A Bph-Like Nitroarene Dioxygenase Catalyzes the Conversion of 3-Nitrotoluene to 3-Methylcatechol by Rhodococcus sp. Strain ZWL3NT

Nitroarenes are synthetic molecules widely used in the chemical industry. Microbial degradation of nitroarenes has attracted extensive attention, not only because this class of xenobiotic compounds is recalcitrant in the environment but also because the microbiologists working in this field are curious about the evolutionary origin and process of the nitroarene dioxygenases catalyzing the initial reaction in the catabolism. In contrast to previously reported nitroarene dioxygenases from Gram-negative strains, which originated from a Nag-like naphthalene dioxygenase, the 3-nitrotoluene (3NT) dioxygenase in this study is from a Gram-positive strain and is an example of a Bph-like nitroarene dioxygenase. The preference of hydroxylation of this enzyme at the 2,3 positions of the benzene ring to produce 3-methylcatechol exclusively from 3NT is also a unique property among the studied nitroarene dioxygenases. These findings will enrich our understanding of the diversity and origin of nitroarene dioxygenase in microorganisms.

ABSTRACT

All nitroarene dioxygenases reported so far originated from Nag-like naphthalene dioxygenase of Gram-negative strains, belonging to group III of aromatic ring-hydroxylating oxygenases (RHOs). Gram-positive Rhodococcus sp. strain ZWL3NT utilizes 3-nitrotoluene (3NT) as the sole source of carbon, nitrogen, and energy for growth. It was also reported that 3NT degradation was constitutive and the intermediate was 3-methylcatechol. In this study, a gene cluster (bndA1A2A3A4) encoding a multicomponent dioxygenase, belonging to group IV of RHOs, was identified. Recombinant Rhodococcus imtechensis RKJ300 carrying bndA1A2A3A4 exhibited 3NT dioxygenase activity, converting 3NT into 3-methylcatechol exclusively, with nitrite release. The identity of the product 3-methylcatechol was confirmed using liquid chromatography-mass spectrometry. A time course of biotransformation showed that the 3NT consumption was almost equal to the 3-methylcatechol accumulation, indicating a stoichiometry conversion of 3NT to 3-methylcatechol. Unlike reported Nag-like dioxygenases transforming 3NT into 4-methylcatechol or both 4-methylcatechol and 3-methylcatechol, this Bph-like dioxygenase (dioxygenases homologous to the biphenyl dioxygenase from Rhodococcus sp. strain RHA1) converts 3NT to 3-methylcatechol without forming 4-methylcatechol. Furthermore, whole-cell biotransformation of strain RKJ300 with bndA1A2A3A4 and strain ZWL3NT exhibited the extended and same substrate specificity against a number of nitrobenzene or substituted nitrobenzenes, suggesting that BndA1A2A3A4 is likely the native form of 3NT dioxygenase in strain ZWL3NT.

IMPORTANCE Nitroarenes are synthetic molecules widely used in the chemical industry. Microbial degradation of nitroarenes has attracted extensive attention, not only because this class of xenobiotic compounds is recalcitrant in the environment but also because the microbiologists working in this field are curious about the evolutionary origin and process of the nitroarene dioxygenases catalyzing the initial reaction in the catabolism. In contrast to previously reported nitroarene dioxygenases from Gram-negative strains, which originated from a Nag-like naphthalene dioxygenase, the 3-nitrotoluene (3NT) dioxygenase in this study is from a Gram-positive strain and is an example of a Bph-like nitroarene dioxygenase. The preference of hydroxylation of this enzyme at the 2,3 positions of the benzene ring to produce 3-methylcatechol exclusively from 3NT is also a unique property among the studied nitroarene dioxygenases. These findings will enrich our understanding of the diversity and origin of nitroarene dioxygenase in microorganisms.

INTRODUCTION

Nitroarenes are synthetic molecules widely used in the synthesis of drugs, dyes, and explosives. They are toxic, mutagenic, and carcinogenic to humans and animals (1–3). So far, a number of bacterial strains have been reported to be able to utilize nitroarenes as a sole carbon source for growth. Based on the different methods of nitro group removal, the metabolic pathways of these substrates can be generally discriminated into reduction and oxidation pathways. The oxidation pathway of nitroarenes is always initiated by dioxygenases (3). For 2-nitrotoluene (2NT) catabolism, 2NT dioxygenase (2NTDO) in Acidovorax sp. strain JS42 (4) converts 2NT to 3-methylcatechol, which is degraded via the meta-cleavage pathway (5). In the nitrobenzene utilizer Comamonas sp. strain JS765, nitrobenzene is converted to catechol by nitrobenzene dioxygenase (6). In 2,4-dinitrotoluene (DNT) utilizers Burkholderia cepacia R34 (7) and Burkholderia sp. strain DNT (8), DNT is also oxidized by dioxygenases. Pseudomonas stutzeri ZWLR2-1 has a 2-chloronitrobenzene (2CNB) metabolic pathway initiated by a dioxygenase via the 3-chlorocatechol pathway (9). Considering the high identity (more than 80%) of all aforementioned dioxygenases to the naphthalene dioxygenase NagAcAd from Ralstonia sp. strain U2 (10), the naphthalene dioxygenase is thought to be the progenitor of these nitroarene dioxygenases (3). These studied nitroarene dioxygenases from Gram-negative strains were collectively called Nag-like or Dnt/Ntd-like dioxygenases (11). In addition, laboratory-evolved mutants of 2NT utilizer Acidovorax sp. JS42 gained the ability to catabolize 3-nitrotoluene (3NT) (12) or 4-nitrotoluene (4NT) via 4-methylcatechol (13). The reported nitroarene dioxygenases always belong to group III ring-hydroxylating dioxygenase, according to the classification system for these enzymes (11).

