Novel Three-Component Phenazine-1-Carboxylic Acid 1,2-Dioxygenase in Sphingomonas wittichii DP58

ABSTRACT

Phenazine-1-carboxylic acid, the main component of shenqinmycin, is widely used in southern China for the prevention of rice sheath blight. However, the fate of phenazine-1-carboxylic acid in soil remains uncertain. Sphingomonas wittichii DP58 can use phenazine-1-carboxylic acid as its sole carbon and nitrogen sources for growth. In this study, dioxygenase-encoding genes, pcaA1A2, were found using transcriptome analysis to be highly upregulated upon phenazine-1-carboxylic acid biodegradation. PcaA1 shares 68% amino acid sequence identity with the large oxygenase subunit of anthranilate 1,2-dioxygenase from Rhodococcus maanshanensis DSM 44675. The dioxygenase was coexpressed in Escherichia coli with its adjacent reductase-encoding gene, pcaA3, and ferredoxin-encoding gene, pcaA4, and showed phenazine-1-carboxylic acid consumption. The dioxygenase-, ferredoxin-, and reductase-encoding genes were expressed in Pseudomonas putida KT2440 or E. coli BL21, and the three recombinant proteins were purified. A phenazine-1-carboxylic acid conversion capability occurred in vitro only when all three components were present. However, P. putida KT2440 transformed with pcaA1A2 obtained phenazine-1-carboxylic acid degradation ability, suggesting that phenazine-1-carboxylic acid 1,2-dioxygenase has low specificities for its ferredoxin and reductase. This was verified by replacing PcaA3 with RedA2 in the in vitro enzyme assay. High-performance liquid chromatography–mass spectrometry (HPLC-MS) and nuclear magnetic resonance (NMR) analysis showed that phenazine-1-carboxylic acid was converted to 1,2-dihydroxyphenazine through decarboxylation and hydroxylation, indicating that PcaA1A2A3A4 constitutes the initial phenazine-1-carboxylic acid 1,2-dioxygenase. This study fills a gap in our understanding of the biodegradation of phenazine-1-carboxylic acid and illustrates a new dioxygenase for decarboxylation.

IMPORTANCE Phenazine-1-carboxylic acid is widely used in southern China as a key fungicide to prevent rice sheath blight. However, the degradation characteristics of phenazine-1-carboxylic acid and the environmental consequences of the long-term application are not clear. S. wittichii DP58 can use phenazine-1-carboxylic acid as its sole carbon and nitrogen sources. In this study, a three-component dioxygenase, PcaA1A2A3A4, was determined to be the initial dioxygenase for phenazine-1-carboxylic acid degradation in S. wittichii DP58. Phenazine-1-carboxylic acid was converted to 1,2-dihydroxyphenazine through decarboxylation and hydroxylation. This finding may help us discover the pathway for phenazine-1-carboxylic acid degradation.

INTRODUCTION

Phenazine-1-carboxylic acid, a type of phenazine compound produced by some Pseudomonas species, has strong suppressive activity toward various plant fungal pathogens, nematodes, and Gram-negative bacteria (1–4). The biocide shenqinmycin, with phenazine-1-carboxylic acid as its main active compound, is widely used in southern China for rice protection because of its high fungicidal efficiency, low toxicity, and environmental compatibility (5). The carboxyl group is a structural feature important for the antifungal activity of phenazine-1-carboxylic acid (5). Phenazines can cause toxicity by producing reactive oxygen species (ROS) and by interfering with the respiratory electron transport chain (6). Phenazines affect microbial communities in both positive and negative ways, and their presence correlates with decreased species richness and diversity (7). Phenazine-1-carboxylic acid is a biologically active factor that has several activities that could alter the host's immune and inflammatory response and thereby contribute to bacterial pathogenesis (8). However, the degradation characteristics of phenazine-1-carboxylic acid and the environmental consequences of the long-term application have not been clarified. To reveal the effects and potential hazards of phenazine-1-carboxylic acid and its intermediates on the environment, it is imperative to study the biodegradation characteristics of phenazine-1-carboxylic acid.

Many bacteria that efficiently degrade polycyclic aromatic hydrocarbons (PAHs) and heterocyclic aromatics have been isolated. Among these, strains from the genus Sphingomonas have attracted wide attention for their diverse environmental adaptations and biodegradative capabilities (9, 10). Strains in this genus have been shown to degrade dioxins (11), carbazole, m-xylene, biphenyl, naphthalene (12), pentachlorophenol, and many other methyl, chloro, hydroxyl, and nitroaromatic compounds (13–17).

Dioxygenases play important roles in the degradation of aromatic compounds, especially in the hydroxylation and benzene ring cleavage reactions. Dioxygenases establish a functional complex with an electron transfer chain (ETC) consisting of, for example, a reductase and ferredoxin. Ring-hydroxylating oxygenases (RHOs) are classified into five groups according to the sequence analyses of ETC components of the standard RHOs (18). To date, many dioxygenases have been found in Sphingomonas strains, e.g., dioxin dioxygenase from Sphingomonas wittichii RW1 (19), which belongs to the type IV RHOs; lignin demethylase from Sphingomonas paucimobilis SYK-6 (16), which belongs to the type I RHOs; and carbazole 1,9a-dioxygenase from Sphingomonas sp. strain CB3 (20), which belongs to the type III RHOs.

