Novel 3,6-Dihydroxypicolinic Acid Decarboxylase-Mediated Picolinic Acid Catabolism in Alcaligenes faecalis JQ135

Picolinic acid is a natural toxic pyridine derived from l-tryptophan metabolism and other aromatic compounds in mammalian and microbial cells. Microorganisms can degrade and utilize picolinic acid for their growth, and thus a microbial degradation pathway of picolinic acid has been proposed. Picolinic acid is converted into 6-hydroxypicolinic acid, 3,6-dihydroxypicolinic acid, and 2,5-dihydroxypyridine in turn. However, there was no physiological and genetic validation for this pathway. This study demonstrated that 3,6-dihydroxypicolinic acid was an intermediate in picolinic acid catabolism and further identified and characterized a novel amidohydrolase 2 family decarboxylase PicC. PicC was also shown to catalyze the decarboxylation of 3,6-dihydroxypicolinic acid into 2,5-dihydroxypyridine. This study provides a basis for understanding picolinic acid degradation and its underlying molecular mechanism.

ABSTRACT

Picolinic acid (PA), a typical C-2-carboxylated pyridine derivative, is a metabolite of l-tryptophan and many other aromatic compounds in mammalian and microbial cells. Microorganisms can degrade and utilize PA for growth. However, the precise mechanism of PA metabolism remains unknown. Alcaligenes faecalis strain JQ135 utilizes PA as its carbon and nitrogen source for growth. In this study, we screened a 6-hydroxypicolinic acid (6HPA) degradation-deficient mutant through random transposon mutagenesis. The mutant hydroxylated 6HPA into an intermediate, identified as 3,6-dihydroxypicolinic acid (3,6DHPA), with no further degradation. A novel decarboxylase, PicC, was identified to be responsible for the decarboxylation of 3,6DHPA to 2,5-dihydroxypyridine. Although, PicC belonged to the amidohydrolase 2 family, it shows low similarity (<45%) compared to other reported amidohydrolase 2 family decarboxylases. Moreover, PicC was found to form a monophyletic group in the phylogenetic tree constructed using PicC and related proteins. Further, the genetic deletion and complementation results demonstrated that picC was essential for PA degradation. The PicC was Zn²⁺-dependent nonoxidative decarboxylase that can specifically catalyze the irreversible decarboxylation of 3,6DHPA to 2,5-dihydroxypyridine. The K_m and k_cat toward 3,6DHPA were observed to be 13.44 μM and 4.77 s⁻¹, respectively. Site-directed mutagenesis showed that His163 and His216 were essential for PicC activity. This study provides new insights into the microbial metabolism of PA at molecular level.

IMPORTANCE Picolinic acid is a natural toxic pyridine derived from l-tryptophan metabolism and other aromatic compounds in mammalian and microbial cells. Microorganisms can degrade and utilize picolinic acid for their growth, and thus a microbial degradation pathway of picolinic acid has been proposed. Picolinic acid is converted into 6-hydroxypicolinic acid, 3,6-dihydroxypicolinic acid, and 2,5-dihydroxypyridine in turn. However, there was no physiological and genetic validation for this pathway. This study demonstrated that 3,6-dihydroxypicolinic acid was an intermediate in picolinic acid catabolism and further identified and characterized a novel amidohydrolase 2 family decarboxylase PicC. PicC was also shown to catalyze the decarboxylation of 3,6-dihydroxypicolinic acid into 2,5-dihydroxypyridine. This study provides a basis for understanding picolinic acid degradation and its underlying molecular mechanism.

INTRODUCTION

Decarboxylation is a fundamental process in nature (1, 2). A variety of organic compounds, including carbohydrates, fatty acids, aromatic compounds, and environmental xenobiotics, are involved in decarboxylation. The decarboxylase family can be subdivided into two groups based on the cofactor involved (1). Some enzymes require organic cofactors, such as flavin or NAD(P)⁺, while others utilize inorganic cofactors, such as Zn²⁺ or Mn²⁺. Recently, the amidohydrolase 2 family decarboxylases that use inorganic ions as cofactors, have been gaining more attention (3–6). The enzymes in this family are usually involved in the catabolism of important natural compounds such as α-amino-β-carboxymuconate-ε-semialdehyde (ACMSD) (7, 8), γ-resorcylate (9), 2,3-dihydroxybenzoate (10), 2,5-dihydroxybenzoate (11), 3,4-dihydroxybenzoate (2), 4-hydroxybenzoate (12), 5-carboxyvanillate (13), vanillate (14), and 2-hydroxy-1-naphthoate (3). However, most of these decarboxylases studied were involved in the decarboxylation of benzene ring derivatives while not in the metabolism of pyridine derivatives.

