The Novel Monooxygenase Gene dipD in the dip Gene Cluster of Alcaligenes faecalis JQ135 Is Essential for the Initial Catabolism of Dipicolinic Acid

ABSTRACT

Dipicolinic acid (DPA), an essential pyridine derivative biosynthesized in Bacillus spores, constitutes a major proportion of global biomass carbon pool. Alcaligenes faecalis strain JQ135 could catabolize DPA through the “3HDPA (3-hydroxydipicolinic acid) pathway.” However, the genes involved in this 3HDPA pathway are still unknown. In this study, a dip gene cluster responsible for DPA degradation was cloned from strain JQ135. The expression of dip genes was induced by DPA and negatively regulated by DipR. A novel monooxygenase gene, dipD, was crucial for the initial hydroxylation of DPA into 3HDPA and proposed to encode the key catalytic component of the multicomponent DPA monooxygenase. The heme binding protein gene dipF, ferredoxin reductase gene dipG, and ferredoxin genes dipJ/dipK/dipL were also involved in the DPA hydroxylation and proposed to encode other components of the multicomponent DPA monooxygenase. The ¹⁸O₂ stable isotope labeling experiments confirmed that the oxygen atom in the hydroxyl group of 3HDPA came from dioxygen molecule rather than water. The protein sequence of DipD exhibits no significant sequence similarities with known oxygenases, suggesting that DipD was a new member of oxygenase family. Moreover, bioinformatic survey suggested that the dip gene cluster was widely distributed in many Alpha-, Beta-, and Gammaproteobacteria, including soil bacteria, aquatic bacteria, and pathogens. This study provides new molecular insights into the catabolism of DPA in bacteria.

IMPORTANCE Dipicolinic acid (DPA) is a natural pyridine derivative that serves as an essential component of the Bacillus spore. DPA accounts for 5 to 15% of the dry weight of spores. Due to the huge number of spores in the environment, DPA is also considered to be an important component of the global biomass carbon pool. DPA could be decomposed by microorganisms and enter the global carbon cycling; however, the underlying molecular mechanisms are rarely studied. In this study, a DPA catabolic gene cluster (dip) was cloned and found to be widespread in Alpha-, Beta-, and Gammaproteobacteria. The genes responsible for the initial hydroxylation of DPA to 3-hydroxyl-dipicolinic acid were investigated in Alcaligenes faecalis strain JQ135. The present study opens a door to elucidate the mechanism of DPA degradation and its possible role in DPA-based carbon biotransformation on earth.

INTRODUCTION

Dipicolinic acid (DPA), a principal “nonnutrient” component of Bacillus spores for stress resistance, constitutes 5 to 15% of the dry weight of spores. DPA is 1:1 chelated with divalent cations (mainly Ca²⁺) in the spore core (1, 2). DPA cannot be decomposed by spores and is released into the environment rapidly during spore germination. Based on different models, the total amount of DPA is estimated between 5.6 × 10¹² mol and 4.2 × 10¹³ mol (approximately 936 million tons to 7.002 billion tons) in the spores of the uppermost kilometer of global marine sediments. In addition, the estimated carbon pool driven from the DPA in the uppermost kilometer of global marine sediments is 0.47 Pg C to 3.53 Pg C (1 Pg C means 1 billion tons of carbon), accounting for about 0.85 to 6.03‰ of the global biomass carbon pool (3–6).

In the chemical synthesis industry, DPA is also widely utilized as an intermediate for the synthesis of pharmaceuticals or chemicals due to its unique compound structure, e.g., (i) the symmetrical carboxyl group substitution on 2′ and 6′ positions of pyridine ring and (ii) the rigidity of the pyridine ring and the flexibility of carboxyl group. The synthesized products include 2,6-diamino-4-chloropyridine, 2,6-pyridinedicarbonyl chloride, 2,6-pyridine dimethanol, etc. (7, 8). In addition, DPA itself or some of the derivatives can be used to inhibit the activity of metallo-β-lactamases and enhance the activity of β-lactam antibiotics (9–11).

DPA is widely distributed and abundant in the environment. Thus, its catabolism by microorganisms would play an important role in the global carbon cycle. However, the microbial degradation of DPA has not yet received considerable attention. The bacterial degradation of DPA has only been reported in few species. Achromobacter sp. strain 1-2 and Alcaligenes faecalis strain JQ135 could catabolize DPA and utilize it as a sole carbon source for growth (12–15). Achromobacter sp. 1-2 catabolizes DPA with formation of several metabolites, such as 3-hydroxydipicolinic acid (3HDPA), oxalic acid, and α-ketoglutaric acid (15). In our previous study, A. faecalis JQ135 was shown to possess a similar DPA catabolic pathway (termed the “3HDPA pathway”) to Achromobacter sp. 1-2, and a metabolite 4-imino-pent-2-enedioic acid (IPEA) was also identified (Fig. 1) (12). However, no genes involved in the DPA degradation in bacteria have been reported to date.

FIG 1.

Several catabolic pathways of representative pyridine derivatives by microorganisms. The maleamate pathway, 3HDPA pathway, and 2,3,6THP pathway are highlighted by blue, yellow, and brown backgrounds, respectively. Compounds in red boxes are the parent compounds of 2,5-dihydroxy-pyridine and 2,3,6-trihydroxypyridine (2,3,6THP). Compound 3 (in red) was detected and identified in this study. Numbered compounds: 1, dipicolinic acid (DPA); 2, 3-hydroxydipicolinic acid (3HDPA); 3, 2-oxalyl-imino-5-oxo-pent-3-enoic acid (OIOPE); 4, 4-oxalyl-imino-pent-2-enedioic acid; 5, 4-imino-pent-2-enedioic acid (IPEA); 6, oxalic acid; 7, 4-oxo-pent-2-enedioic acid; 8, ammonium; 9, α-ketoglutarate; 10, 2-ketoglutaramate; 11, 2,3,6-trihydroxypyridine (2,3,6THP); 12, cis-5,6-dihydro-5,6-dihydroxy-2-pyridone; 13, 2-hydroxypyridine; 14, nicotine; 15, 2,6-dihydroxypyridine; 16, nicotinic acid; 17, 6-hydroxynicotinic acid; 18, 6-hydroxy-3-succinoyl-pyridine; 19, picolinic acid; 20, 6-hydroxypicolinic acid; 21, 3,6-dihydroxypicolinic acid; 22, 2,5-dihydroxy-pyridine; 23, N-formylmaleamic acid; 24, maleamic acid; 25, maleic acid; 26, fumaric acid.