Mycobacterium fortuitum CT6, Mycobacterium septicum DKN1213, Rhodococcus sp. strain JVH1, and M. fortuitum ATCC 6841 can degrade phenazine-1-carboxylic acid (7). MFORT_16319, MFORT_16334, and MFORT_16349 in M. fortuitum ATCC 6841 are involved downstream of phenazine-1-carboxylic acid biodegradation, and the phenazine-1-carboxylic acid consumption ability was lost in this strain when MFORT_16269 was inactivated (7). Unfortunately, no intermediate during phenazine-1-carboxylic acid degradation was found or purified in the four strains. Previously, S. wittichii strain DP58, which can use phenazine-1-carboxylic acid as its sole carbon and nitrogen source, was isolated from pimiento rhizo-soil, and the genome was sequenced (21, 22). Two intermediates of phenazine-1-carboxylic acid degradation, 4-hydroxy-1-(2-carboxyphenyl) azacyclobut-2-ene-2-carbonitrile and 4-hydroxy-1-(2-carboxyphenyl)-2-azetidinecarbonitrile, were found (21). However, the initial enzyme for phenazine-1-carboxylic acid degradation in strain DP58 has not been identified. In this study, a gene cluster for phenazine-1-carboxylic acid degradation was predicted using a transcriptome sequencing strategy. A novel angular dioxygenase system for the decarboxylation and hydroxylation of phenazine-1-carboxylic acid was identified. Phenazine-1-carboxylic acid 1,2-dioxygenase was found to be a three-component dioxygenase without an exclusive ETC. This finding helps us to understand how phenazine-1-carboxylic acid is degraded.

RESULTS

Inductive degradation of phenazine-1-carboxylic acid in S. wittichii DP58.

Our previous work showed that strain DP58 can use phenazine-1-carboxylic acid as its sole carbon and nitrogen sources for growth. To determine whether phenazine-1-carboxylic acid degradation in strain DP58 is inducible, the degradation rate of phenazine-1-carboxylic acid in strain DP58 and its consumption in phenazine-1-carboxylic acid-induced strain DP58 were investigated. As Fig. 1 shows, strain DP58 could degrade 1 mM phenazine-1-carboxylic acid in 48 h, while phenazine-1-carboxylic acid was consumed in 4 h by the phenazine-1-carboxylic acid-induced strain DP58. These results indicate that phenazine-1-carboxylic acid degradation in S. wittichii DP58 is induced by phenazine-1-carboxylic acid. The results allow us to use an RNA sequencing strategy to identify phenazine-1-carboxylic acid degradation-related genes.

FIG 1.

Biotransformation of phenazine-1-carboxylic acid by induced and uninduced S. wittichii strain DP58. The cells were suspended to an OD₆₀₀ of 0.2 in minimal medium containing 1 mM phenazine-1-carboxylic acid. The experiments were performed in triplicate. The results are the averages of three independent experiments, and the error bars show standard deviations.

Transcriptome analysis.

An RNA sequencing strategy was performed to identify genes involved in phenazine-1-carboxylic acid degradation. A total of 12,803,200 and 13,382,662 high-quality reads were obtained from S. wittichii strain DP58 cultured with 1 mM phenazine-1-carboxylic acid and 5 mM glycerol, respectively (Table S1 in the supplemental material). Under these two conditions, 4,461 and 4,462 expressed genes, respectively, were observed. Compared with strain DP58 cultured in minimal medium with glycerol (MMG), 2,531 differentially expressed genes were found in strain DP58 cultured in minimal medium with phenazine-1-carboxylic acid (MMP), with 2,233 and 298 up- and downregulated genes, respectively. Genes involved in the degradation of xenobiotics were upregulated, while genes involved in glycerol metabolism were found to be downregulated (supplemental material). Moreover, culturing strain DP58 with phenazine-1-carboxylic acid resulted in the upregulation of genes in the catechol ortho-pathway, and the taurine dioxygenase- and gentisate 1,2-dioxygenase-coding genes were also found to be upregulated (supplemental material). Thus, the central pathway appeared to be a dynamic network that the strain can adjust according to the substrate(s) present (23, 24).

We hypothesized that there would be many significant differences in gene expression between DP58 cultured with phenazine-1-carboxylic acid and glycerol. These would not be limited to genes in the phenazine-1-carboxylic acid degradation pathway and would likely include genes involved in substrate recognition and transport, metabolic network regulation, motility, and detoxification. pcaA1, pcaA2, and pcaB1 were found to be upregulated by 16-, 8- and 8-fold, respectively, and were thus strongly upregulated genes. pcaA1 and pcaB1 are annotated as the α-subunit of aromatic-ring-hydroxylating dioxygenase, and pcaA2 is annotated as the β-subunit of dioxygenase. In the adjacent sequence of pcaA1, a pyridine nucleotide-disulfide oxidoreductase gene and a Rieske [2Fe-2S] domain-containing protein-encoding gene, named pcaA3 and pcaA4, respectively, were also found to be upregulated. This seems to be a typical three-component dioxygenase system which may be involved in phenazine-1-carboxylic acid degradation. pcaB1B2 and its immediately upstream GntR family transcriptional regulator pcaR were found to be upregulated in the phenazine-1-carboxylic acid cultured strain, suggesting that PcaR might be a positive regulatory element for pcaB1B2 expression in phenazine-1-carboxylic acid degradation. Flagellum-dependent motility and chemotaxis are essential in enabling bacteria to move toward a substrate or away from pernicious compounds (25, 26). In MMP-cultured strain DP58, 22 motility-related fli and flg genes were found to be upregulated by 2- to 5-fold (supplemental material). However, whether the enhancement of motility was caused by pressure from phenazine-1-carboxylic acid or its properties as a substrate is unknown.

Analysis of the putative phenazine-1-carboxylic acid degradation-related gene cluster.

From the transcriptome analysis, a gene cluster comprising 10 genes was predicted to be involved in phenazine-1-carboxylic acid degradation (Fig. 2B). Seven of the 10 genes were upregulated. This cluster contains two sets of dioxygenase genes, pcaA1A2 and pcaB1B2; one pyridine nucleotide-disulfide oxidoreductase gene, pcaA3; one Rieske [2Fe-2S] domain-containing protein gene, pcaA4; one regulator gene, pcaR; one oxidoreductase gene, pcaE; one extradiol dioxygenase gene, pcaC; and one monooxygenase gene, pcaD (Table 1). The gene cluster was also found in S. wittichii strain RW1 and Sphingomonas sp. strain DC-6, with 99% nucleotide sequence similarities. No transposase-encoding gene or insert sequence was found in the adjacent sequence (10 kb up- and downstream), implying that the gene cluster may have integrated into the genome at an early stage of evolution.

FIG 2.

