INTRODUCTION

Dioxygenases catalyzing the fission of benzene rings are key enzymes in the metabolic pathways of aromatic compounds by microorganisms. Most of these kinds of previously reported dioxygenases attack monocyclic aromatic compounds with two adjacent hydroxyl groups such as catechol and protocatechuic acid and open the benzene rings through the intradiol or extradiol fission reaction (1, 2). However, some bacteria have been reported to synthesize dioxygenases that cleave the benzene rings of hydroquinone (3-5) and gentisic acid (6, 7).

In our investigations on the microbial metabolism of anilines, we isolated several microorganisms capable of growing on 2-aminophenol as the sole carbon, nitrogen, and energy source. When one isolate, Pseudomonas sp. AP-3, grows with this substrate, it synthesizes an enzyme acting on 2-aminophenol. This enzyme was partially purified with a 103-fold increase in the specific activity from its cell extracts. We proposed that the enzyme is a dioxygenase catalyzing the ring fission of 2-aminophenol with the consumption of 1 mol of O₂ per mol of substrate (8).

Our aim was to advance the purification of 2-aminophenol 1,6-dioxygenase from Pseudomonas sp. AP-3 and elucidate the molecular and catalytic properties of the purified enzyme. Because the product from 2-aminophenol by catalysis of the enzyme is rapidly and nonenzymatically converted into picolinate (8, 9), the real product has remained unverified. Furthermore, we attempted the cloning and sequencing of the gene of the dioxygenase, which would determine the category of this enzyme in the dioxygenase groups.

Recently, Lendenmann and Spain (10) reported the purification and characterization of the 2-aminophenol 1,6-dioxygenase from nitrobenzene-degrading Pseudomonas pseudoalcaligenes JS45, although they did not refer to the cloning and sequencing of its gene. In this report, the comparison of the dioxygenases from the two pseudomonads growing on 2-aminophenol or nitrobenzene is also described.

MATERIALS AND METHODS

The chemicals used in this study and their sources are as follows. Polypepton, 2-aminophenol, catechol, 4-methylcatechol, methyl chlorocarbonate, and N,O-bis(trimethylsilyl)trifluoroacetamide were from Wako Pure Chemicals, Osaka; 2-amino-p-cresol, 6-amino-m-cresol, 2-amino-4-chlorophenol, 3-methylcatechol, 3-chlorocatechol, 4-chlorocatechol, and 3-fluorocatechol were from Tokyo Kasei, Tokyo; pentafluorophenylhydrazine, 2-amino-m-cresol, and 2-amino-4,5-dimethylphenol were from Aldrich; DE52 cellulose was from Whatman; DEAE-Cellulofine A-500 was from Seikagku Co., Tokyo; and restriction endonucleases were from Takara Shuzo, Otsu.

Pseudomonas sp. AP-3 was used throughout this study as a producer of 2-aminophenol 1,6-dioxygenase and a donor of its gene. Escherichia coli JM109 and E. coli P2392 were used as hosts for the recombinant plasmids and bacteriophages, respectively. A lambda FIX II/XhoI partial fill-in vector (Stratagene, La Jolla) was used for the construction of a gene library. pGEM-T (Promega, Madison) and pBluescript II SK(+) vectors (Stratagene) were used for cloning of the PCR (polymerase chain reaction) products and subcloning of the DNA fragments, respectively.

Pseudomonas sp. AP-3 was cultured in the 2-aminophenol medium (8) containing 0.12% (w/v) of the substrate and supplemented with 1% (w/v) Polypepton. Pseudomonas sp. AP-3 used for isolating its total DNA and E. coli strains were cultured in Luria broth (11) with shaking at 30 and 37 °C, respectively.