Xenobiotic 3-nitrotoluene (3NT) is a by-product in the process of industrial synthesis of 2-nitrotoluene or 4-nitrotoluene (14). Nitrobenzene utilizer Comamonas sp. strain JS765 can grow with 3-nitrotoluene supplied as the sole carbon source. The Nag-like nitrobenzene dioxygenase (NBDO) catalyzes the conversion of 3NT to 4-methylcatechol with nitrite release (6). In contrast, the Nag-like 3-nitrotoluene dioxygenase in Diaphorobacter sp. strain DS2 converts 3NT into both 3-methylcatechol and 4-methylcatechol, which are both easily degraded by strain DS2 (15, 16). To date, several nitroarene dioxygenases have been found to be able to oxidize 3NT, and their dioxygenation products always appear to include 4-methylcatechol (9, 15, 17, 18).

In Gram-positive strains, there is a group of important aromatic ring-hydroxylating dioxygenases belonging to group IV (11), which can initiate the oxidation pathway of certain aromatic compounds. Rhodococcus sp. strain I24 was studied for converting indene into indandiol which is useful as a pharmaceutical synthesis intermediate, where a toluene-induced dioxygenase (TID) was involved (19). In strain Rhodococcus sp. strain RHA1, biphenyl was oxidized into cis-(2R,3S)-dihydroxy-1-phenylcyclohexa-4,6-diene (dihydrodiol) by a biphenyl 2,3-dioxygenase (20, 21), which also showed activity to polychlorinated biphenyl (22). In Rhodococcus erythropolis BD2, isopropylbenzene was oxidized by a 2,3-dioxygenase to 2,3-dihydro-2,3-dihydroxyisopropylbenzene (23, 24). In Rhodococcus sp. strain 065240, benzotrifluoride (BTF) was converted into cis-2,3-dihydro-2,3-dihydroxybenzotrifluoride by a BTF 2,3-dioxygenase (25, 26). The aforementioned group IV aromatic ring-hydroxylating dioxygenases from Rhodococcus sp. share high identities (over 85%) and are collectively named Bph-like dioxygenases in this study for an easy comparison with Nag-like enzymes. Unlike the Nag-like dioxygenases, no Bph dioxygenases have so far been found from any nitroarene utilizers. Although Gram-positive Micrococcus sp. strain SMN-1 was reported to degrade 2-nitrotoluene (27), the enzyme(s) involved is yet to be identified.

Rhodococcus sp. strain ZWL3NT was isolated by its ability to utilize 3NT as the sole source of carbon, nitrogen, and energy for growth. This strain also grows on 2-nitrotoluene, 3-methylcatechol, and catechol (14). Two metabolites in the reaction mixture, namely, 3-methylcatechol and 2-methyl-cis,cis-muconate, were detected after incubation of strain ZWL3NT with 3NT. Both catechol 1,2-dioxygenase and catechol 2,3-dioxygenase were detected in this strain. In addition, the cell extracts of Escherichia coli expressing CatA and CatB from the catRABC cluster exhibited catechol 1,2-dioxygenase activity and cis,cis-muconate cycloisomerase activity, respectively. This evidence suggests the presence of a novel pathway for 3NT degradation with 3-methylcatechol as the ring-cleavage metabolite (at least dominated) by strain ZWL3NT (14). However, the enzyme catalyzing the conversion of 3NT to 3-methylcatechol, which is the initial reaction for the 3NT catabolism, has not been described in this Gram-positive strain at molecular and genetic levels. In this study, we report the characterization of a Bph-like, rather than Nag-like, nitroarene dioxygenase converting 3NT exclusively into 3-methylcatechol from Rhodococcus sp. strain ZWL3NT. This research will enrich our understanding of the diversity and origin of nitroarene dioxygenase in microorganisms.

RESULTS

Cloning of bnd and nid gene clusters from Rhodococcus sp. strain ZWL3NT.

To clone 3NT dioxygenase genes associated with nitro oxidation in strain ZWL3NT, PCR was done with the primer pair bndF/bndR designed according to the conserved sequence of isopropylbenzene dioxygenase genes from Rhodococcus erythropolis BD2 (23) and biphenyl dioxygenase genes from Rhodococcus sp. strain RHA1 (28), using genomic DNA of strain ZWL3NT as PCR templates. With the genome walking operation method, a 700-bp upstream fragment and a 5-kb downstream fragment were obtained as a bnd cluster as shown in Fig. 1. Genes bndA1A2A3A4 in this cluster were very similar to their counterparts encoding biphenyl dioxygenase from Rhodococcus sp. strain RHA1 (89% to 97% protein identities) (see Table S1 in the supplemental material), and here, their products are referred to as a Bph-like dioxygenase (group IV), in contrast to the aforementioned Nag-like nitroarene dioxygenases (group III). In addition, a possible 2,3-dihydroxybiphenyl 1,2-dioxygenase gene, bndC, was also found to be similar to BphC1 (99% identities) from strain RHA1.

FIG 1.