Initial degradation step of phenazine-1-carboxylic acid in strain DP58, and the organization of the genes involved in the pathway. (A) Pattern of phenazine-1-carboxylic acid conversion by PcaA1A2A3A4 from strain DP58. red, reduction; ox, oxidation. (B) Organization of the genes that may be involved in phenazine-1-carboxylic acid catabolism in strain DP58 and the predicted promoter of pcaA1A2.

TABLE 1.

Functional annotations of the hypothetical proteins

Protein	Size (amino acids)	Putative function	Superfamily designation
			Region (positions)	Superfamily (specific hit/conserved domain)	CDD accession no.	E value
PcaA3	409	Pyridine nucleotide-disulphide oxidoreductase	147–228	Pyr_redox	pfam00070	3.27e-15
			322–405	Reductase_C	pfam14759	5.97e-20
PcaE	275	Oxidoreductases	25–260	SDR_c	cd05233	2.21e-62
PcaA1	400	Dioxygenase α-subunit	57–172	Rieske_RO_Alpha_N	cd03469	2.71e-38
			57–150	Rieske	pfam00355	3.78e-22
PcaA2	103	Dioxygenase β-subunit	1–92	Ring_hydroxyl_B	pfam00866	4.53e-10
PcaC	313	Dioxygenase	147–295	BphC1-RGP6_C_like	cd07237	2.35e-71
			14–135	BphC1-RGP6_N_like	cd07252	8.73e-52
			154–269	Glyoxalase	pfam00903	3.36e-21
PcaA4	106	Ferredoxin	4–100	Rieske_RO_ferredoxin	cd03528	5.14e-38
			5–96	Rieske	pfam00355	2.46e-26
PcaD	481	Monooxygenase	75–192	FMO-like	pfam00743	7.55e-12
PcaR	231	Regulator	82–196	FCD (GntR family)	pfam07729	1.96e-25
PcaB1	431	Dioxygenase α-subunit	183–411	RHO_alpha_C_NDO-like	cd08881	6.03e-73
			41–124	Rieske	pfam00355	6.23e-19
PcaB2	175	Dioxygenase β-subunit	10–137	Ring_hydroxylating_dioxygenases_beta	cd00667	1.87e-49
			21–140	Ring_hydroxyl_B	pfam00866	1.36e-46

The BLASTP analysis showed that both PcaA1A2 and PcaB1B2 were members of the RHO family. PcaA1 shared 68% amino acid sequence similarity with the α-subunit of anthranilate 1,2-dioxygenase from Rhodococcus maanshanensis DSM 44675. PcaB1 shared a 75% amino acid sequence similarity with the α-subunit of 3-phenylpropionate dioxygenase from Nocardia farcinica. A flavin adenine dinucleotide (FAD)- and a NAD-binding domain were found in PcaA3, and a reductase conserved domain, which is usually present in reductases and is responsible for interactions with ferredoxin, was found at the C terminus. This suggests that PcaA3, a GR-type reductase associated with PcaA4, may transfer electrons to PcaA1A2 from NAD(P)H. A conserved domain that shared a high similarity with the active site of BphC, 2,3-dihydroxybiphenyl 1,2-dioxygenase from Sphingobium yanoikuyae B1 was found at the C terminus of PcaC, suggesting that PcaC might be an extradiol dioxygenase that cleaves the hydroxylated phenazine or other intermediate.

pcaA1A2A3A4 encode the initial dioxygenase system for phenazine-1-carboxylic acid degradation.

To determine the function of the putative phenazine-1-carboxylic acid degradation dioxygenase, isopropyl β-d-thiogalactopyranoside (IPTG)-induced E. coli DH5α harboring the corresponding vectors, pMDpcaA1A2, pMDpcaA1A2A3A4, pMDpcaB1B2, and pMDpcaB1B2A3A4, were resuspended in MMP. Phenazine-1-carboxylic acid was transformed to a brown complex by IPTG-induced E. coli DH5α/pMDpcaA1A2A3A4 after 2 h of culturing. The strains independently harboring pMDpcaA1A2, pMDpcaB1B2, and pMDpcaB1B2A3A4 did not show phenazine-1-carboxylic acid conversion capabilities. Similarly, pBBRpcaB1B2 was transferred into Pseudomonas putida strain KT2440, but no phenazine-1-carboxylic acid conversion was observed.

As mentioned above, phenazine-1-carboxylic acid degradation in strain DP58 was induced by phenazine-1-carboxylic acid. However, no regulatory gene was found adjacent to pcaA1A2. To find the promoters of the two genes, the upstream sequence of pcaA1A2 was analyzed using a promoter prediction system. As shown in Fig. 2B, a 50-bp promoter sequence was found and named P_pca. The predicted sequence was cloned into pME6015 and then independently electroporated into S. wittichii DP58 and P. putida KT2440. The strains harboring pMEPpcap showed high β-d-galactosidase activity, suggesting that P_pca is the promoter of pcaA1A2 (Fig. 3).

FIG 3.

β-d-Galactosidase activity of S. wittichii DP 58 and P. putida KT2440 harboring pMEpcap and pME6015, respectively.

Phenazine-1-carboxylic acid 1,2-dioxygenase is a type IV dioxygenase.

To test the phenazine-1-carboxylic acid conversion ability of phenazine-1-carboxylic acid dioxygenase in vitro, pcaA1A2, pcaA3, and pcaA4 were cloned into pBBR2S (pcaA1A2) or pET28a (pcaA3 and pcaA4, independently) with a His₆ tag on the N terminus. The purified and desalted recombinant proteins were further analyzed. SDS-PAGE showed that His₆-PcaA1A2 was a complex consisting of two polypeptides with apparent molecular masses of 47 and 17 kDa, while His₆-PcaA3 and His₆-PcaA4 were approximately 49 and 15 kDa in size, respectively (Fig. 4A). A molecular mass of 190 kDa was obtained for the native PcaA1A2, indicating an α₃β₃ subunit conformation (Fig. 4B). The solution containing His₆-PcaA1A2 was light yellow, which may be caused by the binding of ferredoxin and/or Fe²⁺. The proteins did not display any phenazine-1-carboxylic acid metabolism capabilities when tested individually or in pairs. The phenazine-1-carboxylic acid dioxygenase ability was observed only when the mixture contained His₆-PcaA1A2, His₆-PcaA3, and His₆-PcaA4, indicating that this dioxygenase complex contains three components: a nonheme iron dioxygenase, a [2Fe-2S] ferredoxin, and a GR-type reductase (Fig. 2A and Table 2). Thus, phenazine-1-carboxylic acid 1,2-dioxygenase is a type IV dioxygenase. NADH supported the phenazine-1-carboxylic acid 1,2-dioxygenase activity better than NADPH, indicating that NADH was the best electron donor for phenazine-1-carboxylic acid 1,2-dioxygenase (Table S2).

