INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Nitroaromatic compounds are very common environmental pollutants which are used as dyes, pesticides, and explosives (9). Many sites of factories that produce such chemicals are highly contaminated with these substances, including picric acid (2,4,6-trinitrophenol [2,4,6-TNP]) (33). Microbial degradation of these compounds depends on the number of nitro groups on the aromatic ring. Usually, the first step in aromatic degradation is electrophilic attack. Due to the electron-withdrawing effect of the nitro group, this step becomes more difficult with an increase in the number of nitro group substituents. Consequently, very few examples of microbial metabolism of trinitroarenes have been described.

In contrast to studies of 2,4,6-trinitrotoluene transformation (2, 5, 21, 32), there have been few investigations of the metabolism of picric acid (4, 8, 11, 16). The first description of microbial attack on this compound was the description of Erikson (4). Only two recent investigations dealing with picric acid metabolism have been described. Lenke et al. characterized a Rhodococcus erythropolis strain that is able to convert this substance with partial dead-end production of 2,4,6-trinitrocyclohexanone; picric acid is used as a sole source of nitrogen by this organism (11). The only example of complete mineralization of this compound was described by Rajan et al. The Nocardioides strain of these authors was able to grow on picric acid as a sole source of carbon and energy without dead-end product formation (16). The only intermediate metabolite detected in the culture supernatant was 2,4-dinitrophenol (2,4-DNP).

The first step in picric acid metabolism by R. erythropolis HL PM-1 is formation of a [H]-Meisenheimer complex. Generation of Meisenheimer complexes of nitroaromatic compounds is a well-known chemical reaction (13, 22-24). In contrast, there are only two examples of microbial conversion of nitroaromatic compounds to Meisenheimer complexes; both of these involve trinitroarenes (picric acid [11] and trinitrotoluene [31]).

In this communication we describe the first isolation and characterization of a strain which completely mineralizes picric acid with intermediate formation of the [H]-Meisenheimer complexes H-TNP and H-DNP. As this is the first description of a H-DNP complex in microbial metabolism, the complex was characterized in more detail.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Chemicals. Picric acid, 2,4-DNP, 2-nitrophenol, 4-nitrophenol, 4-nitrocatechol, phenol, 2,4-dichloro-6-nitrophenol, 2,6-dichloro-4-nitrophenol, and dry acetonitrile were obtained from Fluka (Neu-Ulm, Germany); 2,6-DNP, 2,4-dinitrotoluene, and 2,6-dinitrotoluene were obtained from Aldrich (Steinheim, Germany); picramic acid was obtained from Tokyo Kasei (Tokyo, Japan); and 2,4,6-trinitrotoluene was a gift from Chemiewerk Schönebeck, Schönebeck, Germany. All other chemicals were A-grade purity and were obtained from Merck (Darmstadt, Germany). Quartz bidistilled water was used throughout this study.

Media and growth conditions. One liter of R₂A medium (17) (pH 7.0) contained 0.5 g of yeast extract, 0.5 g of proteose peptone no. 3, 0.5 g of Casamino Acids, 0.5 g of glucose, 0.5 g of soluble starch, 0.3 g of sodium pyruvate, 0.3 g of K₂HPO₄, and 0.05 g of MgSO₄ · 7H₂O. The glucose was autoclaved separately. One liter of mineral salts medium A (pH 7.0) consisted of 1.53 g of Na₂HPO₄ · 2H₂O, 0.76 g of KH₂PO₄, 0.5 g of (NH₄)₂SO₄, 0.2 g of MgSO₄ · 7H₂O, 0.05 g of CaCl₂, and 10 ml of trace element solution SL-4. Trace element solution SL-4 contained 0.5 g of EDTA, 0.2 g of FeSO₄ · 7H₂O, 100 ml of trace element solution SL-6, and 900 ml of water. Trace element solution SL-6 contained 0.1 g of ZnSO₄ · 7H₂O, 0.03 g of MnCl₂ · 4H₂O, 0.3 g of H₃BO₃, 0.2 g of CoCl₂ · 6H₂O, 0.01 g of CuCl₂ · 2H₂O, 0.02 g of NiCl₂ · 6H₂O, 0.03 g of Na₂MoO₄ · 2H₂O, and 1,000 ml of water. Mineral salts medium B was the same as mineral salts medium A except that it lacked ammonium sulfate. All media were sterilized by heating them at 121°C for 30 min. Agar plates contained 15 g of agar per liter of medium. Cells were grown at 25°C; cultures in Erlenmeyer flasks were incubated on a rotary shaker at 200 rpm. Additional experiments were performed in a 1.5-liter Biostad B fermentor (Braun, Melsungen, Germany).