The activity of 2-aminophenol 1,6-dioxygenase was measured by monitoring the decrease in absorbance at 282 nm according to a previous paper (8). The activities for the 2-aminophenol analogs were measured by scanning changes in the absorbance of each reaction mixture, because all substrates tested had absorption bands in the UV range. Molar extinction coefficients of the substrates attacked by the enzyme were determined in this study as follows: 3100 at 287 nm for 2-amino-p-cresol, 2700 at 287 nm for 6-amino-m-cresol, 3100 at 291 nm for 2-amino-4-chlorophenol, 2100 at 279 for 2-amino-m-cresol, and 2600 at 289 nm for 2-amino-4,5-dimethylphenol. For catechol, catechol 1,2-dioxygenase (12) and catechol 2,3-dioxygenase (13) activities were assayed. Protein concentrations were measured by the method of Lowry et al. (14).

All operations for enzyme purification were done at 0-4 °C, and centrifugation was carried out at 20,000 × g for 10 min. The frozen cells (37 g, wet weight) of Pseudomonas sp. AP-3 were used for the purification. The preparation of the cell extracts (step 1, fraction 1), streptomycin sulfate treatment (step 2, fraction 2), and ammonium sulfate fractionation (step 3, fraction 3) were essentially carried out by the same methods as described previously (8).

After the protein concentration of fraction 3 was adjusted to 7 mg ml¹ by adding buffer A (20 mM Tris-HCl (pH 8.0) containing 10% (v/v) ethanol, 1 mM dithiothreitol, and 0.5 mM L-ascorbate), acetone was added to the diluted solution to give 55% (v/v). The precipitate was removed by centrifugation and then acetone was added to the supernatant to give 65% (v/v). The precipitate was obtained by centrifugation and then dissolved in buffer A. The enzyme solution was dialyzed against buffer A (fraction 4).

Fraction 4 was applied to a column (2.2 × 27 cm) of DE52 cellulose equilibrated with buffer A. Proteins were eluted with a linear gradient (0 to 0.4 M) of NaCl in 1.4 liters of buffer A, and then the protein concentrations and 2-aminophenol 1,6-dioxygenase activities were assayed. Fractions with the specific activity higher than 2.7 units mg¹ were pooled (fraction 5).

Fraction 5 was dialyzed against buffer A. The dialyzed solution was applied to a column (2 × 26 cm) of DEAE-Cellulofine A-500 equilibrated with buffer A. The enzyme was eluted with a linear gradient (0 to 0.35 M) of NaCl in 1.4 liters of buffer A. The enzyme in each fraction was tested by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE)¹ (15). Fractions showing two distinct protein bands (- and -subunits) on the gel were pooled (fraction 6).

The molecular mass of the native enzyme was measured by the two methods of gel filtration and PAGE (16). Those of the enzyme subunits were measured by SDS-PAGE (15).

The reaction mixture consisted of 5 mM 2-aminophenol, 12 ml; 2-aminophenol 1,6-dioxygenase, 10 mg in 10 ml of buffer A; and 100 mM sodium-potassium phosphate buffer (pH 7.5), 330 ml. The reaction started at 24 °C by adding the enzyme solution with shaking. After 1 min, 2.4 ml of methyl chlorocarbonate was added to the mixture, and the reaction was continued for 10 min (17). The reaction mixture was then concentrated to 100 ml with a rotary evaporator. After the pH of the concentrated solution was adjusted to pH 3.0 with 3 N HCl, the solution was extracted with ethyl acetate. The upper layer was then evaporated to dryness. The residues were dissolved in 10 ml of methanol and reacted with 34 mg of pentafluorophenylhydrazine for 1 h at room temperature (18). The hydrazone derivative produced was trimethylsilylated with N,O-bis(trimethylsilyl)trifluoroacetamide and analyzed using a Hitachi M-2500 mass spectrometer at an ionization potential of 70 eV, coupled to a Hitachi G-3000 gas chromatograph. A TC-1 fused silica capillary column (GL Sciences, Tokyo) was used.

Iron in the enzyme was determined using o-phenanthroline after it was reduced to Fe²⁺ with hydroxylamine·HCl (19).

SDS-PAGE was employed to dissociate the purified enzyme into the - and -subunits. The separated subunits were electroblotted using the method of Matsudaira (20) and then sequenced.