Picolinic acid (PA) is a typical C-2-carboxylated pyridine derivate that is widely generated from physiological metabolism in mammalian and microbial cells (15). PA is a natural dead-end metabolite of l-tryptophan produced via kynurenine pathway in humans and other mammals (16–18). Moreover, it can be produced in other biological processes such as the microbial degradation of 2-aminophenol, catechol, and nitrobenzene (19–21). PA was found to be toxic, and it inhibited the growth of normal rat kidney cells and T cell proliferation, enhancing seizure activity in mice and inducing cell death via apoptosis (22–25). PA cannot be metabolized by humans and thus gets excreted through urine or sweat (26). However, PA can be degraded by microorganisms in the natural environment (15). Numerous PA-degrading bacterial strains have been isolated, including strains of Achromobacter (27), Aerococcus (28), Alcaligenes (29), Arthrobacter (30), Bacillus (31), Burkholderia (32), and Streptomyces (33). The metabolic pathway of PA in microorganisms has been partially elucidated in previous studies (15, 28, 32) (Fig. 1). In other studies, the crude enzyme facilitating the conversion of PA to 6-hydroxypicolinic acid (6HPA) has been preliminarily purified in Arthrobacter picolinophilus DSM 20665 and an unidentified Gram-negative bacterium (designated the UGN strain) (30, 34). Nevertheless, the functional genes or enzymes involved in PA degradation have not been cloned or characterized yet.

FIG 1.

Proposed PA degradation pathway in A. faecalis JQ135. Dotted arrows indicate the proposed steps. The 3,6DHPA and 2,5-DHP are shown in blue. TCA, tricarboxylic acid cycle.

In our previous work, we demonstrated that Alcaligenes faecalis strain JQ135 utilizes PA as the sole carbon and nitrogen source and as an energy source and that 6-hydroxypicolinic acid (6HPA) was the first intermediate of PA (35). Further studies showed that the maiA gene was essential for PA catabolism (36). In the present study, we report the fully characterized intermediate compound, 3,6-dihydroxypicolinic acid (3,6DHPA) (Fig. 1). Further, a novel nonoxidative 3,6-dihydroxypicolinic acid decarboxylase gene (picC) was cloned from A. faecalis strain JQ135, and the corresponding product was characterized.

RESULTS

Transposon mutant and identification of the intermediate 3,6DHPA.

A library of A. faecalis JQ135 mutants incapable of 6HPA utilization was constructed by random transposon mutagenesis. More than 30 mutants that could not grow on 6HPA-containing medium were selected from approximately 10,000 clones and their ability to convert 6HPA was examined. High-performance liquid chromatography (HPLC) results showed that one mutant (designated Mut-H4) could convert 6HPA into a new intermediate with no further degradation (Fig. 2). After liquid chromatography/time of flight-mass spectrometry (LC/TOF-MS) analysis, it was found that the molecular ion peak ([M+H]⁺) of this new intermediate was 156.0295 (ion formula, C₆H₆NO₄⁺; calculated molecular weight, 156.0297 with −3.2 ppm error), indicating that one oxygen atom was added to 6HPA (C₆H₅NO₃). According to the previously predicted PA degradation pathway, the intermediate is most likely to be 3,6DHPA (15, 31, 34). In the present study, 3,6DHPA was chemically synthesized and characterized by UV-visible spectroscopy (UV-VIS), LC/TOF-MS, ¹H nuclear magnetic resonance (NMR), and ¹³C NMR spectroscopies (see Fig. S1 and S2 in the supplemental material) and HPLC analysis showed that the retention time of the new intermediate was identical to that of the synthetic sample of 3,6DHPA (Fig. 2). Thus, this intermediate compound was identified as 3,6DHPA.

FIG 2.

HPLC and LC/TOF-MS profiles of the conversion of 6HPA by mutant Mut-H4. (A and C) The authentic samples of 6HPA and 3,6DHPA, respectively. (B) Conversion of 6HPA into 3,6DHPA by mutant Mut-H4. The detection wavelength was set at 310 nm. (D) LC/TOF-MS spectra of 3,6DHPA produced in panel B.