Buffers. Buffer A contained 50 mM potassium phosphate, 1 mM dithioerythritol, and 1 mM EDTA, and the pH was adjusted to 7.0. Buffer B contained 20 mM potassium phosphate, 5 mg of lysozyme per g of cells, 12.5 mg of deoxycholic acid (sodium salt) per g of cells, 1 mM dithioerythritol, 1 mM phenylmethylsulfonyl fluoride, and 1 mM EDTA, and the pH was adjusted to 7.0.

Isolation and selection of picric acid-degrading bacteria. A mixture of soil samples from former German trinitrotoluene production sites contaminated with various nitroaromatic compounds was incubated in mineral salts medium A supplemented with 0.44 mM picric acid for 2 days (1 g of soil per 100 ml of medium). Decolorized cultures were transferred into fresh mineral medium containing 0.44 mM picric acid. After five serial transfers, cultures were streaked onto agar plates containing mineral salts medium A and 0.44 mM picric acid as the only carbon source. Single colonies were picked. In order to obtain pure cultures, the cells were alternately plated onto R₂A agar or mineral salts medium A agar supplemented with different concentrations of picric acid (0.44 to 2.2 mM). Only one type of colonies, designated strain CB 22-1, was obtained.

Taxonomic assignment of colonies. Systematic microbiological assays of CB 22-1 and CB 22-2, including Gram staining, were performed as described previously (30). Strain CB 22-1 was classified by performing biochemical tests and a 16S ribosomal DNA analysis and by determining the composition of the cellular fatty acids, which was done by workers at the Deutsche Sammlung von Mikroorganismen und Zellkulturen GmbH (Braunschweig, Germany). The similarity of CB 22-1 and CB 22-2 was confirmed by biochemical and morphological tests.

Determination of cell density. Bacterial growth was monitored spectrophotometrically by measuring the optical density at 600 nm. As the culture supernatant turned red during fermentation, a cell-free sample of the corresponding supernatant was used as the blank.

Nitrite analysis. The nitrite concentration was determined as previously described (14), with the following modifications. Two hundred microliters of a sample diluted so that the nitrite concentration was up to 0.1 mM was added to 50 µl of 20 mM sulfanilic acid. The reaction mixture was incubated for 5 min at room temperature, and then 50 µl of a 0.04% (wt/vol) N-(1-naphthyl)ethylenediamine hydrochloride solution was added. After 20 min of incubation at room temperature, 700 µl of water was added, and the absorption at 550 nm was determined. Nitrite concentrations were calculated by using nitrite standards.

COD analysis. The chemical oxygen demand (COD) was determined by using a SPEKTROQUANT test (Merck) for a COD range of 4 to 40 mg of O₂. For the COD analysis cells were grown in flasks in mineral salts medium B supplemented with 1 mM picric acid until complete degradation was achieved. Subsequently, each suspension was centrifuged at 5,000 × g, and the nitrite concentration of the supernatant was determined. Various dilutions of the supernatants were analyzed by using the instructions provided with the COD test. Mineral salts medium B supplemented with nitrite at the appropriate concentration was used as the blank.

Preparation of H-TNP and H-DNP. H-TNP was prepared by adding 20 µl of 0.525 M sodium borohydride to 458 µl of 44 mM picric acid in aqueous solution (pH 7.0). H-DNP was synthesized as follows: 5 mmol of dry 2,4-DNP was dissolved in 10 ml of dry acetonitrile and heated to 50°C. At this temperature 4 mmol of dry sodium borohydride was added over a period of 10 s. The reaction mixture spontaneously turned red, and an orange-red solid formed at the bottom of the flask. After 3 min of incubation at 50°C, the supernatant was removed. The remaining solid was washed with 5 ml of cold acetonitrile and dried.

NMR spectra. Samples of the supernatant H-TNP were prepared by using thin-layer chromatography. Samples of H-DNP were prepared as described above. Nuclear magnetic resonance (NMR) spectra were recorded with a Bruker 400-MHz spectrometer by using solutions in D₂O-0.1 M NaOD and tetramethylsilane as the external standard.