The proposed pathway for 3NT degradation by strain ZWL3NT, and the gene cluster that may be involved in the pathway. (A) Proposed catabolic pathway of 3NT degradation in strain ZWL3NT. In a previous study, 3-methylcatechol was identified as the intermediate of 3NT degradation; CatC and BndC were proved to be catechol 1,2-dioxygenase and catechol 2,3-dioxygenase, respectively (14). BndA1A2A3A4 was identified as a 3NT dioxygenase in this study. (B) Gene clusters involved in the 3NT catabolic pathway.

Similarly, an nid gene cluster from strain ZWL3NT was obtained (Fig. 2) by genome walking according to the conserved sequence of naphthalene dioxygenase genes from Rhodococcus sp. strain NCIMB12038. The obtained NidABCD in strain ZWL3NT exhibited 97% to 99% identities to their homologs for naphthalene degradation in Rhodococcus sp. strain P200 or Rhodococcus sp. strain P400 (29). This putative hydroxylating dioxygenase belongs to group Gram⁺ PAH (polycyclic aromatic hydrocarbons)/phthalate according to Parales and Resnick (11).

FIG 2.

Transcription analysis of bnd and nid gene clusters in LB-grown strain ZWL3NT. Lane 1, products amplified by bndARTF/bndARTR primer set (with cDNA template from LB-grown strain ZWL3NT). The expected product is 158 bp in size. Lane 3, no product was amplified by nidARTF/nidARTR primer set. The expected product is 145 bp in size. Lane 2 shows a molecular size marker.

bnd cluster is transcribed in LB-grown strain ZWL3NT.

As mentioned above, two possible dioxygenase gene clusters, namely, bnd and nid, were present in strain ZWL3NT. Considering that the 3NT dioxygenase in strain ZWL3NT is constitutively expressed (14), reverse transcription-PCR (RT-PCR) analysis was then done to determine if both gene clusters or one of them encodes the constitutively expressed 3NT dioxygenase. This was determined using the RNA extracted from the LB-grown strain ZWL3NT as the template. The length of the expected PCR product would be 158 bp as a partial sequence of bndA1 or 145 bp as a partial sequence of nidA. The RT-PCR amplification showed that only a 158-bp fragment from the bnd cluster was generated (Fig. 2), indicating that the bnd cluster rather than the nid cluster was constitutively transcribed. This suggested that the bnd cluster encodes the active 3NT dioxygenase in strain ZWL3NT.

BndA1A2A3A4 is a 3NT dioxygenase converting 3NT into 3-methylcatechol.

The bnd gene cluster was initially cloned into the expression vector pET28a to produce pZWLC (pET28a-bndA1A2A3A4), which was transferred into E. coli Rosetta. The isopropyl-β-d-thiogalactopyranoside (IPTG)-induced E. coli Rosetta (pZWLC) was found to be able to release trace amounts of nitrite ion when 3NT was a substrate in a whole-cell biotransformation assay, while its negative control E. coli Rosetta (pET28a) was not. These results indicate that BndA1A2A3A4 was active on 3NT, but its activity in this heterogeneous host is too low to carry out further assays.

To get adequate dioxygenase activity, the bnd cluster together with its putative promoter was cloned into a shuttle vector, pRESQ (30), to produce pRESQ-bnd, which was transferred into a heterogeneous host, Rhodococcus imtechensis strain RKJ300 (31). Strain RKJ300 (pRESQ-bnd) had 3NT dioxygenase activity while the negative control strain RKJ300 (pRESQ) did not.

To determine the products from 3NT oxidization by Rhodococcus imtechensis strain RKJ300 (pRESQ-bnd), liquid chromatography-mass spectrometry (LC-MS) was used to analyze the supernatants of the biotransformation reaction. The 3NT dioxygenation product would be catechol derivatives, but strain RKJ300 exhibits catechol 1,2-dioxygenase activity (see Fig. S1 in the supplemental material), which will hamper us from detecting the formed products from 3NT dioxygenation. It was reported that substrate analogues were inhibitors for catechol 1,2-dioxygenase (32). It is also true that catechol 1,2-dioxygenase in strain RKJ300 can be effectively suppressed by 3-chlorocatechol, as shown in Fig. S1. Therefore, excessive 3-chlorocatechol was added into the biotransformation mixtures for the 3NT biotransformation analyses in the presence of cells of RKJ300 carrying bndA1A2A3A4. LC-MS analysis indicated that the product formed had the retention time of 20.98 min and an m/z of 123.04 (Fig. 3), which are the same values as those of authentic 3-methylcatechol. The product of 3NT oxidation reaction catalyzed by BndA1A2A3A4 was 3-methylcatechol. In a time course of biotransformation assay which was quantified by high-performance liquid chromatography (HPLC), the substrate 3NT gradually decreased and the product 3-methylcatechol as well as nitrate gradually increased (Fig. 4). The consumption of 3NT was almost equal to the accumulation of 3-methylcatechol after a 120-min incubation, indicating a stoichiometry conversion of 3NT to 3-methylcatechol. The biotransformation activity of Rhodococcus imtechensis strain RKJ300 (pRESQ-bnd) is 0.27 ± 0.02 (nmol mg⁻¹ min⁻¹). Based on these data, bndA1A2A3A4 can be identified to encode a 3NT dioxygenase catalyzing the conversion of 3NT into 3-methylcatechol with nitrite release.

FIG 3.

LC-MS analysis of biotransformation products of strain RKJ300 carrying bnd gene cluster. (A) The retention time of authentic 3-methylcatechol. (B) Negative control: LC results of 3NT biotransformation system with Rhodococcus imtechensis strain RKJ300 (pRESQ). (C) LC results of 3NT biotransformation system with Rhodococcus imtechensis strain RKJ300 (pRESQ-bnd). The dotted line shows the retention time of 3-methylcatechol. (D) The mass spectrum of the product in panel C at 20.98 min, which is the same as authentic 3-methylcatechol.