Screening of the 3,6DHPA decarboxylase gene.

The transposon insertion site of mutant Mut-H4 was identified using the genome walking method (37). The insertion site of the transposon was located in gene AFA_15145 (genome position 3298929). Gene AFA_15145 was a 972-bp length open reading frame (ORF) starting with GTG. AFA_15145 exhibited the highest sequence similarity to several nonoxidative decarboxylases such as γ-resorcylate decarboxylase (γ-RSD; 45% identity) (9), 2,3-dihydroxybenzoate decarboxylase (2,3DHBD; 36% identity) (10), 5-carboxyvanillate decarboxylase (5CVD; 27% identity) (13), and hydroxynaphthoate decarboxylase (HndA; 22% identity) (3) (Fig. 3). All of these decarboxylases belong to the amidohydrolase 2 family proteins (COG2159) that contain a triosephosphate isomerase (TIM)-barrel fold. Based on the phenotype of the mutant Mut-H4 and bioinformatics analysis, it was predicted that the AFA_15145 (designated picC) encoded the 3,6DHPA decarboxylase.

FIG 3.

Amino acid sequence analysis of PicC. (A) Phylogenetic analysis of PicC and related decarboxylases. Each item was arranged in the following order: protein name, accession number, and strain. AmiH_1 and _2 are amidohydrolase 1 and 2 family members, respectively. The phylogenetic tree was constructed using the neighbor-joining method (with a bootstrap of 1,000) with the software MEGA 6.0. The bar represents amino acid substitutions per site. (B) Multiple sequence alignment of PicC and seven decarboxylases. The predicted N-terminal motifs for Zn²⁺ binding are denoted by blue box. The other three Zn²⁺-binding sites are denoted by diamonds. The seven histidine residues for site-directed mutagenesis are denoted by triangles.

Function identification of picC gene in PA degradation in A. faecalis JQ135.

To confirm whether picC is involved in PA degradation, picC-deleted mutant JQ135ΔpicC was constructed. The mutant JQ135ΔpicC lost the ability to grow on PA, 6HPA, or 3,6DHPA. The complemented strain, JQ135ΔpicC/pBBR-picC, completely restored the phenotype of growth on PA, 6HPA, and 3,6DHPA (Fig. S3). These results showed that picC was essential for the degradation of PA in A. faecalis JQ135.

picC encodes a 3,6DHPA decarboxylase.

Recombinant PicC was overexpressed in Escherichia coli BL21(DE3) cells containing the plasmid pET-picC. SDS/PAGE analysis showed the presence of an intense band, identifying with the His₆-tagged PicC (37 kDa) (Fig. 4). The degradation of 3,6DHPA by purified PicC was monitored spectrophotometrically. The maximum absorption was shifted from 340 nm (3,6DHPA) to 320 nm (2,5-dihydroxypyridine [2,5DHP]). LC/TOF-MS analysis suggested that the molecular ion peak of the product was 112.0400 (M+H⁺), which was identical to that of 2,5DHP (36, 38). Further, the HPLC analysis showed that the retention time of the product was identical to that of the authentic sample of 2,5DHP. The 3,6DHPA was degraded completely with the formation of equimolar amounts of 2,5DHP. Moreover, PicC did not catalyze the reverse carboxylation of 2,5DHP in a reaction mixture containing NaHCO₃.

FIG 4.

Characterization of PicC. (A) SDS-PAGE of purified PicC. Lane M, protein marker. Lane 1, purified PicC. (B) Spectrophotometric changes during transformation of 3,6DHPA by purified PicC. The reaction was initiated by adding 3,6DHPA. Spectra were recorded every 1 min. The arrow denotes the biotransformation of 3,6DHPA into 2,5DHP. (C) HPLC analysis of the transformation of 3,6DHPA into 2,5DHP by PicC. The detection wavelength was 310 nm.

Biochemical properties of PicC.

