A Novel Degradation Mechanism for Pyridine Derivatives in Alcaligenes faecalis JQ135

Unlike the benzene ring, the uneven distribution of the electron density of the pyridine ring influences the positional reactivity and interaction with enzymes; e.g., the ortho and para oxidations are more difficult than the meta oxidations. Hydroxylation is an important oxidation process for the pyridine derivative metabolism. In previous reports, the ortho hydroxylations of pyridine derivatives were catalyzed by multicomponent molybdenum-containing monooxygenases, while the meta hydroxylations were catalyzed by monocomponent FAD-dependent monooxygenases. This study identified the new monocomponent FAD-dependent monooxygenase HpaM that catalyzed the ortho decarboxylative hydroxylation of 5HPA. In addition, we found that the maiA gene coding for maleic acid cis-trans isomerase was pivotal for the metabolism of 5HPA, nicotinic acid, and picolinic acid in A. faecalis JQ135. This study provides novel insights into the microbial metabolism of pyridine derivatives.

ABSTRACT

5-Hydroxypicolinic acid (5HPA), a natural pyridine derivative, is microbially degraded in the environment. However, the physiological, biochemical, and genetic foundations of 5HPA metabolism remain unknown. In this study, an operon (hpa), responsible for 5HPA degradation, was cloned from Alcaligenes faecalis JQ135. HpaM was a monocomponent flavin adenine dinucleotide (FAD)-dependent monooxygenase and shared low identity (only 28 to 31%) with reported monooxygenases. HpaM catalyzed the ortho decarboxylative hydroxylation of 5HPA, generating 2,5-dihydroxypyridine (2,5DHP). The monooxygenase activity of HpaM was FAD and NADH dependent. The apparent K_m values of HpaM for 5HPA and NADH were 45.4 μM and 37.8 μM, respectively. The genes hpaX, hpaD, and hpaF were found to encode 2,5DHP dioxygenase, N-formylmaleamic acid deformylase, and maleamate amidohydrolase, respectively; however, the three genes were not essential for 5HPA degradation in A. faecalis JQ135. Furthermore, the gene maiA, which encodes a maleic acid cis-trans isomerase, was essential for the metabolism of 5HPA, nicotinic acid, and picolinic acid in A. faecalis JQ135, indicating that it might be a key gene in the metabolism of pyridine derivatives. The genes and proteins identified in this study showed a novel degradation mechanism of pyridine derivatives.

IMPORTANCE Unlike the benzene ring, the uneven distribution of the electron density of the pyridine ring influences the positional reactivity and interaction with enzymes; e.g., the ortho and para oxidations are more difficult than the meta oxidations. Hydroxylation is an important oxidation process for the pyridine derivative metabolism. In previous reports, the ortho hydroxylations of pyridine derivatives were catalyzed by multicomponent molybdenum-containing monooxygenases, while the meta hydroxylations were catalyzed by monocomponent FAD-dependent monooxygenases. This study identified the new monocomponent FAD-dependent monooxygenase HpaM that catalyzed the ortho decarboxylative hydroxylation of 5HPA. In addition, we found that the maiA gene coding for maleic acid cis-trans isomerase was pivotal for the metabolism of 5HPA, nicotinic acid, and picolinic acid in A. faecalis JQ135. This study provides novel insights into the microbial metabolism of pyridine derivatives.

INTRODUCTION

Pyridine and its derivatives (such as nicotine, nicotinic acid [NA], picolinic acid [PA], vitamin B₆, etc.) are among the most abundant natural N-heterocyclic compounds. Furthermore, artificial pyridine derivatives are widely used in agriculture and pharmaceutical and chemical industries as solvents, dyes, pharmaceuticals, herbicides, and pesticides (1–3). In addition, many of the pyridine derivatives are also toxic and carcinogenic (4, 5). Therefore, the biodegradation, detoxification, and transformation of pyridine derivatives are of significant interest. Unlike the benzene ring, the pyridine ring contains an N atom with stronger electronegativity than that of the C atom. The electron cloud shifted to the N atom, and the cloud density of the C atom on the pyridine ring is thus significantly reduced, especially at the ortho and para positions. In turn, the electrophilic substitution reaction (oxidation reaction) is more difficult to occur in the ortho and para positions than in the meta position.

Hydroxylation increases the hydrophilicity and polarity of the substrate and is thus typically the initial and key degradation step for pyridine derivatives. The hydroxylation could occur in ortho, meta, and para positions of the pyridine ring. Generally, the position and substituted group affect the ring hydroxylation of pyridine derivatives. As previously shown, hydroxylations that occur at the ortho position of the pyridine ring are catalyzed via multicomponent molybdenum-containing monooxygenases, e.g., nicotinate hydroxylase, 3-succinoyl-pyridine monooxygenase, and nicotine hydroxylase (6–8). On the other hand, the hydroxylations at the meta position are catalyzed via monocomponent flavin adenine dinucleotide (FAD)-dependent monooxygenases, e.g., 6-hydroxynicotinate 3-monooxygenase (NicC), 2,6-dihydroxypyridine 3-monooxygenase (2,6-DHPH), and 6-hydroxy-3-succinoyl-pyridine 3-monooxygenase (HspB) (6, 9, 10). However, a monocomponent FAD-dependent monooxygenase that catalyzes the ortho hydroxylation of the pyridine ring and its coding gene have not been identified thus far.

5-Hydroxypicolinic acid (5HPA) is a typical natural pyridine derivative with an ortho-substituted carboxyl group and a meta hydroxy group in the pyridine. 5HPA is produced either by the marine bacterium Nocardia sp. (11) or by microalgae (12). Recently, 5HPA was also found as a biologically active compound in several traditional Chinese medicines, including Hericium erinaceus, Gynura divaricata, and purslane (13–15). The microbial degradation or transformation of 5HPA has not been extensively studied to date. Previously, Karvelis et al. isolated a 5HPA-degrading strain, Pusillimonas sp. strain 5HP, from soil (16). In this strain, the degradation of 5HPA was found to be started by ortho decarboxylative hydroxylation, producing 2,5-dihydroxypyridine (2,5DHP). A 5-hydroxypicolinic acid 2-monooxygenase was partially purified from Pusillimonas sp. strain 5HP. However, both the amino acid sequence of the 5-hydroxypicolinic acid 2-monooxygenase and its coding genes remain unknown. In this study, we cloned an operon (hpa) responsible for 5HPA degradation in Alcaligenes faecalis JQ135, a picolinic acid-degrading bacterium reported previously (17). HpaM (5-hydroxypicolinic acid 2-monooxygenase) was identified as a monocomponent monooxygenase catalyzing the ortho decarboxylative hydroxylation of 5HPA (Fig. 1). In addition, the functions of the hpaDXF and maiA genes in the metabolism of 5HPA were investigated.