FIG 4.

SDS-PAGE analysis of dioxygenase, reductase, and ferredoxin (A) and native molecular weight (MW) of PcaA1A2 (B). Proteins were purified using Ni-NTA agarose and desalted using PD-10 columns. Gels were stained with Coomassie brilliant blue. Native molecular weight of PcaA1A2 was measured using protein-folding liquid chromatography (PFLC).

TABLE 2.

Activities of different combinations of oxygenase, ferredoxin, and reductase in phenazine-1-carboxylic acid degradation

Enzyme	Activity (nmol · min−1 · mg−1)a
PcaA1A2	ND
PcaA1A2A3	ND
PcaA1A2A4	ND
PcaA1A2A3A4	350.71 ± 1.44
PcaA1A2A4-RedA2	63.98 ± 19.60

^aND, no phenazine-1-carboxylic acid conversion was detected.

Phenazine-1-carboxylic acid was converted to 1,2-dihydroxyphenazine by phenazine-1-carboxylic acid 1,2-dioxygenase.

As shown in Fig. 5B, a compound with a retention time of 10.3 min was found in the whole-cell biotransformation of phenazine-1-carboxylic acid performed by E. coli DH5α/pMD1634. An ultrahigh-performance liquid chromatography-mass spectrometry (LC-MS) analysis gave an [M-H]⁺ ion at m/z 213.0659. This metabolite was purified using preparative high-performance LC (HPLC) and identified as 1,2-dihydroxyphenazine using nuclear magnetic resonance (NMR) analysis (Fig. S1). 1,2-Dihydroxyphenazine was unstable and spontaneously transformed to a black precipitate at room temperature.

FIG 5.

Degradation of phenazine-1-carboxylic acid by induced E. coli DH5α harboring pMDpcaA1A2A3A4. (A) Degradation of phenazine-1-carboxylic acid by E. coli DH5α/pMDpcaA1A2A3A4 (■) and E. coli DH5α/pMD18T (●). The cells were suspended to an OD₆₀₀ of 2.0 in MMP. The experiments were performed in triplicate. The results are the averages of three independent experiments, and the error bars show standard deviations. (B) HPLC and MS analyses of the intermediate produced from phenazine-1-carboxylic acid by E. coli DH5α/pMDpcaA1A2A3A4. Chemical structures: left, 1,2-dihydroxyphenazine; right, phenazine-1-carboxylic acid.

Decarboxylation by different mechanisms has been found during the degradation of benzoic acid and substituted benzoate (27–30). Using a conserved domain analysis, a ferredoxin-binding domain and AhdA1_c-like domain were found in the N and C termini, respectively, of PcaA1. We found that six residues for the [2Fe-2S]cluster binding in this domain were conserved, except Lys₈₆ from the salicylate-5-hydroxylase of Novosphingobium aromaticivorans F199 (marked in gray in Fig. S2A). The residues for Fe²⁺ binding and conserved residues in the putative catabolic pocket were marked with red and black asterisks, respectively. The Fe²⁺-binding site, comprising His₂₁₉, His₂₂₄, and Asp₃₄₇ in PcaA1, was conserved in this superfamily. The catabolic pocket, marked with black and red asterisks in Fig. S2A, contained 20 residues and showed some inconsistency. Three residues, 288, 301, and 346, in the catabolic pocket were highly varied, which may allow the binding of various substrates. In PcaA1, compared with other five enzymes, 14 residues were missing between Tyr₂₅₇ and Glu₂₅₈. However, the function of this fragment needs to be investigated through additional experiments. A benzoate 1,2-dioxygenase large subunit-like multidomain was also found in PcaA1, with an E value of 2.29e−40 (Fig. S2B). The multidomain contained Rieske_RO_Alpha_HBDO in its N terminus and RHO_alpha_C_AntDO-like in its C terminus. Six residues for [2Fe-2S] cluster binding in the Rieske_RO_Alpha_HBDO domain were conserved among these five enzymes. While there were conserved residues in the catabolic pocket, the RHO_alpha_C_AntDO-like domain shows a high level of variability. The variation of the residues might endow the ability to convert several substrates. The salicylate 1-hydroxylase and benzoate 1,2-dioxygenase catabolic pocket-like conserved domains may endow phenazine-1-carboxylic acid 1,2-dioxygenase with decarboxylation and dihydroxylation abilities. To confirm this hypothesis, the capability of phenazine-1-carboxylic acid 1,2-dioxygenase to convert benzoate, salicylate, and 2-hydroxyphenazine-1-carboxyl acid was tested. As shown in Fig. S3, benzoate and 2-hydroxyphenazine-1-carboxylic acid cannot be converted by phenazine-1-carboxylic acid 1,2-dioxygenase. Nearly 0.25 mM salicylate was consumed and converted to catechol in the reaction, indicating that phenazine-1-carboxylic acid 1,2-dioxygenase have the same function as salicylate 1-hydroxylase in salicylate degradation, and this result is consistent with the conserved domain analysis.

PcaA1A2 has low specificities for ferredoxin and reductase.

