydrolysis to pyruvate and maleate (5, 22, 39) or by isomerization to fumarylpyruvate and subsequent hydrolysis to fumarate and pyruvate (5, 8, 24, 27, 41, 49).

GDO has been purified from a number of organisms (7, 10, 13, 20, 53), partial N-terminal amino acid sequences have been determined for six (10, 13, 20, 53), and one has been subjected to mechanistic studies (19). However, the first nucleotide sequences for a GDO gene have appeared in the databanks only very recently: nagI in the catabolism of naphthalene by Ralstonia sp. strain U2 (accession number AF036940), gtdA in the catabolism of 3,6-dichloro-2-methoxybenzoate from Sphingomonas sp. strain RW5 (AJ224977), an unnamed gene in the catabolism of 3-phenylpropionate from the extreme halophile Haloferax sp. strain D1227 (AF069949), and xlnE in the xylenol-degrading Pseudomonas alcaligenes P25X (AF173167). The GDO genes from Sphingomonas sp. strain RW5 and from Haloferax sp. strain D1227 have been expressed and partially characterized (13, 53). However, the genetic determinants of the further catabolism of maleylpyruvate have not yet been described, although DNA encoding maleylpyruvate isomerase and fumarylpyruvate hydrolase activities has been cloned from both Klebsiella pneumoniae (41) and Pseudomonas putida (23) but without any nucleotide sequence information.

Ralstonia sp. strain U2 (originally classified as a Pseudomonas sp.) converts naphthalene to central metabolites via gentisate. A 5.7-kb region encoding both naphthalene dioxygenase (nagAaAbAcAd) and salicylate 5-hydroxylase (nagGH) has previously been sequenced and functionally analyzed, revealing a novel gene order of nagAaGHAbAcAd (14). In this paper, we report the cloning and sequencing of overlapping fragments from strain U2 covering a continuous 21.6 kb of the nag genes and the characterization of genes encoding the enzymes involved in catabolism to central metabolites of gentisate formed from naphthalene.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Bacterial plasmids and strains. The bacterial plasmids used and constructed in this study are listed in Table 1. Escherichia coli DH5 [80dlacZM15 recA1 endA1 gyrA96 thi-1 hsdR17(r_K m_K⁺) supE44 relA1 deoR (lacZYA-argF)U169] (Life Technologies, Paisley, United Kingdom) was used routinely as a host in cloning experiments. E. coli BL21(DE3)pLysS [F ompT hsdS_B(r_B m_B) dcm gal (DE3) pLysS (Cm^r)] (48) was purchased from Promega and used as the host for the overexpression of genes cloned in expression plasmid pET5a.

Media and bacterial culture. Liquid Luria-Bertani (LB) medium (32) containing the appropriate antibiotic was used for the cultivation of E. coli strains. Sensitivity test agar (LabM, Bury, United Kingdom) or LB agar were used with added ampicillin for the selection of strains carrying plasmids derived from pUC18, pET5a, or pGEM-TEasy. Minimal medium was prepared according to the description in reference 56. LB and minimal medium plates contained 1.5% agar (LabM). Cultures of strain U2 were grown on 10 mM Na succinate or 0.5% (wt/vol) naphthalene as a carbon source. The latter was added as powdered solid directly to liquid medium or spread on the lids of inverted petri dishes. Ampicillin was used at 100 µg/ml and kanamycin was used at 50 µg/ml where appropriate.

Plasmid extraction and DNA manipulation. Plasmid DNA was extracted from strain U2 by the method of Wheatcroft and Williams (54). Restriction endonuclease digestions and ligations with T4 ligase were done in accordance with the manufacturer's instructions. E. coli DH5 was transformed by standard procedures (42). Plasmids were purified by using the Concert Rapid Plasmid Mini System (Life Technologies).

Cloning of the 16S rRNA gene (rDNA) from Ralstonia sp. strain U2. The 27f and the 1492r universal primers (28) were used to amplify the 16S rDNA from strain U2 by PCR. The fragment generated was purified by agarose gel electrophoresis and band extraction before it was cloned into the pGEM-T vector.