FIG 1.

Metabolism of 5HPA in A. faecalis JQ135. (A) Proposed 5HPA degradation pathway in A. faecalis JQ135. HpaM, 5-hydroxypicolinic acid 2-monooxygenase; HpaX, 2,5-dihydroxypyridine 5,6-dioxygenase; HpaD, N-formylmaleamic acid deformylase; HpaF, maleamic acid amidohydrolase; MaiA, maleic acid cis-trans isomerase; TCA, tricarboxylic acid cycle. (B) Organization of the hpa operon. Numbers above each gene represent the locus tags for each gene. The numbered horizontal lines below the hpa operon show the location and size of the PCR fragments in panel C. (C) Agarose gel electrophoresis of RT-PCR products. Plus and minus indicate that the cells were cultured with 5HPA and glycerol, respectively.

RESULTS

Degradation of 5HPA, NA, and nicotine by A. faecalis JQ135.

A. faecalis JQ135 was tested for its ability to degrade 5HPA, NA, and nicotine in mineral salts medium (MSM) supplemented with each substrate as sole carbon and nitrogen sources. The results showed that A. faecalis JQ135 completely degraded 1.0 mM 5HPA and NA but not nicotine within 8 h; correspondingly, the optical density at 600 nm (OD₆₀₀) of the culture increased from approximately 0.3 to 0.5 (see Fig. S1 in the supplemental material). These results indicated that A. faecalis JQ135 could degrade and utilize 5HPA and NA, but not nicotine, as sole carbon and nitrogen sources for growth. Moreover, attempts to detect metabolic intermediates of 5HPA within the culture failed, which might be due to little or no intermediate excretion from strain JQ135 cells during the degradation of 5HPA.

Transposon mutagenesis and cloning of an operon involved in 5HPA degradation.

To clone the genes involved in 5HPA degradation, a transposon mutant library of A. faecalis JQ135 was constructed. One mutant (Z10) that could not grow in MSM agar containing 1.0 mM 5HPA was identified from approximately 5,000 mutants (Fig. S2). The genome of A. faecalis JQ135 was determined via PacBio sequencing. The complete genome of the strain contained one circular chromosome (4,078,346 bp), and no plasmid was found. A total of 3,723 open reading frames (ORFs) were predicted. The insertion position of the transposon, determined via DNA walking (18), was located in gene AFA_18575 (genome position 4070825). AFA_18575 (named hpaM) is 1,218 bp in length and has a G+C content of 56.08%. The predicted protein was used to search the NCBI database (http://blast.ncbi.nlm.nih.gov/) with BLASTP (Table 1). HpaM shared low identities with reported flavin-containing monooxygenases, e.g., 28% identity with salicylate 1-monooxygenase (SalM) from Pseudomonas putida KF715 (19) and 31% identity with NicC from P. putida KT2440 (6) (Fig. S3 and S4). NicC and SalM catalyze the decarboxylative hydroxylation of 6-hydroxynicotinic acid (6HNA) and salicylate, respectively. Downstream of hpaM, a gene (AFA_18580) coding for the DUF2236 family enzyme and a TetR-type regulator gene (AFA_18585) were predicted (Fig. 1; Table 1). Upstream of hpaM, eight genes were identified. AFA_18535, AFA_18540, and AFA_18545 were predicted to be branched-chain amino acid ABC transporter substrate-binding proteins, while AFA_18550 and AFA_18555 were predicted to be branched-chain amino acid ABC transporter ATP-binding proteins. Based on the above analysis, the five consecutive genes putatively encoded an ABC-type transporter. Genes hpaD, hpaX, and hpaF showed 57%, 53%, and 41% identities, respectively, with NicD (N-formylmaleamic acid deformylase), NicX (2,5DHP dioxygenase), and NicF (maleamate amidohydrolase) from P. putida KT2440 at the amino acid level (Fig. S3). Interestingly, NicC, NicX, NicD, and NicF formed a complete metabolic pathway for the transformation of 6HNA into maleic acid (6). Considering the structural similarity between 5HPA and 6HNA, we presumed that these contiguous genes were organized in an operon and were involved in the degradation of 5HPA in A. faecalis JQ135. In this study, this operon was designated hpa (5-hydroxypicolinic acid).

TABLE 1.