Chromosomal DNA of Pseudomonas sp. AP-3 was prepared according to the protocol described by DiLella and Woo (21). The purified DNA (30 µg) was partially digested with Sau3AI, and DNA fragments larger than 23 kb were ligated into a lambda FIX II/XhoI partial fill-in vector after fill-in of the first two nucleotides of the Sau3AI-compatible site, according to the manufacturer's instructions. Gigapack III gold (Stratagene) was used to package the recombinant lambda phages. The infection of phage particles into E. coli P2392 and recovery of the recombinant phage DNA from the top agar were performed using the standard method (11). Subcloning experiments were performed using conventional techniques (11).

On the basis of NH₂-terminal amino acid sequences of - and -subunits of 2-aminophenol 1,6-dioxygenase, the four primers 1, 2, 1, and 2 were synthesized (Fig. 3). 2 and 2 had complementary sequences to 1 and 1, respectively. For amplifying fragments containing an 2-aminophenol 1,6-dioxygenase gene (amn gene) by PCR, Takara Ex Taq was used with various combinations of the primers. The cycling of PCR was performed at 94 °C (1 min), 51 °C (2 min), and 72 °C (2 min) for 25 cycles.

NH₂-terminal amino acid sequences of the - and -subunits and their corresponding primer sequences.

Amplified DNA fragments were ligated into a vector and partially sequenced to ensure that the ligated fragments contained an amn gene. After sequencing, the recombinant plasmid containing the amn gene was digested with SacII and PstI. The inserted DNA fragments were separated from the vector DNA using 1% agarose gel electrophoresis, recovered from gel slices by GeneClean II (Bio 101, La Jolla), and radiolabeled with [-³²P]dCTP (Amersham Corp., Buckinghamshire) and a random primer DNA labeling kit version 2 (Takara).

Genomic DNA from Pseudomonas sp. AP-3 was digested with various restriction endonucleases. The DNA fragments were transferred onto a Hybond-N membrane (Amersham Corp.) with a probe tech 2 (Oncor Inc., Gaithersburg) according to the manufacturer's instructions after 1% agarose gel electrophoresis. Plaque and Southern blot hybridizations for DNA were performed by the standard methods (11). Autoradiograms were analyzed with a Fuji BAS2000 bio-image analyzer (Fuji Film, Tokyo) after 3-4 h of exposure at room temperature.

A Qiagen plasmid mini kit (Qiagen, Hilden) was used for preparing the double-stranded DNAs for sequencing. Sequencing reactions were performed by using a dye terminator cycle sequencing FS ready reaction kit or dye primer cycle sequencing FS ready reaction kit (Applied Biosystems). Reaction mixtures were run on a 373 DNA sequencer (Applied Biosystems).

RESULTS

Table I shows a summary of a typical enzyme purification. The specific activity of the final preparation of 2-aminophenol 1,6-dioxygenase was 4.8 units mg¹ with an overall recovery of 36%. A 120-fold increase in the specific activity was observed at the final step of the purification procedure. The final enzyme preparation showed one major protein band and two indistinct bands on a polyacrylamide gel without SDS. The molecular mass of the major band was 146 kDa on the gel (16). Those of other two bands were lower than 85 kDa. However, on a SDS-polyacrylamide gel, the final preparation showed two distinct protein bands with molecular masses of 32 kDa (-subunit) and 40 kDa (-subunit) (Fig. 1).

Table I. Purification of 2-aminophenol 1,6-dioxygenase Fraction Volume Total activity Total protein Specific activity Recovery ml units mg units/mg % 1. Cell extract 460 160 4300 0.04 100 2. Streptomycin sulfate 463 150 4300 0.03 94 3. Ammonium sulfate 97 230 1500 0.15 140 4. Acetone 58 260 500 0.52 160 5. DE52 49 74 26 2.8 46 6. DEAE-cellulofine 25 58 12 4.8 36

Densitometric analyses of the bands revealed that the molar ratio of the two subunits was one to one on the basis of these molecular sizes. In addition, the molecular mass of the native enzyme was 140 kDa by gel filtration. These findings indicate that the enzyme was made up of four heterogeneous subunits with the structure of ₂₂.

