INTRODUCTION

Top Abstract Introduction Materials & Methods Results & Discussion References

Large amounts of aromatic compounds have been released into the environment during the last decades as a result of extensive production of industrial chemicals and agricultural applications of pesticides. Many of these compounds, particularly the chlorinated derivatives, are toxic, even at low concentrations, and persist in the environment (14, 39). Numerous microorganisms have been isolated which degrade xenobiotic aromatic compounds through aerobic or anaerobic degradative reactions (16, 17, 34, 46, 47). A wide variety of polycyclic and homocyclic aromatic compounds are aerobically transformed to a limited number of central dihydroxylated intermediates like catechol, protocatechuate, or gentisate. Whereas catabolic pathways for catechol and protocatechuate have been extensively characterized (22, 35), little is known about gentisate degradation.

Gentisic acid (2,5-dihydroxybenzoic acid) is a key intermediate in the aerobic degradation of such aromatic compounds as dibenzofuran (15), naphthalene (18, 48), salicylate (40, 55), anthranilate (32), and 3-hydroxybenzoate (26). Degradation of gentisate is initiated by gentisate 1,2-dioxygenase (GDO; EC 1.13.11.4, gentisate:oxygen oxidoreductase), which cleaves the aromatic ring between the carboxyl and the vicinal hydroxyl group to form maleylpyruvate (30). Maleylpyruvate can be converted to central metabolites of the Krebs cycle either by cleavage to pyruvate and maleate (5, 24) or by isomerization to fumarylpyruvate and subsequent cleavage to fumarate and pyruvate (10, 31, 51).

Of the two well-studied classes of ring cleavage dioxygenases, intradiol enzymes, such as catechol 1,2-dioxygenase or protocatechuate 3,4-dioxygenase, contain an Fe³⁺ atom in the catalytic center and cleave the aromatic substrate between two vicinal hydroxyl groups (7, 37, 38), whereas dioxygenases of the extradiol class, such as catechol 2,3-dioxygenase or protocatechuate 4,5-dioxygenase, contain Fe²⁺ and cleave the aromatic substrate adjacent to two vicinal hydroxyl groups (1, 13). Gentisate 1,2-dioxygenase cleaves aromatic rings containing hydroxyl groups situated para to one another. Although the mechanism of oxygen activation was proposed to be similar to that of enzymes of the extradiol dioxygenase class (20), and the active center contains Fe²⁺ (11, 21, 29, 49, 51), the Fe²⁺ is not bound to the enzyme by electron-donating ligands such as cysteine or tyrosine (21) as is the case for extradiol-cleaving dioxygenases (19). It is being assumed, therefore, that GDO represents a novel class of ring-cleaving dioxygenases.

GDOs have been purified and characterized from gram-positive bacteria of the genera Bacillus and Rhodococcus (29, 50) and gram-negative bacteria of the genera Klebsiella, Comamonas, and Moraxella (11, 21, 49), and amino-terminal sequences of GDOs from Comamonas testosteroni and Comamonas acidovorans have been determined (21), but until now, no complete sequence of any GDO or of a gene encoding GDO has been reported. Here we describe the cloning and sequencing of the gene encoding the GDO from Sphingomonas sp. strain RW5 and its partial characterization. This GDO represents a novel class of dioxygenases with very low similarity to any other known ring-cleaving dioxygenases.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results & Discussion References

Bacterial strains and plasmids. Sphingomonas sp. strain RW5 was isolated from aerobic sediment samples obtained from the Elbe River in Hamburg, Germany, on the basis of its ability to grow with Dicamba (3,6-dichloro-2-methoxybenzoate) as the sole source of carbon and energy. Other bacterial strains used in this study were Escherichia coli BL21(DE3)[pLys] (host for the T7 expression system, obtained from Stratagene, Heidelberg, Germany) F dcm ompT hsdS(r_B m_B) gal(DE3) and E. coli DH5 endA1 hsdR17(r_K m_K⁺) supE44 thi-1 recA1 gyrA96 relA1 (lacZYA-argFV169) 80lacZM15 F (obtained from Clontech, Palo Alto, Calif.). Plasmids used were pBluescript II SK(+) (Ap^r) and pCR-Script SK(+) (Ap^r), both from Stratagene, and pCR2.1 (Ap^r Km^r) from Invitrogen. Plasmid pT7-7 (Ap^r Cm^r) was kindly supplied by Stan Tabor, Harvard Medical School, Boston, Mass.