Cloning of Ralstonia sp. strain U2 plasmid DNA. In order to isolate DNA overlapping the 3' end of pWWF6 (Fig. 1), digested plasmid DNA from strain U2 was ligated into pUC18. Transformants were screened for expression of 1,2-dihydroxynaphthalene dioxygenase (NagC) activity by spraying with 3-methylcatechol and selecting colonies producing the yellow product 2-hydroxy-6-oxohepta-2,4-dienoate. pWWF19 with a 7.6-kb BglII insert (Fig. 1) was thus selected. To identify fragments overlapping the 3' end of pWWF19 (Fig. 1), Southern hybridizations (46) using the ECL direct nucleic acid labeling and detection system (Amersham) were made against digests of plasmid DNA from U2. The probe was a 401-bp PCR-generated fragment (bases 16580 to 16980) obtained from pWWF19 by using two primers, 5'-TCGGCTTGACGAAAAATACG and 5'-GGATGCAGGCGTCAGCAGAA. The region of the gel that hybridized to the probe was excised, and its DNA was extracted and cloned into pUC18. Transformant colonies carrying DNA homologous to the probe were identified by colony hybridization, and pWWF60, carrying an 8.9-kb EcoRI insert, was finally selected (Fig. 1).

View larger version (23K): [in this window] [in a new window]   FIG. 1.   Physical map of the nag gene cluster from the large catabolic plasmid of Ralstonia sp. strain U2 and inserts of recombinant plasmids constructed from it. ORFs are marked by the open arrows, with the direction of the arrowheads indicating the direction of transcription. The restriction sites are designated as follows: B, BamHI; Bg, BglII; E, EcoRI; H, HindIII; M, MspA1 I; P, PvuII; Sa, SalI; Sg, SgfI; Sm, SmaI; X, XhoI. Nd* represents an NdeI site and E* represents an EcoRI site which were engineered in the PCR primer used to generate the inserts of pET5a constructs. Sg represents the modified SgfI site causing a 2-bp deletion and a frameshift in nagK. Only restriction sites relevant to the clones and constructs indicated in this figure are shown.

Expression of nagJ, nagI, nagK, nagL, nagM, and nagN. Pairs of oligonucleotide primers were designed to produce PCR fragments of each nag gene singly and nagM and nagN together. These were designed with (i) an NdeI site introduced at the putative ATG start codon of each reading frame, (ii) an EcoRI restriction site upstream of the NdeI site, and (iii) an EcoRI restriction site downstream of the gene. Initially the amplified fragment was cut at the two EcoRI sites and inserted first into pUC18. In this way, the nagJ, nagI, nagK, nagL, nagMN, nagM, and nagN genes were amplified from pWWF60 by using Pfu polymerase (Promega) to create pWWF41, pWWF19-24, pWWF42, pWWF43, pWWF61, pWWF62, and pWWF63, respectively (Table 1). The inserts of these clones were sequenced on a single strand to ensure that no mutation had been incorporated during the PCR. Fragments from each of these clones were then excised with NdeI and EcoRI, religated into the expression vector pET5a, and transformed into E. coli DH5 to produce plasmids pWWF51 (nagJ), pWWF19-25 (nagI), pWWF52 (nagK), pWWF53 (nagL), pWWF71 (nagMN), pWWF72 (nagM), and pWWF73 (nagN). All the pET5a constructs were subsequently transformed into E. coli BL21(DE3)pLysS. In the following list of the PCR primers used to make these plasmids, the NdeI site is italicized, the EcoRI site used for cloning into pUC18 is underlined, and the bases that differ from those in the wild-type sequence are in boldface: nagJ (forward), 5'-TTGATTTTAAGAATTCATATGAAGCTTTATTACAGC-3'; nagJ (reverse), 5'-GATACGCTCGAATTCCATCAAGCATGTGGATAGG-3'; nagI (forward), 5'-AGCGAATTCATATGAGTCACGAACTTGGCCG-3'; nagI (reverse), 5'-ACCGAATTCCACCAGGGTTTGTGGCGATTT-3'; nagK (forward), 5'-ACGCGAATTCGGTGCATATGGGCCGTCCG-3'; nagK (reverse), 5'-AGCGCGGGAATTCACTGCTGTGGGTTCAG-3'; nagL (forward), 5'-TGAGCCGAATTCATATGAAGCTGTACAACTTCTGG-3'; nagL (reverse), 5'-CCTTACGAATTCACTATTGCTCATTGATACTTGTC-3'; nagM (forward), 5'-CAGTGAGAATTCCATATGAGCAATAGTG-3'; nagM (reverse), 5'-CGGTTGGAATTCGACGCTTCTTAC-3'; nagN (forward), 5'-GAGGGAATTCCATATGATCAAACGCC-3'; nagN (reverse), 5'-TAGCCGAATTCAACTGCGCCAGGT-3'; nagMN (forward), 5'-CAGTGAGAATTCCATATGAGCAATAGTG-3'; and nagMN (reverse), 5'-AAGTGGGCGAATTCAACGAGAGGTTA'-3.