The purified enzyme was stable in buffer A at 4 °C for a week without any decrease in activity. However, it lost activity within 24 h in the absence of ethanol, dithiothreitol, and L-ascorbate.

Fig. 2 shows the mass spectra of the N-acylated and trimethylsilylated hydrazone derivative of compound I. There was a molecular ion at m/z 451 (M⁺), which agreed with the empirical formula of C₁₇H₁₈F₅N₃O₄Si. Major fragment ions appeared at m/z 436 (M⁺-CH₃), 419 (M⁺-OCH₃-H), 404 (M⁺-OCH₃-CH₃-H), 391 (M⁺-COOCH₃-H), 377 (M⁺-COOCH₃-NH), 361 (M⁺-COOCH₃-2CH₃-H), 302 (M⁺-COOCH₃-OSi(CH₃)₃-H), 274(M⁺-COOCH₃-H-COOSi(CH₃)₃), 255 (M⁺-C₆F₅-NH-N), 242 (M⁺-C₆F₅-NH-N-CH), 229 (M⁺-C₆F₅-NH-N-2CH), 195 ([C₆F₅N₂]⁺), and 73 ([Si(CH₃)₃]⁺). These data showed that compound I was 2-aminomuconic 6-semialdehyde.

The enzyme did not have any absorption band in the visible range. However, it had an absorption peak at 280 nm and a small shoulder at 287-290 nm: E^1% =13.7 × 10⁴ at 280 nm. The E₂₈₀/E₂₆₀ ratio of the enzyme was 1.8.

The enzyme contained 0.98 mol of Fe²⁺ per mol of protein on the basis of the molecular mass of 140 kDa.

The substrate specificity of 2-aminophenol 1,6-dioxygenase was examined with 39 aromatic compounds consisting of catechol, phenol, and aniline compounds (Table II). Besides 2-aminophenol, the enzyme was active toward 2-amino-p-cresol, 6-amino-m-cresol, 2-amino-m-cresol, 2-amino-4,5-dimethylphenol, 2-amino-4-chlorophenol, and catechol. However, it did not act on 2-aminophenol analogs that were substituted by a carboxyl or nitro group at the 3-, 4-, and 5-positions. In addition, catechol compounds except catechol were not substrates of the enzyme.

Table II. Substrate specificity of 2-aminophenol 1,6-dioxygenase The enzyme did not act on phenol, o-, m-, or p-cresol, aniline, o-, m-, or p-aminobenzoic acid, o-, m-, or p-phenylenediamine, protocatechuic acid, 2,3-dihydroxybenzoic acid, 3-aminophenol, 4-aminophenol, 2-amino-4-isopropylphenol, 2-amino-tert-butylphenol, 3-aminophenazin-2-ol, 2-amino-4-nitrophenol, 2-amino-5-nitrophenol, 3-amino-4-hydroxybenzoic acid, or 4-amino-3-hydroxybenzoic acid. These compounds also did not inhibit the enzyme for 2-aminophenol as the substrate. Substrate Relative activity Inhibition of 2-aminophenol turnover Ki (µM)/type of inhibition % 2-Aminophenol 100 2-Amino-p-cresol 10.0 +a 6-Amino-m-cresol 19.1 +a 2-Amino-m-cresol 4.4 +a 2-Amino-4,5-dimethylphenol 1.3 +a 2-Amino-4-chlorophenol 6.8 +a Amidol 0 ±b 4-Aminoresorcinol 0 9.1 /Noncompetitive 3-Hydroxyanthralinic acid 0 ±b Catechol 2 10.4 /Noncompetitive 3-Methylcatechol 0 6.8 /Noncompetitive 4-Methylcatechol 0 9.5 /Uncompetitive 3-Fluorocatechol 0 5.4 /Noncompetitive 3-Chlorocatechol 0 6.7 /Competitive 4-Chlorocatechol 0 10.4 /Competitive Pyrogaroll 0 8.4 /Noncompetitive 1,2,4-Trihydroxybenzene 0 63.0 /Noncompetitive a The numerical values were not calculated, because maximal absorption wavelengths of 2-aminophenol and its analogs used for the measurement of the enzyme activity were overlapped. b Slightly inhibited.

