Chemoselective Nitro Group Reduction and Reductive Dechlorination Initiate Degradation of 2-Chloro-5-Nitrophenol by Ralstonia eutropha JMP134

Abstract

Ralstonia eutropha JMP134 utilizes 2-chloro-5-nitrophenol as a sole source of nitrogen, carbon, and energy. The initial steps for degradation of 2-chloro-5-nitrophenol are analogous to those of 3-nitrophenol degradation in R. eutropha JMP134. 2-Chloro-5-nitrophenol is initially reduced to 2-chloro-5-hydroxylaminophenol, which is subject to an enzymatic Bamberger rearrangement yielding 2-amino-5-chlorohydroquinone. The chlorine of 2-amino-5-chlorohydroquinone is removed by a reductive mechanism, and aminohydroquinone is formed. 2-Chloro-5-nitrophenol and 3-nitrophenol induce the expression of 3-nitrophenol nitroreductase, of 3-hydroxylaminophenol mutase, and of the dechlorinating activity. 3-Nitrophenol nitroreductase catalyzes chemoselective reduction of aromatic nitro groups to hydroxylamino groups in the presence of NADPH. 3-Nitrophenol nitroreductase is active with a variety of mono-, di-, and trinitroaromatic compounds, demonstrating a relaxed substrate specificity of the enzyme. Nitrosobenzene serves as a substrate for the enzyme and is converted faster than nitrobenzene.

All living organisms are able to reduce nitroaromatic as well as nitroheteroaromatic compounds. Research on the fate of the compounds in eucaryotic systems has revealed cytotoxic and mutagenic effects which are caused by reduction and further transformation of the reduction products (14–17, 22, 24, 41). The majority of nitroreductases found in mammalian tissues are oxygen sensitive (type II) (11). The enzymes, which are also present in other organisms, initially catalyze a one-electron reduction yielding a nitro anion radical (29, 36). The radical spontaneously reacts with elementary oxygen to form the superoxide radical, and the nitro group is regenerated. The phenomenon is also called a “futile cycle” because reduced pyridine nucleotides are oxidized without net reduction of the nitro group. In the absence of oxygen, further reduction of the nitro anion radical, yielding nitroso, hydroxylamino, and amino derivatives, takes place (11, 29, 36). In contrast, oxygen-insensitive reductases (type I) reduce nitroaromatic compounds by two-electron transfers irrespective of the presence of oxygen. Thus, nitroso, hydroxylamino, or amino derivatives can be products of the enzymatic reduction.

Oxygen-insensitive nitroreductases have been reported to be involved in aerobic degradative pathways of nitrobenzene (34), 4-nitrotoluene (21, 37, 43), 3-nitrophenol (3NP) (31, 39), and 4-nitrobenzoate (19). The common characteristic of these pathways is that the unique product of nitro group reduction is the hydroxylamino derivative rather than the aminoaromatic derivative. Characteristically, the arylhydroxylamines are directly converted to ring cleavage substrates.

Following this mechanism, Ralstonia eutropha JMP134 degrades 3NP via 3-hydroxylaminophenol (3HAP) and aminohydroquinone. The reaction is catalyzed by a nitroreductase and a 3HAP mutase, respectively (39, 40). Additionally, R. eutropha JMP134 can grow with 2-chloro-5-nitrophenol (2C5NP) as the sole source of nitrogen, carbon, and energy.

Most investigations on the degradation of chloronitroarenes have revealed cometabolic transformations (33, 38, 45). In contrast, Rhodococcus erythropolis HL 24-1 was shown to utilize 2-chloro-4,6-dinitrophenol as a sole nitrogen, carbon, and energy source (27). Another exception is 4-chloro-2-nitrophenol, which was mineralized by a mixed culture in a coupled anaerobic-aerobic process (5). Interestingly, Bruhn et al. (8) constructed 4-chloro-2-nitrophenol-assimilatory bacteria by transferring the plasmid-encoded haloaromatic degrading sequences from either R. eutropha JMP134 or Pseudomonas sp. strain B13 into Pseudomonas sp. strain N31. Before the conjugation experiment, the recipient strain was able to remove the nitrogen from 4-chloro-2-nitrophenol as nitrite by a monooxygenase but failed to further degrade the resultant 4-chlorocatechol.

In order to elucidate how R. eutropha JMP134 metabolizes 2C5NP and how the nitrogen and the chlorine are eliminated, physiological investigations have been undertaken. This study reveals the initial reactions of the degradative pathway and specifies the nitroreductase involved.