In the present study, a dip gene cluster responsible for DPA catabolism was cloned in A. faecalis JQ135. The gene elements involved in the initial hydroxylation of DPA to form 3HDPA and the oxygen source of the hydroxyl group were investigated. Furthermore, the homologous dip genes were analyzed in the newly identified DPA degraders and many other Alpha-, Beta-, and Gammaproteobacteria. This study reveals the underlying genetic determinants of DPA degradation by bacteria and provides new insights into the potential fate of the massive DPA in nature.

RESULTS

Cloning of the dip gene cluster.

A mutagenesis library containing approximately 6,300 mutants (D0001 to D6300) was constructed using the pSC123 based transposon to locate the genes responsible for DPA catabolism. The mutant D3347 failed to degrade DPA or utilize it as a sole carbon source for growth. The transposon insertion site was determined using self-formed adaptor PCR (SEFA-PCR) and finally located in gene AFA_17815 (designated dipD). Subsequently, an 11-kb gene cluster (termed the dip gene cluster) comprised of 13 predicted open reading frames (ORFs), AFA_17775 to AFA_17835, was proposed to be involved in DPA catabolism. The 13 ORFs were designated dipR, dipA, dipB, dipC, dipD, dipE, dipF, dipG, dipH, dipJ, dipK, dipL, and dipM, respectively. The predicted functions of dip genes are summarized in Table 1.

TABLE 1.

Annotation of ORFs in the dip gene cluster in strain JQ135

Gene locus	Product sizea	Databaseb	Homologous proteinc (accession no.)	Identity (%)	Proposed function
AFA_17775 (dipM)	315	NR	Ornithine cyclodeaminase (HBQ90117.1)	99.68
		Swiss-Prot	Ketamine reductase mu-crystallin (O54983.1)	30.98
AFA_17780 (dipL)	92	NR	(2Fe-2S)-binding protein (WP_009457378.1)	29.63	Ferredoxin
		Swiss-Prot	1,2-Phenylacetyl-CoA epoxidase, subunit E (P76081.1)
AFA_17785(dipK)	114	NR	Ferredoxin (WP_094198104.1)	98.25	Ferredoxin
		Swiss-Prot	*
AFA_17790 (dipJ)	102	NR	(2Fe-2S)-binding protein (WP_094198105.1)	100	Ferredoxin
		Swiss-Prot	Na(+)-NQR subunit F (Q9HZL1.1)	37.5
AFA_17795 (dipH)	63	NR	Hypothetical protein (WP_094198106.1)	100
		Swiss-Prot	*
AFA_17800 (dipG)	344	NR	2Fe-2S iron-sulfur cluster-binding protein (WP_108728340.1)	100	Ferredoxin reductase
		Swiss-Prot	CDP-6-deoxy-l-threo-d-glycero-4-hexulose-3-dehydrase reductase (Q66DP5.3)	35.03
AFA_17805 (dipF)	159	NR	Heme-binding protein (WP_094198108.1)	100	Coenzyme for DipD
		Swiss-Prot	Uncharacterized protein (P45517.1)	40.46
AFA_17810 (dipE)	486	NR	Aldehyde dehydrogenase (HBQ90110.1)	99.59	Dehydrogenase of OIOPE
		Swiss-Prot	Thermostable aldehyde dehydrogenase (P42329.1)	48.01
AFA_17815 (dipD)	244	NR	Iron-containing redox enzyme family protein (WP_094198110.1)	100	Key subunit of DPA monooxygenase
		Swiss-Prot	Probable oxidoreductase CT_610 (O84616.1)	29.32
AFA_17820 (dipC)	493	NR	Gamma-glutamyltransferase (AWG36174.1)	99.19
		Swiss-Prot	Glutathione hydrolase-like YwrD proenzyme (O05218.1)	30.7
AFA_17825 (dipB)	505	NR	Hypothetical protein (WP_094198112.1)	100
		Swiss-Prot	Membrane protein YhfA (O07599.1)	33.6
AFA_17830 (dipA)	246	NR	TSUP family transporter (WP_094198464.1)	100	Transporter for DPA
		Swiss-Prot	*
AFA_17835 (dipR)	235	NR	GntR family transcriptional regulator (WP_125277224.1)	99.57	Transcriptional regulator
		Swiss-Prot	HTH-type transcriptional regulator YdhC (O05494.1)	29.7

^aNumber of amino acids. *, No significant similarity was found.

^bNR, NCBI Nonredundant Protein Sequence Database.

^cThe top BLASTP hit (except sequences from strain JQ135) was selected.

To determine the role of the DPA on the transcriptions of the dip gene cluster, the mRNA level of the 13 dip genes in JQ135 grown on DPA and citrate were compared using quantitative reverse transcription-PCR (RT-qPCR). The transcription levels of the 13 dip genes are increased by 7.40-fold (dipR), 27.16-fold (dipA), 57.47-fold (dipB), 1.94-fold (dipC), 1,169.76-fold (dipD), 267.97-fold (dipE), 227.25-fold (dipF), 546.16-fold (dipG), 16.08-fold (dipH), 240.03-fold (dipJ), 257.19-fold (dipK), 175.22-fold (dipL), and 27.38-fold (dipM), respectively, in wild-type JQ135 cells grown on DPA compared to citrate. These results indicated that the expression of dip gene cluster was induced by DPA (Fig. 2). Moreover, a dipR disruption strain designated JQ135-idipR was constructed to investigate the effect of DipR on the transcriptions of the dip gene cluster. The RT-qPCR results showed that when grown with citrate, the transcription levels of the 13 dip genes are increased by 6.92-fold (dipR), 7.95-fold (dipA), 356.87-fold (dipB), 64.68-fold (dipC), 5,903.45-fold (dipD), 1,022.44-fold (dipE), 504.34-fold (dipF), 473.82-fold (dipG), 34.46-fold (dipH), 192.65-fold (dipJ), 253.08-fold (dipK), 148.35-fold (dipL), and 21.80-fold (dipM), respectively, in uninduced mutant strain JQ135-idipR compared to uninduced wild-type JQ135. The result confirmed that DipR negatively regulated the transcription of the dip cluster (Fig. 2). In addition, the lag phase of DPA catabolism in mutant JQ135-idipR was disappeared (Fig. 3A). Mutant JQ135-idipR (the optical density at 600 nm [OD₆₀₀] of resting cells was 0.5) degraded 0.6 mM DPA completely in 4 h, whereas the degrading period of wild-type JQ135 was 20 h (a nearly 12-h lag phase). Thus, the transcription of dip gene cluster is induced by DPA and negatively regulated by DipR (16).