Sequence comparison of the hpa operon of A. faecalis JQ135 with database entries

Gene locus	Product sizea	Databaseb	Homologous proteinc	Accession no.	Identity (%)	Proposed function
AFA_18535	369	NR	Branched-chain amino acid ABC transporter, substrate-binding protein LivK	CUI69837	99	ATP binding and 5-hydroxypicolinic acid transportation
		Swiss-Prot	Leu/Ile/Val/Thr/Ala-binding protein	P21175	30
AFA_18540	291	NR	Branched-chain amino acid ABC transporter, substrate-binding protein LivH	WP_035269511	100
		Swiss-Prot	High-affinity branched-chain amino acid transport system permease protein LivH	P0AEX8	35
AFA_18545	319	NR	Branched-chain amino acid ABC transporter, substrate-binding protein LivM	WP_083055103	99
		Swiss-Prot	High-affinity branched-chain amino acid transport system permease protein LivM	P22729	30
AFA_18550	252	NR	Branched-chain amino acid ABC transporter, ATP-binding protein LivG	WP_042485129	100
		Swiss-Prot	High-affinity branched-chain amino acid transport system permease protein LivG	P0A9S8	41
AFA_18555	242	NR	Branched-chain amino acid ABC transporter, ATP-binding protein LivF	WP_035269507	100
		Swiss-Prot	High-affinity branched-chain amino acid transport system permease protein LivF	P0A191	46
AFA_18560 (hpaD)	275	NR	Alpha/beta hydrolase	WP_042485331	98	N-Formylmaleamic acid deformylase HpaD
		Swiss-Prot	N-Formylmaleamate deformylase NicD	Q88FY3	57
AFA_18565 (hpaX)	346	NR	2,5-Dihydroxypyridine 5,6-dioxygenase	WP_083055106	98	2,5-Dihydroxypyridine 5,6-dioxygenase HpaX
		Swiss-Prot	2,5-Dihydroxypyridine 5,6-dioxygenase NicX	Q88FY1	53
AFA_18570 (hpaF)	205	NR	N-Carbamoylsarcosine amidase	WP_060186798	100	Maleamic acid amidohydrolase HpaF
		Swiss-Prot	Maleamate amidohydrolase NicF	Q88FY5	41
AFA_18575 (hpaM)	405	NR	Salicylate 1-monooxygenase	ARP55611	100	5-Hydroxypicolinic acid 2-monooxygenase HpaM
		Swiss-Prot	6-Hydroxynicotinate 3-monooxygenase NicC	Q88FY2	35
AFA_18580	291	NR	DUF2236 domain-containing protein	WP_086060754	99	Unknown function
		Swiss-Prot	Ubiquitin-activating enzyme E1C	Q99MI7	25
AFA_18585	235	NR	TetR family transcriptional regulator	WP_035269502	100	Transcriptional regulator
		Swiss-Prot	HTH type transcriptional regulator BetI	Q13NG5	41

^aNumber of amino acids.

^bNR, NCBI nonredundant protein sequence database.

^cThe top BLASTP hit was selected.

Heterologous expression and purification of HpaM and function determination.

The gene hpaM was cloned into pET29a(+) and expressed in Escherichia coli BL21(DE3). The N-terminal 6×His-tagged HpaM was purified using nickel affinity chromatography. The purified enzyme migrated as a single band, which was in agreement with its theoretical value (45.2 kDa) (Fig. S5). The result of gel filtration chromatography indicated that HpaM was a dimer (Fig. S6).

HpaM was predicted to contain FAD binding domains (such as GXGXXG) (Fig. S4). However, purified HpaM was colorless, suggesting that no flavin was associated with the enzyme after purification. In addition, the UV-visible spectrum of HpaM also did not show the typical absorption of a FAD cofactor at 380 nm to 480 nm (Fig. S7). Purified HpaM showed enzymatic activity only after the addition of external FAD (but not flavin mononucleotide [FMN]). NADH (but not NADPH) was used as an electron donor, and no activity was detected under anaerobic conditions. These results indicate that HpaM was FAD, NADH, and O₂ dependent, which resembles 5-hydroxypicolinic acid 2-monooxygenase purified from Pusillimonas sp. strain 5HP (16). High-performance liquid chromatography (HPLC) results showed that 5HPA (with a retention time of 5.52 min) decreased and a product (with a retention time of 5.22 min) accumulated (Fig. S8). The retention time of the product was equal to that of the authentic 2,5DHP standard, and liquid chromatography-mass spectrometry (LC-MS) analysis showed that the product had a molecular ion at m/z 112.0400 [M+H]⁺, which was identical to that of 2,5DHP (16, 20, 21). Thus, the product was identified as 2,5DHP. Based on the above analysis, HpaM was identified as a monocomponent FAD-dependent monooxygenase catalyzing the ortho decarboxylative hydroxylation of 5HPA producing 2,5DHP.

Analysis of kinetic constants and biochemical properties.

The K_m and k_cat values of HpaM for 5HPA were 45.4 ± 4.2 μM and 10.2 ± 0.3 s⁻¹, respectively (Fig. S9). The apparent catalytic efficiency (k_cat/K_m) of HpaM was 225.5 s⁻¹ mM⁻¹, and the apparent K_m value of HpaM for NADH was 37.8 ± 3.4 μM. The optimal pH of HpaM was 7.0; when the pH exceeded 8.0, the enzyme activity decreased dramatically (Fig. S10).

The following structural analogs of 5HPA were tested as the substrates in a standard activity test (Table S1): PA, NA, 5-aminopicolinic acid, 5-methylpicolinic acid, 5-chloropicolinic acid, 5-bromoopicolinic acid, 3-hydroxypicolinic acid, 6-hydroxypicolinic acid, 2,3-pyridinedicarboxylic acid, 2,5-pyridinedicarboxylic acid, 2,6-pyridinedicarboxylic acid, 2-hydroxynicotinic acid, 4-hydroxynicotinic acid, 5-hydroxynicotinic acid, 6-hydroxynicotinic acid, 3-hydroxyisonicotinic, and 6-hydroxy-3-succinoylpyridine (HSP). HpaM showed no activity against any of these compounds, indicating that the hydroxylase activity was 5HPA specific.

Functional study of the genes hpaD, hpaX, and hpaF.

As mentioned above, HpaD, HpaX, and HpaF shared the highest homology with NicD, NicX, and NicF from P. putida KT2440, respectively, suggesting that HpaX, HpaD, and HpaF convert 2,5DHP into maleic acid. To study the function of hpaX, the gene was cloned into pET29a(+) and expressed in E. coli BL21(DE3) (Fig. S11). HPLC results showed that 2,5DHP was degraded by purified HpaX (Fig. S12). When Fe²⁺ was absent or under anaerobic conditions, no 2,5DHP dioxygenase activity was detected. These characteristics were similar to those of previously reported Fe²⁺- and O₂-dependent 2,5DHP dioxygenases (6, 16, 20, 22). HpaX degraded 2,5DHP with K_m and k_cat values of 77.9 ± 6.2 μM and 11.6 ± 0.3 s⁻¹, respectively (Fig. S13).