MATERIALS AND METHODS

Growth of bacteria.

R. eutropha JMP134 (35) was maintained on solid mineral media as described previously (39). Additionally, the strain was grown on solid nitrogen-free mineral medium (9) containing 3NP (0.5 mM) as a nitrogen source and succinate (10 mM) as a carbon and energy source. During tests of 2C5NP as a sole source of nitrogen, carbon, and energy, the starting concentration of 2C5NP was 0.46 mM. After the substrate was utilized, an additional 0.35 mM 2C5NP was added. In the control medium, 2C5NP was omitted. The cultivation conditions reported for the growth experiment with 3NP were chosen (39). Cell growth, protein content in the growing cells, and ammonia concentrations were determined as described previously (39).

Induction of cells.

R. eutropha JMP134 was grown with 0.5 mM 3NP and 10 mM succinate in nitrogen-free mineral medium (9). Additional portions of 3NP (0.25 mM) were added after exponential growth began. 2C5NP-induced cells were obtained in the same way except that 3NP was replaced by 2C5NP in the growth medium. Uninduced cells were obtained as described above but ammonium (1 mM) replaced the nitroaromatic compounds. When the culture reached an A₅₄₆ of ≥0.5, the cells were harvested by centrifugation, washed twice with 50 mM phosphate buffer (pH 7), and suspended in the same buffer. The cells were used for resting-cell experiments or for preparation of cell extracts.

Preparation of cell extracts.

Cells were disrupted as described previously (39). For preparation of cell extracts containing membrane-bound proteins, cell debris was removed by centrifugation at 30,000 × g for 30 min at 4°C. Cell extracts without membrane-bound proteins were obtained by centrifugation at 100,000 × g for 60 min at 4°C. Cell extracts were stored either on ice or frozen at −20°C until they were used. The protein content of lysates or enzymes was determined by the method of Bradford (7).

Partial purification of 3NP nitroreductase.

3NP nitroreductase was partially purified by the same procedure and with the same starting material (in 55 ml of phosphate buffer) as those described for the purification of 3HAP mutase (40). 3NP nitroreductase coeluted (30 ml) with 3HAP mutase from the DEAE CL-6B (weak anion-exchange resin) column at a concentration of 0.21 M NaCl. The enzymes were separated by using a butyl agarose (resin for hydrophobic interactions with proteins) column that was equilibrated with buffer containing ammonium sulfate (1 M). Whereas the mutase adhered, the nitroreductase passed through the column. Fractions containing 3NP nitroreductase activity (45 ml) were concentrated and desalted by ultrafiltration with a Centriprep 30 filter unit (Amicon). The filtrate containing 3NP nitroreductase (1.15 ml) was stored frozen at −20°C until it was used for the experiments described below.

Experiments with resting cells, cell extracts, and enzyme preparations.

Resting-cell experiments and experiments with cell extracts or enzyme preparations were performed as described previously (39). Concentrations of substrates and cofactors, as well as cell densities or protein contents, are stated in the text. 3HAP mutase was purified to homogeneity, and its activity was determined as previously described (40). The activity of 3NP nitroreductase was measured spectrophotometrically as described previously (39).

Analytical methods.

Chloride ion concentration was determined by ion chromatography (DX-100 Ion Chromatograph; Dionex, Idstein, Germany) with a anion-exchange column (Ionpac AS14, guard column AG14; Dionex) using a conductivity detector with a suppression technique. The mobile phase consisted of an aqueous solution of Na₂CO₃ (3.5 mM) and NaHCO₃ (1.0 mM), and the flow rate was 1.2 ml/min. Samples for chloride analyses (1 to 2 ml) were incubated in a water bath at 80°C for at least 15 min. Precipitated proteins or cells were removed by centrifugation prior to analysis by high-pressure liquid chromatography (HPLC).