HPLC. High-pressure liquid chromatography (HPLC) was performed with a Shimadzu instrument equipped with a photodiode array UV-visible light detector. A C₁₈ reversed-phase column (ET 250/8/4, 10-µm-diameter particles in packing; Macherey-Nagel, Düren, Germany) was used. The solvent systems used were water-1 M potassium phosphate (pH 7.0)-1 M sodium azide (979:20:1, vol/vol/vol) (solvent A) and methanol (solvent B). Compounds were eluted at a flow rate of 1 ml min¹ by using a gradient starting with 15% solvent B (0 to 5 min), followed by a gradual increase to 90% solvent B (5 to 15 min) and 90% solvent B (15 to 17.5 min). Commercially available picric acid, 2,4-DNP, 4-nitrophenol, 2-nitrophenol, 2-amino-4-nitrophenol, 4-amino-2-nitrophenol, and picramic acid were used as references. For quantification of H-DNP, a linear extinction coefficient at 420 nm (₄₂₀) (9 mM¹ cm¹) was estimated by using the UV spectrum of the synthetic substance.

Preparation of cell extracts. Cell extracts were prepared as follows. Cells were harvested by centrifugation at 5,000 × g for 20 min, washed with buffer B, and resuspended in 2 ml of buffer B per g of cells. After 30 min of gentle stirring at room temperature, the mixture was centrifuged at 30,000 × g for 30 min. The concentration of the resulting crude cell extract was determined as described by Bradford by using bovine serum albumin as the protein standard (2).

Preparation of substrates for enzymatic conversions. As the synthetic Meisenheimer complexes of picric acid and 2,4-DNP contained traces of sodium borohydride, the substrates used for transformation studies were prepared as follows. One hundred milliliters of culture supernatant from an incomplete batch fermentation containing H-TNP and H-DNP was applied to 1 ml of Q-Sepharose (Pharmacia, Uppsala, Sweden) previously equilibrated with buffer A. Ten-milliliter portions of the supernatant were subsequently added to the same 1 ml of Q-Sepharose, and each preparation was shaken until total decolorization was achieved. The saturated Q-Sepharose was applied to 4 ml of fresh Q-Sepharose (previously equilibrated with buffer A) in a 10-ml column connected to a fast-protein liquid chromatography system. Elution was accomplished by using a linear 0 to 2 M NaCl gradient in buffer A. The Meisenheimer complexes of 2,4-DNP and picric acid eluted in separate fractions at NaCl concentrations of 0.5 and 1.4 M, respectively.

Enzyme assays. The H-TNP-synthesizing enzyme was assayed by monitoring the increase in absorbance at 490 nm in buffer A containing 75 µM NADPH and 5 to 20 µl of sample in a total volume of 950 µl. The assay was started by adding 50 µl of 1 mM picric acid. Specific activities were calculated by using an ₄₉₀(H-TNP) of 14 mM¹ cm¹. The linear extinction coefficient was determined as follows. A 50 µM picric acid solution was enzymatically converted to H-TNP by using partially purified H-TNP-synthesizing enzyme. An HPLC analysis was performed to confirm the purity of the resulting H-TNP. The extinction coefficient was calculated by using the difference between the absorption of the completed reaction and the original absorption at 490 nm. The enzyme that converts H-TNP to 2,4-DNP was assayed by monitoring the decrease in absorbance at 490 nm in buffer A containing the Meisenheimer complex at a concentration of 50 µM and 20 µl of sample. The H-DNP-synthesizing enzyme was assayed by determining the absorbance at 460 nm in buffer A containing 75 µM NADPH and 5 to 20 µl of sample in a 950-µl (total volume) reaction mixture. The assay was started by adding 50 µl of 1 mM 2,4-DNP. H-DNP transformations were measured by monitoring the changes in the UV spectra in buffer A containing 50 µM Meisenheimer complex, 20 µl of sample, and 0.1 mM NADPH.

Partial purification of the H-TNP-synthesizing enzyme. Eight milliliters of crude cell extract with a protein concentration of 2 mg/ml was applied to a 10-ml DEAE Trisacryl column (Pharmacia) previously equilibrated with buffer A. The column was washed with 30 ml of buffer A. Elution was accomplished by using a linear gradient of sodium chloride (0 to 2 M) in 60 ml of buffer A at a concentration of 1.4 M of NaCl.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Taxonomic assignment of colonies. Cells of the strain that was isolated initially, strain CB 22-1, and its derivative, strain CB 22-2, were gram-positive, strictly aerobic, motile, nonsporulating, club-shaped rods. They were catalase positive. No acid or gas was produced from glucose. The peptidoglycan type of CB 22-1 was A3, LL-diaminopimelic acid-Gly. The 16S ribosomal DNA analysis of the section with the highest variability revealed that this strain exhibited 94.3% similarity to the type strain of Nocardioides simplex (the highest value obtained). As data for all previously described Nocardioides strains were available for comparison, we concluded that strain CB 22-1 represents a new species in this genus. Therefore, it was designated a Nocardioides sp. strain.