The activities of HpaD and HpaF were more difficult to determine. To investigate the function of hpaD and hpaF, a DNA fragment containing the hpaDXFM genes plus a maleic acid cis-trans isomerase gene (maiA) was cloned into the 5HPA-nondegrading bacterium Sphingomonas wittichii DC-6 (23). The reason for inclusion of the maiA gene was that the MaiA converts maleic acid to fumaric acid (6, 22) and the hpa operon lacks the maiA gene. Bioinformatic analysis showed that the gene AFA_16520, which was physically distant from the hpa operon, encoded a putative protein with 99% identity to a previously reported MaiA from A. faecalis IFO13111 (24). Thus, AFA_16520 (maiA) was predicted to convert maleic acid to fumaric acid in A. faecalis JQ135. The whole-cell transformation experiments showed that recombinant DC-6/pBBR-hpaDXFM-maiA acquired the ability to degrade and utilize 5HPA for growth while DC-6/pBBR-hpaDXFM and DC-6/pBBR-maiA did not (Fig. S14). These results suggest that hpaD, hpaX, and hpaF were responsible for the conversion of 2,5DHP to maleic acid.

Deletion of the hpa genes and maiA.

To determine the physiological functions of hpaD, hpaX, hpaF, and hpaM in vivo, the in-frame deletion mutants JQ135 ΔhpaD, JQ135 ΔhpaX, JQ135 ΔhpaF, and JQ135 ΔhpaM, respectively, were constructed. These four mutants could still grow on MSM plates supplemented with glycerol as the carbon source, indicating that these four genes did not affect the basic metabolism (Fig. 2). However, mutant JQ135 ΔhpaM lost the ability to utilize 5HPA for growth, while the complemented strain JQ135 ΔhpaM/pBBR-hpaM recovered the ability to grow on 5HPA. The result clearly demonstrated that hpaM was essential for the catabolism of 5HPA in A. faecalis JQ135. Furthermore, mutant JQ135 ΔhpaM could grow on NA and PA, indicating that hpaM was not involved in the degradation of NA or PA in A. faecalis JQ135. Interestingly, the mutants JQ135 ΔhpaD, JQ135 ΔhpaX, and JQ135 ΔhpaF could still grow on 5HPA, which might be because the other isoenzymes of HpaX, HpaD, and HpaF were present in A. faecalis JQ135.

FIG 2.

Growth phenotype of wild-type A. faecalis JQ135 and mutant derivatives on MSM plates with 1.0 mM glycerol, 5HPA, PA, and NA as sole carbon sources.

To investigate the function of maiA, this gene was also deleted from A. faecalis JQ135. The mutant JQ135 ΔmaiA could still degrade 5HPA, NA, and PA but lost the ability to utilize the three substrates for growth (Fig. 2). These results indicated maiA to be essential for 5HPA, NA, and PA metabolism in A. faecalis JQ135.

Transcription of hpa genes in response to 5HPA, NA and PA.

Reverse transcription-PCR (RT-PCR) was conducted to investigate whether the transcription of hpa genes was induced by 5HPA. The results showed that the five putative transporter genes (AFA_18535 to AFA_18555), hpaD, hpaX, hpaF, and hpaM comprised a transcriptional operon and were transcribed in cells grown on 5HPA but not on glycerol (Fig. 1C). RT-PCR results showed that hpaM was not transcribed in cells grown on NA or PA (Fig. 3). Moreover, in mutant JQ135 ΔhpaM, the transcription of the hpa operon was still induced by 5HPA, suggesting that 5HPA but not its metabolites was the inducer.

FIG 3.

Agarose gel electrophoresis of RT-PCR products generated using RNA of A. faecalis JQ135 or JQ135 ΔhpaM. Lane M, ladder DNA marker; lanes 1, 2, 3, and 4, RT-PCR products of A. faecalis JQ135 grown in MSM supplemented with 1.0 mM glycerol, 5HPA, PA, and NA, respectively; lane 5, RT-PCR products of JQ135 ΔhpaM incubated in MSM supplemented with 1.0 mM 5HPA.

DISCUSSION

5HPA is a natural pyridine derivative produced by many marine bacteria, microalgae, and plants. The microbial metabolism of 5HPA has been studied only preliminarily in Pusillimonas sp. 5HP (16). In this study, we cloned the operon hpa from strain A. faecalis JQ135, which was involved in the metabolism of 5HPA. 5HPA was initially decarboxylatively hydroxylated at the ortho position of the pyridine ring by the monocomponent FAD-dependent monooxygenase HpaM, yielding 2,5DHP. Furthermore, 2,5DHP was further decomposed by HpaX, HpaD, HpaF, and MaiA. Thus, this study provided a physiological and genetic base of the microbial metabolism of 5HPA.

Flavin-dependent monooxygenases constitute an important class of enzymes that utilize a reduced flavin to oxidize their substrates. Flavin-dependent monooxygenases catalyze a variety of reactions, such as hydroxylation, Baeyer-Villiger oxidation, sulfoxidation, halogenation, and epoxidation reactions, and play an important role in the metabolism of natural and xenobiotic compounds (25). So far, a number of flavin-dependent monooxygenases catalyzing the hydroxylation of aromatic compounds have been identified. Most of them catalyzed the hydroxylation of the substituted/unsubstituted carbon atom in the benzene ring (26–28). Only three of them (NicC, 2,6-DHPH, and HspB) catalyzed the hydroxylation of the pyridine ring, and their hydroxylation happened at the meta position (6, 10, 29). To the best of our knowledge, HpaM is the only reported monocomponent FAD-dependent monooxygenase that catalyzes the ortho decarboxylative hydroxylation of the pyridine ring. Based on their structural features, protein sequence motifs, electron donor, and type of oxygenation reaction, flavin-dependent monooxygenases can be divided into eight groups (A to H) (25). Group A flavin monooxygenases are monocomponent, contain a glutathione reductase (GR-2) type Rossmann fold for FAD binding, and use NAD(P)H as the electron donor. HpaM was monocomponent and FAD dependent and used NADH as the electron donor. Sequence alignment and structure prediction results suggested that HpaM has a classical GR-2 type Rossmann fold (GXGXXG) (see Fig. S4 in the supplemental material). A phylogenetic tree of related monocomponent flavin-dependent monooxygenases showed that HpaM was clustered with group A enzymes and formed a subclade with NicC (Fig. 4) (25, 30). These indications suggest that HpaM belongs to group A flavin-dependent monooxygenases. However, HpaM shared very low identities with reported group A flavin-dependent monooxygenases, showed strict substrate specificity for 5HPA, and did not hydroxylate PA, NA, 6HNA, or HSP. Strict substrate specificity appears to be a common feature for flavin monooxygenases in group A, which may be due to the diversity of substrate binding sites among these enzymes; e.g., the 6HNA binding sites His47, Cys202, Met213, and His302 (NicC numbering) in NicC of P. putida KT2440 were not found in HpaM (Fig. S4). The diverse binding sites of group A flavin monooxygenases have led to their adaptation to different substrates. Based on the above analysis, HpaM was a novel group A flavin-dependent monooxygenase.