Analyses of 3NP, 2C5NP, nitrobenzene, and their metabolites were performed with an HPLC system (Sykam, Gilching, Germany) equipped with a diode array detector (Philips, Gilching, Germany). A reversed-phase column (Lichrospher 100 RP8; 4.6 by 125 mm; Merck, Darmstadt, Germany) was used for the separation of the compounds, and the flow rate was 1 ml/min. Conversion of 3NP was measured with a gradient system that started with 10% solvent A (methanol containing hexane sulfonate [Pic B6; Waters, Milford, Conn.]) and 90% solvent B (water containing hexane sulfonate) and remained constant for 6 min. Then the ratio was stepped up to 40% solvent A and 60% solvent B and remained constant for 6 min. Conversion of 2C5NP or nitrobenzene was measured with a gradient system that started with 10% solvent A and 90% solvent B and remained constant for 7 min. Then the composition changed linearly within 3 min to 50% solvent A and 50% solvent B, and the solvents remained at this ratio for another 5 min. An isocratic system consisting of 10% solvent A and 90% solvent B was used to compare retention volumes and UV spectra of standards and metabolites that were formed upon experiments with 2C5NP and 3NP.

Chemicals.

2C5NP was a gift from Bayer AG (Leverkusen, Germany). Hydroxylaminobenzene was kindly provided by Shirley Nishino (Tyndall Air Force Base, Fla.). 2-Amino-5-chlorophenol was synthesized by reduction of 2C5NP. For that, 20 μl of HCl (18% [vol/vol]) was added to 1 ml of aqueous 2C5NP (1 mM), and the reaction was started by adding approximately 10 mg of zinc powder. The mixture was mixed vigorously at room temperature until the yellow color completely disappeared. The residual powder was removed by centrifugation, and the supernatant was used as a standard solution for HPLC analysis. 2-Chloro-5-hydroxylaminophenol was synthesized by a method analogous to that described for synthesis of 3HAP, aminohydroquinone was prepared from dimethoxyaniline, and N-acetylamino-hydroquinone was obtained by transformation of 3NP by resting cells of R. eutropha JMP134 as described previously (39). All other chemicals were of the highest purity commercially available.

RESULTS

Induction of 3NP nitroreductase during growth on 2C5NP or 3NP.

The growth of R. eutropha JMP134 on 2C5NP as a sole source of nitrogen, carbon, and energy in liquid culture was comparable with that on 3NP (39). After a lag phase (2 h), the A₅₄₆ increased from 0.028 to 0.074 (for the control without 2C5NP, the A₅₄₆ increased from 0.028 to 0.035) within 8 h, which synced with an increase in cell protein (11.5 μg/ml; control, 0.8 μg/ml) on the one hand and a complete consumption of 2C5NP (1.1 mM) on the other hand. Higher cell densities were obtained when more 2C5NP was added during exponential growth or when 2C5NP only served as a nitrogen source for the cells.

Ammonia release during growth on 2C5NP indicated an initial reduction of the nitro group. A 2C5NP-dependent NADPH-oxidizing activity comparable to that existent in 3NP-grown cells (39) was found in lysates of 2C5NP-grown cells but not in extracts from NH₄⁺–succinate-grown cells. A cross-experiment was carried out to determine whether both activities resulted from the induction of the same nitroreductase (Table 1). The identical relative rates indicated that R. eutropha JMP134 expressed the same nitroreductase during growth on either 3NP or 2C5NP.

TABLE 1.

Comparison of NAD(P)H-oxidizing activities

Substrate and cosubstratea	Activity (%)b in extract from:
	3NP-induced cells	2C5NP-induced cells
3NP
NADPH	100	100
NADH	6	7
2C5NP
NADPH	110	114
NADH	7	8

^aThe substrate and cosubstrate were used at 0.2 mM concentrations in 50 mM phosphate buffer (pH 7.5) containing the cell extract. Reactions were started by addition of the substrate. 

^bMeasured spectrophotometrically at 340 nm. One unit was defined as the oxidation of 1 μmol of NAD(P)H per min. One hundred percent activity was 176 U/g for the extract from 3NP-induced cells and 211 U/g for the extract from 2C5NP-induced cells. 

Chemoselective reduction of aromatic nitro groups to the corresponding hydroxylamino groups by 3NP nitroreductase.

3NP nitroreductase was separated from 3HAP mutase to facilitate detailed investigation of the action of the reductase (39, 40). Separation of 3NP nitroreductase from 3HAP mutase could be achieved by gel filtration or hydrophobic interaction chromatography but not by anion-exchange chromatography. The activity of 3NP nitroreductase always completely eluted as a single homogeneous band by all purification methods used. This is consistent with only one NADPH-dependent 3NP nitroreductase being present in extracts of induced cells of R. eutropha JMP134.