The K_m and V_max values for 2-aminophenol of 2-aminophenol 1,6-dioxygenase were 46.7 µM and 0.10 µM s¹ mg¹, respectively. The enzyme for 2-aminophenol as the substrate was inhibited by the catechols and 4-aminoresorcinol listed in Table II. In addition, the 2-aminophenol analogs that could be degraded by the enzyme also inhibited it from the action on 2-aminophenol, although their types of inhibition were unmeasured. The activity of the enzyme for 2-aminophenol was not affected by the 2-aminophenol analogs such as phenols and anilines that were not substrates of this enzyme.

The effects of metal salts, chelating and sulfhydryl agents on the enzyme activity were tested using 2-aminophenol as the substrate (Table III). Among the metal ions tested, the enzyme was strongly inhibited by CuSO₄, FeCl₃, K₃Fe(CN)₆, AgNO₃, HgCl₂, or MnCl₂. FeSO₄ and MgSO₄ did not inhibit the enzyme very much, and FeSO₄(NH₄)₂SO₄ slightly increased the activity for 2-aminophenol. Chelating agents and NaN_3, completely repressed the enzyme activity.

Table III. Effects of various compounds on the enzyme activity The enzyme (25 µg) was incubated with 0.1 or 0.5 mM of each compound in 3 ml of 0.1 M sodium/potassium phosphate buffer (pH 7.5) at 24 °C for 10 min. The enzyme reaction was started by adding 0.1 ml of 10 mM 2-aminophenol. After incubation for 10 min, absorbance at 282 nm was monitored. Compound Concentration Remaining activity mM % None 100 CuSO4 · 5H2O 0.1 0 FeSO4 · 7H2O 0.5 78 FeSO4(NH4)2SO4 · 6H2O 0.5 105 FeCl3 0.5 0 K3Fe(CN)6 0.1 14 AgNO3 0.5 6 MgSO4 · 7H2O 0.5 79 HgCl2 0.1 12 MnCl2 · 4H2O 0.5 53 CH3ICOOH 0.5 40 PCMB 0.5 65 DTNB 0.5 0  ,-Dipyridyl 0.5 0 N-Ethylmaleimide 0.5 83 o-Phenanthroline 0.5 26 Tiron 0.5 0 EDTA 0.5 0 NaN3 0.5 0

The amino acid sequences of 30 and 20 residues of the - and -subunits, respectively, of the enzyme were determined. On the basis of the two sequences, the four primers 1, 1, 2, and 2 were synthesized (Fig. 3).

When the 1 and 2 primers and DNA purified from Pseudomonas sp. AP-3 as a template were incubated, a 1-kb DNA fragment was amplified. The sequencing of both termini of the amplified fragment showed that this fragment encoded a large portion of the -subunit and an NH₂-terminal region of the -subunit and that the -subunit gene was located upstream of the -subunit gene. The sequenced fragment was labeled with [-³²P]dCTP and was used as a probe for Southern hybridization. This DNA probe hybridized to the DNA fragments from the AP-3 strain digested with several restriction endonucleases (Fig. 4). These results showed that the PCR product was amplified on the basis of the DNA sequence from the AP-3 strain. The appearance of two positive bands for the KpnI- (lane 7) and SacI- (lane 8) digested DNAs suggests that the PCR product contained recognition sites for KpnI and SacI in its sequence.

The genomic library of Pseudomonas sp. AP-3 was constructed in a lambda FIXII phage vector. The hybridization to about 3000 plaques was performed using the probe for the amn gene mentioned above. After screening twice, we obtained five positive clones, p3-1, p4-3, p5-2, p12-2, and p12-7. The DNAs purified from these phage clones were digested with ApaI, BanIII, EcoRI, EcoRV, KpnI, and SacI. Agarose gel electrophoretic analyses revealed that restriction fragments obtained from these DNAs contained several fragments with the same size as the positive bands detected in the Southern hybridization (Fig. 4). The p4-3 DNA was selected for subcloning of the amn gene, because it was recovered with the greatest yield of the obtained fragments. SacI (1.7-kb) and EcoRI (1.4-kb) fragments encoding whole - and -subunit genes, respectively, were separated from each other and ligated into a pBluescript II SK(+) vector. The obtained pS1 and pE1 plasmids carried 1.7-kb SacI and 1.4-kb EcoRI fragments, respectively.