The recombinant PicC was highly active at pH 7.0 and 40°C (Fig. S4). The K_m and k_cat values for 3,6DHPA were found to be 13.44 μM and 4.77 s⁻¹, respectively (Table 1 and Fig. S5). The enzyme was unstable at room temperature and could retain only 50% of initial activity when incubated at 30°C for 24 h. In addition, PicC could not convert the structural analogues of 3,6DHPA, including 3-hydroxy-picolinic acid, gentisic acid, 2,3-dihydroxybenzoic acid, and 2,6-dihydroxybenzoic acid. This can be attributed to the substrate specificity of PicC toward 3,6DHPA.

TABLE 1.

Kinetic constants of wild-type PicC and mutants^a

Enzyme	Mean ± SD		kcat (s−1)	kcat/Km (s−1 mM−1)
	Km (μM)	Vmax (μmol min−1 mg−1)
PicC	13.44 ± 2.91	7.73 ± 0.05	4.77	354.54
PicCH12A	53.06 ± 4.05	2.83 ± 0.09	1.75	32.94
PicCH135A	9.38 ± 0.68	6.14 ± 0.11	3.78	403.40
PicCH163A	ND	ND	ND	ND
PicCH172A	15.80 ± 1.21	3.33 ± 0.08	2.06	130.14
PicCH177A	110.81 ± 14.70	2.18 ± 0.01	1.35	12.14
PicCH194A	11.38 ± 0.64	5.15 ± 0.07	3.18	279.05
PicCH216A	ND	ND	ND	ND

^aThe kinetic constants were determined at pH 7.0 and 40°C. ND, not detected.

The effects of various metal ions and inhibitors on decarboxylase activity are shown in Table 2. PicC activity was not affected by metal ions such as Ca²⁺, Cd²⁺, Co²⁺, Fe³⁺, Mg²⁺, Mn²⁺, and Zn²⁺ but was strongly inhibited by Ag⁺, Co²⁺, and Hg²⁺ ions. In addition, several inhibitors such as EDTA, 8-hydroxy-quinoline-5-sulfonic acid (8-HQSA; a zinc metal-specific inhibitor), phenylmethylsulfonyl fluoride (PMSF; serine- and cysteine-specific inhibitor), and sodium iodoacetate (cysteine-specific inhibitor) showed relatively low effects on PicC activity. However, diethylpyrocarbonate (DEPC), a histidine residue modifier, strongly inhibited PicC activity, indicating the presence of active-site histidine residues.

TABLE 2.

Effect of metal ions and inhibitors on PicC activity^a

Ion or inhibitor	Mean relative activity (%) ± SD
Blank	100
CaCl2	90.68 ± 1.49
CoCl2	84.47 ± 2.21
FeCl3	97.59 ± 1.44
MgSO4	87.59 ± 1.95
ZnSO4	97.11 ± 1.84
EDTA	60.94 ± 1.64
PMSF	97.27 ± 0.32
AgNO3	4.63 ± 0.53
CdCl2	99.98 ± 2.43
CuCl2	23.89 ± 0.67
HgCl2	2.36 ± 0.40
MnSO4	91.30 ± 1.13
DEPC	4.05 ± 0.19
8-HQSA	55.21 ± 0.91
Sodium iodoacetate	81.68 ± 0.99

^aMetal ions were tested at 0.1 mM. DEPC, 0.1%; PMSF, 1%; sodium iodoacetate, 0.025 mM; 8-HQSA, 100 μg/ml; EDTA, 100 μg/ml.

Further, inductively coupled plasma optical emission spectrometry (ICP-OES) analysis revealed the presence of Zn²⁺ at 0.85 ± 0.1 mol per mol of protein that is similar to several other nonoxidative decarboxylases of the amidohydrolase 2 superfamily (1, 3). The moderate effects, exhibited by additional Zn²⁺ or EDTA in the reaction system, indicated the presence of Zn²⁺ in the center of PicC.

Site-directed mutagenesis.

In order to assess their roles in the function of PicC, seven histidine residues (H12, H135, H163, H172, H177, H194, and H216) were replaced with Ala residues through site-directed mutagenesis (Fig. 3B). The seven PicC mutants obtained were expressed and purified and their activities were determined (Fig. S6; Table 1). PicC^H135A showed a slight increase in decarboxylase activity, while PicC^H172A and PicC^H194A showed a slight reduction in decarboxylase activities (ca. 10 to 50%), and PicC^H12A and PicC^H177A strongly reduced the decarboxylase activities (>90%). Moreover, the mutant proteins PicC^H163A and PicC^H216A completely lost decarboxylase activity.