FIG 2.

dip gene cluster in A. faecalis strain JQ135 and its transcriptions. (A) Organization of the dip gene cluster. (B) RT-qPCR analysis of transcriptional levels of 13 dip genes in wild-type JQ135 and JQ135-idipR. The gene expression levels are normalized to 16S rRNA. ****, Significant differences (P < 0.0001, two-sided Welch’s t test); *, P < 0.5.

FIG 3.

Degradation of DPA by strain JQ135 and its derivatives. (A) Degradation of DPA by uninduced strain JQ135 and the transcriptional regulator gene dipR inactivated mutant JQ135-idipR. (B) Degradation of DPA by mutant JQ135-idipR, the dipR dipD inactivated mutant JQ135-idipRD, and the dipD gene complemented strain JQ135-idipRD/dipD. (C) Degradation of DPA by mutant JQ135-idipR, the gene dipR dipEFG inactivated mutant JQ135-idipREFG, and the dipEFG gene complemented strain JQ135-idipREFG/dipEFG. (D) Degradation of DPA by mutants JQ135-idipR, JQ135-idipREFG, and the dipDEF gene complemented strain JQ135-idipREFG/dipDEF. (E) Degradation of DPA by mutants JQ135-idipR and JQ135-idipREFG and the dipG gene complemented strain JQ135-idipREFG/dipG. (F) Degradation of DPA by the mutants JQ135-idipR and JQ135-idipREFG and the dipFG gene complemented strain JQ135-idipREFG/dipFG. The experiments were performed in BM containing 0.6 mM DPA. In panels B, C, D, E, and F, mutant JQ135-idipR instead of the wild-type strain JQ135 was used as the control. Plasmid MCS-dipDEF was used for gene dipEF complementation in mutant JQ135-idipREFG. The initial OD₆₀₀ of the inoculated cells was 0.5.

The genes dipD, dipF, and dipG are involved in the initial hydroxylation of DPA.

In the dip gene cluster, the genes dipD, dipE, dipF, and dipG encode monooxygenase, aldehyde dehydrogenase, heme-binding protein, and ferredoxin reductase, respectively. Since the dipD transposon mutant was screened out in the above mutagenesis library, dipD was most like the key genes for initial hydroxylation of DPA. Two mutants, JQ135-idipRD (with dipR and dipD genes inactivated) and JQ135-idipREFG (with dipR, dipE, dipF, and dipG genes inactivated), were constructed. The mutant JQ135-idipRD completely lost the capacity to degrade DPA compared to the positive-control JQ135-idipR. The complemented strain JQ135-idipRD/dipD recovered the degradation capacity to DPA (Fig. 3B). Moreover, the mutant JQ135-idipREFG showed a very low degradation capacity for DPA, while its complemented strain, JQ135-idipREFG/dipEFG, regained a high-efficiency degradation capacity (Fig. 3C). In addition, 3HDPA, the first metabolite of DPA, was detectable in the mutant JQ135-idipREFG but not in the mutant JQ135-idipRD during the DPA catabolic assay by resting cells (see Fig. S1 in the supplemental material). These results indicated that the gene dipD was crucial for the hydroxylation of DPA to 3HDPA. One or more dipEFG genes played significant roles in the initial hydroxylation of DPA.

For further study of the function of dipEFG genes, DPA catabolic assays using the resting cells of various gene-complemented strains of mutant JQ135-idipREFG were carried out. Mutant JQ135-idipR and JQ135-idipREFG were used as positive and negative controls, respectively. The mutant JQ135-idipREFG complemented with dipDEF (JQ135-idipREFG/dipDEF) could not improve the DPA degradation rate (Fig. 3D), while mutants complemented with dipG (strain JQ135-idipREFG/dipG) or dipFG (strain JQ135-idipREFG/dipFG) significantly improved the DPA degradation rate. Furthermore, the strain JQ135-idipREFG/dipFG had a significantly higher degradation rate than the strain JQ135-idipREFG/dipG (Fig. 3E and F). In addition, gene complementation of dipE or dipF individually cannot improve the DPA degradation rate compared to the negative control JQ135-idipREFG; gene complementation of dipE and dipG together cannot further improve the DPA degradation rate compared to the gene-complemented strain JQ135-idipREFG/dipG. These results suggested that dipF and dipG genes rather than dipE were involved in the initial hydroxylation of DPA.

The dipJKL genes are also important for the initial hydroxylation of DPA.

Ferredoxins act as the electron carriers for some oxygenases (17–19). In the dip gene cluster, transcription of three consecutive ferredoxin genes dipJ, dipK, and dipL was induced by DPA (Fig. 2). Deletion of dipJKL genes from strain JQ135 (mutant JQ135-idipJKL) resulted in significant extension (from 12 to 20 h) of the lag phase of DPA degradation (Fig. 4A). Moreover, when the dipJKL genes were deleted in the mutant JQ135-idipR (mutant JQ135-idipRJKL), the time for DPA complete degradation extended from 4 to 24 h, and the lag phase increased from 0 h to nearly 16 h (Fig. 4B). For the dipJKL gene complemented strains JQ135-idipJKL/dipJKL and JQ135-idipRJKL/dipJKL, the phenotypes were recovered, respectively. Therefore, one or more of the three ferredoxin-encoding genes dipJ, dipK, and dipL are involved in the initial hydroxylation of DPA, while other inducible ferredoxin genes on the genome could partially substitute their function.