The failure of E. coli DH5α/pMDpcaA1A2 to degrade phenazine-1-carboxylic acid may be caused by the lack of proper ETC components. To check if the reductases from other bacteria can transfer electrons to the phenazine-1-carboxylic acid degradation dioxygenase, pBBRpcaA1A2 was electroporated into P. putida strain KT2440 and S. yanoikuyae strain XLDN2-5. Interestingly, strain KT2440 with pBBRpcaA1A2 could degrade phenazine-1-carboxylic acid, while strain XLDN2-5 with pBBRpcaA1A2 could not. The amino acid sequences of PcaA3 and PcaA4 were aligned with the genomes of strains KT2440 and XLDN2-5. No proteins showed homology (>10%) with PcaA3 and PcaA4, suggesting that PcaA3 and PcaA4 are not required as the specific reductase and ferredoxin in the transfer of electrons to PcaA1A2, and that strain KT2440 has a reductase and ferredoxin that are compatible with PcaA1A2. In addition, reductase RedA2 from S. wittichii RW1 was purified and used for an in vitro analysis. In combination with PcaA1A2A4, the enzyme mixture showed phenazine-1-carboxylic acid degradation ability (Table 2). This result also proved that PcaA3 is not the exclusive reductase for S. wittichii phenazine-1-carboxylic acid 1,2-dioxygenase.

DISCUSSION

In the present study, we identified and characterized an RHO-type dioxygenase catalyzing the decarboxylation and dihydroxylation of phenazine-1-carboxylic acid, which is different from the previously reported PAHs or heterocyclic aromatic dioxygenases, such as biphenyl 1,2-dioxygenase (31), carbazole 1,9a-dioxygenase (32), and dioxin dioxygenase (19). The phenazine-1-carboxylic acid 1,2-dioxygenase is a typical three-component dioxygenase that consists of a hexamer oxygenase (PcaA1A2), a [2Fe-2S] ferredoxin, and a GR-type reductase. Moreover, phenazine-1-carboxylic acid 1,2-dioxygenase can utilize both NADH and NADPH as cofactors for reduction and, like many ring-hydroxylating oxygenases, the preferred reductant is NADH (33). Phylogenetic analysis was performed on the basis of the amino acid alignment of PcaA1, PcaB1, and PcaC with the catabolic subunits of characterized RHOs (Fig. 6). In the constructed phylogenetic tree, PcaA1, indicated by filled triangles in Fig. 6, clustered with the oxygenase components of RHOs responsible for the degradation of benzoates and substituted benzoates and forms a subclade with BphA1_e (31), AhdA1_e (34), NagG (35), HybB (36), AhdA1_d (34), AhdA1_c (34), and PhnA1_b (30). However, PcaA1 shares only 27 to 44% amino acid sequence identities with these oxygenases. In this subclade, PcaA1 was the only oxygenase that was reported to be involved in heterocyclic aromatic degradation. Although clustered with PcaA1, the S. yanoikuyae strain B1 harboring bphA1_e cannot degrade phenazine-1-carboxylic acid. Phenazine-1-carboxylic acid 1,2-dioxygenase has the same function as AhdA1_c and PhnA1_b in salicylate conversion (Fig. S3).

FIG 6.

Phylogenetic tree constructed based on the alignment of the PcaA1 protein with the α-subunits of characterized RHOs. The multiple-alignment analysis was performed with ClustalX 2.1 software. The trees were constructed by the neighbor-joining method using MEGA 5.0. The numbers on some branches refer to the percent confidence estimated by a bootstrap analysis with 1.000 replications. The names of the proteins and strains are displayed in the phylogenetic tree.

Generally, bacteria adapt to the surroundings through the evolution of regulation system or catabolic enzymes. The degradation of xenobiotics is usually activated by transcriptional regulators in the presence of the substrate or its catabolic intermediates (37, 38). However, the constitutive expression of degradation-related genes is also found in some strains (35, 39). Phenazine-1-carboxylic acid is a microbial antibiotic that is mainly produced by Pseudomonas spp., especially rhizosphere Pseudomonas species (5). The wide distribution of phenazine-1-carboxylic acid would facilitate the microbes' evolution of a pathway to catabolize phenazine-1-carboxylic acid. The inducible expression of the phenazine-1-carboxylic acid degradation genes would enhance the survival of microbes in a barren environment with finite phenazine-1-carboxylic acid and other sources. The low specificity of the reductase in the dioxygenase system has been reported in the degradation of 3-phenoxybenzoate (14), 2-chloronitrobenzene (40), and naphthalene (41, 42). The low specificity of the reductases and ferredoxins allows the microbes to obtain new catabolic abilities faster and more efficiently through gene recruitment, which may be helpful for survival in new environments. Moreover, the low specificity of the ETC would remove the block between microbes during gene rearrangement and horizontal gene transfer and facilitate the evolution of bacterial pathways to metabolize xenobiotic compounds and elude selective pressure (32, 40).

It is known that genes involved in the same pathway usually form a cluster. Two other oxygenase-encoding genes, pcaC and pcaB1, were found in the phenazine-1-carboxylic acid degradation gene cluster. PcaC was annotated to be extradiol dioxygenase and shared 38% amino acid sequence similarity with BphC from S. yanoikuyae B1. To test the function of PcaC, pcaC was cloned into pET-28a(+), resulting in pETpcaC, and expressed in E. coli BL21(DE3) (Fig. S4). We also coexpressed pETpcaC with pMDpcaA1A2A3A4 in E. coli BL21(DE3). Neither of the strains showed any 1,2-dihydroxyphenazine conversion capabilities after IPTG induction, suggesting that pcaC might not be the gene encoding the second-step enzyme in phenazine-1-carboxylic acid degradation. The disruption of MFORT_16319 and MFORT_16334 in M. fortuitum ATCC 6841, which are homologous with pcaC and pcaB1, respectively, in strain DP58, prevented the strain from growing on plates having phenazine-1-carboxylic acid as the sole carbon source (7). However, the mutant strain still consumed phenazine-1-carboxylic acid in LB medium supplemented with 100 μM phenazine-1-carboxylic acid (7). In the constructed phylogenetic tree, PcaB1 clustered with type IV dioxygenase, which is involved in the degradation of biphenyl, benzene, and other PAHs. Combined with the upregulation of pcaC and pcaB1 in the MMP cultured strain DP58, it is speculated that pcaC and pcaB1 might be involved in the downstream pathway of phenazine-1-carboxylic acid degradation. Moreover, no similar gene was found by a BLASTN algorithm-based analysis using the nucleotide sequence of pcaA1 as the query against the genomes of M. fortuitum ATCC 6841 and CT6, suggesting that these strains might have initial phenazine-1-carboxylic acid conversion enzymes that are different from that of strain DP58.