FIG 4.

Phylogenetic analysis of HpaM. (A) The phylogenetic tree was constructed based on alignment of HpaM with related monocomponent flavin-dependent monooxygenases. Multiple-alignment analysis was performed using ClustalX v2.0. The phylogenetic tree was constructed via the neighbor-joining algorithm using MEGA 6.0, and bootstrap values (based on 1,000 replications) are indicated at branch nodes. Bar, 0.20 substitutions per amino acid position. Each item was arranged in the following order: protein name, UniProtKB/Swiss-Prot accession number, and EC number. (B) Comparison of reactions catalyzed via HpaM (red, ortho hydroxylation) and three monooxygenases (blue, meta hydroxylation), NicC, HspB, and 2,6-DHPH. (C) Five C—C bond cleavage hydroxylation reactions of the benzene ring. SalM, salicylate 1-monooxygenase; pHB1M, p-hydroxybenzoate 1-monooxygenase; 4AB1M, 4-aminobenzoate 1-monooxygenase; HipH, 4-hydroxyisophthalate 3-monooxygenase; PhzS, 5-methylphenazine-1-carboxylate 1-monooxygenase. Proteins in panels B and C are indicated in bold in the phylogenetic tree.

2,5DHP is a common catabolic intermediate of several pyridine derivatives, such as 6HNA (from NA) in P. putida KT2440, HSP (from nicotine) in P. putida S16, 2-hydroxypyridine in Arthrobacter sp., and 3-methyl-pyridine in Pseudomonas sp. KM-3 (31). This study showed that 2,5DHP was also an intermediate of 5HPA degradation in A. faecalis JQ135. The metabolism of 2,5DHP involves four steps: 2,5DHP is initially ring cleaved into N-formylmaleamic acid by a 2,5DHP dioxygenase; N-formylmaleamic acid is subsequently transformed to maleamic acid by N-formylmaleamate deformylase, which is then converted to maleic acid by maleamate amidohydrolase; and maleic acid is finally transformed into fumaric acid, a metabolite of citric acid cycle, by maleic acid cis-trans isomerase (6, 20, 22). Our study demonstrated that the metabolism of 2,5DHP produced from 5HPA degradation also followed the same pathway (Fig. 1A). The degradation of 6HNA, HSP, and 5HPA differed only at the initial steps, which started with three different flavin monooxygenases, NicC, HspB, and HpaM (Fig. S3). However, it is interesting that the JQ135 ΔhpaD, JQ135 ΔhpaX, and JQ135 ΔhpaF mutants could still grow on 5HPA. Bioinformatic analysis revealed that there were two other gene clusters (AFA_14960 to AFA_14970 and AFA_15130 to AFA_15140) that contained homologous genes, sharing 54 to 69% identities with hpaD, hpaX, and hpaF predicted in the genome of A. faecalis JQ135. It is possible that these homologous genes were involved in the metabolism of other pyridine derivatives (such as NA or PA) with 2,5DHP as an intermediate. It is interesting that only one copy of the maleic acid cis-trans isomerase gene maiA (AFA_16520) was found in the genome of A. faecalis JQ135 and that maiA was not located in the two clusters and the hpa operon. In addition, gene deletion results demonstrated that the maiA gene was essential for the degradation of 5HPA, PA, and NA in A. faecalis JQ135. These analyses suggest that MaiA was a pivotal enzyme of the metabolism of the pyridine derivatives in A. faecalis JQ135, i.e., each pyridine derivative was converted to maleic acid, and then MaiA isomerized maleic acid to fumaric acid, which is a metabolite of citric acid cycle. The possible reason for this arrangement is that it is more advantageous for the flexible regulation of the metabolism of various pyridine derivatives and saves genetic resources. Moreover, maleic acid can also be hydrated to d-malic acid by some microorganisms (32–34). However, the inability to utilize 5HPA, NA, and PA of mutant JQ135 ΔmaiA suggests that this d-malic acid shunt pathway does not exist in A. faecalis JQ135.

MATERIALS AND METHODS

Chemicals and media.

5HPA (99%) and its structural analogs were purchased from J&K Scientific Ltd. (Shanghai, China). 2,5DHP (98%) was purchased from SynChem OHG (Altenburg, Germany). All other chemicals and solvents used for this experiment were commercially available. Enzymes used in this study were purchased from Vazyme Biotech Co., Ltd. (Nanjing, China). Mineral salts medium (MSM) and Luria-Bertani medium (LB) have been described previously (17). To prepare nitrogen-free MSM, NH₄NO₃ was omitted.

Bacterial strains, plasmids, and growth conditions.

Bacterial strains and plasmids used in this study are listed in Table 2. A. faecalis JQ135 has previously been identified as a picolinic acid-degrading bacterium (17). E. coli strains were grown at 37°C, while other strains were grown at 30°C. Media were supplemented with chloramphenicol (Cm; 34 μg/ml), kanamycin (Km; 50 μg/ml), gentamicin (Gm; 50 μg/ml), or streptomycin (Str; 50 μg/ml) as required.

TABLE 2.