Culture conditions. Strain RW5 was routinely grown in mineral salts medium, pH 7.3, containing (per liter): 14.0 g of Na₂HPO₄ · 12H₂O, 2.0 g of KH₂PO₄, 50 mg of Ca(NO₃)₂ · 4H₂O, 0.5 g of (NH₄)₂SO₄, 0.1 g of MgCl₂ · 6H₂O, 20 mg of Fe₂NH₄-citrate, and 0.1 ml of a trace-element solution (44). Agar plates contained 1.5% (wt/vol) purified agar (Oxoid, Hampshire, United Kingdom). Liquid cultures were incubated at 30°C on a rotary shaker at 160 rpm. E. coli strains were grown on Luria broth (LB) plates containing appropriate antibiotics. Liquid cultures were grown in baffled Erlenmeyer flasks.

Enzyme assays, kinetic measurements, and protein determination. Gentisate 1,2-dioxygenase activity was assayed spectrophotometrically at 25°C by measuring the formation of maleylpyruvate at 330 nm (30). Cleavage of 3,6-dichlorogentisate by gentisate 1,2-dioxygenase was assayed by measuring product formation at 361 nm under the same conditions. The molar extinction coefficient used for maleylpyruvate was 10,200 M¹ cm¹ (54), and that for the chlorinated derivative was estimated to be 6,620 M¹ cm¹ (this work). K_m values were determined from initial reaction rates by fitting initial velocities to the Michaelis-Menten equation by nonlinear regression. One unit of activity was defined as the turnover of 1 µmol of (chloro)gentisate per min. Protein concentrations were determined by the method of Bradford (6), with bovine serum albumin as the standard.

Enzyme purification. All purification steps were performed at 4°C with a Fast Protein Liquid Chromatography system from Pharmacia, Uppsala, Sweden.

(i) Preparation of cell extracts. Cultures (3 liters each) of strain RW5 were grown to mid-exponential phase with Dicamba as the sole source of carbon and energy and harvested by centrifugation (13,000 × g). Cell pellets were washed three times with 20 mM Tris-HCl buffer (pH 7.0) and resuspended in 3 ml of the same buffer. The suspension was passed twice through a precooled French pressure cell (Aminco, Silver Spring, Md.) at a pressure of 76 MPa. Cell debris was removed by centrifugation at 35,000 × g for 30 min, and the resulting supernatant fluid was immediately used. Cell extracts of E. coli BL21(DE3) containing plasmid pJW48 were induced with 0.4 mM isopropyl--D-thiogalactopyranoside (IPTG) and disrupted by the freeze-thaw method previously described (2).

(ii) Ion-exchange chromatography I. The cell extract was applied to a DEAE-Sepharose column prepared as described by the manufacturer (Pharmacia). The column was equilibrated in 20 mM Tris-HCl buffer, pH 7.0 (buffer A), and developed with a linear gradient of 0 to 1 M NaCl in buffer A. Fractions were assayed for GDO activity, and active fractions were pooled and concentrated to less than 1 ml with the Centriprep-30 system (Amicon Inc., Beverly, Mass.).

(iii) Ion-exchange chromatography II. Pooled fractions from the DEAE Sepharose column were applied to a prepacked MonoQ HR10/10 column (Pharmacia) equilibrated with buffer A. Proteins were eluted with NaCl in buffer A, first with a 40-ml linear 0 to 0.25 M gradient, then with a 160-ml linear 0.25 to 0.35 M gradient, and finally with 40 ml of 1 M NaCl. Active fractions were pooled and concentrated to less than 1 ml with the Centriprep-30 system.