FIG 4.

Time course of 3NT transformation by Rhodococcus imtechensis RKJ300 (pRESQ-bnd). The cells were grown on LB liquid medium containing kanamycin and shaken under the conditions of 30°C and 200 rpm for 36 hours. Then, the cells were suspended to an OD₆₀₀ of 15 with 50 mM PB (pH 7.4). The experiments were performed in triplicate, the results show the average value of three independent experiments, and error bars show standard deviations.

Substrate specificity of 3NT dioxygenase.

Like other nitroarene dioxygenases that always have extended substrate specificity (33, 34), BndA1A2A3A4 also exhibited extended substrate specificity to nitrobenzene and other halogenated or methylated nitroarenes (Table 1). Furthermore, whole-cell biotransformation of Rhodococcus imtechensis strain RKJ300 (pRESQ-bnd) and Rhodococcus sp. strain ZWL3NT exhibited similar relative activities of dioxygenase against nitrobenzene or substituted nitrobenzenes, as shown in Table 1. Both Rhodococcus imtechensis strain RKJ300 (pRESQ-bnd) and strain ZWL3NT showed activity to 3-nitrotoluene, 2-nitrotoluene, nitrobenzene, 2-chloronitrobenzene, and 3-chloronitrobenzene, and they were inactive to 4-nitrotoluene and 4-chloronitrobenzene. It is worth noting that 3-nitrotoluene and 3-chloronitrobenzene were attacked at the same position on the benzene ring to produce 3-methylcatechol and 3-chlorocatechol, respectively. These data suggested that BndA1A2A3A4 is likely the native form of 3NT dioxygenase in strain ZWL3NT.

TABLE 1.

Substrate specificity of 3NT dioxygenase^a

^aFrom whole-cell biotransformation system of Rhodococcus imtechensis RKJ300 (pRESQ-bnd) and strain ZWL3NT. Both exhibited the same substrate specificity.

^bThe wavelength of maximum absorbance of corresponding substrate determined by DAD detector of the Dionex UltiMate 3000 RS HPLC system and PE Lambda 25 UV/VIS spectrophotometer.

^cND, no product was detected.

Considering the high similarity between this 3NT dioxygenase and biphenyl dioxygenase, their activities on biphenyl and 4-chlorobiphenyl (4-CB) were tested using a previously reported method (21) but were not detected, based on the following observations. Neither cis-(2R,3S)-dihydroxy-1-phenylcyclohexa-4,6-diene (dihydrodiol) nor 2,3-dihydroxybiphenyl (2,3-DHBP) and their chlorinated derivatives, possible products from the dioxygenation of biphenyl and 4-chlorobiphenyl, were detected by gas chromatography-mass spectrometry (GC-MS). On the other hand, no obvious decreasing trend of biphenyl or 4-chlorobiphenyl treated with strain RKJ300 (pRESQ-bnd) was observed.

The three-dimensional (3D) structure model of BndA1.

In a reaction catalyzed by biphenyl dioxygenase, C-2 and C-3 of biphenyl are oxidized to produce (+)-biphenyl cis-(2R,3S)-dihydrodiol (35). With biphenyl or 3NT as a substrate, the Nag-like naphthalene dioxygenase targets C-3 and C-4, producing (+)-biphenyl-3,4-dihydrodiol or 4-methylcatechol (36).

Considering BndA1 from Rhodococcus sp. strain RKJ300 has a 95% amino acid sequence identity with biphenyl dioxygenase BphA1 from Rhodococcus sp. strain RHA1, the binding of biphenyl dioxygenase with biphenyl is compared with that of BndA1 with 3NT. As shown in Fig. 5, their active site residues are considerably similar and their substrates adopted similar orientations. The substrate-binding pocket of BndA1 is lined by hydrophobic residues. The substrate 3NT is located beneath the mononuclear iron with its nitro group orienting His313. The aromatic ring of 3NT forms a π-π interaction with Phe218, Phe368, and Phe374 and is sandwiched by His224 and Leu323. The nitro group forms hydrogen bonds with the main chain oxygen atom of Gln217 and the imidazole-nitrogen atom of His313. This suggested that the hydroxylation of 3NT catalyzed by BndA1 also occurs at C-2 and C-3, like biphenyl dioxygenase BphA1, producing 3-methylcatechol.

FIG 5.

The active sites of the BndA1·3NT model (A) and BphA1·biphenyl complex (B). The 3NT and biphenyl molecules are shown in yellow. The mononuclear iron center (large orange sphere) is coordinated with Asp378, His224, and His230 (shown in cyan). Gln217 and His313 are shown in green. This model explains why 3NT was hydroxylated at C-2 and C-3, like biphenyl dioxygenase BphA1, producing 3-methylcatechol.

DISCUSSION

There has been a long-standing mystery of why nitroarene dioxygenases, including nitrotoluene dioxygenases, in general seem to have evolved from Nag naphthalene dioxygenase when the structures of the nitro compounds seem more similar to toluene. The findings presented here indicate that there is nothing inherent in the reaction mechanism that requires naphthalene dioxygenase as the progenitor of nitroarene dioxygenases. In contrast to the reported nitroarene dioxygenases originating from Nag-like naphthalene dioxygenase of Gram-negative strains, the 3NT dioxygenase in this study is the only example of Bph-like nitroarene dioxygenases from biphenyl dioxygenase of Gram-positive strains. The preference of hydroxylation of this enzyme at the 2,3 positions of the benzene ring to produce 3-methylcatechol exclusively from 3NT is also a unique property among the studied nitroarene dioxygenases.