Partially purified 3NP nitroreductase (Table 2) was used to analyze the selectivity of the nitro group reduction of 3-NP and 2C5NP and to identify the products of 2C5NP and 3NP reduction. The enzyme readily reduced 3NP to 3HAP in the presence of excess NADPH under anaerobic conditions (Fig. 1A). Neither 3-nitrosophenol nor 3-aminophenol could be detected in the medium, and the 3HAP remained stable after 3NP (0.51 mM) was completely transformed. The reaction consumed 2.2 mol of NADPH per mol of 3NP. The enzyme was still active after 40 min, as indicated by the fact that when additional 0.5 mM 3NP was added, it was instantly reduced to 3HAP (data not shown). The results indicated that 3NP nitroreductase specifically reduced 3NP by the transfer of four electrons to yield 3HAP.

TABLE 2.

Partial purification of 3NP nitroreductase

Purification step	Total protein (mg)	Total activity (U)a	Sp act (U/mg)	Yield (%)
Cell extract	1,102	19.3	0.018	100
Lysate after ultracentrifugation	941	20.9	0.022	108
DEAE chromatography	162	14.9	0.092	77
Butyl agarose chromatography	3	5.5	1.8	28

^aOne unit is defined as the oxidation of 1 μmol of NADPH per min. 

FIG. 1.

Conversion of 3NP (A) and 2C5NP (B) to the corresponding hydroxylamino derivatives by 3NP nitroreductase. (A) 3NP (0.51 mM), NADPH (2.7 mM), and partially purified 3NP nitroreductase (63 mU/ml [44 ng/ml of protein]) were incubated in 50 mM phosphate buffer (pH 7.5). (B) 2C5NP (0.96 mM), NADPH (3 mM), NADH (3 mM), and partially purified 3NP nitroreductase (56 mU/ml [44 ng/ml of protein]) were incubated in 62 mM phosphate buffer (pH 7). Both experiments were carried out under an argon atmosphere at 30°C. Samples were analyzed by HPLC.

When 2C5NP and NADPH were incubated with the partially purified 3NP nitroreductase (Fig. 1B), a single product (metabolite A) was formed. Metabolite A had a UV spectrum similar to that of 3HAP (maxima at 203, 234, and 279 nm) but a larger retention volume. Authentic 2-chloro-5-hydroxylaminophenol showed the same UV absorption (maxima at 204, 240, and 288 nm) and chromatographic properties as metabolite A. Reduction of both metabolite A and the chemically synthesized standard by zinc in HCl acidic solution yielded 2-amino-5-chlorophenol, which was identified by HPLC.

Enzymes in cell extracts of R. eutropha JMP134 did not catalyze the reduction of 3NP, 2C5NP, or the hydroxylamino derivatives to the corresponding amino derivatives. In contrast, traces of 3-aminophenol and 2-amino-5-chlorophenol were produced from 3NP and 2C5NP by intact cells of JMP134. The constitutive activity was insignificant compared to that of 3NP nitroreductase in induced cells.

3NP nitroreductase showed 94% of the NADPH-oxidizing activity when nitrobenzene replaced 3NP as a substrate (Table 3). In contrast, 3HAP mutase was much less active (2.3%) toward hydroxylaminobenzene than toward 3HAP as a substrate (40). In fact, anaerobic conversion of nitrobenzene (0.5 mM) by an extract from induced cells of R. eutropha JMP134 led to a fast accumulation of hydroxylaminobenzene, which was slowly converted further to 2-aminophenol and 4-aminophenol. Aniline was not formed although excessive NADPH (2 mM) was added into the reaction mixture, which indicated that 3NP nitroreductase specifically formed hydroxylaminobenzene from nitrobenzene.

TABLE 3.

Substrate specificity of the 3NP nitroreductase from R. eutropha JMP134

Substratea	Relative activity (%) in cell extracts of:
	R. eutropha JMP134b	P. putida B2c
Nitrobenzene	94	122
1,3-Dinitrobenzene	12	—d
1,3,5-Trinitrobenzene	21	—
2-Nitrophenol	62	106
3-Nitrophenol	100	100
4-Nitrophenol	18	33
2-Methyl-3-nitrophenol	25	—
4-Methyl-3-nitrophenol	15	—
2-Methyl-5-nitrophenol	98	—
4-Chloro-3-nitrophenol	52	—
2-Chloro-5-nitrophenol	114	—
2,3-Dinitrophenol	14	—
2,4-Dinitrophenol	<5	—
2,5-Dinitrophenol	53	—
2,6-Dinitrophenol	22	—
3,4-Dinitrophenol	6	—
3-Nitrocatechol	29	—
4-Nitrocatechol	<5	6
Nitrohydroquinone	<5	—
Picric acid	135	—
2-Nitrobenzoate	<5	10
3-Nitrobenzoate	70	83
4-Nitrobenzoate	81	67
2-Nitrotoluene	8	31
3-Nitrotoluene	95	—
4-Nitrotoluene	86	131
2,3-Dinitrotoluene	93	—
2,4-Dinitrotoluene	114	—
2,6-Dinitrotoluene	103	—
3,4-Dinitrotoluene	100	—
3-Nitroaniline	71	—
2-Amino-4-nitrotoluene	219	—
2-Amino-6-nitrotoluene	<10	—
3-Amino-4-nitrotoluene	<10	—
4-Amino-2-nitrotoluene	<10	—
4-Amino-3-nitrotoluene	<10	—
2,6-Diamino-4-nitrotoluene	20	—
2,4,6-Trinitrotoluene	99	—