(iv) Gel filtration chromatography. The pooled fractions obtained from the Mono Q column were loaded onto a prepacked Superdex 200 column (Pharmacia) equilibrated with 50 mM Tris-HCl buffer (pH 7.0) containing 100 mM NaCl and eluted with the same buffer and analyzed for activity. Active fractions were pooled and used for further studies.

Amino acid sequencing. For N-terminal sequencing, the purified protein was transferred onto a ProBlot polyvinylidene difluoride membrane (Applied Biosystems, Weiterstadt, Germany) for 70 min at a constant current of 120 mA as described by the manufacturer. The polyvinylidene difluoride-blotted protein was sequenced by Edman degradation in the blot cartridge of an Applied Biosystems 494A Procise sequencer with the standard gas-phase cycles recommended by the manufacturer (Applied Biosystems). Peptide fragments were generated by carboxyamidomethylation of the protein and digestion in situ with endoproteinase Lys-C (sequencing grade; Promega, Heidelberg, Germany).

Determination of molecular mass. (i) Denatured polypeptide. The subunit molecular mass of the GDO was estimated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) under denaturating conditions with 12% (wt/vol) acrylamide gels according to the method of Laemmli (33). Gels were stained with Coomassie brilliant blue R-250. Broad-range molecular standards from Pharmacia were used to determine the sizes of the subunits.

(ii) Native polypeptide. The native molecular mass of the enzyme was estimated by gel filtration chromatography on a prepacked Superdex-200 column (Pharmacia) with protein standards of the gel filtration calibration kit from Pharmacia.

Genetic methods and sequence analysis. Standard procedures such as Southern blot experiments, DNA subcloning, and DNA manipulations were performed by the method of Sambrook et al. (45). Hybridization was performed at 68°C with an RPN2510 oven (Amersham, Buckinghamshire, United Kingdom). Total DNA from strain RW5 was extracted according to the CTAB protocol (4). [-³²P]dCTP and the Megaprime DNA labeling system used for the labeling of gene probes were purchased from Amersham. Biodyne B membrane (Pall, Glen Cove, N.Y.) and uncharged Qiabrane membrane (Qiagen, Hilden, Germany) were used for DNA blotting experiments. Plasmid DNA for sequencing was extracted with the Plasmid Maxi kit (Qiagen). Sequencing reactions on both strands were performed with the Applied Biosystems 373A DNA sequencer according to the protocols of the manufacturer for Taq cycle sequencing with fluorescent dye-labeled dideoxynucleotides, as described previously (28). Sequences were compiled and analyzed with the GeneWorks software package V2.45 (IntelliGenetics, Mountain View, Calif.). The computational resources of the European Molecular Biology Laboratory (Heidelberg, Germany) and the National Center for Biotechnology Information (Bethesda, Md.) were used for initial similarity searches. Sequence comparisons, calculations of evolutionary distances, and drawing of unrooted phylogenetic trees were done with GeneWorks software and the Phylips programs package.

Cloning of the gentisate 1,2-dioxygenase gene. Two degenerate oligonucleotide primers, 5'-AARCCIYTITGGGARGT-3' and 5'-GTICCIACCATICCRTC-3', were designed that corresponded to the amino acid sequences representing the lowest-possible nucleotide redundancy of the N-terminal fragment and several Lys-C-generated fragments thought to be derived from the C-terminal region of the GDO. The primers were used for PCR under low-stringency conditions, with total DNA from RW5 as a template. A resulting 1.2-kb PCR product containing the determined N-terminal sequence of the GDO was cloned into the pCR-Script SK(+) vector and used as a gene probe for colony blots. Southern blots of SfuI-digested DNA of RW5 were hybridized under stringent conditions with the probe, and signals were analyzed on Biomax MR X-ray films (Kodak, Rochester, N.Y.).