DISCUSSION

PA is a natural and toxic monocarboxylated pyridine derivative. Studies on the microbial degradation mechanism of PA began 50 years ago (28). A partial catabolic pathway of PA has been proposed (Fig. 1) (15, 34): PA was dehydrogenated to 6HPA and then converted into 3,6DHPA via hydroxylation, leading to the decarboxylation of 3,6DHPA into 2,5DHP. The intermediate, 6HPA, has been substantially identified in most strains, including Aerococcus sp. (28), Alcaligenes faecalis DSM 6269 (29), Arthrobacter picolinophilus DSM 20665 (30, 39), Burkholderia sp. strain ZD1 (32), Streptomyces sp. strain Z2 (33), and the UGN strain (34). Further, another intermediate compound, 2,5DHP has been detected in the media during PA degradation in a few strains (32, 34). However, the intermediate 3,6DHPA, a key link between 6HPA and 2,5DHP, was hardly detectable. This could be mostly attributed to its immediate degradation before excretion out of the cells. Previously, 3,6DHPA has been only theoretically proposed in Bacillus sp. (31) and the UGN strain (34). In this study, to the best of our knowledge, we demonstrated the chemical properties (UV-visible, LC/TOF-MS, ¹H NMR, and ¹³C NMR spectroscopies) of 3,6DHPA for the first time (see Fig. S2 in the supplemental material) and detected it in the media from the transposon mutant strain, thus confirming that 3,6DHPA is a catabolic intermediate of PA.

Some previous studies have attempted to identify the genes and enzymes involved in PA degradation, such as the PA dehydrogenase in Arthrobacter (39) and 2,5DHP dioxygenase in UGN strain (34). However, their amino acid sequences and respective coding genes remained unknown, with no biochemical, physiological, or genetic evidence to explain the decarboxylation of 3,6DHPA to 2,5DHP. In this context, we cloned the decarboxylase gene picC through random transposon mutagenesis and ascertained that PicC was responsible for 3,6DHPA decarboxylation to form 2,5DHP. We found that PicC shared homology with the amidohydrolase 2 family proteins and contained the conserved TIM-barrel fold of amidohydrolase 2 family (Fig. 3). It has been previously reported that the amidohydrolase 2 family protein (PF04909) catalyzes the decarboxylation reaction (C-C bond) of several benzene derivatives, whereas the amidohydrolase 1 family protein (PF01979) catalyzes the hydrolytic reactions (C-N, C-Cl, or C-P bond) (4). The ACSMD was the first member of amidohydrolase 2 family to be reported (8), followed by other members, including γ-RSD (9), 2,3DHBD (10), 5CVD (13), and HndA (3). A phylogenetic tree of PicC and related proteins showed that PicC was clustered with amidohydrolase 2 but not amidohydrolase 1 family proteins (Fig. 3). However, the identities between PicC and reported decarboxylases were low (<45%) and PicC formed a separate branch in the phylogenetic tree (Fig. 3). In addition, PicC was found to be specific toward its substrate 3,6DHPA. Thus, it can be concluded that PicC could be a novel amidohydrolase 2 family decarboxylase.

The amidohydrolase 2 family proteins contain a few conserved amino acid residues, which are usually the active sites. In ACSMD, the His177, His228, and D294 sites were the Zn²⁺-binding sites (4). These three residues have been found in all reported amidohydrolase 2 family proteins, including PicC (His163, His216, and D283) (Fig. 3B). In addition, the results of site-directed mutagenesis of PicC confirmed that H163 and H216 also played essential roles in PicC-mediated catalysis. Another Zn²⁺-binding motif, “HXH,” has been found in the N terminus of ACSMD (4) and HndA (3), whereas this motif was replaced by “EEH” in γ-RSD (9) or “EEA” in 5CVD (13) (Fig. 3B). In PicC, the corresponding motif has been found to be similar to that of γ-RSD. After substituting His12 by Ala, the resultant enzyme PicC^H12A still exhibited 10% activity, suggesting a variation in the third residue of this motif. In addition, site-directed mutagenesis results demonstrated that several other histidine residues, His172, His177, and His194, were important for PicC activity.

In conclusion, this study revealed that 3,6DHPA was a catabolic intermediate in PA degradation by bacteria. The 3,6DHPA decarboxylase (PicC) was identified and characterized. To the best of our knowledge, PicC is also the first nonoxidative decarboxylase belonging to the amidohydrolase 2 family that catalyzes the irreversible decarboxylation of a pyridine derivative. This study will expand our understanding of the bacterial degradation mechanisms of pyridine derivatives.