The DNA fragments inserted into pS1 and pE1 were sequenced. Fig. 5a shows that 884 base pairs of the two fragments overlapped. Two open reading frames containing each primer sequence used in the PCR were found in the connected sequence and were designated as amnB for the first open reading frame and amnA for the second one. Fig. 5b shows a DNA sequence connected with pE1 and pS1, and amino acid sequences deduced from amnB and amnA. The amnB encoded 305 amino acid residues. The deduced amino acid sequences from positions 2 to 31 completely agreed with the NH₂-terminal amino acid sequences of the -subunit determined from the purified enzyme. The initiation codon (ATG) of amnA started 32 base pairs downstream from the termination codon (TAA) of amnB. AmnA was made up of 271 amino acid residues and agreed with the NH₂-terminal amino acid sequences of the -subunit from positions 2 to 21.

DISCUSSION

The 2-aminophenol 1,6-dioxygenase from Pseudomonas sp. AP-3 growing on 2-aminophenol was purified with a 120-fold increase in the specific activity. Although the final preparations of the enzyme did not show one band on PAGE under nondenaturing conditions, these showed distinct two protein bands on the SDS-PAGE gel (Fig. 1). The two components were made up of an equal molar ratio on the basis of the molecular masses of 32 (-subunit) and 40 kDa (-subunit). In addition, each protein band electroblotted onto the transfer membrane for sequencing exhibited only one amino acid residue at the NH₂ terminus. Therefore, we consider that the final preparations of 2-aminophenol 1,6-dioxygenase were homogeneous and dissociated into several components by native PAGE. It was also observed that the dioxygenase from the nitrobenzene-degrading P. pseudoalcaligenes JS45 shows one main band and two diffuse bands on a PAGE gel at room temperature, although it exhibited a single band at 10 °C (10).

Since the product from 2-aminophenol by the reaction of 2-aminophenol 1,6-dioxygenase was labile, the direct evidence for the chemical structure of the product has not been reported. In this study, after the amino group of the product was acylated with methyl chlorocarbonate, its aldehyde group was modified with pentafluorophenylhydrazine to prevent its cyclization to picolinate. Based on the gas chromatography-mass spectrometry analyses of the modified compound, we could prove that the real product from 2-aminophenol was 2-aminomuconic 6-semialdehyde. On the basis of the fact together with our previous report (8), it was concluded that the dioxygenase catalyzes the ring fission between the 1- and 6-positions of 2-aminophenol with the consumption of 1 mol of O₂ per mol of substrate.

The 2-aminophenol 1,6-dioxygenase from P. pseudoalcaligenes JS45 was described to be stable in 50 mM MOPS (pH 7.3) containing 10% (v/v) glycerol (10). However, it is inactivated in the buffer containing 10% (v/v) ethanol. When it was chromatographed on MonoQ (Pharmacia Biotech Inc.) using the buffer without Fe²⁺ or cysteine, the activity of this enzyme is completely lost. The enzyme contains 2.2 mol of Fe²⁺ per mol of enzyme. On the other hand, our dioxygenase from Pseudomonas sp. AP-3 was stable in buffer A containing 10% (v/v) ethanol and was not stabilized in the presence of 10% (v/v) glycerol. The enzyme was inactivated in the presence of Fe²⁺ or cysteine, or both. In addition, the enzyme had 0.98 mol of Fe²⁺ per mol of enzyme. Protocatechuate 4,5-dioxygenase was also reported to have approximately 1 mol of Fe²⁺ per mol of enzyme, although the enzyme is a heterotetramer of ₂₂ (22). These data showed that the two 2-aminophenol 1,6-dioxygenases are distinctly different in stability against various factors and in Fe²⁺ content, although these are essentially similar to each other with respect to subunit structure, substrate specificity, and inhibited properties.

