The plasmids and bacteria are in Table 1, and the primers are in Table S1. Cupriavidus strains were grown at 30°C in minimal salt medium (MSM, omitting the CaCl₂) (Xiao et al., 2006) with different substrates (2C5NP was dissolved in MSM with initial concentration of 5 mM). E. coli strains were grown in lysogeny broth (LB) medium at 37°C. When required, chloramphenicol (34 μg/ml), tetracycline hydrochloride (10 μg/ml) and kanamycin (50 μg/ml) was added to the medium. All reagents used here were purchased from Sigma Chemical Co. (St. Louis, MO).

Strain CNP-8 was isolated from the pesticide-contaminated soil collected from Yantai, China, by enrichment culture method (Liu et al., 2005). This strain was identified by 16S rRNA sequencing with universal primers 27F and 1492R (Polz and Cavanaugh, 1998). Its morphology was observed by scanning electron microscopy (Model S-4800, Hitachi Ltd., Japan).

Degradation of 2C5NP by strain CNP-8 was investigated by monitoring the OD₆₀₀ and the consumption of substrate. Ammonia concentration was quantified colorimetrically by the Nessler reaction (Krug et al., 1979), and chloride ion was determined by using an ion-selective chloride electrode (Model 96-17, Orion). Degradation of different concentrations of 2C5NP (0.3–0.7 mM) by strain CNP-8 was carried out to study the degradation kinetics. The values of kinetic parameters were derived from modified Gompertz model (Fang et al., 2007; Deng et al., 2016) which could be expressed as follows:

S is the substrate concentration. S₀ is the initial substrate concentration. A is the biodegradation potential. μ_m is the maximum biodegradation rate. λ is the lag phase time.

Different carbon source (0.5 and 5 g/L of glucose, succinate or lactate) was added to MSM containing 0.4 mM of 2C5NP to investigate their effect on 2C5NP degradation by strain CNP-8. Biotransformation of meta-nitrophenol (MNP) and 2C5NP by strain CNP-8 was carried out as described (Chen et al., 2014). Quantification of 2C5NP and MNP were performed by HPLC analysis. In order to identify the intermediates of 2C5NP degraded by strain CNP-8, 2C5NP was incubated under anaerobic condition with 2C5NP-induced cells, and the metabolites were identified by GC-MS analysis.

HPLC assay was performed as previously described (Min et al., 2016a). MNP and 2C5NP were quantified at 280. The authentic MNP and 2C5NP had retention times of 10.5 and 13.4 min, respectively. The HPLC-MS condition was the same as described previously (Min et al., 2016a) except the gradient mobile phase was different. The mobile phase consisted of A (acetonitrile) and B (H₂O) with the following gradients: 0–5 min, 5% A; 5–10 min, 5–20% A; 10–20 min, 20–40% A; 20–25 min, 40% A. The condition of GC-MS analysis was the same as described previously (Zhang J. J. et al., 2009).

Whole-genome sequencing and annotation of strain CNP-8 was performed in the Novogene Bioinformatics Institute (Beijing, China). The transcriptional levels of mnp genes were performed by real-time quantitative PCR (RT-qPCR) as described (Min et al., 2016a) with primers described in Table S1. The relative expression levels of target genes were calculated according to the 2-ΔΔCT method with the 16S rRNA as a reference gene (Livak and Schmittgen, 2001). mnpA, mnpB, mnpC, and mnpD genes amplified from the genomic DNA (the primers are listed in Table S1), were respectively cloned into pET-28a to obtain the expression plasmids (Table 1). The expression plasmids were subsequently introduced into E. coli Rosetta(DE3)pLysS, and the protein expression and purification was performed as previously described (Liu and Zhou, 2012).

The catalytic activity of MnpA against 2C5NP and MNP was analyzed in the reaction mixture (1 ml) containing, 0.2 mM NADPH, 0.1–5 μg of purified H₆-MnpA, 50 μM substrate (2C5NP or MNP) and 50 mM phosphate buffer (pH 7.5). The reference cuvette contained all components except the substrate, and the activity assay was initiated with the addition of substrate. During identification of products, the experiment was performed under anaerobic conditions to prevent autoxidation of the products. The reaction mixture was extracted with ethyl acetate after HCl acidification, and the ethyl acetate phase was then collected and dried using anhydrous sodium sulfate before HPLC-MS analysis. In the kinetics assays, three independent experiments were carried out with 7 substrate concentrations (2C5NP: 5–50 μM; MNP: 2–20 μM), while the concentration of NADPH was fixed at 200 μM. The Michaelis-Menten equation was fitted to the data by using OriginPro 8. Bradford method (Bradford, 1976) was used to determine the protein concentration. One unit of enzyme activity was defined as the amount of enzyme consuming 1 μmol of substrate (2C5NP or MNP) in 1 min. The dioxygenase activity of MnpC toward hydroquinone and chlorohydroquinone was measured as described for aminohydroquinone dioxygenase (Yin and Zhou, 2010).