Construction of overexpressing plasmid pJW48. The synthetic oligonucleotide primers 5'-GGACATATGCAGCCAGTATTGGCCAATGATCAGC-3', which contains an engineered NdeI site (underlined), and 5'-AAAGGATCCTCATAGATTTTCCTCTCGGAACAGC-3', which contains an engineered BamHI site (underlined), were used for PCR, with pJW39 as a template. In order to minimize nucleotide misincorporation errors by DNA polymerase, the 1,052-bp fragment was amplified with VENT polymerase from New England Biolabs (Beverly, Mass.). Plasmid pJW45 was constructed by cloning the PCR product into the pCR-Script SK(+) vector. The plasmid insert was resequenced to control possible polymerase errors. The NdeI-BamHI fragment containing the complete sequence of the GDO gene was prepared by digesting pJW45 with NdeI and BamHI. Expression vector pJW48 was constructed by ligating this fragment into plasmid pT7-7, which had been previously digested with NdeI and BamHI. Hyperexpression of the recombinant GDO gene of pJW48 in E. coli BL21(DE3) was induced with 0.4 mM IPTG.

Production and identification of maleylpyruvate. Purified GDO was used for the production of maleylpyruvate from gentisate in an enzyme assay containing 1 mM gentisate. The formation of a single product was monitored by reversed-phase high-pressure liquid chromatography (LC-10AD liquid chromatograph, DGU-3A degasser, SPD-M10A diode array detector, FCV-10AL solvent mixer, all from Shimadzu, Kyoto, Japan) on a reversed-phase type SC 125 RP8 column (125 by 4.6 cm; Bischoff, Leonberg, Germany) (Lichrospher 100 beads with a diameter of 5.0 µm). The aqueous solvent system (flow rate, 1.0 ml/min) contained 0.1% (vol/vol) ortho-phosphoric acid and 20% (vol/vol) methanol. The UV spectrum from 200 to 400 nm was monitored. Direct derivatization of the reaction mix (100 µl) for gas chromatography-mass spectrometry (GC-MS) analysis was performed with trimethylsulfonium hydroxide (Machery-Nagel, Düren, Germany). GC-MS analysis was performed with a GC-17A gas chromatograph (Shimadzu, Kyoto, Japan) equipped with an XTI-5 column from Resteck (Bellefonte, Pa.) and coupled with the QP-5000 mass spectrometer (Shimadzu) operating in the electron impact mode at 70 eV. The ion source temperature was 320°C. Helium was used as carrier gas, with a flow rate of 1.0 ml/min. The oven temperature was maintained at 60°C for 2 min and then increased to 150°C at a rate of 20°C/min, followed by an increase to 320°C at a rate of 6°C/min. Samples (1.0 µl each) were injected into the GC, operating in the splitless mode with an injector temperature of 270°C.

Chemicals, enzymes, and antibiotics. 3,6-Dichlorogentisate was synthesized as follows: 3,6-dichloro-2-methoxybenzoic acid (obtained from Promochem, Wesel, Germany) was nitrated in position 5 with nitric acid in concentrated sulfuric acid. The resulting nitro group was reduced to the amino group with Raney nickel in ethanolic solution. Subsequent diazotization with isoamylnitrite, followed by hydrolysis with dilute sulfuric acid at elevated temperatures, resulted in 3,6-dichlorogentisic acid whose structure was confirmed by nuclear magnetic resonance and MS analysis. All other chemicals were of the highest purity commercially available. Restriction enzymes were purchased from Amersham, Boehringer (Mannheim, Germany), MBI Fermentas (Vilnius, Lithuania), New England Biolabs, and United States Biochemical. T4 ligase was purchased from New England Biolabs. IPTG was obtained from Roth (Karlsruhe, Germany). Antibiotics were purchased from Sigma (St. Louis, Mo.). X-Gal was obtained from Biomol (Hamburg, Germany). Taq DNA polymerase and calf intestine phosphatase were obtained from Boehringer.

Nucleotide sequence accession number. The nucleotide sequence presented in this article is deposited in EMBL/GenBank under accession no. AJ224977.