MATERIALS AND METHODS

Chemicals.

Phenazine-1-carboxylic acid (>93% purity) was obtained from Nonglebio Co., Ltd., Shanghai, China. 2-Hydroxyphenazine-1-carboxylic acid was a gift from Mingmin Chen, Shanghai Jiao Tong University. Benzoate, catechol, salicylate, ethyl acetate, and chromatography-grade methanol and acetonitrile were purchased from Shanghai Chemical Reagent Co., Ltd., Shanghai, China.

Strains, primers, plasmids, and culture conditions.

The bacterial strains and plasmids used in this study are listed in Table 3. The primers used in this study are listed in Table 4. S. wittichii strain DP58, S. wittichii strain RW1, S. yanoikuyae XLDN2-5, and P. putida strain KT2440 were grown at 28°C in Luria-Bertani (LB) broth or minimal medium supplemented with 1 mM phenazine-1-carboxylic acid (MMP) or 5 mM glycerol (MMG) (43). Escherichia coli strains were grown at 37°C in LB broth or LB plus 1.5% agar, supplemented with appropriate concentrations of antibiotics.

TABLE 3.

Strains and plasmids used in this study

Strain or plasmid	Characteristic(s)a	Source or reference
Strains
S. wittichii DP58	Degrades phenazine-1-carboxylic acid; Smr	Laboratory stock
S. wittichii RW1	Degrades dibenzo-p-dioxin and phenazine-1-carboxylic acid; Smr	11
P. putida KT2440	Wild type	ATCC 12633
S. yanoikuyae XLDN2-5	Degrades carbazole	55
E. coli DH5α	Host strain for cloning vectors, F− recA1 endA1 thi-1 supE44 relA1 deoR Δ(lacZYA-argF)U169 Ф80dlacZΔM15	Invitrogen
E. coli BL21(DE3)	Host strain for pET28a	Invitrogen
Plasmids
pMD18T	Cloning vector; Ampr	TaKaRa
pET28a(+)	Expression vector; Kmr	Novagen
pBBRMCS-5	Broad-host-range cloning vector; Gmr	56
pME6015	pVS1-p15A E. coli_Pseudomonas shuttle vector for translational lacZ fusions and promoter probing; Tcr	49
pHN1257	Vector with Ptrc promoter; Kmr	57
pMDpcaA1A2A3A4	pMD18T derivative carrying pcaA1A2A3A4; Ampr	This study
pMDpcaA1A2	pMD18T derivative carrying pcaA1A2; Ampr	This study
PMDpcaA1A2A4	pMD18T derivative carrying pcaA1A2A4; Ampr	This study
pBBRpcaA1A2	pBBR1MCS-5 derivative carrying pcaA1A2; Gmr	This study
pBBR2S	Broad-host-range cloning vector with Ptrc promoter from pHN1257; Kmr	This study
pBBR2SHpcaA1A2	pBBR2S derivative carrying pcaA1A2 and His6 coding sequence; Kmr	This study
pETpcaA1A2	pET28a(+) derivative carrying pcaA1A2; Kmr	This study
pETpcaA3	pET28a(+) derivative carrying pcaA3; Kmr	This study
pETpcaA4	pET28a(+) derivative carrying pcaA4; Kmr	This study
pMDpcaB1B2	pMD18T derivative carrying pcaB1B2; Ampr	This study
pMEpcap	pME6015 with the promoter sequence of pcaA1A2; Tcr	This study
pETRedA2	pET-29a(+) derivative carrying redA2; Kmr	14
pETpcaC	pET-29a(+) derivative carrying pcaC; Kmr	This study

^aSm^r, streptomycin resistance; Amp^r, ampicillin resistance; Km^r, kanamycin resistance; Gm^r, gentamicin resistance; Tc^r, tetracycline resistance.

TABLE 4.