Gene knockout and complementation were performed as previously described (Min et al., 2014). Plasmid pEX18Tc-mnpA for mnpA knockout was constructed by fusing kanamycin resistance gene (nptII) and the upstream and downstream fragments of mnpA to EcoRI/HindIII digested pEX18Tc by using in-fusion HD cloning kit (Takara) (the primers are listed in Table S1, and the plasmids are in Table 1). Plasmid pRK415-mnpA used for mnpA complementation was obtained by cloning mnpA to the HindIII/KpnI digested pRK415. The ability of strains CNP-8ΔmnpA and CNP-8ΔmnpA[pRK415-mnpA] to utilize substrate (2C5NP or MNP) was examined by determining bacterial growth and substrate consumption.

The GenBank accession numbers for the nucleotide sequence of 16S rRNA, mnp gene cluster and mnpB gene reported in this paper are KY643479, KY937901, and KY937902, respectively.

A 2C5NP-degrading bacterium, designated strain CNP-8, was isolated from pesticide-contaminated soil with 2C5NP as a sole source of carbon and nitrogen. It was characterized as Gram-negative by Gram staining, and observed as short-rod-shaped by scanning electron microscopy (Figure 1A). It was identified as a member of Cupriavidus based on 16S rRNA sequence analysis. The phylogenetic relationships of the 16S rRNA gene sequences of strain CNP-8 and other representative Cupriavidus strains are shown in Figure 1B. Before this study, there is only one bacterium, Cupriavidus pinatubonensis JMP134 (Schenzle et al., 1999b), was reported to be able to degrade 2C5NP.

Considering that strain CNP-8 degraded both MNP and 2C5NP via partial reductive pathways; therefore, biotransformation analysis was initially carried out to determine whether the metabolism of 2C5NP and MNP share the enzymes. The un-induced strain CNP-8 had no activity for both 2C5NP and MNP (Figures 3A,B), whereas the 2C5NP-induced cells degraded 2C5NP with a rate of 5.22 ± 0.47 μM/OD₆₀₀ of cell/min and MNP-induced cells degraded MNP with a rate of 10.34 ± 1.21 μM/OD₆₀₀ of cell/min, indicating that the genes responsible for 2C5NP and MNP degradation in strain CNP-8 were inducible. Furthermore, MNP-induced cells can degrade 2C5NP (8.35 ± 0.87 μM/OD₆₀₀ of cell/min), while 2C5NP-induced strain CNP-8 was found to be able to transform MNP (6.01 ± 0.54 μM/OD₆₀₀ of cell/min). This suggests that the enzymes involved in 2C5NP degradation are likely also responsible for the catabolism of MNP in strain CNP-8. No intermediate was detected when strain CNP-8 was grown in MSM with 2C5NP as sole sources of carbon and nitrogen. However, a metabolite was captured by GC-MS analysis when 2C5NP was incubated under anaerobic condition with 2C5NP-induced cells. This intermediate was proposed as aminohydroquinone based on the molecular ion peak at m/z 125 (Figure S1).

The above finding revealed that metabolism of 2C5NP and MNP in strain CNP-8 may share identical enzymes, which motivate us to identify their coding genes. This knowledge is very significant because the molecular mechanism of microbial 2C5NP degradation remains unknown. Initially, the draft genome of strain CNP-8 was sequenced. A gene cluster designated as mnp (Figure 4A) was identified from the contig 80 by comparative genomics analysis with Cupriavidus pinatubonensis JMP134. The proteins encoded by mnpA and mnpC exhibit high homology with MNP nitroreductase (93% identity) (Yin et al., 2010) and aminohydroquinone dioxygenase (94% identity) (Yin and Zhou, 2010), respectively, which were reported to be involved in MNP catabolism in strain JMP134. mnpD appears to encode a reductive dechlorinase belong to the glutathione S-transferase family as the level of identity of MnpD with the 2,5-dichlorohydroquinone reductive dehalogenase from Sphingomonas paucimobilis UT26 is 40% (Kumari et al., 2002). MnpR, MnpE, and MnpF were proposed to be LysR regulatory protein, amidase and maleylacetate reductase, respectively, by BLAST analysis with the available genome sequence of strain JMP134. In addition, a gene (designed as mnpB) encoding a protein with 99% of identity with the 3-hydroxylaminophenol mutase involved in MNP catabolism in strain JMP134 (Schenzle et al., 1999a) was identified from the contig 35 by comparative genome analysis.