FIG 4.

Degradation of DPA by strain JQ135, mutant JQ135-idipR, the dipJKL gene mutants, and their complemented strains. (A) Degradation of DPA by strain JQ135, mutant JQ135-idipJKL, and the complemented strain JQ135-idipJKL/dipJKL. (B) Degradation of DPA by mutants JQ135-idipR, JQ135-idipRJKL, the complemented strain JQ135-idipRJKL/dipRJKL. The experiments were performed in the BM containing 0.6 mM DPA. The initial OD₆₀₀ of the inoculated cells was 0.5.

Hydroxylation of DPA is an oxidative process proved by ¹⁸O₂ stable isotope labeling.

To study the property of the initial hydroxylation of DPA, an ¹⁸O₂ stable isotope labeling experiment was performed in a sealed bottle. The resting cells of the mutant JQ135-idipREFG/dipG were used to convert DPA, and the metabolites produced were analyzed by liquid chromatography-mass spectrometry (LC-MS). Two compounds, C₇H₄O₄NO^– (m/z 182.0098) and C₇H₄O₄N¹⁸O^– (m/z 184.0147) were detected and identified as 3HDPA and O¹⁸-labeled 3HDPA, respectively. Meanwhile, only one no-isotope peak (C₇H₄O₄NO^–, m/z 182.0097) was detected in the negative control (supplied with air only) (Fig. 5). These results proved that the oxygen of the 3′-OH group of 3HDPA is derived from oxygen gas. Thus, the enzyme catalyzing initial hydroxylation of DPA to 3HDPA was a monooxygenase rather than a hydrolase. According to the previous studies, hydroxylation on the pyridine ring is usually catalyzed by multicomponent monooxygenases (20–25), and in the present study the genes dipD, dipF, dipG, and dipJKL were proposed to be involved in the DPA hydroxylation. Hence, the DPA hydroxylation was proposed to be catalyzed by a multicomponent monooxygenase, and DipD was proposed to be the crucial catalytic component of the multicomponent DPA monooxygenase, while the heme-binding protein DipF, ferredoxin reductase DipG, and ferredoxins DipJ/DipK/DipL were proposed to be other components of the DPA monooxygenase.

FIG 5.

LC-MS analysis of the ¹⁸O₂ stable isotope labeling experiment. (A) Negative ion mass spectrometry of the metabolite from DPA degradation by the resting cells of strain JQ135-idipREFG/dipG. The compound with an m/z of 182.0097 (C₇H₄O₄NO^–) was 3HDPA (error, 0.9 ppm). (B) Negative ion mass spectrometry of the metabolite from DPA by the same resting cells supplied with ¹⁸O₂ in an anaerobic bottle. Two compounds with an m/z of 182.0098 (C₇H₄O₄NO^–) and with an m/z of 184.0147 (C₇H₄O₄N¹⁸O^–) were 3HDPA (error, 1.6 ppm) and ¹⁸O-labeled 3HDPA, respectively.

In addition, a ring cleavage product of 3HDPA was found by LC-MS using the mutant JQ135-idipREFG/dipG. Its molecular ion was m/z 198.0044 (C₇H₅O₆N^–) in the negative ion mass spectrometry (see Fig. S2). This compound was suggested to be 2-oxalyl-imino-5-oxo-pent-3-enoic acid (OIOPE; error, 0.1 ppm) (Fig. 1), the ring lysis product of 3HDPA.

Phylogenetic analysis of the monooxygenase DipD.

As mentioned above, DipD was the possible crucial catalytic component of the multicomponent DPA monooxygenase. Swiss-Prot database survey showed that only four proteins exhibited <30% sequence identities (query coverage, 76%) with DipD were found, indicating that no homologous protein was characterized before. Bioinformatic analyses suggested that DipD contains a Haem_oxygenas_2 domain (see Fig. S3). The domain analyses in the NCBI database indicated that DipD was predicted as an unknown Haem_oxygenas_2 family protein, which is an iron-containing redox enzyme with a heme oxygenase-like fold. Furthermore, DipD was aligned with oxygenases from five representative families, including 8 heme iron-dependent cytochrome P450 monooxygenases, 7 Fe(II)/α-ketoglutarate-dependent dioxygenases, 12 flavin-dependent monooxygenases, 13 Rieske non-heme iron-dependent oxygenases, and 6 ring cleavage oxygenases, and no oxygenases with a query coverage of >30% and a percent identity of >30% to DipD were found. Thirteen oxygenases with a relative higher identity and coverage and a lower E value were selected for the phylogenetic analysis with DipD and, notably, DipD formed a separated branch on the phylogenetic tree (tree scale is 1) (Fig. 6; see also Table S1 and Fig. S3 in the supplemental material). These results suggested that DipD was a new monooxygenase clearly differing form known oxygenases.

FIG 6.

Phylogenetic analysis of DipD and related oxygenases. Accession numbers are shown in brackets after each oxygenase name. Oxygenases in the same color belong to the same family. The family name is mentioned under the tree. The phylogenetic tree was constructed by the maximum-likelihood method (with a bootstrap of 500) using the software MEGA 5.2.

Diversity and distribution of 3HDPA pathways and dip gene clusters.