MATERIALS AND METHODS

Chemicals.

PA, 6HPA, and 2,5DHP were purchased from J&K Scientific, Ltd. (Shanghai, China). EDTA, 8-HQSA, PMSF, sodium iodoacetate, DEPC, and other reagents of analytical grade were purchased from Sangon Biotech Co., Ltd. (Shanghai, China). 3,6DHPA was chemically synthesized (as detailed in the supplemental material). The structure of 3,6DHPA was confirmed by UV-VIS, LC/TOF-MS, and NMR spectroscopy (Fig. S1 and S2).

Strains, plasmids, and primers.

All bacterial strains and plasmids used in this study are listed in Table 3. Alcaligenes faecalis JQ135 (CCTCC M 2015812) is the wild-type PA-degrading strain (35). E. coli DH5α was used as the host for the construction of plasmids. E. coli BL21(DE3) was used to overexpress the proteins. Bacteria were cultivated in Luria-Bertani (LB) medium at 37°C (E. coli) or 30°C (Alcaligenes and their derivatives). Antibiotics were added at the following concentrations (as required): chloramphenicol (Cm), 34 mg/liter; gentamicin (Gm), 50 mg/liter; kanamycin (Km), 50 mg/liter; and streptomycin (Str), 50 mg/liter. Primer synthesis and the sequencing of PCR products or plasmids were performed by GenScript Biotech (Nanjing, China) (40). The primers used in this study are listed in Table 4.

TABLE 3.

Strains and plasmids used in this study

Strain or plasmid	Description	Source
Strains
Alcaligenes faecalis
JQ135	Strr; PA-degrading bacterium; Gram negative; wild type	CCTCC M 2015812
Mut-H4	Strr Kmr; picC (AFA_15145) mutant of A. faecalis JQ135 inserted by transposon	This study
JQ135ΔpicC	Strr; picC deletion mutant of JQ135	This study
JQ135ΔpicC/pBBR-picC	Strr Gmr; JQ135ΔpicC complementation with pBBR-picC	This study
Escherichia coli
DH5α	F− recA1 endA1 thi-1 hsrdR17 supE44 relA1 deoRΔ(lacZYA-argF)U169 ϕ80lacZΔM15	TaKaRa
BL21(DE3)	F− ompT hsdS(rB− mB−) gal dcm lacY1(DE3)	TaKaRa
SM10λpir	Donor strain for biparental mating	Lab stock
Plasmids
pET29a(+)	Kmr; expression plasmid	Novagen
pSC123	Cmr Kmr; suicide plasmid, mariner transposon	Lab stock
pJQ200SK	Gmr mob+ orip15A lacZα+ sacB; suicide plasmid	Lab stock
pBBR1MCS-5	Gmr; broad-host-range cloning plasmid	Lab stock
pJQ-ΔpicC	Gmr; picC gene deletion plasmid	This study
pBBR-picC	Gmr; the fragment containing the picC gene inserted into XhoI/HindIII-digested pBBR1MCS-5	This study
pET-PicC	Kmr; NdeI-XhoI fragment containing picC gene inserted into pET29a(+)	This study
pET-PicCH12A	Kmr; NdeI-XhoI fragment containing picCH12A gene inserted into pET29a(+)	This study
pET-PicCH135A	Kmr; NdeI-XhoI fragment containing picCH135A gene inserted into pET29a(+)	This study
pET-PicCH163A	Kmr; NdeI-XhoI fragment containing picCH163A gene inserted into pET29a(+)	This study
pET-PicCH172A	Kmr; NdeI-XhoI fragment containing picCH172A gene inserted into pET29a(+)	This study
pET-PicCH177A	Kmr; NdeI-XhoI fragment containing picCH177A gene inserted into pET29a(+)	This study
pET-PicCH194A	Kmr; NdeI-XhoI fragment containing picCH194A gene inserted into pET29a(+)	This study
pET-PicCH216A	Kmr; NdeI-XhoI fragment containing picCH216A gene inserted into pET29a(+)	This study

TABLE 4.