Aromatic ring-hydroxylating oxygenases are classified into six categories according to the sequence of the large subunit. They are phthalate (α_n) (group I), benzoate (group II), naphthalene (group III), toluene/biphenyl (group IV), Gram⁺ PAH/phthalate group, and salicylate group (11). As far as we know, all studied dioxygenases that can release nitrite from nitroarene come from group III, and they are from Gram-negative strains. Although Gram-positive Micrococcus sp. strain SMN-1 (27) and Rhodococcus pyridinivorans NT2 (37) were reported to degrade 2-nitrotoluene and 4-nitrotoluene, respectively, the enzymes and genes involved have yet to be identified. To study the evolutionary origin of the 3NT dioxygenase of strain ZWL3NT, phylogenetic analysis of the oxygenase large subunit (BndA1) of 3NTDO was performed using the neighbor-joining method with MEGA5. BndA1 was selected to construct a phylogenetic tree because the oxygenase large subunit always plays a key role in the catalytic reaction and nitroarene substrate selection (17). The phylogenetic tree in Fig. 6 shows that BndA1 laid in the branch of isopropylbenzene and biphenyl dioxygenases from Rhodococcus sp., belonging to group IV aromatic ring-hydroxylating dioxygenases (11), and is a Bph-like nitroarene dioxygenase. However, other classic nitroarene dioxygenases, such as NBDO from Comamonas sp. strain JS765 (6), 2NT dioxygenase from Acidovorax sp. JS42 (34), 2,4-dinitrotoluene dioxygenase from Burkholderia sp. strain DNT and Burkholderia cepacia R34 (7), 2-chloronitrobenzene dioxygenase from Pseudomonas stutzeri ZWLR2 (9), and 3NT dioxygenase from Diaphorobacter sp. DS2 (16), all belong to group III aromatic ring-hydroxylating dioxygenases. The naphthalene dioxygenase Nag system is the sole ancestor for all of these Nag-like nitroarene dioxygenases. The above comparison clearly indicates that the Bph-like 3NT dioxygenase from Rhodococcus sp. strain RKJ300 is a unique one among the identified nitroarene dioxygenases.

FIG 6.

The differences between group III and group IV aromatic ring-hydroxylating dioxygenases. (A) Physical map of the mnt gene cluster from 3NT-utilizer Diaphorobacter sp. strain DS2 and nag gene cluster from Ralstonia sp. strain U2 (top panel); bnd gene cluster in 3NT-utilizer Rhodococcus sp. strain ZWL3NT (this study) and bph gene cluster from Rhodococcus jostii RHA1 (bottom panel). (B) Neighbor-joining phylogenetic tree based on amino acid sequences of BndA1 (oxygenase large subunit of 3NTDO from Rhodococcus sp. strain ZWL3NT) with homologous sequences of other strains. The name, accession number, and source organism of these proteins are as follows: CnbAc (ADQ90222) 2-chloronitrobenzene dioxygenase from Pseudomonas stutzeri ZWLR2, MntAc (AGH09226) 3-NT dioxygenase from Diaphorobacter sp. DS2, NtdAc (AAB40383) 2-NT dioxygenase from Acidovorax sp. JS42, NbzAc (AAL76202) NBDO from Comamonas sp. strain JS765, DntAc (AAB09766) 2,4-dintrotoluene dioxygenase from Burkholderia sp. strain DNT, DntAc (AAL50021) 2,4-dintrotoluene dioxygenase from Burkholderia cepacia R34, PahAc (AAF72976) polyaromatic hydrocarbon dioxygenase from Comamonas testosteroni, NagAc (AAD12610) naphthalene dioxygenase from Ralstonia sp. U2, NagAc (AAZ93388) naphthalene dioxygenase from Polaromonas naphthalenivorans CJ2, PahAc (ACT53249) polyaromatic hydrocarbon dioxygenase from Burkholderia sp. C3, NagAc (AEV91670) naphthalene dioxygenase from Comamonas sp. MQ, BndA1 (AFW98858) 3-NT dioxygenase from Rhodococcus sp. strain ZWL3NT, IpbA1 (AAB08025) isopropylbenzene 2,3-dioxygenase from Rhodococcus erythropolis BD2, IpbA1 (AAB08025) isopropylbenzene dioxygenase from Rhodococcus wratislaviensis strain IFP2016, BphA1 (BAA06868) biphenyl dioxygenase from Rhodococcus sp. RHA1, TidA (AAL61663) toluene inducible dioxygenase from Rhodococcus aetherivorans strain I24, BnzA1 (BAD95523) benzene dioxygenase from Rhodococcus opacus B4, and BtfA1 (BAQ00536) benzotrifluoride 2,3-dioxygenase from Rhodococcus sp. 065240. The chemical names in red are the nitroarene substrates for their corresponding dioxygenases listed before the strain names; the chemical names in blue are the nonnitroarene substrates for their corresponding dioxygenases.