We cloned the amnA and amnB genes encoding the - and -subunits, respectively, of the 2-aminophenol 1,6-dioxygenase from Pseudomonas sp. AP-3 and determined their nucleotide sequences. This is the first report describing the cloning and sequencing of the 2-aminophenol 1,6-dioxygenase genes. Lendenmann and Spain (10) reported a part of the NH₂-terminal amino acid sequences of the - and -subunits of the 2-aminophenol 1,6-dioxygenase from P. pseudoalcaligenes JS45. The sequence comparison of AmnA (-subunit) from the AP-3 strain with the -subunit from the JS45 strain revealed that they contained 25 identical amino acid residues of the 30 comparable residues (83% identity); AmnB (-subunit) from the AP-3 strain and the -subunit from the JS45 strain showed 21 identical amino acid residues of the 30 comparable residues (70% identity). The comparison of the amino acid sequences of AmnB with those deduced from DNA data bases revealed that it had amino acid sequence identities of 25.5 and 24.4% with HpaD (23) and HpcB (24), respectively, which are homoprotocatechuate 2,3-dioxygenases from E. coli strains (Fig. 6). However, AmnB did not show any identity with other extradiol dioxygenases such as XylE (catechol 2,3-dioxygenase) (25) encoded on the TOL plasmid and several 2,3-dihydroxybiphenyl 1,2-dioxygenases (26, 27).

HpaD and HpcB belong to class III in the extradiol dioxygenases proposed by Spence et al. (28). They described that the enzymes of class III possess an NH₂-terminal domain containing the active center consisting of a Fe²⁺ cofactor and four histidine residues that are conserved in these enzymes (shaded area in Fig. 6). Multiple alignments of AmnB and the class III enzymes revealed that AmnB retains three histidine residues, His-14, His-63, and His-196, of the four histidine residues conserved in the class III enzymes; one residual histidine residue is replaced by an arginine residue at the corresponding position in AmnB (Fig. 6). At the same position in human 3-hydroxyanthranilic-acid dioxygenase belonging to class III, the histidine residue conserved in the class III enzymes was replaced by a threonine residue (28). In addition, we found the identical amino acid residues with those in the class III enzymes, which are located in the vicinity of the three conserved histidine residues. They were Pro-16, Glu-51, Leu-57, Ser-191, and Ser-195 in AmnB (Fig. 6). These results suggest that the amnB gene has evolved from a common ancestor gene for the class III enzymes and that amnB is classified into class III.

On the other hand, AmnA showed 23.3% identity with only AmnB among the enzymes examined (Fig. 6). The amino acid sequences common to the two proteins were dispersed in their sequences. This suggests that amnA and amnB would have diverged by the duplication of an ancestor gene. Multiple alignments of AmnA and the class III enzymes revealed that the three amino acid residues Pro-16, Val-147, and Ser-195 are conserved in every sequence of the enzymes listed in Fig. 6, although they are not catalytically active histidine residues (marked by ! in Fig. 6). The similarity of AmnA with AmnB and the existence of the conserved amino acid residues in AmnA may support the fact that AmnA also belongs to class III in extradiol dioxygenases. Because AmnA lacks the functional histidine residues conserved in other class III enzymes, it may be independent of the catalytic process of 2-aminophenol 1,6-dioxygenase and therefore play other roles, such as the stabilization of AmnB in the enzymatic reaction.

Among the class III enzymes whose amino acid sequences have been reported, only protocatechuate 4,5-dioxygenase from Pseudomonas paucimobilis SYK6 consists of two distinct subunits, LigA and LigB (29). However, it is noted that LigA, the small subunit of protocatechuate 4,5-dioxygenase, showed no identity with AmnA, the small subunit of 2-aminophenol 1,6-dioxygenase, although LigB aligned with AmnB. This fact suggests that LigA and AmnA are different in the process of molecular evolution and in the function of enzymatic catalysis.

Further efforts are now in progress to identify other genes responsible for the 2-aminophenol metabolism using the cloned DNA fragments from Pseudomonas sp. AP-3.