Inactivation of nagK. pWWF100 was digested at its single SgfI site within nagK. The 3' overhangs of the SgfI site were converted to blunt ends by 3'-to-5' exonuclease action of T4 DNA polymerase (Promega), and the blunt ends were religated to produce pWWF105 (Fig. 1; Table 1.)

Preparation of cell extract. Cell extracts were prepared by resuspending the bacterial pellets in ice-cold 100 mM phosphate buffer (pH 7.4; approximately 0.1 g [wet weight]/ml) and disrupting them by sonication in an ice-water bath for three periods of 30 s with 30-s intervals, after which cell debris was removed by centrifugation at 100,000 × g for 1 h at 4°C.

Sequence determination and analysis. Nucleotide sequences were determined by MWG-Biotech Ltd. (Ebersberg, Germany). Sequences were analyzed with the Lasergene software package (DNASTAR). BLASTP was used for the deduced amino acid identity search, and BLASTN was used for the nucleotide sequence identity search (1). Pfam was used for the protein family search (4).

Enzyme assays. All assays were performed in 100 mM phosphate buffer (pH 7.4). GDO was assayed by measuring the increase in absorbance at 330 nm due to conversion of gentisate to maleylpyruvate (26), the molar extinction coefficient of which was taken as 13,000 M¹ cm¹ (26). GDO oxidation of substituted gentisates was assayed by measuring the increase in absorbance at 324 nm (for 3-methylgentisate), 334 nm (for 3-bromogentisate), and 324 nm (for 3-isopropylgentisate) due to their corresponding products (20): their molar extinction coefficients were taken as 13,500 M¹ cm¹, 11,800 M¹ cm¹, and 14,100 M¹ cm^1, respectively (20). The substituted gentisates were kindly supplied by C. L. Poh, National University of Singapore. Maleylpyruvate isomerase was qualitatively monitored by measuring the change in absorbance at 330 nm due to maleylpyruvate disappearance in the presence of glutathione (GSH). For quantitative determination of specific activity, the rate of change of A₃₃₀ was measured in the presence of excess fumarylpyruvate hydrolase (NagK). Fumarylpyruvate hydrolase was assayed by measuring the decrease in A₃₄₀ due to fumarylpyruvate disappearance: the molar extinction coefficient of fumarylpyruvate was taken as 9,400 M¹ cm¹ (49). Protein concentrations were determined by the biuret procedure. One unit of enzyme activity is defined as the amount required for the disappearance (or production) of 1 µmol of substrate (or product) per min at 30°C. Specific activities are expressed as units per milligram of protein.

Determination of kinetic parameters. To determine K_m values, initial velocities were measured at several nonsaturating concentrations of each compound. Preliminary experiments determined the approximate value of K_m, and the accurate rate determinations were then performed with 10 substrate concentrations spanning the approximate K_m value. Initial velocities were analyzed by the direct linear method using the program EnzPack, which calculates the most probable values for the kinetic parameters with their 68% confidence limits (55).