Primers used in this study

Primer	Sequence (5′–3′)	Description
kopicC-UF	AGCTTGATATCGAATTCCTGCAGTTCTTCTTGCTTTAGCTCGGC	To construct plasmid pJQ-ΔpicC
kopicC-UR	TCTAGAACTAGTGGATCCAGTCTGCGACAGCACGACGTAC
kopicC-DF	GGATCCACTAGTTCTAGAGTGTTGTTCTCGGTGGACTACC
kopicC-DR	AGGGAACAAAAGCTGGAGCTCAGAAGGCGCGCATATCCTCAG
picC-F	CAGGAATTCGATATCAAGCTTAGTGGCTTTCAGCCTGTTGCC	To construct plasmid pBBR-picC
picC-R	GGTACCGGGCCCCCCCTCGAGGTAATGCTGACGAATGCCATTGG
expPicC-F	CTTTAAGAAGGAGATATACATATGAAACGTATTAAAAAAATAGC	To construct plasmid pET-PicC
expPicC-R	GTGGTGGTGGTGGTGGTGCTCGAGGCTCCGGTCCAGCTTGAACAG
Mut-H12A-1	CTTTAAGAAGGAGATATACATATGAAACGTATTAAAAAAATAGCACTGGAGG	To construct plasmid pET-PicCH12A
Mut-H12A-2	AAAAAAATAGCACTGGAGGAGGCATTCAACGCCGTTGG
Mut-H12A-3	TCAGTGGTGGTGGTGGTGGTGCTCGAGGCTCCGGTCCAGCTTGAACAG
Mut-H135A-1	GCTTTGGTCAACGGTGCTACGCATGGTGTGTAC	To construct plasmid pET-PicCH135A
Mut-H135A-2	GTACACACCATGCGTAGCACCGTTGACCAAAGC
Mut-H163A-1	GTGCCGTTCTATCTGGCTCCCTTTGATGCTTACG	To construct plasmid pET-PicCH163A
Mut-H163A-2	CGTAAGCATCAAAGGGAGCCAGATAGAACGGCAC
Mut-H172A-1	GCTTACGAAATGCCAGCTGCTTACACAGGCCAC	To construct plasmid pET-PicCH172A
Mut-H172A-2	GTGGCCTGTGTAAGCAGCTGGCATTTCGTAAGC
Mut-H177A-1	CACGCTTACACAGGCGCTCCGGAGCTGGTTGGG	To construct plasmid pET-PicCH177A
Mut-H177A-2	CCCAACCAGCTCCGGAGCGCCTGTGTAAGCGTG
Mut-H194A-1	GTAGAAACCGGCACCGCAGCGCTGCGCATGTTG	To construct plasmid pET-PicCH194A
Mut-H194A-2	CAACATGCGCAGCGCTGCGGTGCCGGTTTCTAC
Mut-H216A-1	AAGCTGGTGCTGGGTGCAATGGGTGAAGGCCTG	To construct plasmid pET-PicCH216A
Mut-H216A-2	CAGGCCTTCACCCATTGCACCCAGCACCAGCTT

Transposon mutagenesis, mutant screening, and gene cloning.

A transposon mutant library of A. faecalis JQ135 was constructed using the transposon-based plasmid pSC123 (kanamycin resistance gene) as described previously (35). In this study, the PA was replaced by its intermediate 6HPA. Mutants that could not utilize 6HPA as the sole carbon source were selected. The ability of the mutant to utilize 6HPA was determined as follows: the concentration of 6HPA was 1 mM and the inoculum of mutant was set at a final optical density at 600 nm (OD₆₀₀) of 2.0. Degradation was measured by HPLC.

The flanking sequences of the transposon in the mutants were amplified using the DNA walking method (37). The amplified PCR products were sequenced and analyzed. The insertion sites were confirmed by comparison with the genome sequence of A. faecalis JQ135.

Gene knockout and genetic complementation of A. faecalis JQ135.