Strains and plasmids used in this study

Strain or plasmid	Description	Source or reference
Strains
Alcaligenes faecalis
JQ135	Strr; wild type; 5HPA-degrading strain; G−	17
Z10	Strr Kmr; transposon mutant of JQ135; 5HPA growth deficient	This study
JQ135 ΔhpaM	Strr; hpaM deletion mutant of JQ135	This study
JQ135 ΔhpaM/pBBR-hpaM	Strr Gmr; JQ135 ΔhpaM containing pBBR-hpaM	This study
JQ135 ΔhpaD	Strr; hpaD deletion mutant of JQ135	This study
JQ135 ΔhpaX	Strr; hpaX deletion mutant of JQ135	This study
JQ135 ΔhpaF	Strr; hpaF deletion mutant of JQ135	This study
JQ135 ΔmaiA	Strr; maiA deletion mutant of JQ135	This study
JQ135 ΔmaiA/pBBR-maiA	Strr Gmr; JQ135 ΔmaiA containing pBBR-maiA	This study
Sphingomonas wittichii
DC-6	Chloroacetanilide herbicide degrader; cannot degrade 5HPA	21
DC-6/pBBR-hpaDXFM-maiA	DC-6 containing pBBR-hpaDXFM-maiA	This study
DC-6/pBBR-hpaDXFM	DC-6 containing pBBR-hpaDXFM	This study
DC-6/pBBR-maiA	DC-6 containing pBBR-maiA	This study
Escherichia coli
DH5α	F− recA1 endA1 thi-1 hsdR17 supE44 relA1 deoRΔ(lacZYA-argF) U169 ϕ80lacZΔM15	TaKaRa
SM10λpir	Donor strain for biparental mating	40
BL21(DE3)	F− ompT hsdS(rB− mB−) gal dcm lacY1 (DE3)	Novagen
Plasmids
pET29a(+)	Kmr; expression plasmid	Novagen
pSC123	Cmr Kmr; suicide plasmid, mariner transposon	40
pJQ200SK	Gmr; mob+ orip15A lacZα+ sacB; suicide plasmid	41
pBBR1-MCS5	Gmr; broad-host-range cloning plasmid	42
pJQΔhpaM	Gmr; hpaM gene deletion plasmid; the upstream and downstream regions of hpaM gene fused into SacI/PstI-digested pJQ200SK	This study
pJQΔhpaD	Gmr; hpaD gene deletion plasmid; the upstream and downstream regions of hpaD gene fused into SacI/PstI-digested pJQ200SK	This study
pJQΔhpaX	Gmr; hpaX gene deletion plasmid; the upstream and downstream regions of hpaX gene fused into SacI/PstI-digested pJQ200SK	This study
pJQΔhpaF	Gmr; hpaF gene deletion plasmid; the upstream and downstream regions of hpaF gene fused into SacI/PstI-digested pJQ200SK	This study
pJQΔmaiA	Gmr; maiA gene deletion plasmid; the upstream and downstream regions of maiA gene fused into SacI/PstI-digested pJQ200SK	This study
pBBR-hpaM	Gmr; hpaM gene complementation plasmid; the hpaM gene fused into XhoI/HindIII-digested pBBR1-MCS5	This study
pBBR-maiA	Gmr; hpaM gene complementation plasmid; the maiA gene fused into XhoI/HindIII-digested pBBR1-MCS5	This study
pBBR-hpaDXFM	Gmr; hpaDXFM genes fused into XhoI/HindIII-digested pBBR1-MCS5	This study
pBBR-hpaDXFM-maiA	Gmr; hpaDXFM and maiA genes fused into XhoI/HindIII-digested pBBR1-MCS5	This study
pET-hpaM	Kmr; NdeI-XhoI fragment containing hpaM inserted into pET29a(+)	This study
pET-hpaX	Kmr; NdeI-XhoI fragment containing hpaX inserted into pET29a(+)	This study

Isolation of 5HPA growth-deficient mutants and determination of the transposon insertion site.

Both generation and selection of 5HPA growth-deficient mutants of A. faecalis JQ135 were performed according to a selection method previously used to identify picolinic acid (PA) growth-deficient mutants (17). The plasmid pSC123 (carrying the mariner transposon) was used as a transposon delivery plasmid. 5HPA was substituted for PA. The insertion site of the transposon was determined via self-formed adaptor PCR, as previously described (17, 18).

Gene deletion and complementation.

Standard DNA manipulation techniques were performed as previously described (35). All primers used in this study are listed in Table 3. Deletion of the hpaM, hpaD, hpaX, hpaF, and maiA genes in A. faecalis JQ135 was performed via the two-step homogenetic recombination method using the suicide plasmid pJQ200SK. Using the deletion of hpaM as an example, two primer pairs (kohpaM-UF/kohpaM-UR and kohpaM-DF/kohpaM-DR) were used to amplify the homologous recombination-directing sequences. Both PCR fragments were then cloned into SacI/PstI-digested pJQ200SK using the ClonExpress MultiS one-step cloning kit (Vazyme Biotech Co., Ltd., Nanjing, China). The resulting plasmid, pJQ-ΔhpaM, was then introduced into A. faecalis JQ135 cells. Single-crossover mutants were screened on LB plates containing Str and Gm. A double-crossover mutant (JQ135 ΔhpaM) was selected on LB plates containing Str and 10% (wt/vol) sucrose.

TABLE 3.