Analysis of fumarate and pyruvate production from fumarylpyruvate by fumarylpyruvate hydrolase using HPLC. The fumarylpyruvate hydrolase reaction was initiated by adding 160 µl of cell extract of E. coli DH5(pWWF96 [nagK]) to 20 ml of 100 µM fumarylpyruvate in 100 mM phosphate buffer (pH 7.4), produced in situ from gentisate by NagI and NagL. Samples (1.8 ml) were withdrawn from the reaction mixture at regular intervals, and the enzyme reaction was stopped by adding H₂SO₄ to each to a final concentration of 50 mM. Fumarate and pyruvate in each sample were measured by high-performance liquid chromatography (HPLC) using a CECIL-1100 chromatograph (Cecil Instruments, Cambridge, United Kingdom). Two hundred microliters of each sample was injected on to an Aminex HPX-87H column (300 by 7.8 mm) (Bio-Rad) run at ambient temperature with 2.5 mM H₂SO₄ as the eluant at a flow rate of 0.6 ml/min. Detection was by UV at 210 nm. Fumarate and pyruvate concentrations were determined by reference to standards of known concentrations.

Analysis of pyruvate production from fumarylpyruvate by fumarylpyruvate hydrolase using LDH. The concentration of pyruvate was determined by measuring the decrease of absorbance at 340 nm due to NADH oxidation in the presence of excess lactic dehydrogenase (LDH). The sample cuvette (1 ml) contained the products from hydrolysis of fumarylpyruvate (or, in the standards, pyruvate alone) plus 0.16 µmol of NADH in 100 mM phosphate buffer (pH 7.4). The reference cell contained only buffer. The assay was initiated by adding 0.64 mg (0.54 U) of LDH from rabbit muscle (EC 1.1.1.27) (Sigma) to both cuvettes: 1 U of LDH reduces 1.0 µmol of pyruvate to lactate per min. The molar extinction coefficient for NADH at 340 nm was taken to be 6,220 M¹ cm¹. The stoichiometry of the reaction was calculated from the change in A₃₄₀ resulting from the total conversion of amounts of substrate varying from 10 to 100 nmol.

Nucleotide sequence accession numbers. The DNA sequence of 21,160 bp of the nag genes has been submitted to GenBank under accession number AF036940. The DNA sequence of the 16S rDNA has been submitted to GenBank under accession number AF301897.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Reclassification of strain U2. The almost complete 16S rDNA of strain U2 was cloned and sequenced. The gene from this strain exhibited the highest identity (98%) to the 16S rDNA from three strains of the genus Ralstonia (accession numbers AF067657 [38], AF085226 [50], and AF098288 [33]), indicating that it should be reclassified as Ralstonia sp. strain U2 rather than as a Pseudomonas sp.

Cloning of naphthalene catabolic genes. The genes for naphthalene dioxygenase (nagAaAbAcAd), together with those for the 5-salicylate hydroxylase (nagGH), naphthalene cis-dihydrodiol dehydrogenase (nagB), and salicylaldehyde dehydrogenase (nagF) were originally located on an 8.3-kb BamHI fragment (pWWF6) with the novel gene order nagAaGHAbAcAdBF (14). With the important exception of the insertion of nagGH within the naphthalene dioxygenase cluster (nagAaAbAcAd), the gene order is identical to the nahAaAbAcAdBF found on the archetypal plasmid NAH7 (9, 45). We therefore hypothesized that the analogous nagC in U2 is downstream of nagF in pWWF6, as is the case in NAH7. Plasmid DNA of U2 was digested with BglII and ligated into BamHI-digested vector pUC18. Transformants in E. coli DH5 were selected by screening for the yellow product (2-hydroxy-6-oxohepta-2,4-dienoate) produced from 3-methylcatechol (6) by 1,2-dihydroxynaphthalene dioxygenase (NagC). We identified a positive clone, designated pWWF19, that contained a 7.6-kb BglII insert. Sequence alignment confirmed its overlap with pWWF6 (Fig. 1). Further cloning of DNA downstream of pWWF19, as described in Materials and Methods, resulted in a plasmid, pWWF60, which contained an 8.9-kb EcoRI insert which overlapped with pWWF19 (Fig. 1).