^aThe substrate and NADPH were each provided at a 0.2 mM concentration. 

^bNADPH-oxidizing activities shown are relative to that of 3NP (taken as 100%) and varied between 73 and 157 U/g, depending on the induction state of the cells prior to harvesting. Activities were measured in extracts from 3NP-succinate-grown (induced) cells. Relative activities in the negative control (extracts from NH₄⁺–succinate-grown cells) never exceeded 6%. 

^cData were taken from Meulenberg et al. (31). An activity level of 100% corresponded to 120 U/g. The nitroaryl concentration was 0.1 mM. 

^d—, not determined. 

Conversion of nitrosobenzene by cell extracts.

The proposed two-electron-transfer mechanism of oxygen-insensitive nitroreductases presupposes a transient formation of the corresponding nitrosoarenes during the enzymatic reaction (14). Therefore, nitrosoarenes should serve as potential substrates for the enzymes. Although 3HAP was the only detectable product during reduction of 3NP by 3NP nitroreductase of R. eutropha JMP134, an intermediate formation of 3-nitrosophenol could not be excluded. Since a standard of 3-nitrosophenol was not available and nitrobenzene was an alternative substrate for 3NP nitroreductase, nitrosobenzene was tested as a substrate for the enzyme in cell extracts.

The spontaneous reduction rate of nitrosobenzene (0.2 mM) by NADPH (0.2 mM) in 50 mM phosphate buffer (pH 7.2) was 42 μM min⁻¹. The corresponding reduction rate in the presence of an extract from uninduced cells of R. eutropha JMP134 (0.115 mg of protein per ml) was 53 μM min⁻¹. In the presence of an extract from 3NP-induced cells with the same protein content, the rate was 124 μM min⁻¹, demonstrating that nitrosobenzene served as a substrate for 3NP nitroreductase. After the unspecific reduction rate was subtracted, the rate for conversion of nitrosobenzene was 3.4-fold higher than that for nitrobenzene as a substrate in the extract from induced cells of R. eutropha JMP134.

Substrate specificity and characteristics of 3NP nitroreductase.

In Table 3, the nitroaromatic compounds tested as substrates for 3NP nitroreductase are listed. Of 39 compounds, only 10 were not reduced at measurable rates, 9 were reduced slowly, and 20 were reduced at significant rates, demonstrating a relaxed substrate specificity of the enzyme. Some of the compounds were previously tested as substrates for the NADPH-oxidizing activity in cell extracts from the 3NP-degrading Pseudomonas putida B2 (31), and the resulting rates were comparable with those found in extracts of R. eutropha JMP134. Both nitroreductases were inducible by 3NP and used NADPH as a cosubstrate. Therefore, the two enzymes seem to be similar.

Storage of cell extracts in buffer systems other than phosphate did not improve the stability of 3NP nitroreductase. The partially purified enzyme retained 74% of its original activity after 31 days when it was stored frozen at −20°C in 50 mM phosphate buffer (pH 7.5) or 89% after storage on ice for 3 days. High concentrations of ammonium sulfate (≤30% [wt/vol]) and low-pH conditions (<pH 7) significantly decreased 3NP nitroreductase activity.