	RESULTS AND DISCUSSION

Top Abstract Introduction Materials & Methods Results & Discussion References

Isolation of strain RW5. Single strains were isolated from enrichment cultures with 3,6-dichloro-2-methoxybenzoate as the sole energy and carbon source by plating on agar medium containing the target compound. Strain RW5 is an aerobic, gram-negative, nonmotile rod and was identified as a Sphingomonas sp. strain on the basis of 16S ribosomal DNA sequencing.

Purification of gentisate 1,2-dioxygenase from strain RW5. Cell extracts of strain RW5 bacteria, pregrown on 3,6-dichloro-2-methoxybenzoate, exhibited a GDO with specific activities of 0.426 U/mg with gentisate and 0.023 U/mg with 3,6-dichlorogentisate as the substrates.

Amino-terminal sequence of gentisate 1,2-dioxygenase. The 43-residue amino-terminal sequence of purified GDO was determined by automated Edman degradation to be Met-Gln- Pro-Val-Leu-Ala-Asn-Asp-Gln-Gln-Ala-Gln-Leu-Thr-Ala-Leu-Tyr-Asp-Glu-Met-Arg-Pro-Ala-Gly-Leu-Lys-Pro-Leu-Trp-Glu-Val-Leu-His-Ala-Leu-Val-Leu-Ala-Glu-Gln-Pro-Val-Leu-Ala-Asn-Asp-Gln-Gln-Ala-Gln-Leu-Thr-Ala-Leu-Tyr-Asp-Glu-Met-Arg-Pro-Ala-Gly-Leu-Lys-Pro-Leu-Trp-Glu-Val-Leu-His-Ala-Leu-Val-Leu-Ala-Glu-Pro-Ala-Glu- Leu. This amino-terminal sequence showed no homology to the known amino-terminal sequences of the strains Comamonas testosteroni and Comamonas acidovorans (21).

Cloning and sequencing of the gtdA gene encoding gentisate 1,2-dioxygenase. A gene probe was constructed with degenerate primers based on the amino acid sequences of the N-terminal region and with peptides assumed to be derived from the C-terminal region of GDO. A clone that hybridized strongly with the gene probe was isolated from a partial gene library of strain RW5 DNA constructed in E. coli. The hybrid plasmid containing a 4,103-bp SfuI fragment cloned into pBluescript was termed pJW39. Cell suspensions of E. coli DH5 harboring plasmid pJW39 showed high activities of GDO. Both strands of the 4,103-bp insert of plasmid pJW39 were sequenced, and four open reading frames (ORFs) were identified. The first 43-amino-acid residues deduced from the sequence of ORF 1 were identical to those of the N terminus of the native enzyme purified from strain RW5. This gene of 1,052 bp in length, designated gtdA (Fig. 1), specifies a 350-amino-acid polypeptide with a predicted size of 38.85 kDa. The probable Shine-Dalgarno sequence is located six nucleotides upstream of the initiation codon. No obvious E. coli ⁷⁰ promoter consensus sequence upstream of the start codon of gtdA was detected.

View larger version (36K): [in this window] [in a new window]   FIG. 1.   Nucleotide and deduced amino acid sequence of the putative 1,053-bp gene gtdA. The sense DNA strand is shown. The putative Shine-Dalgarno sequence is underlined. A point indicates the stop codon.

Analysis of sequences downstream of the gentisate 1,2-dioxygenase gene. Three additional ORFs, all of which are accompanied by upstream sequences exhibiting potential ribosomal binding sites, were found downstream of the gtdA gene. ORF 2, which overlaps the last four nucleotides of the gtdA gene, is predicted to encode a protein of 25 kDa. Its deduced amino acid sequence shows significant homology to a number of ORFs coding for putative isomerases and decarboxylases which may also be associated with genes of degradative pathways, for example, 45 and 34% identity to two ORFs of the dibenzo-p-dioxin- and dibenzofuran-degrading Sphingomonas sp. strain RW1 (3), 29% to an ORF of Methanococcus janashii (8), and 28% to HpcE of Synechocystis sp. (27). The gene product of ORF 2 may thus be involved in an isomerization reaction of the gentisate pathway, such as the isomerization of maleylpyruvate to fumarylpyruvate (10), but the real function remains unknown so far.