Primers used in this study

Primer	Sequence (5′–3′)a	Purpose
pMDpcaA1-2F	TAGAAGGAGACGTAGATGAAAGGACAATGCGACATGG	Forward primer to amplify pcaA1 and pcaA2
pMDpcaA1-2R	CTGCAGGTCGACGATTCAAAGCGGATAGGCGAGGTAG	Reverse primer to amplify pcaA1 and pcaA2
pMDpcaA3F	GGATCCTCTAGAGATATGACCAGGAAGACCTTTGCGA	Forward primer to amplify pcaA3
pMDpcaA3R	TCGAAATCTCCTTGTTTCAGCGTTTCATATATTTCTGGATGTT	Reverse primer to amplify pcaA3
pMDpcaA4F	AACAAGGAGATTTCGATGGCGGAAAAGATTCGGGT	Forward primer to amplify pcaA4
pMDpcaA4R	CTACGTCTCCTTCTATCATTCGGGATCGGGAATGA	Reverse primer to amplify pcaA4
pMDpcaB1-2F	TAGAAGGAGACGTAGTCAGGGGCGACCAAATCACT	Forward primer to amplify pcaB1 and pcaB2
pMDpcaB1-2R	CTGCAGGTCGACGATGCTGTTCCACCGAGCCCAC	Reverse primer to amplify pcaB1 and pcaB2
pETpcaA1-2F	AGGCACCATATGAAAGGACAATGCGACATGG	Forward primer to amplify pcaA1 and pcaA2 with an NdeI site
pETpcaA1-2R	TTTAAGCTTTCAAAGCGGATAGGCGAGGTA	Reverse primer to amplify pcaA1 and pcaA2 with an XhoI site
pETpcaA3F	AGGCACCATATGACCAGGAAGACCTTTGCG	Forward primer to amplify pcaA3 with an NdeI site
pETpcaA3R	TTTAAGCTTTCAGCGTTTCATATATTTCTGGATG	Reverse primer to amplify pcaA3 with an XhoI site
pETpcaA4F	AGGCACCATATGGCGGAAAAGATTCGGGT	Forward primer to amplify pcaA4 with an NdeI site
pETpcaA4R	TTTAAGCTTGCACGGTGCAGGTGGAAATC	Reverse primer to amplify pcaA4 with an XhoI site
pBBR2SF	GGGAACAAAAGCTGGCGCTCACTGCCCGCTTTC	Forward primer to amplify Ptrc
pBBR2SR	TGGCGGCCGCTCTAGTCTGCGGACTGGCTTTCTACG	Reverse primer to amplify Ptrc
pBR2SHF	TTTAAGCTTAGCGGATAACAATTCCCCTCTAGA	Forward primer to amplify pcaA1 and pcaA2 from pETpcaA1A2 with a HindIII site
pBR2SHR	CCGCTCGAGTCAAAGCGGATAGGCGAGGTA	Reverse primer to amplify pcaA1 and pcaA2 from pETpcaA1A2 with an XhoI site
pBBRpcaA1-2CF	AAGGAGATATACATACCTTAGCCGCTTGACACCAGA	Forward primer to amplify pcaA1 and pcaA2
pBBRpcaA1-2CR	TCGCTACTCGCCATATTGGTCTGTCATCATTCAAAGCG	Reverse primer to amplify pcaA1 and pcaA2
pBBRpcaB1-2CF	GGGAACAAAAGCTGGTCAGGGGCGACCAAATCACT	Forward primer to amplify pcaB1 and pcaB2
pBBRpcaB1-2CR	TGGCGGCCGCTCTAGGCTGTTCCACCGAGCCCAC	Reverse primer to amplify pcaB1 and pcaB2
pcapF	CAAAAGCTTCGAATTCCATGACGCGGGTCTTTCCG	Forward primer to amplify promoter of pcaA1A2 with an EcoRI site
pcapR	CGAAGCTTGGCTGCAGTGCCATGTCGCATTGTCCTTT	Reverse primer to amplify promoter of pcaA1A2 with an PstI site
pETpcaCF	AGGCACCATATGATGACAGACCAACAGTCGAATAT	Forward primer to amplify pcaA3 with an NdeI site
pETpcaCR	TTTAAGCTTTCAAATTCCTTCGCGTGGAGA	Reverse primer to amplify pcaA3 with an XhoI site

^aUnderlined bases indicate that they overlapped and were used to construct plasmids by the In-Fusion method. The bold bases indicate restriction enzyme sequences.

Transcriptome and conserved domain analysis.

S. wittichii strain DP58 was cultured in MMP or MMG. Cells were collected when approximately 50% of the phenazine-1-carboxylic acid was degraded. Total RNA was extracted using an RNA isolation kit (Qiagen, China), according to the manufacturer's instructions. RNA integrity was checked on an Agilent 2100 Bioanalyzer (Agilent Technologies, Palo Alto, CA) and an ABI StepOnePlus real-time PCR system (ABI). Transcriptome sequencing of strain DP58 was performed on an Illumina HiSeq 2000 system. Reads were mapped to the reference genome of S. wittichii RW1 using SOAPaligner/SOAP2 (44). The mapped genes were used for pathway discovery using the Kyoto Encyclopedia of Genes and Genomes (45). Reads per kilobase of transcript per million mapped reads (46) were used to calculate the gene expression level. The gene ontology and differential expression of genes in the two samples were obtained according to the methods reported by Audic and Claverie (47). To assess conserved domains in PcaA1, we queried the amino sequence of PcaA1 against the Conserved Domain Database using a BLASTP algorithm-based analysis (48). The amino acid sequence of PcaA1 was aligned with the corresponding domains of naphthalene dioxygenase (NahA_c) from Pseudomonas sp. strain NCIB 9816-4, salicylate 1-hydroxylase (AhdA1_c) from Sphingomonas sp. strain P2, salicylate 1-hydroxylase (AndA1_c) from P. aeruginosa DB01, ortho-halobenzoate 1,2-dioxygenase (OhbA) from Burkholderia cepacia JB2, and salicylate-5-hydroxylase (BphA1_d) from N. aromaticivorans F199. The amino acid sequence of PcaA1 was also aligned with the corresponding domain of benzoate 1,2-dioxygenase (BenA) from Acinetobacter pittii PHEA-2, benzoate 1,2-dioxygenase (BenA) from P. putida GJ31*, anthranilate 1,2-dioxygenase (AntA) from Pseudomonas resinovorans CA10, and 2-halobenzoate 1,2-dioxygenase (CbdC) from B. cepacia 2CBS.

Expression of initial dioxygenase in E. coli.

The entire genes pcaA1A2, pcaA3, pcaA4, and pcaB1B2 were amplified from the genomic DNA of S. wittichii strain DP58 using the primer pairs pMDpcaA1-2F/R, pMDpcaA3F/R, pMDpcaA4F/R, and pMDpcaB1B2F/R, respectively. The fragments were cloned into pMD18-T using the In-Fusion HD cloning system (Clontech) in the order pcaA3-pcaA4-pcaA1A2 or pcaB1B2, according to the manufacturer's instructions. The entire genes of pcaA1A2 and pcaB1B2 were also amplified and cloned into pMD18-T individually. E. coli DH5α harboring each of the constructed plasmids was grown at 37°C to an optical density at 600 nm (OD₆₀₀) of 0.4, and isopropyl β-d-thiogalactopyranoside (IPTG) was then added to a final concentration of 0.4 mM. After a 16 h of incubation at 16°C, the cells were collected by centrifugation and used for the whole-cell transformation of phenazine-1-carboxylic acid.

Whole-cell transformation.

IPTG-induced E. coli DH5α (100 ml) harboring pMDpcaA1A2, pMDpcaA1A2A3A4, pMDpcaB1B2, or pMDpcaB1B2A3A4 was harvested by centrifugation, washed with 20 ml of MMP, and resuspended in MMP to a final OD₆₀₀ of 2.0. E. coli DH5α harboring pMD18-T was used as the control. Cell suspensions were incubated at 28°C with shaking at 180 rpm. Each hour, a 1-ml sample was taken and 20 μl of 6 M HCl was added. Then, the intermediates of phenazine-1-carboxylic acid were extracted by ethyl acetate. The extraction solutions were vacuum-dried and resuspended in 1 ml of methanol for high-performance liquid chromatography (HPLC) or HPLC-mass spectrometry (MS) analysis.