Sequence analysis. Approximately 6 kb of DNA containing nagAa to nagAd of plasmid pWWF6 had been previously cloned and sequenced (14). Using pWWF6, pWWF19, pWW24, and pWWF60, sequencing was extended upstream of the 5' end of nagAa and downstream of the 3' end of nagAd, providing a continuous sequence of 21,160 bp, from which the presence of 20 complete open reading frames (ORFs) was deduced (Fig. 1; Table 2).

NagI catalyzes gentisate transformation. A PCR fragment containing the complete reading frame of nagI was ligated into expression vector pET5a as plasmid pWWF19-25 and transformed into E. coli BL21(DE3). After induction with IPTG, cell extracts were found to contain GDO with a specific activity of 2.52 U/mg against gentisate as the substrate. SDS-PAGE of the same extracts (data not shown) showed elevated levels of a polypeptide of ~40 kDa, as expected from the deduced amino acid composition. No activity or enhanced 40-kDa polypeptide band was detectable in controls where expression of the protein was not induced or where the expression vector contained no insert. Figure 2 shows the rapid transformation by the cell extract of gentisate (_max = 320 nm) to maleylpyruvate (_max = 330 nm) as described originally by Lack (26). GDO activity was also detected in the cell extract of strain U2 grown on naphthalene (0.078 U/mg). NagI exhibited an extended substrate specificity towards available alkyl and halogenated gentisates (Table 3). From these very limited data, K_m values for 3-substituted gentisates were found to be lower than those for gentisate, indicating a higher substrate affinity. However, the specific activities towards the substituted gentisates were also lower than towards gentisate, indicating a lower turnover number. A similar situation was reported for the GDOs in two Pseudomonas strains (10).

View larger version (20K): [in this window] [in a new window]   FIG. 2.   Formation of maleylpyruvate from gentisate by cell extracts of E. coli DH5(pWWF19-25 [nagI]) grown on LB medium containing 100 µg of ampicillin per ml, induced with 0.4 mM IPTG. Sample and reference cuvettes contained 100 mM phosphate buffer (pH 7.4) in 1-ml volumes. The sample cuvette also contained 0.05 µmol of gentisate. Spectra were recorded before and every 20 s after the addition of cell extract (5 µl containing 25 µg of protein). Cell extract was added to the reference cuvette in an equivalent amount.

NagL catalyzes maleylpyruvate isomerization. pWWF53 carries the complete reading frame of nagL in expression vector pET5a (Table 1). When cell extracts of IPTG-induced E. coli(pWWF53) were incubated with maleylpyruvate, generated from gentisate by the action of NagI (as above), no change in the absorption spectrum between 250 and 400 nm took place unless GSH was introduced into the reaction. When GSH was present, the spectrum changed from a peak with a _max of 330 nm, characteristic of maleylpyruvate, to one with a _max of 340 nm and with a lower extinction coefficient (Fig. 3), as reported for that of fumarylpyruvate (26, 49). No equivalent reaction was seen in the presence of GSH without added cell extract or with added cell extract of E. coli(pET5a). We have noted that the change from maleylpyruvate to fumarylpyruvate occurs without a true isobestic point in the vicinity of 350 nm (Fig. 3), as has similarly been reported for the isomerization catalyzed by 4-oxalocrotonate tautomerase (29). This suggests that there might be a similar explanation, namely, that the reaction mechanism is more complex than just a direct interconversion of the two isomers. If the fumarylpyruvate was allowed to stand in the presence of excess GSH, then a slow change in spectrum occurred, forming a product with a _max of 308 nm, as was described by Lack (27); this is assumed to be the result of a spontaneous nonenzymatic reaction between the two.