Most bacterial oxygen-insensitive nitroreductases contain flavin mononucleotide (FMN) as a prosthetic group and utilize NAD(P)H as an electron donor (14). Incubation of an extract (0.56 mg/ml) from induced cells of R. eutropha JMP134 together with FMN, flavin adenine dinucleotide (FAD), or Fe²⁺ (each at 0.1 mM) for 20 min at room temperature did not affect the activity of 3NP nitroreductase. Fractions of the partially purified enzyme exhibited no characteristic visible absorption, and no significant loss of activity was observed during the purification procedure. This, however, does not rule out the possibility that tightly bound cofactors exist. Some nitroreductases additionally require a metal cation as a cofactor (6). Incubation of a cell extract containing 3NP nitroreductase with Cu²⁺ (1 mM) inhibited the activity completely. Inhibition by Cu²⁺ was also noticed for 2,4-dinitrophenol nitroreductase from Rhodobacter capsulatus (6) and for 1-nitropyrene nitroreductase (NRase I) from Bacteroides fragilis (25), which indicates the presence of a metal cofactor.

Identification of aminohydroquinone and its acetyl derivative as metabolites of 2C5NP.

When 2C5NP (1 mM) was incubated aerobically with 2C5NP-induced resting cells (A₅₄₆ = 9.8), no organic metabolites were detected by HPLC analysis but chloride (0.82 mM) was released. Therefore, the same experiment was repeated under anaerobic conditions (A₅₄₆ = 27), as shown in Fig. 2. Here, complete conversion of 2C5NP (0.65 mM) was also accompanied with a nearly stoichiometric increase in the concentration of chloride (0.57 mM) in the medium. One dominant metabolite, which exhibited the same chromatographic properties and UV absorption as authentic aminohydroquinone, was detected by HPLC. In fact, aminohydroquinone was slowly converted further to N-acetylaminohydroquinone (data not shown), which could be identified by comparing its chromatographic properties and UV absorption with those of a biologically obtained standard.

FIG. 2.

Conversion of 2C5NP by induced resting cells of R. eutropha JMP134. 2C5NP (0.65 mM) and resting cells (A₅₄₆ = 27) were incubated in 50 mM phosphate buffer (pH 7) under an argon atmosphere at 30°C. The decrease in 2C5NP (□) and the formation of aminohydroquinone (○) and chloride (▵) were determined by HPLC.

Characterization of metabolites formed from 2C5NP by R. eutropha JMP134.

An unknown metabolite, designated metabolite B, accumulated when an extract from 2C5NP-induced cells (0.47 mg of protein per ml) was incubated anaerobically with 2C5NP (0.5 mM) and NADPH (3 mM). Aminohydroquinone was not formed even when the corresponding extract (0.88 mg of protein per ml) contained membrane-bound protein and NADH (1 mM) or glutathione (2.5 mM) as alternative electron donors were added to the buffer. When samples with metabolite B were exposed to air, an increasingly red color appeared in the solution concomitant with the formation of a new product. The red color disappeared instantly when NADH was added to the samples, and metabolite B was regenerated. Metabolite B could be stabilized by acidification of the sample with HCl. Exactly the same characteristics were observed with a sample containing aminohydroquinone and cell extract. The UV-absorption spectra of aminohydroquinone (maxima at 218 and 291 nm) and metabolite B (maxima at 200 and 299 nm) were similar, but metabolite B eluted later from the reversed-phase column. The same was noticed for the red oxidation products, which were probably the quinone derivatives of aminohydroquinone (maxima at 210, 263, and 470 nm) and of metabolite B (maxima at 212, 285, and 485 nm). The results suggested that metabolite B was 2-amino-5-chlorohydroquinone. Isolation of metabolite B failed due its instability, which is comparable to that of aminohydroquinone (39). Therefore, no mass spectrometry and or nuclear magnetic resonance data are available for the compound.

When purified 3HAP mutase (40) was added to a reaction mixture in which 2C5NP had previously been converted to 2-chloro-5-hydroxylaminophenol (Fig. 3) it was instantly converted to metabolite B. 3-NP-grown cells readily dechlorinated metabolite B to aminohydroquinone, confirming that it was 2-amino-5-chlorohydroquinone (Fig. 3, step 3). The observations suggested that 3NP nitroreductase formed 2-chloro-5-hydroxylaminophenol, which was rearranged to 2-amino-5-chlorohydroquinone by 3HAP mutase, and the dechlorination was identified as the third step in the pathway.

FIG. 3.