Promoter prediction and expression of pcaA1A2 in P. putida KT2440 and S. yanoikuyae XLDN2-5.

Promoters were predicted by the Neural Network Promoter Prediction's online tools (http://www.fruitfly.org/seq_tools/promoter.html). The predicted promoter of pcaA1A2 was cloned into pME6015 (49) using primer pair pcapF/R, resulting in pMEPpca. The constructed vector was then introduced into S. wittichii strain DP58 and P. putida strain KT2440 by electroporation according to the protocol described by Coronado et al. (50). β-Galactosidase activity was detected and calculated as described previously (51). The entire pcaA1A2 with its native promoter region was cloned into vector pBBR1MCS-5. The resulting plasmid pBBRpcaA1A2 was transferred into P. putida KT2440 and S. yanoikuyae XLDN2-5 by electroporation.

HPLC, LC-MS, and nuclear magnetic resonance analyses.

Phenazine-1-carboxylic acid, 2-hydroxy-phenazine-1-carboxylic acid, and its conversion products were measured using an Agilent series 1200 system (Agilent Technologies) equipped with a C₁₈ reversed-phase column (4.6 by 250 mm, 5 μm; Agilent Technologies) at 25°C with a monitoring wavelength of 254 nm. The mobile phase consisted of acetonitrile (A) and water (B) at a 1.0 ml/min flow rate. A stepped solvent gradient was used as follows: 0 to 4 min, 20% A; 4.01 to 20 min, 40% A; and 20.01 to 30 min, 20% A. For purification of 1,2-dihydroxy-phenazine, an Agilent series 1200 system equipped with a C₁₈ reversed-phase column (50 by 250 mm, 5 μm; Agilent Technologies) was used at 25°C. The mobile phase consisted of 30% methanol and 70% water at a 10.0 ml/min flow rate. Benzoate, catechol, and salicylate were detected using methods previously described (52, 53).

Expression and purification of oxygenase, ferredoxin, and reductase.

The genes encoding oxygenase, reductase, and ferredoxin were amplified from the genomic DNA of strain DP58 with the primers listed in Table 4 using KOD polymerase. The NdeI and XhoI doubly digested PCR products were purified and subcloned into the corresponding site of vector pET28a(+). The P_trc promoter from pHN1257 was amplified and doubly digested with KpnI and XbaI and cloned into the corresponding site of vector pBBR1MCS-2, resulting in pBBR2S. The genes pcaA1A2 with an N-terminal His₆ tag coding sequence were amplified, digested with HindIII and XhoI, and cloned into pBBR2S in the correct orientation with respect to the P_trc promoter. All of the recombinant plasmids were sequenced to verify that the coding sequence of each gene was in frame with the vector sequence that encodes the N terminus His₆ tag. His₆-tagged PcaA1A2 was overproduced in P. putida KT2440 carrying pBBR2SHpcaA1A2. His₆-tagged PcaA3 and PcaA4 were overproduced in E. coli BL21(DE3) carrying pETpcaA3 and pETpcaA4, respectively. P. putida KT2440 harboring the appropriate expression vector was grown at 28°C to an optical density at 600 nm of 0.4, and then 0.5 mM IPTG was added. E. coli BL21(DE3) harboring the appropriate expression vector was grown at 37°C to an optical density at 600 nm of 0.4, and then 0.1 mM IPTG was added. After a 16 h of incubation at 16°C, cells were collected and used for protein purification. Cells were resuspended in 10 ml of binding buffer (50 mM disodium hydrogen phosphate-sodium dihydrogen phosphate buffer [pH 8.0] containing 10% glycerol and 0.3 M NaCl), lysed by sonication on ice, and centrifuged at 10,000 × g for 30 min. The supernatant was applied to a 5-ml nickel-nitrilotriacetic acid (Ni-NTA) agarose column (GE Healthcare). After washing with 10 ml of washing buffer 1 (binding buffer with 50 mM imidazole) and 10 ml of washing buffer 2 (binding buffer with 100 mM imidazole), His₆-tagged proteins were eluted from the column with the elution buffer (binding buffer with 250 mM imidazole). Desalination was performed using PD-10 columns (GE Healthcare), according to the manufacturer's instructions. The protein concentration was determined by the Bradford method with bovine serum albumin as the standard. The native molecular weights of expressed proteins were determined as described previously (54). Briefly, the molecular weight of PcaA1A2 was measured using fast protein LC and detected at 280 nm. The mobile phase was eluted in phosphate buffer (pH 7.0) at a 1.0 ml/min flow rate. The standard proteins used were as follows: thyroglobulin (bovine, 670 kDa), γ-globulin (bovine, 158 kDa), ovalbumin (chicken, 44 kDa), myoglobin (horse, 17 kDa), and vitamin B₁₂ (1.35 kDa). The linear relationship between the log(molecular weight of protein) and retention time was used to calculate the molecular weight of PcaA1A2.

Enzyme assays.

The activity of phenazine-1-carboxylic acid 1,2-dioxygenase was determined in a 1-ml mixture containing 50 mM acetate buffer (pH 7.0), 0.4 μg of oxygenase, 0.1 μg of reductase, 0.1 μg of ferredoxin, 0.1, 1, or 10 mM NADH/NADPH, and 0.5 mM Fe²⁺. The assay was initiated by the addition of phenazine-1-carboxylic acid, 2-hydroxyphenazine-1-carboxylic acid, benzoate, or salicylate at a final concentration of 0.5 mM to the enzyme mixture. The reaction was performed at 28°C for 30 min and then terminated by the addition of 20 μl of 6 M HCl. One unit of phenazine-1-carboxylic acid dioxygenase activity was defined as the amount of enzyme needed to consume 1 nmol substrate per minute at 28°C. The products were identified by HPLC-MS, as described above.