View larger version (18K): [in this window] [in a new window]   FIG. 3.   Isomerization of maleylpyruvate to fumarylpyruvate by cell extracts of E. coli DH5(pWWF53 [nagL]) grown on LB medium containing 100 µg of ampicillin per ml, induced with 0.4 mM IPTG. Sample and reference cuvettes contained 100 mM phosphate buffer (pH 7.4) and 1.0 µmol of reduced GSH in 1-ml volumes. The sample cuvette also contained maleylpyruvate transformed from 0.05 µmol of gentisate by gentisate dioxygenase (NagI) in situ. Spectra were recorded before and every 20 s after the addition of cell extract (5 µl containing 45 µg of protein). Cell extract was added to the reference cuvette in an equivalent amount.

NagK catalyzes fumarylpyruvate hydrolysis. Cell extracts of E. coli(pWWF52) carrying nagK inserted in pET5a (Table 1) showed no activity against fumarylpyruvate, formed from gentisate by the joint action of NagI and NagK, even though SDS-PAGE of induced cells showed a polypeptide of the right size (~21 kDa) expressed from this construct (data not shown); we assume that the high levels of expression produced an insoluble product in inclusion bodies. In order to reduce the expression, a 0.8-kb MspA1 I fragment containing nagK was subcloned from pWW86 (Fig. 1; Table 1) into the HincII site of pUC18 in the correct orientation to the lac promoter, forming plasmid pWWF96. In contrast to the absence of activity in the pET5a-based construct, cell extracts of E. coli(pWW96) incubated with fumarylpyruvate effected the complete disappearance of the 340-nm peak (Fig. 4). The measured specific activity of fumarylpyruvate hydrolysis was 0.32 U/mg. The products of degradation from fumarylpyruvate were confirmed as pyruvate and fumarate by HPLC. Their retention times, under the conditions used, were 12 min 50 s and 22 min 30 s, respectively, and could be distinguished from those of other short-chain carboxylic acids, particularly maleate (retention time, 10 min 30 s). From the peak areas compared with those of standard amounts of the two products, it was calculated that equimolar amounts of pyruvate and fumarate were produced and that the stoichiometry of production of each was ultimately between 80 and 90% of the original gentisate added (Fig. 5). There was no maleate detected in the reaction. As additional confirmation that pyruvate was a product of fumarylpyruvate hydrolysis, NADH oxidation in the presence of LDH was used to measure pyruvate in the reaction mixture; 0.9 mol of NADH was oxidized per mol of fumarylpyruvate hydrolyzed, confirming the stoichiometry of pyruvate production. Fumarylpyruvate hydrolase activity was also detected in cell extracts of the U2 strain grown on naphthalene (0.075 U/mg).

View larger version (22K): [in this window] [in a new window]   FIG. 4.   Hydrolysis of fumarylpyruvate by cell extracts of E. coli DH5(pWWF96 [nagK]) grown on LB medium containing 100 µg of ampicillin per ml, induced with 0.4 mM IPTG. Sample and reference cuvettes contained 100 mM phosphate buffer (pH 7.4) in 1-ml volumes. The sample cuvette also contained fumarylpyruvate formed from maleylpyruvate by the action of NagL (see the legend to Fig. 3), itself previously transformed from 0.05 µmol of gentisate by the gentisate dioxygenase (NagI) in situ. Spectra were recorded before and every 20 s after addition of cell extract (8 µl containing 90 µg of protein). Cell extract was added to the reference cuvette in an equivalent amount.

View larger version (16K): [in this window] [in a new window]   FIG. 5.   Time course of production of fumarate and pyruvate from fumarylpyruvate by cell extracts of E. coli DH5(pWWF96 [nagK]) grown on LB medium containing 100 µg of ampicillin per ml, induced with 0.4 mM IPTG. Reaction was initiated by adding 160 µl of cell extract to 20 ml of 100 µM fumarylpyruvate in 100 mM phosphate buffer (pH 7.4), produced in situ from gentisate by NagI and NagL. Fumarate and pyruvate were detected and quantified by HPLC on an Aminex HPX-87H column.