In addition to A. faecalis JQ135, a new DPA degrader, Achromobacter sp. strain MY14, was isolated from soil samples (see Fig. S4). Notably, a homologous dip gene cluster was also annotated in the draft genome of strain MY14 (Fig. 7). Next, using DipD as a query against the nonredundant protein sequences database (NR) in NCBI, more than 330 strains containing homologous dip cluster (identity >80% and coverage >90% of the protein sequences) were collected (see Table S2). Most of them were Betaproteobacteria (255 strains in 21 genera) and Alphaproteobacteria (74 strains in 14 genera), while four strains in two genera of Gammaproteobacteria were found. These strains were reported from various habitats with diverse functions, including pollutant-degrading bacteria (e.g., Aminobacter aminovorans DSM 7048), soil bacteria with important functions in the environment (Achromobacter xylosoxidans A8 and Cupriavidus necator N-1), plant rhizosphere bacteria (Bradyrhizobium diazoefficiens CB1809), and pathogens (Bordetella bronchiseptica RB50 and Pandoraea pnomenusa NCTC13160). In particular, B. bronchiseptica RB50 is a well-known opportunistic pathogen causing respiratory diseases in a variety of animals, including humans (26–33). Alternatively, a congeneric nonpathogenic strain, Bordetella petrii MY10, was obtained from a public culture center, and its capacity to catabolize DPA was tested. As expected, strain MY10 was could degrade DPA and utilize it as the sole carbon source for growth (see Fig. S4). Furthermore, bioinformatic analysis showed that the dip clusters of these bacteria were highly diverse and mainly characterized by eight representative arrangements (Fig. 7). These arrangements could be further divided into two types. In type 1, the gene dipD was adjacent to dipEFG, and in type 2, the gene dipD was clustered with dipJKL. The dip gene clusters of strains JQ135 and MY14 belonged to type 1 and type 2, respectively. Based on these results, the dip gene clusters were considered to be widely distributed in Alpha-, Beta-, and Gammaproteobacteria.

FIG 7.

Comparative analysis of dip gene clusters and its homologs in other bacteria. Arrows of the same color represent homologous genes. White arrows indicate hypothetical proteins. The number below each gene indicates the amino acid sequence identities compared to the related Dip proteins of strain JQ135. Alcaligenes faecalis JQ135, Bordetella petrii MY10, and Achromobacter sp. MY14 are from this study. Bordetella bronchiseptica RB50 is a pathogen. Cupriavidus necator N-1 is a soil bacterium and the type strain of the genus Cupriavidus. Roseicella is a newly reported genus and is rarely studied. Hoeflea sp. 108 is a methylotroph. Bradyrhizobium sp. cir1 is a nitrogen fixing bacteria.

DISCUSSION

DPA, as a biogenic pyridine derivative, could be decomposed by microorganisms. A study of microbial degradation of DPA in an Achromobacter strain was reported 60 years ago (15). Our previous study also found that DPA could be utilized as carbon source for cell growth of A. faecalis strain JQ135 (12). Although the metabolic pathway (termed the 3HDPA pathway) was proposed in previous reports, the underlying genes responsible for DPA degradation have not been investigated. Here, a dip gene cluster responsible for DPA degradation was cloned from strain JQ135. A novel monooxygenase gene, dipD, was crucial for the initial hydroxylation of DPA and was proposed to be the key catalytic component of the multicomponent DPA monooxygenase. In addition, the heme-binding protein gene dipF, the ferredoxin reductase gene dipG, and the ferredoxin gene cluster dipJKL were also involved in the initiation of DPA catabolism, which were proposed to be other components of the multicomponent DPA monooxygenase. This study thus provides new insight into DPA degradation at the genetic level.

Hydroxylation processes are always involved in the degradation of pyridine derivatives, which are usually mediated by three-component molybdenum-containing hydroxylases (21, 22, 25). These hydroxyl groups of products are derived from water, such as nicotine dehydrogenase from Arthrobacter nicotinovorans pAO1 and cotinine hydroxylase from Nocardioides sp. strain JQ2195 (20, 21). The present study confirmed that the hydroxyl group of 3HDPA came from oxygen gas rather than water molecules as demonstrated by a ¹⁸O₂ stable isotope labeling experiment. This hydroxylation step of DPA degradation was mediated by monooxygenase rather than hydroxylase. This reaction seems to be more similar to the hydroxylation of 6HPA (compounds 20 and 21 in Fig. 1) catalyzing by a four-component Rieske family monooxygenase PicB (22). Our results suggested that DPA was hydroxylated by a multicomponent monooxygenase in which DipD was the key catalytic component and the heme-binding protein DipF, ferredoxin reductase DipG, and ferredoxin DipJ/K/L might be other components of the multicomponent DPA monooxygenase. However, DipD showed relatively low protein sequence homology (<30% coverage or <30% identity) with Rieske family oxygenases, as well as flavin-dependent monooxygenases, α-ketoglutarate dependent dioxygenases, and ring cleavage oxygenases (Fig. 6). Additionally, DipD forms a separate branch on the phylogenetic tree with known oxygenase. Therefore, DipD was considered a new member of the oxygenase family.

When the pyridine or its derivatives are served as carbon or nitrogen sources for the growth of microorganisms, the pyridine ring is usually cleaved (34). To date, only a few ring cleavage substrates for pyridine derivatives have been reported (Fig. 1). In Gram-negative bacteria, pyridine derivatives such as picolinic acid, nicotinic acid, nicotine, and 2-hydroxypyridine usually converge to 2,5-dihydroxypyridine, followed by ring cleavage to form fumaric acid through the maleamate pathway (23, 35–38). In Gram-positive bacteria, pyridine derivatives such as nicotine, cotinine, and 2-hydroxypyridine are converted to 2,3,6THP, followed by ring cleavage to form α-ketoglutarate through the 2,3,6THP pathway (20, 24, 39, 40). Recently, pyridine was reported to be directly ring cleaved in Arthrobacter sp. strain 68b (41). In the present study, we provide new evidence on the ring cleavage substrate 3HDPA, which is lysed between the 2′ and 3′ positions of the pyridine ring, forming the product OIOPE (compound 3 in Fig. 1) (42–46). These results expand the understandings on diversity of degradation pyridine derivatives by bacteria.

One point should be mentioned here was that our attempts to determination of enzymatic activites of DipD failed. For example, the dipD gene or a long fragment containing dipD to dipM genes were heterologously expressed in many hosts, including E. coli DH5α, Alcaligenaceae strains, Pigmentiphaga sp. H8, and Pseudomonas putida KT2440, but no functions were found from whole cells or cell extracts (data not shown) (43, 47). These might due to lack of unknown crucial cofactors of DipD (47). Our further research will focus on the functional identification of dip genes and determination of the activities of corresponding enzymes.