Picric acid degradation. In experiments performed in flasks, picric acid was completely transformed by Nocardioides sp. strain CB 22-2 at concentrations up to 5 mM in mineral salts medium A, and transformation was accompanied by stoichiometric nitrite release (Table 1). The COD of culture supernatants after degradation of 1 mM picric acid was 5 ± 2 mg of O₂ liter¹, which corresponded to an average of 3% of the initial COD plus the COD due to nitrite.

Characterization of metabolites. During growth on picric acid a change in color from yellow to orange-red was observed. A bathochrome effect was detected in the corresponding UV-visible light spectra (data not shown).

Examination of cell extracts. Figure 1 shows the time-dependent UV-visible light spectra related to the conversion of picric acid after addition of NADPH and cell extract. The characteristic increases in absorbance at 420 and 490 nm indicate that H-TNP was formed, as described previously for R. erythropolis HL PM-1 (11, 18). HPLC analysis confirmed this result. No activity was detected in the controls which did not contain either crude extract, picric acid, or NADPH. NADH could not be used instead of NADPH. The specific activities were 0.15 µmol min¹ mg of protein¹.

View larger version (15K): [in this window] [in a new window]   FIG. 1.   [H]-Meisenheimer complex formation during conversion of picric acid by cell extracts of Nocardioides sp. strain CB 22-2. UV spectra were recorded at zero time (), 5 min (----), 15 min (····), and 25 min (-··-).

1

1

View larger version (15K): [in this window] [in a new window]   FIG. 2.   Degradation of the [H]-Meisenheimer complex of picric acid by cell extracts of Nocardioides sp. strain CB 22-2 with formation of 2,4-DNP. UV spectra were recorded at zero time (), 2 min (----), 4 min (····), and 8 min (-··-).

View larger version (11K): [in this window] [in a new window]

View larger version (13K): [in this window] [in a new window]   FIG. 3.   Formation of the [H]-Meisenheimer complex of 2,4-DNP during conversion of 2,4-DNP by cell extracts of Nocardioides sp. strain CB 22-2. UV spectra were recorded at zero time (), 4 min (----), and 12 min (····).

View larger version (11K): [in this window] [in a new window]   FIG. 4.   UV-visible light spectra after HPLC analysis for a retention time of 2.07 min, showing the curves for the Meisenheimer complex of 2,4-DNP. The dashed line is the curve for the synthetic complex, and the solid line is the curve for the culture supernatant.

View this table: [in this window] [in a new window]   TABLE 2.   1H NMR chemical shift and coupling constants for H-DNP, 1,5-dinitro-3-methyl-3-aza-bicyclo[3.3.1]-nonene-(6)-one-(8) (a derivative of H-DNP) (compound 1), and 3-[H]-2,4-dinitroaniline (compound 2)

Partial purification of the H-TNP-synthesizing enzyme. In order to determine if H-DNP is formed by the first enzyme in a side reaction or is produced by a separate enzyme, a chromatographic fractionation experiment was performed. After this the H-TNP-forming enzyme was recovered at a low activity. The fractions did not exhibit any H-DNP-synthesizing or H-TNP-degrading activity.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

Nocardioides sp. strain CB 22-2 was enriched by using picric acid as the sole source of carbon, nitrogen, and energy. The release of approximately 3 mol of nitrite per mol of picric acid in mineral salts medium A containing an additional nitrogen source and the lack of dead-end product formation, as verified by the low residual COD, provided strong evidence that mineralization occurred. The yield, 0.29 g (dry weight) of cells per g of picric acid, seemed to be high for the following two reasons: (i) a considerable portion of the molecular weight of picric acid was provided by the three nitro groups; and (ii) at least two NADPH molecules had to be used to remove the nitro groups from the aromatic system.

Compared to the picric acid-degrading strains described previously, which were limited to picric acid concentrations lower than 4.4 mM, the strain which we investigated exhibited nearly 10 times greater picric acid resistance. In addition, the degradation rates during cultivation in flasks were about three times higher than the degradation rates reported by Rajan et al. (16). Therefore, Nocardioides sp. strain CB 22-2 has greater potential for picric acid degradation in waste streams or soils containing high concentrations of picric acid than other strains have.