Primers used in the study

Primer	Sequence (5′–3′)	Description
kohpaM-UF	AGCTTGATATCGAATTCCTGCAGCCATCATCATGACGGTCGAGAACG	To construct plasmid pJQΔhpaM
kohpaM-UR	TCTAGAACTAGTGGATCCGACCGCATTCGGAGTCATTTGAATC
kohpaM-DF	GGATCCACTAGTTCTAGAGGCATCAAGACGGACTGGGTTTA
kohpaM-DR	CTAAAGGGAACAAAAGCTGGAGCTCCAATGTGGTACCGGCATAGCCAG
kohpaD-UF	AGCTTGATATCGAATTCCTGCAGAAACCCAGATTGTTGATGCTGG	To construct plasmid pJQΔhpaD
kohpaD-UR	TCTAGAACTAGTGGATCCGCCATAGCTAAGTCGCGATAGG
kohpaD-DF	GGATCCACTAGTTCTAGATCTTACGTGTACCAGACGCTGG
kohpaD-DR	CTAAAGGGAACAAAAGCTGGAGCTACTTCATTATTGGGGCCAAAGG
kohpaX-UF	AGCTTGATATCGAATTCCTGCAGATGGCGTGTCCATTCTGGTGG	To construct plasmid pJQΔhpaX
kohpaX-UR	TCTAGAACTAGTGGATCCAGACTCAGCACCTCATTCCAGG
kohpaX-DF	GGATCCACTAGTTCTAGAAAAGAGCAAACCCTGGGCATGG
kohpaX-DR	CTAAAGGGAACAAAAGCTGGAGCTCTCGGGCAAGAAACCAACCTTGG
kohpaF-UF	AGCTTGATATCGAATTCCTGCAGTCGGGCACCAAAATTCTCCTGG	To construct plasmid pJQΔhpaF
kohpaF-UR	TCTAGAACTAGTGGATCCCTGTCGCTCGTAAACACTCTGG
kohpaF-DF	GGATCCACTAGTTCTAGATTTGTCCTGGCAGATTGCGTGG
kohpaF-DR	CTAAAGGGAACAAAAGCTGGAGCTCGGTTCTTCCTCCCCAAAGATGG
komaiA-UF	AGCTTGATATCGAATTCCTGCAGGCGAAACGGGTAGGCTTCTATG	To construct plasmid pJQΔmaiA
komaiA-UR	TCTAGAACTAGTGGATCCAGTGGGTCTGTTGAACCCGGAAAAC
komaiA-DF	GGATCCACTAGTTCTAGAGCAACCAGGCAAGCGTAGGCCATCACG
komaiA-DR	TAAAGGGAACAAAAGCTGGAGCTCGACATGATTTTTCAGCAACAC
hpaM-F	GGTACCGGGCCCCCCCTCGAGGTACGTGCTTCCGTCGTGGA	To construct plasmid pBBR-hpaM
hpaM-R	CAGGAATTCGATATCAAGCTTGCAAGCACAATGCAGAAATCC
maiA-F	GGTACCGGGCCCCCCCTCGAGGTACGTGCTTCCGTCGTGGA	To construct plasmid pBBR-maiA
maiA-R	CAGGAATTCGATATCAAGCTTGCAAGCACAATGCAGAAATCC
hpaDXFM-F	CGAGGTCGACGGTATCGATAAGCTTAGCATGGCGTGTCCATTCTGG	To construct plasmid pBBR-hpaDXFM-maiA
hpaDXFM-R	GGCCGATGCGGTAGGTTTTCATTTAAACCTGTAATGCTTTCGGCTG
maiA-F2	CAGCCGAAAGCATTACAGGTTTAAATCTCTGCCACTCCTGCCATCG
maiA-R2	ACTCACTATAGGGCGAATTGGAGCTCTTAGGCTTGTGGCTTGGCACCCGA
exphpaMF	TAAGAAGGAGATATACATATGAAAAAAACGTTGAAAGTAGGTGTGAT	To construct plasmid pET-hpaM
exphpaMR	GTGGTGGTGGTGGTGCTCGAGAACCTGTAATGCTTTCGGCTGA
exphpaXF	TAAGAAGGAGATATACATATGGCTGTATCTGATCATCAAATGGTTCA	To construct plasmid pET-hpaX
exphpaXR	GTGGTGGTGGTGGTGCTCGAGTGCCAACTCCTCTACCACTTTGC
RT1F	GAGATGGTCAAACAGTTCCAGG	To amplify fragment 1 in Fig. 1
RT1R	TACGCTCTTTGACAGCCAACTTGG
RT2F	GATCCGGACAAGATTCGTCTGG	To amplify fragment 2 in Fig. 1
RT2R	TTGTCCCGGATACGAGCAATCAGG
RT3F	AGCTCCATTTATGTCGCGCAGG	To amplify fragment 3 in Fig. 1
RT3R	CCAGCGTCACAATGACAAAGAAGG
RT4F	GCTACTCGCAGTTTGTCTTTGTGG	To amplify fragment 4 in Fig. 1
RT4R	AAACACTCAGAGCCCGCAGCAC
RT5F	TGGCCAGCTCCTTGTCTTGTGG	To amplify fragment 5 in Fig. 1
RT5R	TGTCTTCAACGGTCATTTCAGG
RT6F	TGAAGTGGAGCGGCGTAAAGG	To amplify fragment 6 in Fig. 1
RT6R	ATGGAGTGCCCCACATAAATGG
RT7F	TATCCGGCGAATATCGCCTGG	To amplify fragment 7 in Fig. 1
RT7R	TGTTCGGGAGAGAACAGCAAGG
RT8F	GGCTTACGCCATGTCCCACATAGG	To amplify fragment 8 in Fig. 1
RT8R	CAGTACCAAAGAAGGCCGATGG
RT9F	TATTCGGTCTGAAAGTCCCCGG	To amplify fragment 9 in Fig. 1
RT9R	TACTTCCGAACGCACTCCATCGG
RT10F	GGTGAATGAAACAGCCAGCTGG	To amplify fragment 10 in Fig. 1
RT10R	CGGGTCATTGGCATCGTAAGG
RT11F	TCTGCAATCCCATATGCGCTGG	To amplify fragment 11 in Fig. 1
RT11R	GTAGACGCTCATCGGAAATGG
RT-ΔhpaMR	TAAACCTGTAATGCTTTCGGCTG	To amplify fragment 5 with RT9F in Fig. 3

The plasmid pBBR-hpaM was constructed for gene complementation. hpaM was amplified using the primers hpaM-F and hpaM-R and then ligated to XhoI/HindIII-digested pBBR1-MCS5, thus generating pBBR-hpaM. The pBBR-hpaM plasmid was then transferred into the JQ135 ΔhpaM mutant via biparental mating to generate the complemented strain JQ135 ΔhpaM/pBBR-hpaM. The plasmid pBBR-maiA and complementation strain JQ135 ΔmaiA/pBBR-maiA were constructed in a similar manner.