DPA is widely distributed and abundant in environment (5, 6). Thus, microorganisms utilizing this natural biogenic pyridine derivative as carbon, nitrogen, or energy sources must exist widely in soil, water, and other environments. In this study, the dip gene cluster responsible for DPA catabolism was cloned from A. faecalis JQ135 and further identified in other two bacteria, Bordetella petrii MY10 and Achromobacter sp. strain MY14. Subsequent bioinformatic surveys indicated the dip gene clusters were found in many Alpha-, Beta-, and Gammaproteobacteria, including soil bacteria, aquatic bacteria, and pathogens, despite demonstrating different gene arrangements. Therefore, our present study opens a door to elucidate the microbial mechanism of DPA degradation and its possible role in DPA-based carbon biotransformation on earth (5, 13, 48–52).

MATERIALS AND METHODS

Chemicals, cultural conditions, primers, strains, and plasmids.

DPA was purchased from Makclin Biochemical (Shanghai, China). 3HDPA was synthesized by AiTuo Chemical (Shenzhen, China), and the stable isotope ¹⁸O₂ was purchased from Nanjing Qian’ao Electronic Technology Co., Ltd. (Nanjing, China). DNA polymerases, RNA isolation kits, and PrimeScript RT reagent kits were purchased from TaKaRa Bio, Inc. (Nanjing, China). All chromatographic-grade organic solvents were purchased from Merck KGaA (Darmstadt, Germany). Lysogeny broth (LB) medium and basic medium (BM) were prepared as described in a previous study (12).

Primers used in this study are listed in Table 2. The strains and plasmids used or constructed in this study are listed in Table 3. A. faecalis JQ135 is a wild-type strain and could degrade DPA and utilize DPA as the carbon, nitrogen, and energy sources. Achromobacter sp. MY14 was isolated in this study; Bordetella petrii MY10 was purchased from the Marine Culture Collection of China (MCCC). Strains JQ135, MY14, and MY10 were cultured aerobically at 30°C, while the E. coli strains were cultured aerobically at 37°C.

TABLE 2.

Primers used in this study

Primer	Sequence (5′–3′)	Purpose
q16S-F	CGCGGTAATACGTAGGGTGCAA	RT-qPCR analysis
q16S-R	AACTTCACGCGTTAGCTGCGC	RT-qPCR analysis
qdipR-F	CAACCTTCCCAATTCCCTTGTCGA	RT-qPCR analysis
qdipR-R	TATTGGCCGGGTCCTGGCGTTGA	RT-qPCR analysis
qdipA-F	TGCCATCTTGGGCGCAGTCGC	RT-qPCR analysis
qdipA-R	ATTAATGGCAGGCTGACTTGATGG	RT-qPCR analysis
qdipB-F	CACGGCAACATCAGCACCGACA	RT-qPCR analysis
qdipB-R	ACTGTTCAACATCTTCCTGGT	RT-qPCR analysis
qdipC-F	TTCCATTGCCTGTTCAGCCTGC	RT-qPCR analysis
qdipC-R	TGTGCTGTCCATTACCACGC	RT-qPCR analysis
qdipD-F	ATGGCACTTGGGACTCCAGGC	RT-qPCR analysis
qdipD-R	AGCGGCAAGCTAAGCCGTGAG	RT-qPCR analysis
qdipE-F	GCGCTGGGTATACACCAGTT	RT-qPCR analysis
qdipE-R	TCCGGCCGATCGGGACGATGT	RT-qPCR analysis
qdipF-F	TGATGTCCGCCTGAACCAGC	RT-qPCR analysis
qdipF-R	TATCCCATCGTGATGGAGCG	RT-qPCR analysis
qdipG-F	AGCCACACAAGTAAATGGCG	RT-qPCR analysis
qdipG-R	GACGAGCTGCTGCTGGAGAT	RT-qPCR analysis
qdipH-F	CGAACTTCGCTACTTGCAGCGT	RT-qPCR analysis
qdipH-R	TTAGGCGCTGTGTGGGAGCTG	RT-qPCR analysis
qdipJ-F	TCACTGCACCTCCTGTATCTCTGC	RT-qPCR analysis
qdipJ-R	ATGCCAACCATTACTTTTCATA	RT-qPCR analysis
qdipK-F	AGCTCACATTGGCGCAACTC	RT-qPCR analysis
qdipK-R	GTATGTGGTGCTAACGAGCA	RT-qPCR analysis
qdipL-F	AGCTCACCGTCACGTCACCG	RT-qPCR analysis
qdipL-R	GTTCAGATTACATTTGCGTCG	RT-qPCR analysis
qdipM-F	AATGCCGCAATCGGCTGCCG	RT-qPCR analysis
qdipM-R	CTCAGGGCACCGCGCATGTCCA	RT-qPCR analysis
dipR-arm-F	AGCTTGATATCGAATTCCTGCAGCAATTGCGTGAACGTATTCTG	To construct plasmid pJQ-dipR
dipR-arm-R	AGGGAACAAAAGCTGGAGCTCTTCCCAGTTCTGGCGGATGCGC	To construct plasmid pJQ-dipR
kodipD-UF	AGCTTGATATCGAATTCCTGCAGTGCACAAGCGGTGGAGCGCT	To construct plasmid pJQ-dipD
kodipD-UR	TCTTACGTGCCGATCAACGTCTCAAGCGGATCAGCAGATCTGTGTG	To construct plasmid pJQ-dipD
kodipD-Cm-F	CACACAGATCTGCTGATCCGCTTGAGACGTTGATCGGCACGTAAGA	To construct plasmid pJQ-dipD
kodipD-Cm-R	TTCGTCCGACACAATGTGCAGGTGCGTTTAAGGGCACCAATAACTG	To construct plasmid pJQ-dipD
kodipD-DF	CAGTTATTGGTGCCCTTAAACGCACCTGCACATTGTGTCGGACGAA	To construct plasmid pJQ-dipD
kodipD-DR	AGGGAACAAAAGCTGGAGCTCTACGATCAACATTGGCATCCA	To construct plasmid pJQ-dipD
kodipEFG-UF	AGCTTGATATCGAATTCCTGCAGAGACCTTTCCGCAGGATGATGCGCAGC	To construct plasmid pJQ-dipEFG
kodipEFG-UR	TACGTGCCGATCAACGTCTCGATGCACCAGCACACGGCTG	To construct plasmid pJQ-dipEFG
kodipEFG-Cm-F	CAGCCGTGTGCTGGTGCATCGAGACGTTGATCGGCACGTA	To construct plasmid pJQ-dipEFG
kodipEFG-Cm-R	ACCTGCGCCCTGACCTGCTGCGTGTCAGCGTTTAAGGGCACCAATAACTGCCT	To construct plasmid pJQ-dipEFG
kodipEFG-DF	AGGCAGTTATTGGTGCCCTTAAACGCTGACACGCAGCAGGTCAGGGCGCAGGT	To construct plasmid pJQ-dipEFG
kodipEFG-DR	AGGGAACAAAAGCTGGAGCTCACGAAACGGAATTCGCACAAGCGGTCT	To construct plasmid pJQ-dipEFG
kodipJKL-UF	AGCTTGATATCGAATTCCTGCAGAGTATCTGAACATCTTGTTCGAGGA	To construct plasmid pJQ-dipJKL
kodipJKL-UR	CCTTCCGTAATTCGGCACTTGCAGAAGACCTCCTGCTTGCTGCCAATG	To construct plasmid pJQ-dipJKL
kodipJKL-DF	CATTGGCAGCAAGCAGGAGGTCTTCTGCAAGTGCCGAATTACGGAAGG	To construct plasmid pJQ-dipJKL
kodipJKL-DR	AGGGAACAAAAGCTGGAGCTCGCCATGCTGATAACGGCGATAC	To construct plasmid pJQ-dipJKL
dipD-F	TCGATAAGCTTGATATCGAATTCAAACAGGCAGGGTTATTCCCTAC	To construct plasmid MCS-dipD
dipD-R	CTCACTATAGGGCGAATTGGAGCTCTTACGCTGCTTCCAATTGCAGATCCTGCA	To construct plasmid MCS-dipD
dipEFG-F	ATCGATAAGCTTGATATCGAATTCGTCGGCCAGCCGGGCGGGCAGATA	To construct plasmid MCS-dipEFG
dipEFG-R	CTCACTATAGGGCGAATTGGAGCTCCTAAGCAAGAGAAGGCTCTGCCGA	To construct plasmid MCS-dipEFG
dipDEF-F	ATCGATAAGCTTGATATCGAATTCAACAGGCAGGGTTATTCCCTACTT	To construct plasmid MCS-dipDEF
dipDEF-R	CTCACTATAGGGCGAATTGGAGCTCCTATTTCTTGATGTCCGCCTGAACCA	To construct plasmid MCS-dipDEF
dipG-F	TATCGATAAGCTTGATATCGAATTCTGGCATCAGGAGGACGGGTCATGGC	To construct plasmid MCS-dipG
dipG-R	CTCACTATAGGGCGAATTGGAGCTCCTAAGCAAGAGAAGGCTCTGCCGA	To construct plasmid MCS-dipG
dipFG-F	TATCGATAAGCTTGATATCGAATTCATTTTATTAACCGGAAATACTGGGAAG	To construct plasmid MCS-dipFG
dipFG-R	CTCACTATAGGGCGAATTGGAGCTCCTAAGCAAGAGAAGGCTCTGCCGA	To construct plasmid MCS-dipFG
dipJKL-F	TATCGATAAGCTTGATATCGAATTCTGTTGTGGGGAAAGTGCAATTTTTAAC	To construct plasmid MCS-dipJKL
dipJKL-R	CTCACTATAGGGCGAATTGGAGCTCTTACCAGCTCACCGTCACGTCACCGC	To construct plasmid MCS-dipJKL