Conversion of 2C5NP by 3NP nitroreductase (step 1), 3HAP mutase (step 2), and 3NP-grown cells (step 3) of R. eutropha JMP134. At time zero, partially purified 3NP nitroreductase (0.56 U) was added to 62 mM phosphate buffer (pH 7) with 2C5NP (0.96 mM), NADPH (3 mM), and NADH (3 mM). After 61.5 min purified 3HAP mutase (0.72 U) was added, and after 170 min 3NP-grown cells (A₅₄₆ = 13) were added. Incubation was carried out at 30°C under anaerobic conditions. Concentrations of 2C5NP (□), 2-chloro-5-hydroxylaminophenol (◊), 2-amino-5-chlorohydroquinone (○), and aminohydroquinone (▵) were analyzed by HPLC. In order to estimate the concentration of 2-amino-5-chlorohydroquinone, aminohydroquinone was used as a standard.

DISCUSSION

In natural microbial communities, nitroaromatic compounds are subject to gratuitous and unspecific reductions of nitro groups. The extent of reduction and the complexity of the reduction products formed depend on the number of the nitro groups on the aromatic ring and the redox potential of the culture. In contrast, nitro group reduction in axenic cultures that harbor a productive and complete catabolic sequence is highly selective. Thus, in R. eutropha JMP134, an inducible and oxygen-insensitive nitroreductase uses NADPH as an electron donor and catalyzes the reduction of aromatic nitro groups of 3NP and 2C5NP, which selectively stops at the level of the hydroxylamino group. This unique chemoselective reaction was shown with 3NP, 2C5NP, and nitrobenzene as substrates. Previously, it was reported that cell extracts from R. eutropha JMP134 containing 3NP nitroreductase reduced 4-nitrobenzoate exclusively to 4-hydroxylaminobenzoate. Also, 4-nitrotoluene was converted to 4-hydroxylaminotoluene and 6-amino-m-cresol but not to 4-aminotoluene (40). These results confirm the high chemoselectivity, albeit relaxed substrate specificity, of the 3NP nitroreductase. An analogous nitroreductase from Pseudomonas pseudoalcaligenes JS45 was purified and characterized (42). This enzyme catalyzed the reduction of nitrobenzene to hydroxylaminobenzene, which is a metabolite of a complete degradative pathway of nitrobenzene in this bacterium (23, 34). The enzyme also transformed 2,4,6-trinitrotoluene to 4-hydroxylamino-2,6-dinitrotoluene and subsequently to 2,4-dihydroxylamino-6-nitrotoluene, demonstrating that the enzyme, although highly chemoselective, exhibits relaxed substrate specificity (18). Usually, nitroreductases from enteric bacteria have not been tested for chemoselectivity. This was done only with NfnB from Escherichia coli B, which specifically reduces the antitumor agent 5-(aziridin-1-yl)-2,4-dinitrobenzamide (CB1954) to equimolar amounts of its 2- and 4-hydroxylamino derivatives but not to the corresponding amino derivatives (1).

Nitrosoaromatic derivatives have been proposed as metabolites in catabolic pathways of mononitroaromatic compounds (19, 31, 42). It is unlikely that oxygen-insensitive nitroreductases release the nitroso intermediates during the reduction of nitroaromatic compounds. The reduction of a nitroso group yielding a hydroxylamino group does not require much activation energy, and therefore most nitrosoaromatic compounds spontaneously react with reduced pyridine and flavin nucleotide cofactors (4, 28). Nitrosobenzene also served as a substrate for 3NP nitroreductase from R. eutropha JMP134 and was converted faster than nitrobenzene, indicating that the initial two-electron transfer forming the nitroso intermediate is the rate-limiting step of the complete reductive sequence. Likewise, nitrosobenzene was shown to be a substrate for nitrobenzene nitroreductase (42). In contrast to the enzyme from R. eutropha JMP134, the corresponding rates for conversion of nitrosobenzene and nitrobenzene were approximately the same.

Nitroreductases from Salmonella typhimurium (47), Enterobacter cloacae (10, 12), and E. coli (32, 52) and a flavin reductase from Vibrio fischeri (53) show high similarity in their amino acid sequences and form a protein family. In contrast, the amino acid sequences of the nitroreductase NfsA from E. coli and of the flavin oxidoreductase from Vibrio harveyi are highly similar, but they are different from those of other enteric nitroreductases (51) mentioned above. Nitrobenzene nitroreductase from P. pseudoalcaligenes JS45 possesses an N-terminal amino acid sequence that is different from all known sequences of other nitroreductases (42). It remains to be clarified whether this enzyme and other nitroreductases involved in the catabolism of nitroaromatic compounds, including 3NP nitroreductase from R. eutropha JMP134, represent an independent class of specific nitroreductases. Nitroreductases presumably derive from different flavin nucleotide-containing enzymes and are therefore a heterogenous group of enzymes.