Enzyme activities encoded on recombinant plasmids. By the enzyme assays for GDO, maleylpyruvate isomerase, and fumarylpyruvate hydrolase it was shown that all three activities could be detected in cell extracts of E. coli carrying any of the large cloned fragments on plasmids pWWF60, pWWF80, and pWWF86 and also on pWWF100, which carries only the three intact genes nagIKL (Fig. 1; Table 1). However, in E. coli carrying pWWF105, a derivative of pWWF100 with a frameshift within nagK, fumarylpyruvate hydrolase activity was not detected but both GDO and maleylpyruvate isomerase activities remained.

nagJ, nagM, and nagN appear to have no enzymatic role in gentisate metabolism. The remaining genes, nagJ, nagM, and nagN, were also expressed from their insertions in both pET5a (pWWF51, pWWF71, pWWF72, and pWWF73 [Table 1]) and in pUC18 (pWWF98, pWWF83, and pWWF84 [Table 1]). SDS-PAGE showed polypeptides of the expected sizes from all of the pET5a constructs (data not shown).

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

Previous biochemical evidence suggests that there may be more than one route by which gentisate is further metabolized. One metabolic pathway appears to be as we have described in this paper for Ralstonia sp. strain U2 and involves the cleavage of the aromatic ring by GDO to form maleylpyruvate, its isomerization to fumarylpyruvate, and subsequent hydrolysis to fumarate and pyruvate. This has been described for a number of different bacteria (5, 8, 24, 26, 27, 41, 49) and was the pathway first elucidated in 1959 by Lack (26). In some bacteria maleylpyruvate isomerization appears to require reduced GSH (8, 27, 41), whereas in the others, particularly gram-positive species, this reaction appears to be GSH independent (8, 18). A third variant of the pathway, in which maleylpyruvate undergoes direct hydrolysis to yield pyruvate and maleate without fumarylpyruvate as an intermediate, has been reported (5, 22). In P. alcaligenes P25X1 both a fumarylpyruvate hydrolase and two isofunctional maleylpyruvate hydrolases appear to coexist (5, 39).

Among naphthalene degraders which have been investigated at the genetic level, Ralstonia sp. strain U2 appears unique. Within the gene cluster for naphthalene dioxygenase, nagAaAbAcAd, there is, between nagAa and nagAb, an insertion of nagGH for salicylate 5-hydroxylase, which converts salicylate to gentisate (14), and downstream of nagD is the cluster nagJIKLMN described in this paper. The rest of the genes for conversion of naphthalene to salicylate are closely related to the classical upper pathway operon typified by the NAH7 plasmid (9), being homologous and in the same order. A further difference between the nag and nah clusters is in the regulatory and putative chemotaxis genes R and Y. In strain U2 both nagR and nagY are upstream of nagAa, whereas in NAH7 both nahR (43) and nahY (16) are downstream of the upper pathway operon and more closely linked to the meta-pathway operon. The similarities between the nag and nah genes for the conversion of naphthalene to salicylate suggest a common ancestry, but both appear to have independently acquired separate modules of genes for the terminal reactions from salicylate to central metabolites. In NAH7, the genes for salicylate hydroxylase and the meta-cleavage pathway are present as a distinct but adjacent operon, whereas in strain U2 genes for the conversion of salicylate through gentisate have been incorporated as part of the same operon by way of an upstream insertion of nagGH and a downstream addition of nagJIKLMN.

In this paper we have concentrated on elucidating the genetic apparatus for further metabolism of gentisate and have identified the main enzyme components. Our approach was to carry out preliminary experiments, unreported here, to check that the catabolism of gentisate was encoded entirely by the nagJIKLMN region and then to clone each gene separately both into pUC18 and into the high-expression vector pET5a. NagI is the GDO converting gentisate to maleylpyruvate, which is then subjected to GSH-dependent isomerization to fumarylpyruvate catalyzed by NagL. We were unable to obtain any evidence that there was a direct hydrolysis of maleylpyruvate to maleate and pyruvate, as has been indicated in other bacteria (5, 22, 39), as none of the nag gene products appeared to catalyze this reaction. The hydrolysis of fumarylpyruvate is catalyzed by NagK, and we have demonstrated that pyruvate and fumarate are produced stoichiometrically in this reaction. Together or sequentially, NagI, NagL, and NagK plus added GSH are all that is required to produce pyruvate and fumarate from gentisate in vitro, and it is reasonable to deduce that this is also true in vivo (Fig. 6)

View larger version (9K): [in this window] [in a new window]   FIG. 6.   Reactions of gentisate catabolism catalyzed by the nag gene products of Ralstonia sp. strain U2.