TABLE 3.

Strains and plasmids used in this study

Strain or plasmid	Description	Source or reference
Strains
Alcaligenes faecalis
JQ135	Strr; DPA-degrading strain; wild type	CCTCC M 2015812
D3347	Strr; mutant of JQ135; dipD inactivated by transposon	This study
JQ135-idipR	Strr Gmr; mutant of JQ135; dipR inactivated	This study
JQ135-idipRD	Strr Gmr Cmr; mutant of JQ135; dipR and dipD inactivated	This study
JQ135-idipREFG	Strr Gmr Cmr; mutant of JQ135; dipR, dipE, dipF, and dipG inactivated	This study
JQ135-idipJKL	Strr; mutant of JQ135; dipJ, dipK and dipL inactivated	This study
JQ135-idipRJKL	Strr Gmr; mutant of JQ135; dipR, dipJ, dipK, and dipL inactivated	This study
JQ135-idipRD/dipD	Strr Gmr Cmr Kmr; JQ135-idipRD containing the plasmid MCS-dipD	This study
JQ135-idipREFG/dipEFG	Strr Gmr Cmr Kmr; JQ135-idipREFG containing the plasmid MCS-dipEFG	This study
JQ135-idipREFG/dipDEF	Strr Gmr Cmr Kmr; JQ135-idipREFG containing the plasmid MCS-dipDEF	This study
JQ135-idipREFG/dipG	Strr Gmr Cmr Kmr; JQ135-idipREFG containing the plasmid MCS-dipG	This study
JQ135-idipREFG/dipFG	Strr Gmr Cmr Kmr; JQ135-idipREFG containing the plasmid MCS-dipFG	This study
JQ135-idipJKL/dipJKL	Strr Kmr; JQ135-idipJKL containing the plasmid MCS-dipJKL	This study
JQ135-idipRJKL/dipJKL	Strr Gmr Kmr; JQ135-idipRJKL containing the plasmid MCS-dipJKL	This study
Achromobacter sp.
MY14	Strr; newly isolated DPA-degrading strain	This study
Bordetella petrii
MY10	Strr Sper Ampr; a nonpathogenic strain with DPA-degrading capacity, which is obtained from public culture collection center	MCCC 1A16274
Escherichia coli
DH5αλpir	Donor strain for triparental mating	Lab stored
HB101(pRK2013)	Kmr; conjugation helper strain	Lab stored
Plasmids
pJQ200SK	Gmr; mob+ orip15A lacZα+ sacB; suicide plasmid	Lab stored
pSC123	Cmr Kmr; suicide plasmid, mariner transposon	Lab stored
pBBR1MCS-2	Kmr; Broad-host-range cloning vector	Lab stored
pJQ-dipR	Gmr; plasmid for gene dipR inactivation	This study
pJQ-dipD	Gmr Cmr; plasmid for gene dipD knocking out	This study
pJQ-dipEFG	Gmr Cmr; plasmid for genes dipE, dipF, and dipG knocking out	This study
pJQ-dipJKL	Gmr; plasmid for genes dipJ, dipK, and dipL knocking out	This study
MCS-dipD	Kmr; pBBR1MCS-2 containing gene dipD	This study
MCS-dipEFG	Kmr; pBBR1MCS-2 containing genes dipE, dipF, and dipG	This study
MCS-dipDEF	Kmr; pBBR1MCS-2 containing genes dipD, dipE, and dipF	This study
MCS-dipG	Kmr; pBBR1MCS-2 containing gene dipG	This study
MCS-dipFG	Kmr; pBBR1MCS-2 containing genes dipF and dipG	This study
MCS-dipJKL	Kmr; pBBR1MCS-2 containing genes dipJ, dipK, and dipL	This study