The nitroarene dioxygenases are diverse in regiospecificity. To remove the nitro group from the nitrotoluene by dioxygenases, the nitro-substituted benzene ring should be oxidatively attacked. Normally, a dioxygenase oxidizes an aromatic compound to a dihydrodiol (10, 21). Due to the presence of a nitro group, the reaction removes the nitro group as nitrite produces a catechol instead of a dihydrodiol (6, 9). When 3-nitrotoluene is attacked at the 2,3 positions by the Bph-like nitroarene dioxygenase in this study, the product should be 3-methylcatechol, which is supported by the three-dimensional structure model of BndA1. For 2-nitrotoluene, the attack can only occur at 2,3 positions to form 3-methylcatechol. For 4-nitrotoluene, the attack can only occur at 3,4 positions to form 4-methylcatechol. Theoretically, the oxidative attack for 3-nitrotoluene may occur at both 2,3 positions and 3,4 positions. According to a previous study, several nitroarene dioxygenases from Gram-negative strains can oxidize 3NT (33). For example, products formed by 3NTDO from Diaphorobacter sp. strain DS2 include 3-methylcatechol, 4-methylcatechol, and 3-nitrobenzyl alcohol (16). The 2NTDO from Acidovorax sp. strain JS42 can transform 3-nitrotoluene into 3-methylcatechol and 4-methylcatechol at the same time (17). NBDO from Comamonas sp. strain JS765 appears to favor hydroxylation at the 3,4 positions and produces 4-methylcatechol as the sole product when oxidizing 3NT (6). Both 3-methylcatechol and 4-methylcatechol from 3NT transformation were also detected with the 2-chloronitrobenzene (2CNB) dioxygenase from Pseudomonas stutzeri ZWLR2-1 (9). However, the 3NT dioxygenase in this study from Rhodococcus sp. strain ZWL3NT appears to favor hydroxylation at the 2,3 positions and produces only 3-methylcatechol and 3-chlorocatechol when transforming 3NT and 3CNB, respectively. The relationship between the spatial structures and the regiospecificities for these nitroarene dioxygenases should be elucidated in a future study.

MATERIALS AND METHODS

Bacterial strains, plasmids, materials, chemicals, and culture conditions.

The bacterial strains and plasmids used in this study are listed in Table 2. Rhodococcus sp. strain ZWL3NT and Rhodococcus imtechensis strain RKJ300 were grown at 30°C in lysogeny broth (LB) medium or minimum medium (MM) (38), supplemented with appropriate substrates. Escherichia coli strains were grown in LB at 37°C. Kanamycin (50 mg/liter) was added into the medium if necessary. Chemicals 2-chloronitrobenzene, 3-chloronitrobenzene, 4-chloronitrobenzene, nitrobenzene, 2,4-dinitrotoluene, 2-nitrobenzyl alcohol, 3-nitrobenzyl alcohol, 2-nitrotoluene, 3-nitrotoluene, and 4-nitrotoluene were purchased from Sinopharm Chemical Reagent Co. Ltd. (SCRC; China) with purities of more than 99%. Catechol, 3-methylcatechol, and 4-chlorocatechol were purchased from Sigma (St. Louis, MO, USA); 4-methylcatechol was from Fluka Chemical (Buchs, Switzerland). 3-Chlorocatechol was from TCI (Tokyo). For solid media, agar was added to a final concentration at 12 g/liter.

TABLE 2.

Bacterial strains and plasmids in this study

Strain or plasmid	Relevant genotype or characteristic(s)	Reference or source
Strains
Rhodococcus sp. strain ZWL3NT	3-Nitrotoluene utilizer	45
Escherichia coli DH5α	λ– endA1 glnV44 thi-1 recA1 relA1 gyrA96 deoR nupG Φ80dlacZΔM15 Δ(lacZYA-argF)U169 hsdR17(rK− mK−)	Novagen
Rhodococcus imtechensis strain RKJ300	4-Nitrophenol degrader	46
Plasmids
pRESQ	Broad-host-range vector, Kanr	30
pRESQ-bnd	bnd fragment inserted into pRESQ	This study
pZWLC	bnd fragment insert into pET28a	This study

Plasmids from E. coli were purified using the EZNA plasmid mini kit I (Omega Bio-Tek Inc.). Purification of PCR products and DNA fragments from agarose gel was done using the Omega cycle-pure kit and gel extraction kit (Norcross, GA). The genomic DNA purification kit was also purchased from Omega, and the RNA preparation kit was purchased from Tiangen (Beijing, China) and used according to the manufacturer’s specifications. Other molecular biology techniques were based on standard procedures, as previously described (39). Restriction enzymes were purchased from TaKaRa (Dalian, China). DNA polymerases were purchased from TransGen Biotech (Beijing, China). DNA fragments were ligated by the NOVA hi-fusion cloning mix purchased from Yugong BioLabs (Lianyungang, China) or T4 DNA ligase purchased from New England BioLabs (Ipswich, MA). Plasmid DNA was transferred into E. coli by chemical transformation and was transferred into Rhodococcus sp. strains by electroporation. The cloned fragments were verified by DNA sequencing by Tsingke (Wuhan, China).

Gene detection and genome walking.

To search for the genes related to the oxidative removal of the 3NT nitro group from strain ZWL3NT, the primer pair bndF/bndR was designed according to the conserved regions of the α subunit of cumene dioxygenase in strain Rhodococcus erythropolis BD2 (23) and biphenyl dioxygenase in Rhodococcus jostii RHA1 (28). Also, the primer pair nidF/nidR was designed according to the sequence of the alpha subunit of naphthalene dioxygenase from Rhodococcus sp. strain NCIMB12038 (40). The sequences of these primer pairs are listed in Table 3. The genomic DNA of strain ZWL3NT was the PCR template to amplify the potential dioxygenase genes.