RT-qPCR showed that the transcription levels of mnpA, mnpB, mnpC, and mnpD under 2C5NP-induced condition were significantly increased in comparison with the un-induced condition (Figure 5), with 518-, 38-, 319-, and 99-fold increase, respectively. The differences of transcription level among mnpAC, mnpB and mnpD is likely due to that they were located on different operons and transcribed independently. Similarly, the mnp genes are also up-regulated under MNP-induced condition, with 407-, 53-, 279-, and 136-fold increase, respectively. This finding indicated that the mnp genes are likely responsible for both 2C5NP and MNP catabolism in strain CNP-8.

A total of 34 mg recombinant MnpA with specific activity of 33.6 U mg⁻¹ for 2C5NP (47.5 U mg⁻¹ for MNP) was purified from 500 ml of culture. Purified fractions of H₆-MnpA were yellow, and the enzyme has absorption peaks at 370 and 450 nm, consistent with previously reported flavoprotein nitroreductases (Somerville et al., 1995; Yin et al., 2010). For H₆-MnpB purification, 25.4 mg enzyme was obtained from 500 ml of culture. Although a certain amount of MnpC was inclusion body, 4.2 mg of H₆-MnpC was purified from 3,000 ml of culture. Unfortunately, only a very small amount of H₆-MnpD was soluble, and extremely little H₆-MnpD was purified even 9,000 ml of culture was collected. SDS-PAGE analysis of the purified Mnp proteins showed that the molecular masses of H₆-MnpA, H₆-MnpB and H₆-MnpC are about 26, 52, and 35 kDa (Figure S2), respectively, consistent with their deduced molecular masses.

E. coli Rosetta(DE3) carrying pET-mnpA had the ability to degrade 2C5NP and MNP by HPLC analysis, whereas neither 2C5NP nor MNP consumption was detected when the E. coli cells only containing plasmid pET-28a. Furthermore, the purified H₆-MnpA transformed MNP rapidly, together with the consumption of NADPH (λ_max = 340 nm) and accumulation of a metabolite with a λ_max of 234 nm (Figure 6C), in accord with the spectral property of 3-hydroxylaminophenol as reported (Schenzle et al., 1999b). Previously, the MNP nitroreductase from Cupriavidus pinatubonensis JMP134 have been reported to catalyze the reduction of MNP (Yin et al., 2010), but its catalytic activity for 2C5NP has not been characterized. In this study, H₆-MnpA was found to be able to catalyze the degradation of 2C5NP, together with consumption of NADPH (Figure 6A). Two isobestic points at 232 and 254 nm, respectively, were observed, suggesting the conversion of 2C5NP to a new product (λ_max ≈240 nm). In contrast, no spectral change occurred when His₆-MnpA was omitted from the reaction mixtures (Figures 6B,D).

By HPLC-MS analysis, two products with retention times of 12.84 and 13.16 min, respectively, were detected when 2C5NP was transformed by purified H₆-MnpA under anaerobic conditions (Figure 7A). Metabolite A (λ_max = 230, 298 nm) was suggested as 2-chloro-5-nitrosophenol, with a deprotonated ion at m/z 156.14 and its fragments at m/z 125.89 (loss of −NO) and at m/z 139.23 (loss of −HO) (Figures 7B,C). Metabolite B (λ_max = 240, 288 nm) has the same spectral property with synthetic 2-chloro-5-hydroxylaminophenol (Schenzle et al., 1999b), which has a deprotonated ion at m/z 158.03 and the fragments at m/z 126.09 (loss of −NHOH) and at m/z 141.23 (loss of −HO) (Figures 7D,E). On the basis of the products identification of MnpA, strain CNP-8 was further proved to degrade 2C5NP via a partial reductive pathway (Figure 4B), apart from the release of ammonium during 2C5NP degradation. The partialy-purified nitroreductase from Cupriavidus pinatubonensis JMP134 was previously proved to catalyze the transformation of 2C5NP to 2-chloro-5-hydroxylaminophenol (Schenzle et al., 1999b), but its protein sequence and coding gene were not reported. Moreover, 2-chloro-5-nitrosophenol, the initial intermediate of microbial 2C5NP degradation, was detected for the first time in this study.