Transposon mutagenesis and gene cloning.

Plasmid pSC123 was used to construct the transposon mutant library of strain JQ135 according to previous methods (53). Conjugated cultures were spread on a BM plate containing 5 mM citrate and 50 mg L⁻¹ kanamycin. The colonies were screened by replica plating. Mutants that grew on citrate plates but not on DPA plates were selected. SEFA-PCR was used for DNA walking to clone the genes inactivated by the transposon (54).

RT-qPCR.

The wild-type strain JQ135 and the mutant JQ135-idipR were cultured in LB, and the cells were collected and washed with fresh BM. The cells were added to the BM at a cell OD₆₀₀ of 0.5 and supplied with appropriate carbon sources. The JQ135 incubated with citrate was set as the negative control. RNA was extracted using an RNA extraction kit (MiniBEST; TaKaRa, Dalian, China). Elimination of residual genomic DNA and reverse transcription of total RNA were performed according to the kit instructions. The cDNA obtained by reverse transcription was diluted 100 times for RT-qPCR, using the 16S rRNA gene as an internal reference. RT-qPCR was carried out using a fluorescent quantitative PCR kit (SYBR Premix Ex Taq; TaKaRa) (55). The primers are listed in Table 2.

Gene inactivation and complementation.

Gene dipR was inactivated by insertional mutagenesis in the mutants JQ135-idipR, JQ135-idipRD, JQ135-idipREFG, and JQ135-RJKL. The gene dipD and the fragment dipEFG were replaced by the chloramphenicol-resistant gene in the mutants JQ135-idipRD and JQ135-idipREFG, respectively. The fragment dipJKL was deleted in the mutants JQ135-idipJKL and JQ135-idipRJKL. The DNA fragments used for gene knockout were amplified using the primers listed in Table 2 (49). Plasmid pJQ200SK was digested by PstI/SacI. The DNA fragments and digested plasmid were fused by using a ClonExpress MultiS one-step cloning kit (Vazyme Biotech Co., Ltd., Nanjing, China). The recombined plasmids were transformed into recipient strains by triparental mating. The double crossover mutants were screened on the LB plates containing 20% (wt/vol) sucrose. The recombinant mutants were confirmed by PCR and DNA sequencing. The plasmid pBBR1MCS-2 was used for gene complementation. The DNA fragments for gene complementation were amplified using the primers listed in Table 2 (56, 57).

Degradation assays of DPA and 3HDPA by strain JQ135 and its mutants.

The bacteria were cultured in 20 mL of LB overnight with the required antibiotics. The cells were harvested by centrifugation, washed twice with BM, and resuspended in fresh BM to prepare the resting cells. The resting cells were inoculated into 20 mL of fresh BM containing 0.6 mM concentrations of substrates at a final cell OD₆₀₀ of 0.5. The degradation assays were performed at 30°C in a rotary shaker at 200 rpm. Samples were taken at regular intervals and measured by high-performance liquid chromatography (HPLC).

Identification of metabolic intermediates by ¹⁸O₂ stable isotope labeling experiment.

The ¹⁸O₂ stable isotope labeling experiment was performed in a 1-L sealed bottle. Oxygen in the bottle was removed with high-purity nitrogen gas. Then, approximately 210 mL of ¹⁸O₂ gas and 20 mL of mutant JQ135-idipREFG/dipG resting cells were injected into the bottle. The bottle was incubated in a rotary shaker at 200 rpm and 30°C. After 2 h, the degradation mixture was harvested by centrifugation, and the cell suspension was freeze-dried by using a vacuum freeze dryer. The dried product was dissolved in 2 mL of methanol, followed by detection using LC-MS. The LC-MS was performed on an LC-20AD high-performance liquid chromatograph (Shimadzu, Japan) with a Phenomenex Kinetex C₁₈ column (2.6 mm, 2.1 mm by 100 mm). The degradation mixture using normal oxygen gas was set as a blank control, and its biotransformation product was processed as described above.

Analytical methods.

The DPA and its catabolic intermediates were determined by HPLC on a Shimadzu AD20 system equipped with a Phecda C₁₈ reversed-phase column (250 mm × 4.60 mm × 5 μm) with array detection at 270 nm for DPA and 300 nm for 3HDPA. The mobile phase consisted of 12.5% methanol, 0.5% acetic acid, and 87% H₂O (vol/vol/vol). The flow rate was 1 mL min⁻¹ at 40°C. All assays were performed in triplicate, and the means and standard errors for all measurements were calculated. LC-MS was performed as described previously (12).

Data availability.

The genome sequences of A. faecalis JQ135, Achromobacter sp. MY14, and Bordetella petrii MY10 have been deposited in the GenBank database under accession numbers CP021641, JAJMLY000000000, and JAJMLX000000000, respectively.