The overall amino acid similarity between NagI and the other complete derived GDO sequences is surprisingly low (35%), perhaps indicating wide divergence within gentisate pathways, possibly as the result of selection for accommodating different substituents on the ring. There is little else in the literature with which to compare the NagL and NagK sequences. One other maleylpyruvate isomerase has been purified, but an N-terminal sequence of only 11 amino acids has been published (41), and only 5 of these are in common with NagL. Although a fumarylpyruvate hydrolase has been purified, no amino acid sequence was determined (5). However, both NagK and NagL show sequence similarities to enzymes with related activities. In both bacteria and eukaryotes the catabolism of phenylalanine and tyrosine proceeds through an analogous pathway in which homogentisate (2,5-dihydroxyphenylacetate) is the ring cleavage substrate and is converted to maleylacetoacetate (11). NagL shows significant similarity to maleylacetoacetate isomerases on this pathway (catalyzing the GSH-dependent conversion to fumarylacetoacetate) in Sinorhizobium meliloti (45% identity) (31) and in Homo sapiens (41% identity) (12). From Pfam database searches, NagK can be clearly seen to be part of the same family as fumarylacetoacetate hydrolase, which catalyzes the next reaction in the same pathway to produce fumarate and acetoacetate (47).

As has been found in other peripheral catabolic pathways, we have found genes (nagJ and nagMN) which appear to play no role in the metabolic reactions but which may encode proteins essential to the overall functioning of the pathway in vivo. Interestingly, they each appear to have homologs within the nag operon. From database searches NagJ, like NagL, is a member of the GST superfamily, although they are substantially different from each other since the GST family encompasses wide diversity (51). GST-like proteins have been found in other aromatic catabolic pathways (25, 30, 52), of which the most similar to NagJ is BphK (58% identity) in biphenyl catabolism (21). Like NagJ, these have had no function allocated to them. The two other proteins, NagM and NagN, are closer to each other than to any other proteins in the present database and thus constitute another homologous pair: a similar situation has been reported in C. testosteroni TA441 (accession no. AB024335), in which there is an adjacent pair of homologous genes (orf4 and orf5) which are also homologous to nagMN. The similarity between the NagM and NagN suggested to us that they might encode subunits of a heteromultimeric protein. For this reason, we cloned them in tandem (as well as singly) into both pUC18 and pET5a in case they require coexpression for activity, but we could detect no activity against any of the pathway substrates.

Comparisons with sequences in the databases make it apparent that the gentisate pathway genes we have characterized in strain U2 are present in other bacteria but their functions have not been previously identified. The gene order nagIKL corresponds exactly to the homologs gtdA-orf2-orf3 in Sphingomonas sp. strain RW5, which metabolizes 3,6-dichloro-2-methoxybenzoate via gentisate, but the function of only GtdA as a GDO has been demonstrated (53). Based on the gene order and protein similarity between the genes in these two clusters, we predict that the enzymes encoded by orf2 and orf3 in strain RW5 are a fumarylpyruvate hydrolase and a GSH-dependent maleylpyruvate isomerase, respectively. If this is correct, the metabolism of maleylpyruvate in RW5 is also via a GSH-dependent isomerization to fumarylpyruvate and then to fumarate and pyruvate by hydrolysis. A similar situation may well apply to Sphingomonas sp. strain RW1, where two adjacent ORFs corresponding to nagKL have been reported (3). In K. pneumoniae, a similar gene order was deduced but without sequence determination: this was mhbD (for GDO), mhbH (for fumarylpyruvate hydrolase), and mhbI (for GSH-dependent maleylpyruvate isomerase isomerase) but with the difference of the insertion of mhbM (for 3-hydroxybenzoate monooxygenase, catalyzing the conversion of 3-hydroxybenzoate to gentisate), between mhbH and mhbI (41).