TABLE 3.

Primers used in this study

Primer name	Primer sequencea	Primer function
bndF	5′-CCCATCTTTCCGGCATTCTC-3′	bnd gene cluster detecting
bndR	5′-GCTGGCGGTATTCCTCTTTCA-3′	bnd gene cluster detecting
nidF	5′-GGATCTTTGGTCATGCGTGGGT-3′	nid gene cluster detecting
nidR	5′-AGAACTTCAGGTCGCCGAGGTAG-3′	nid gene cluster detecting
bndH	5′-TCCCAACCTGTTCGTTCCTGCCA-3′	bnd gene cluster walking
bndT	5′-GTTGGGATCGCCGATGTAGAAGC-3′	bnd gene cluster walking
nidH	5′-CACCTACAAGGGCCTGATCTTCG -3’	nid gene cluster walking
nidT	5′-GAGATGTACCGCACGACGTAATC-3′	nid gene cluster walking
bndARTF	5′-GCAGGACGACAGCGAGAACTATG-3′	bndA1 gene RT-PCR
bndARTR	5′-CCGCTTCTTCGCTGTAGACGTAG-3′	bndA1 gene RT-PCR
nidARTF	5′-CCAAGACAGGGGAGCTCAACTAC-3’	nidA gene RT-PCR
nidARTR	5′-AGCATGAGCTGCATCCAGTATTC-3′	nidA gene RT-PCR
bndAF	5′-GAACCATGGATGACTGACGTGCAATGTGAACCC-3′	bndA gene expression
bndAR	5′-GGCGAATTCTCATGCCGCTGTTGTTCTCTCGCG-3′	bndA gene expression
bndHindIIIs	5′-GCTATGCATCAAGCTTCGCCGCGGCAGTGTCCTCGCAG-3′	bndA gene expression
bndSpeIa	5′-GGCGGCCGTTACTAGTTCATGCCGCTGTTGTTCTCTCGC-3′	bndA gene expression

^aUnderlining indicates the enzyme cutting sites in the primers.

To explore the unknown sequence adjacent to the conserved regions, genome walking was done according to the method of Siebert et al. (41). The following operations were done in this process: (i) genomic DNA extraction and purification of Rhodococcus sp. strain ZWL3NT; (ii) digestion and purification of genomic DNA by four restriction endonucleases (EcoRV, SmaI, StuI, and PvuII) which produced blunt ends; (iii) construction of genome walking library by attaching the linkers to four kinds of digested genomic DNA, respectively; (iv) PCR using this diluted genome walking library as the template DNA (for each PCR, one of the primers was designed according to the sequence of the linker and the other walking primers were specific primers, i.e., bndH, bndT, nidH, and nidT [Table 3], respectively); (v) recovery of appropriate specific PCR products by agarose electrophoresis and then sequencing.

Transcription analysis of bndA1 and nidA genes.

To demonstrate the possibility of bnd and nid gene clusters encoding the 3NT dioxygenase, a transcription analysis of the bndA1 and nidA genes was done as follows. (i) Strain ZWL3NT was shaken in LB medium at 30°C with a 1% incubation volume. (ii) Total RNA was extracted by using an RNAprep pure kit for bacteria (Tiangen Biotech, Beijing, China). (iii) To avoid the influence of genomic DNA, DNase I was used to digest the total RNA at 37°C for 30 min, and then the termination buffer was added and the mixture incubated at 65°C for 10 min to inactivate DNase I. (iv) The downstream primers bndARTR and nidARTR, respectively, were added. The template/primer mixture was incubated at 70°C for 5 min, and then the mixtures were set on an ice bath immediately; the secondary structure of the RNA templates was melted and the reverse transcription primers annealed appropriately. (v) The reverse transcription reaction mixture was as follows: 10 μl RNA, 5 μl murine leukemia virus (MLV) 5× buffer, 2 μl deoxynucleoside triphosphate (dNTP), 0.5 μl RNase inhibitor, 1 μl MLV reverse transcriptase (MLV-RT), and 6.5 μl RNase-free water. The mixtures were incubated at 42°C for 60 min. With the exception of lacking MLV-RT for negative controls, the reaction mixtures were the same. Inactivation of the reaction was achieved by heating at 70°C for 15 min. (vi) PCR mixtures were 5 μl 10× rTaq buffer, 1.5 μl dNTP, 1 μl upstream primers, 1 μl downstream primers (Table 3), 3 μl cDNA (abovementioned reverse transcription mixtures), 0.5 μl rTaq, and 38 μl water. PCR conditions were 94°C for 4 min; 30 cycles of 94°C for 45 s, 58°C for 45 s, and 72°C for 30 s; and 72°C for 7 min.

Plasmid construction and DNA manipulation.

To determine the potential role of the bnd gene cluster in the 3NT metabolic pathway, the primer pair bndAF/bndAR (see Table 3) was used to amplify the bndA1A2A3A4 gene cluster and then it was purified by the E.Z.N.A. gel extraction kit. Both the purified PCR product and plasmid pET28a were digested by NcoI and EcoRI and then were linked by T4 DNA ligase to produce plasmid pZWLC (pET28a-bndA1A2A3A4) (Table 3). This construct and its negative control pET28a were introduced into E. coli Rosetta.