A 3.9-kbp DNA fragment containing hpaDXFM was amplified and fused into XhoI/HindIII-digested pBBR1-MCS5, thus yielding the plasmid pBBR-hpaDXFM. The 3.9-kbp DNA fragment containing hpaDXFM and a 0.9-kbp DNA fragment containing maiA were amplified and fused into XhoI/HindIII-digested pBBR1-MCS5, thus yielding plasmid pBBR-hpaDXFM-maiA. These plasmids were transferred into Sphingomonas wittichii DC-6 by biparental mating (36), generating several recombinant strains (Table 2).

RNA extraction and RT-PCR.

The A. faecalis JQ135 wild-type and JQ135 ΔhpaM mutant strains were cultured in glycerol-containing medium and harvested at mid-exponential phase, washed twice with MSM, and resuspended in MSM. The cell suspension (OD₆₀₀ of 0.6) was transferred into a 50-ml flask containing 20 ml MSM supplemented with 1 mM substrate (glycerol, 5HPA, NA, or PA). After incubation at 30°C for 6 h, the cells were harvested. Total RNA was isolated using an RNA isolation kit (TaKaRa). Reverse transcription (RT)-PCR was conducted with a PrimeScript RT reagent kit (TaKaRa). All RT-PCR primers are listed in Table 3. All samples were run in triplicate.

Cloning and expression of hpaM and hpaX genes.

The hpaM gene (excluding the stop codon) was PCR amplified with primers exphpaMF and exphpaMR and fused into NdeI/XhoI-digested pET29a(+), thus producing pET-hpaM. The hpaX gene was PCR amplified with primers exphpaXF and exphpaXR and fused into NdeI/XhoI-digested pET29a(+), thus producing pET-hpaX. The recombinant plasmids pET-hpaM and pET-hpaX were transformed into E. coli BL21(DE3), and the respective genes were overexpressed. Cells were cultured in LB at 37°C to an OD₆₀₀ of 0.5 and then induced for 16 h via the addition of 0.1 mM isopropyl-β-d-thiogalactopyranoside (IPTG) at 16°C. His-tagged proteins were purified via Ni²⁺-nitrilotriacetic acid agarose chromatography (Novagen), eluted using 100 mM imidazole, dialyzed against phosphate-buffered saline (PBS) (50 mM, pH 7.0) at 4°C for 24 h, and analyzed via 12.5% SDS-PAGE. The oligomeric state of the native protein was determined via gel filtration chromatography (37). HpaM was loaded onto a column of Superdex 2000 (GE Healthcare AKTA Prime liquid system) that was equilibrated with 10 mM Tris-HCl buffer (pH 7.5, with 0.1 M NaCl and 5% glycerol) at a flow rate of 1 ml min⁻¹. The standard proteins myosin (200 kDa), β-galactosidase (116 kDa), phosphorylase b (97 kDa), bovine serum albumin (66 kDa), and ovalbumin (45 kDa) were used.

Enzyme assays.

All enzyme activities were measured at 25°C in 50 mM PBS (pH 7.0), in a total volume of 1 ml using a UV2450 spectrophotometer (Shimadzu). HpaM activity was analyzed by measuring NADH oxidation at 370 nm (ε = 2,470 M⁻¹ cm⁻¹) (38) instead of at its λ_max of 340 nm to avoid interference with the absorption spectra of 5HPA and 2,5DHP. The standard reaction mixture contained 0.2 mM FAD, 0.5 mM NADH, 0.2 mM 5HPA, and 1 μg purified HpaM. The reference cuvette contained all of these components except 5HPA. The assay was initiated via the addition of the substrate 5HPA. One unit of HpaM activity was defined as the amount of enzyme required for the oxidation of 1 μmol of NADH per min. HpaM activity toward 5HPA analogs was determined by replacing 5HPA with the equimolar analogs based on the standard reaction. The reactions were monitored for 10 min. To determine the K_m for 5HPA, a range of 5HPA concentrations (3.6 to 360 μM) were employed at a fixed concentration of the other components in the standard reaction mixture. Similarly, when the K_m for NADH was determined, a range of NADH concentrations (5 to 500 μM) were employed with fixed concentrations of the other components in the standard reaction mixture. The values were calculated via nonlinear regression fitting to the Michaelis-Menten equation using ORIGIN software version 8.5. Buffers at pH values from 4.0 to 6.0 (citric acid-sodium phosphate), 6.0 to 8.0 (KH₂PO₄-K₂HPO₄ buffer), and 8.5 to 10.0 (glycine-NaOH) were used in pH dependency experiments (39). To determine enzyme activity under anaerobic conditions, PBS was purged with high-purity nitrogen (99.99% N₂) for 20 min to remove oxygen before the substrate was added.

HpaX (2,5DHP dioxygenase) activity was measured by assaying for 2,5DHP degradation at 320 nm (ε = 5,200 M⁻¹ cm⁻¹), according to several previously published reports (6, 16, 20, 22). The reaction mixture contained 1 mM 2,5DHP, 1 mM Fe²⁺, and 10 ng purified HpaX protein in 1 ml of 50 mM PBS buffer (pH 7.0). One unit of activity was defined as the amount of enzyme that catalyzed the degradation of 1 μmol of 2,5DHP in 1 min. For determination of kinetic constants, 2,5DHP was appropriately diluted at several concentrations (18 to 315 μM).

Analytical methods.

Determinations of 5HPA and 2,5DHP were performed via HPLC analysis on a Shimadzu AD20 system equipped with a Phecda C₁₈ reversed-phase column (250 mm by 4.60 mm, 5 μm) with array detection at 310 nm. The mobile phase consisted of methanol-water-formic acid (12.5:87.5:0.2, vol/vol/vol) at a flow rate of 0.6 ml/min at 30°C. All assays in this study were independently performed three times, and the means and standard errors of measurements were calculated. LC-MS analysis was performed in a Thermo (America) DECA-60000 XLCQ Deca XP Plus instrument as previously described (21).

Accession number(s).

The hpa operon sequence and the complete genome sequence of A. faecalis JQ135 have been deposited in the GenBank database under accession numbers KY230187 and CP021641, respectively.