In addition to 3NP nitroreductase, cells of R. eutropha JMP134 possess a constitutive, marginal activity for nitro group reduction which converts 3NP and 2C5NP to the corresponding aminophenols. The presence of several different nitroreductases in a bacterial strain is not unusual. Kinouchi and Ohnishi (25) likewise separated four different nitroreductases from a strain of B. fragilis, each reducing 1-nitropyrene to 1-aminopyrene. Kitamura et al. (26) found three enzymes in E. coli B/r that reduced methyl p-nitrobenzoate to unidentified products. Bryant et al. (13) found three type I enzymes in E. coli K-12, each reducing nitrofurazone to the open-chain nitrile. Peterson et al. (36) described a type I and a type II nitroreductase in E. coli K-12, each reducing the nitro group of nitrofurazone, forming the open-chain nitrile and aminofurazone, respectively.

Characteristically, no organism has yet been found that degrades a nitroaromatic compound via complete reduction of the nitroaryl to the corresponding aniline, which is subsequently oxidized and assimilated via catechol. This sequence would require at least one more reducing equivalent than the pathways that include arylhydroxylamines as key metabolites and is therefore less efficient for bacterial growth.

The hydroxylamino compound as the first metabolite of 2C5NP catabolism is subject to an isomerization analogous to the acid-catalyzed Bamberger rearrangement. As recently described, the purification and characterization of the 3-hydroxylaminophenol mutase from R. eutropha JMP134 clearly revealed that the conversion of 3-hydroxylaminophenol to aminohydroquinone is catalyzed by the single enzyme. Correspondingly, the enzymatic reaction of 2-chloro-5-hydroxylaminophenol by the mutase leads to the formation of 2-amino-5-chlorohydroquinone as the second step in the degradative pathway of 2C5NP.

As the third step of 2C5NP degradation by R. eutropha JMP134, a reductive dechlorination of 2-amino-5-chlorohydroquinone to aminohydroquinone was observed. Reductive dechlorination at the aromatic ring by aerobic bacteria is rarely observed. The involvement of a hydride-Meisenheimer complex was proposed for reductive dechlorination of 2-chloro-4,6-dinitrophenol to 2,4-dinitrophenol by R. erythropolis HL 24-1 and R. erythropolis HL PM-1 (27). Azotobacter chroococcum MSB-1 formed 4-chlorophenoxyacetate and chloride from 2,4-dichlorophenoxyacetate (3). Two different bacterial species were reported to dehalogenate 2,4-dichlorobenzoate to yield 4-chlorobenzoate (46, 50). Different aerobic bacterial strains were shown to degrade pentachlorophenol partly by reductive dechlorination (2, 20, 30, 44, 48, 49).

The proposed initial reactions of the degradative pathways of 3NP and 2C5NP are shown in Fig. 4. The first two steps of 3NP and 2C5NP degradation by R. eutropha JMP134 are analogous. The nitrophenolic compounds are reduced to the hydroxylaminophenols, which then undergo an enzymatic Bamberger rearrangement to yield aminohydroquinone or its chloro analogue. In the case of 2-amino-5-chlorohydroquinone, a reductive dechlorination is involved and aminohydroquinone is formed, which is therefore a common metabolite in both pathways.

FIG. 4.

Initial steps of 3NP and 2C5NP degradation by R. eutropha JMP134.

Recently, a catabolic sequence similar to 3NP was reported in Mycobacterium sp. strain HL 4-NT-1, which degrades 4-nitrotoluene via 4-hydroxylaminotoluene and 6-amino-m-cresol by a nitroreductase and a mutase, respectively. 6-Amino-m-cresol is subject to ring cleavage, yielding 2-amino-5-methylmuconic semialdehyde, which is oxidized to 2-amino-5-methylmuconic acid (43). Meanwhile, the deamination reaction in the pathway of nitrobenzene degradation in P. pseudoalcaligenes JS45 has been characterized (23). Here, nitrobenzene is reduced to hydroxylaminobenzene, followed by rearrangement to 2-aminophenol, which then undergoes meta ring cleavage to 2-aminomuconic semialdehyde. The semialdehyde is oxidized to 2-aminomuconate, which is subsequently deaminated to 2-hydroxymuconic acid. Considering these pathways, aminohydroquinone is the most likely substrate for ring cleavage during the degradation of 3NP and 2C5NP by R. eutropha JMP134.