A problem in this context is picric acid metabolism that is limited to about 15 mM nitrite. The ability of the strain to grow on picric acid without an additional nitrogen source does not solve this problem; the amount of nitrite which is necessary for biomass production is small compared to the amount of nitrite liberated. One possible solution, which would be useful in continuous chemostat fermentations, is adding nitrifying bacteria. We are currently investigating whether Nitrobacter strains can remove nitrite from fermentation broth.

Lenke and Knackmuss (11) postulated and later Rieger and Knackmuss (19) proved that the first two steps in picric acid catabolism by R. erythropolis HL PM-1 are the formation of H-TNP and the formation of 2,4-DNP (Fig. 5). The second reaction leads to the liberation of nitrite, which is used by the strain as the sole nitrogen source. These first two steps are the same in picric acid metabolism by Nocardioides sp. strain CB 22-2. Consequently, this mechanism seems to be a common mechanism in picric acid degradation.

View larger version (19K): [in this window] [in a new window]   FIG. 5.   First three steps in picric acid metabolism by Nocardioides sp. strain CB 22-2.

Reduction of the aromatic ring was proposed by Lenke et al. for the metabolism of 2,4-DNP (12). A reduced ring fission product containing two nitro groups, 4,6-dinitrohexanoate, was isolated as a dead-end product of 2,4-DNP metabolism by R. erythropolis. To our knowledge, no further details of this reaction are known. In this report we describe additional possible portions of the reductive pathway for catabolism of 2,4-DNP. We found a novel reduced metabolite, the 3-[H]-Meisenheimer complex of 2,4-DNP, which seems to be physiologically relevant for the following reasons: (i) H-DNP is produced by cell extracts of Nocardioides sp. strain CB 22-2 and also occurs in the culture supernatant; and (ii) an enzyme, which differs from the H-TNP-synthesizing activity, seems to be responsible for the formation of H-DNP.

In contrast to 2,4-DNP, which results from the elimination of nitrite from the H-TNP complex, no 4- or 2-nitrophenol was detected after transformation of H-DNP with cell extracts. Additionally, neither 2-nitrophenol nor 4-nitrophenol was used as a carbon source. It has been reported that uptake of nitrophenols into cells limits the reaction rates of these chemicals in Pseudomonas putida B2 (6). Consequently, it is possible that the lack of degradation capacity for mononitrophenols is caused by the lack of a transport mechanism. This explanation could not be confirmed for Nocardioides sp. strain CB 22-2, because the compounds were not transformed by crude cell extracts.

An alternative hypothetical pathway for H-DNP transformation is presented in Fig. 6. The results of the experiments performed with crude extracts imply that NADPH is necessary as a cofactor. The first reaction step is common for monooxygenases. There are some examples of hydroxylation of phenols in the ortho position, such as the phenol hydroxylase reaction (3, 15), and there are other examples of substitution at the para carbon atom, such as the 3-hydroxybenzoate monooxygenase reaction (10). Monooxygenase-catalyzed reactions are known for nitroaromatic compounds as well (25); p-nitrophenol is para-hydroxylated to hydrochinone by an enzyme of a Moraxella sp. strain in an NADPH- and oxygen-dependent reaction with concomitant release of nitrite (26, 27). The nitrophenol oxygenase of P. putida B2 converts 2-nitrophenol (ortho hydroxylation) to catechol in a corresponding reaction (34, 35). In the case of H-DNP, an analogous mechanism could lead to the following products: para-hydroxylation of H-DNP could result in 2-nitrohydrochinone, as shown in Fig. 6; and ortho hydroxylation could generate 4-nitrocatechol. As we did not find a 4-nitrocatechol-converting activity in crude extracts (data not shown) and as picric acid metabolism was inhibited by this compound, we postulated that the initial attack occurs at position 4. Compared to the examples described previously, there is the following important difference in our proposed mechanism: a Meisenheimer complex is the substrate for the monooxygenase. We argue that the initial electrophilic attack on 2,4-DNP should be more difficult than the attack on the mononitrophenols described previously, because 2,4-DNP contains one supplementary electron-withdrawing nitro group. To our knowledge, electrophilic attack of 2,4-DNP has not been described. In contrast, 2,4-dinitrotoluene is degraded via an initial dioxygenase reaction (28, 29). However, this reaction should be more facile due to the positive inductive effect of the methyl group, in contrast to the negative inductive effect of the hydroxyl group in 2,4-DNP. The addition of a hydride to the 2,4-DNP aromatic system should facilitate the electrophilic attack, because an additional negative charge is brought into the system. Workers are currently trying to prove or reject these mechanistic considerations.

View larger version (14K): [in this window] [in a new window]   FIG. 6.   Hypothetical mechanism for NADPH-dependent conversion of H-DNP.