Expression of recombinant gentisate 1,2-dioxygenase in E. coli. A pair of synthetic oligonucleotide primers was designed to amplify the gene encoding GDO as described in Materials and Methods. The resulting 1,060-bp DNA fragment was subcloned into the expression vector pT7-7 to construct pJW48. E. coli BL21(DE3) cells containing pJW48 were cultivated in the presence of 0.4 mM IPTG in LB medium at 30°C. Cell extracts of these bacteria showed a specific activity of GDO of 2.83 U/mg of protein. Induced cells of E. coli BL21(DE3) without pJW48 exhibited a specific activity of less than 0.001 U/mg of protein. These results confirm that the 1,052-bp ORF represents the structural gene of GDO.

Purification and characterization of gentisate 1,2-dioxygenase hyperexpressed in E. coli BL21. GDO was purified to homogeneity as indicated in Materials and Methods (see also Fig. 2). The results of the purification are summarized in Table 1 and resulted in a 42-fold purification of the specific activity. The molecular mass of GDO estimated by gel filtration chromatography was 178 ± 7.0 kDa (mean ± standard deviation). On SDS-PAGE, GDO migrated as a single band of approximately 39 ± 1.0 kDa (Fig. 2). The molecular mass calculated from the nucleotide sequence is 38.85 kDa. GDO therefore appears to be a homotetramer. Similar results have been reported for the GDOs of Arthrobacter sp. strain GFB 100 (9), C. testosteroni ATCC 49249, C. acidovorans ATCC 17438 (21), and Moraxella osloensis (11). The isoelectric point of our GDO, calculated on the basis of the nucleotide sequence with the GeneWorks software package, was 5.78, very similar to those reported for the GDOs from C. testosteroni ATCC 49249 (21) and Arthrobacter sp. strain GFB100 (9). The pH optimum of the enzyme activity was around 7.0, with a significant decrease in activity below pH 6.0 and above pH 8.0. The enzyme showed a broad activity maximum from pH 6.8 to 8.0 in different buffers such as Tris-HCl or phosphate buffers, as has been reported for other GDOs (11, 21, 29, 49).

View larger version (108K): [in this window] [in a new window]   FIG. 2.   SDS-PAGE analysis of purified GDO from hyperexpressed gtdA. Protein samples were separated by electrophoresis on a 12% polyacrylamide gel and stained with Coomassie blue. Lane 1, protein standards (in kilodaltons); lane 2, E. coli BL21 (DE3, pJW48) cell extract from induced whole cells; lane 3, GDO fraction from fast-flow DEAE-Sepharose chromatography; lane 4, GDO fraction from Mono Q separation; lane 5, GDO fraction after gel filtration.

Identification of maleylpyruvate as the product of cleavage of gentisate. The ring cleavage product from gentisate formed by GDO has previously been proposed on the basis of a UV spectrum to be maleylpyruvate (30), but so far this has not been confirmed unambiguously by determination of the structure of the product, presumably because maleylpyruvate is labile during extraction with organic solvents from acidified aqueous buffer systems.

View larger version (11K): [in this window] [in a new window]   FIG. 3.   Mass spectrum (70eV) of trimethylsulfonium hydroxide-derivatized maleylpyruvate.

Turnover characteristics. GDO activity was assayed at different substrate concentrations in an air-saturated buffer for determination of the K_m value (0.5 to 200 µM for gentisate and 30 to 3,000 µM for 3,6-dichlorogentisate). The K_m value for gentisate was estimated to be 15 µM, the V_max was 0.119 U/mg, and the K_cat/K_m value was 511 s¹ M¹ × 10⁴. The K_m value for 3,6-dichlorogentisate was 754 µM and thus 50-fold higher than for the unchlorinated gentisate. The V_max with this substrate was 0.204 U/mg, and the K_cat/K_m value was 20 s¹ M¹ × 10⁴, i.e., 30-fold lower than with the unchlorinated substrate.