To express the bnd cluster in Rhodococcus imtechensis strain RKJ300, primer pair bndHindIIIs/bndSpeIa was used to amplify bndA1A2A3A4 and its putative promoter. The purified PCR product and the HindIII- and SpeI-digested plasmid pRESQ (30) were linked by the NovoRec PCR one-step directional seamless cloning kit (Novoprotein, Shanghai, China) to construct plasmid pRESQ-bnd. All cloning procedures were verified by DNA sequencing (Tsingke, Wuhan, China). The constructs were introduced into Rhodococcus imtechensis strain RKJ300 by electroporation at 1,800 V with Eppendorf electroporator 2510.

Biotransformation of 3NT.

Rhodococcus imtechensis RKJ300 is a well-studied para-nitrophenol degrader. The successful gene deletion and complementation in this strain led us to use it as an expression host in this study (31).

Rhodococcus imtechensis RKJ300 (pRESQ-bnd) and its negative control strain RKJ300 (pRESQ) were inoculated into LB liquid medium containing kanamycin and shaken under the conditions of 30°C and 200 rpm for 36 h. Then, the cells were harvested and washed twice with 50 mM phosphate buffer (PB) (pH 7.4).

The harvested cells were resuspended in 50 mM PB (pH 7.4) and diluted to an optical density at 600 nm (OD₆₀₀) of 15. Also, 0.2 mM substrate 3NT was added. Biotransformation was performed under the conditions of 30°C and 200 rpm. The samples were harvested at appropriate intervals and centrifuged to remove the cells. The supernatants of the biotransformation reaction mixture were recovered for dioxygenase activity analysis by measuring the concentration of nitrite released from the substrate. The concentration of nitrite was assayed as described previously (42), using sodium nitrite as a standard. The 3NT biotransformation activity was defined as the amount of nitrite produced (nmol) by 1 mg of cells in 1 minute.

The presence of catechol 1,2 dioxygenase activity in strain RKJ300 prevented us from detecting the catechols formed from 3NT dioxygenation, as shown in Fig. S1. Considering the catechol 1,2 dioxygenase can be inhibited by substrate analogues (32), 5 mM 3-chlorocatechol was added into the biotransformation system to inhibit the catechol 1,2 dioxygenase activity derived from strain RKJ300.

Analytical methods.

The biotransformation supernatants after 1-h incubation were centrifuged twice at 16,000 × g for 10 min to remove the cell debris and then were analyzed by HPLC and HPLC-MS. HPLC analysis of nitroaromatic substrates and their oxidation products was performed with a Dionex UltiMate 3000 RS (rapid separation) HPLC system equipped with a diode array detector (DAD) and a Welch Ultimate LP-C₁₈ column (4.6 × 250 mm, 5-μm particle size) maintained at 30°C. The mobile phase consisted of solvents A (0.1% [vol/vol] acetic acid in water) and B (methanol). A stepped solvent gradient was used as follows: 0 to 7 min, 5% to 15% B; 7 to 15 min, 15% to 37% B; 15 to 45 min, 37% to 80% B; 45 to 48 min, 80% B; 48 to 48.5 min, 80% to 5% B; and 48.5 to 55 min, 5% B. The flow rate was 1.0 ml/min. Under these conditions, retention times of nitroaromatic substrates and their oxidation products (listed in Table 1) are as follows: 3-nitrotoluene, 34.71 min; 3-methylcatechol, 21.01 min; 2-nitrotoluene, 33.40 min; 2-nitrobenzyl alcohol, 21.38 min; 4-nitrotoluene, 35.26 min; nitrobenzene, 27.44 min; catechol, 11.10 min; 2-chloronitrobenzene, 32.33 min; 3-chlorocatechol, 22.45 min; 3-chloronitrobenzene, 35.80 min; 4-chloronitrobenzene, 36.21 min.

LC-MS measurements of the oxidation product of 3NT were performed with a Dionex UltiMate3000 LCQ Fleet system (Thermo Fisher Scientific, MA). An Agilent Zorbax Eclipse XDB-C₁₈ column (250 by 4.6 mm, 5-μm particle size) was used with a column temperature of 30°C and a sample injection volume of 20 μl. The monitoring wavelength was 280 nm, and the DAD acquisition wavelength was 200 to 400 nm. The mobile phase consisted of 0.1% (vol/vol) acetic acid in water (A) and methanol (B). A stepped solvent gradient was used as follows: 0 to 30 min, 5% to 98% B; 30 to 40 min, 98% B; 40 to 41 min, 98% to 5% B; and 41 to 50 min, 5% B. The flow rate was 0.3 ml/min. The optimized mass spectrometric parameters were as follows: capillary temperature, 300°C; source heater temperature, 300°C; source voltage, 4 kV; source current, 100 μA; capillary voltage, 30 V; tube lens, 40 V; and negative ion mode. Spectra were recorded within m/z from 50 to 2,000.

Homology modeling.

The amino acid sequence was submitted to the Swiss-model server (43) for homology modeling. The terminal oxygenase component of biphenyl dioxygenase BphA1 from Rhodococcus sp. strain RHA1 (PDB number, 1uli), with a 95% sequence identity with BndA1, was used as the modeling template. The molecular docking process was performed with Autodock vina (44). Based on the above analysis, PyMOL v. 1.0 was used to view and analyze the binding of the enzyme and 3NT.

Data availability.

The DNA sequences of the 7.4-kb bnd (formerly ntd) gene cluster and 5.4-kb nid gene cluster have been deposited in GenBank under accession numbers JX625147 and MF401592, respectively. GenBank accession numbers have been given in the text for all mentioned proteins and genes.