Neglectable activity of H₆-MnpA for both 2C5NP and MNP was observed when NADPH in the reaction mixture was substituted by NADH, indicating that MnpA prefers NADPH than NADH as the co-substrate, consistent with other reported nitroaromatic nitroreductases (Somerville et al., 1995; Yin et al., 2010). Enzymatic kinetics assays revealed that H₆-MnpA exhibit a higher affinity and catalytic efficiency for MNP (K_m = 3.4 ± 1.63 μM, k_cat/K_m = 342 ± 47.6 μM⁻¹ min⁻¹) than 2C5NP (K_m = 16.3 ± 2.87 μM, k_cat/K_m = 158 ± 31.2 μM⁻¹ min⁻¹), suggested that MNP is the physiological substrate for MnpA in strain CNP-8.

The enzymatic analysis of MnpB was not carried out in this study since the commercial standard of its substrate is unavailable. However, considering that MnpB exhibits extremely high degree of identity with the 3-hydroxylaminophenol mutase (accession number WP_011298219), which was capable to catalyze the transformation of synthetic 2-chloro-5-hydroxylaminophenol to 2-amino-5-chlorohydroquinone (Schenzle et al., 1999a), and its coding gene is highly transcribed in 2C5NP-induced strain CNP-8; therefore, it is reasonable to conclude that MnpB is responsible for transformation of 2-chloro-5-hydroxylaminophenol to 2-amino-5-chlorohydroquinone during 2C5NP degradation by strain CNP-8 (Figure 4B). Purification of H₆-MnpD failed, which hampered its enzymatic analysis in vitro. However, MnpD shares moderate identity (40%) with LinD from Sphingomonas paucimobilis UT26 which was reported to catalyze dechlorination of 2,5-dichlorohydroquinone, a structure analog of 2-amino-5-chlorohydroquinone, to 2-chlorohydroquinone (Kumari et al., 2002). Moreover, mnpD, located upstream of the 2C5NP nitroreductase-encoding mnpA, is up-regulated in 2C5NP-induced strain CNP-8. These combined data suggested that MnpD is lilely responsible for dechlorination of 2-amino-5-chlorohydroquinone to aminohydroquinone (Figure 4B).

Aminohydroquinone is extremely unstable (Schenzle et al., 1997; Yin and Zhou, 2010); therefore, its structure analogs chlorohydroquinone (CHQ) and hydroquinone (HQ) were used to identify the ring-cleavage function of MnpC. H₆-MnpC catalyzed rapid degradation of both CHQ and HQ, together with the accumulation of respective product with a λ_max of 320 nm (Figure S3). In contrast, neither substrate consumption nor product accumulation was observed when His₆-MnpC was omitted from the reaction mixtures. Moreover, mnpC was co-transcribed with mnpA in 2C5NP-induced strain CNP-8. Therefore, MnpC was likely involved in the ring-cleavage reaction of 2C5NP degradation. So far, two kinds of (chloro)hydroquinone dioxygenase were reported. The linE-like single-subunit dioxygenases were reported to split the ring of CHQ between C1 and C2 (Miyauchi et al., 1999; Ohtsubo et al., 1999), whereas the HapCD-like two-subunit dioxygenase catalyzed the ring cleavage of CHQ between C1 and C6 (Moonen et al., 2008; Min et al., 2014). MnpC, a single-subunit aminohydroquinone dioxygenases reported here exhibits moderate identity (42%) with LinE, but has no sequence homology to HapCD. Therefore, the ring-cleavage position of aminohydroquinone catalyzed by MnpC during 2C5NP degradation was proposed between C1 and C2 with formation of 6-amino-4-hydroxy-6-oxohexa-2,4-dienoic acid (Figure 4B).

To determine the involvement of MnpA in 2C5NP catabolism in vivo, a mutant of strain CNP-8 with substitution of mnpA by a kanamycin resistant gene nptII was constructed through homologous recombination. Functional analysis showed that strain CNP-8ΔmnpA (with mnpA1 deleted) was no longer able to utilize 2C5NP as well as MNP (Figure 8). This indicated that mnpA is necessary for strain CNP-8 to grow on both 2C5NP and MNP. Although the mnpA1-complemented strain CNP-8ΔmnpA[pRK415-mnpA] regained the ability to utilize these two nitrophenols, it exhibited lower maximum degradation rate compared to the wild-type strain CNP-8 (Table 2).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.