The genes or DNA fragments from A. faecalis JQ135 were amplified by PCR using corresponding primers (Table 4). The fusion of DNA fragments and cut plasmids was carried out using a ClonExpress MultiS one-step cloning kit (Vazyme Biotech Co., Ltd., Nanjing, China). Gene deletion mutant of the picC in A. faecalis JQ135 was constructed using a two-step homogenetic recombination method with the suicide plasmid pJQ200SK (41). Two homologous recombination-directing sequences were amplified using the primers kopicC-UF/kopicC-UR and kopicC-DF/kopicC-DR, respectively. Two PCR fragments were subsequently ligated into SacI/PstI-digested pJQ200SK, generating pJQ-ΔpicC. The pJQ-ΔpicC plasmid was then introduced into A. faecalis JQ135 cells. The single-crossover mutants were screened on a LB plate containing Str and Gm. The gentamicin-resistant strains were then subjected to repeated cultivation in LB medium containing 10% sucrose with no gentamicin. Double-crossover mutants that had lost their plasmid backbone and were sensitive to gentamicin were selected on LB Str plates. Deletion of the picC gene was confirmed by PCR. This procedure resulted in the construction of the deletion mutant strain JQ135ΔpicC.

Knockout mutants were complemented as follows. The intact picC gene was amplified using the primers picC-F and picC-R and then ligated with the XhoI/HindIII-digested pBBR1-MCS5, generating pBBR-picC. The pBBR-picC was then transferred into the mutant strain JQ135ΔpicC to generate the complemented strain JQ135ΔpicC/pBBR-picC.

Expression and purification of the His-tagged PicC and its mutations.

For the overexpression of picC in E. coli BL21(DE3), the complete ORF without the stop codon (genome positions 3298274 to 3299242) were amplified using genomic DNA of strain JQ135 and inserted into the NdeI/XhoI-digested plasmid pET29a(+), resulting in the plasmid pET-PicC. E. coli BL21(DE3) cells (containing pET-PicC) were initiated by the addition of 0.3 mM IPTG when the optical density of the culture (OD₆₀₀) reached 0.5 to 0.8 and was incubated for an additional 12 h at 16°C. Cells were harvested by centrifugation at 4°C, sonicated, and then centrifuged again to remove cell debris. The supernatant was used for recombinant protein purification using an Ni-nitrilotriacetic acid (Ni-NTA)–agarose column (Sangon, Shanghai, China). The purified His₆-tagged proteins were then analyzed using 12.5% SDS-PAGE. The protein concentrations were determined using the Bradford method (42).

For site-directed mutagenesis of PicC, the picC fragments were amplified from plasmid pET-PicC through overlap PCR using the primers carrying point mutations (Table 4). Amplified fragments were fused into plasmid pET29a(+), resulting in pET-PicC^H12A, pET-PicC^H135A, pET-PicC^H163A, pET-PicC^H172A, pET-PicC^H177A, pET-PicC^H194A, and pET-PicC^H216A. Resultant constructs were confirmed by sequencing. The expression and purification of the mutants were performed as described above.

Enzymatic assays of 3,6DHPA decarboxylase.

For the decarboxylase activity, the enzyme reaction mixture contained 50 mM PBS (pH 7.0), 0.3 mM 3,6DHPA, and 5 μg of purified PicC (in 1 ml) and was incubated at 40°C. The enzymatic activities were determined spectrophotometrically by the disappearance of 3,6DHPA at 360 nm (ε = 4.4 cm⁻¹mM⁻¹). To determine the effect of one condition, other conditions were kept at fixed concentration of the standard reaction mixture. The optimum pH of the PicC protein was determined using various buffers such as 50 mM citric acid-sodium citrate (pH 4 to 6), 50 mM KH₂PO₄-K₂HPO₄ (pH 6 to 8), and 50 mM glycine-NaOH (pH 8.0 to 9.8) at 40°C. The optimum temperature of the PicC protein was determined to be between 10 and 50°C in PBS (pH 7.0). Purified PicC was preincubated with various metal ions and inhibitors at 4°C for 30 min to study their effects on the enzyme. The activity was expressed as a percentage of the activity obtained in the absence of the added compound. To determine the kinetic constants for 3,6DHPA, a range of 3,6DHPA concentrations (2 to 300 μM) was used. The values were calculated through nonlinear regression fitting to the Michaelis-Menten equation. One unit of the activity was defined as the amount of enzyme that catalyzed 1 μmol of 3,6DHPA in 1 min at pH 7.0 and 40°C. The decarboxylase activities of PicC point mutants were determined as described above.

The measurement of carboxylase activity of PicC was similar to that in a previous study (43). The reaction mixture contained 50 mM PBS (pH 7.0), 0.3 mM 2,5DHP, 5.0 mM NaHCO₃, and 5 μg of purified PicC in 1 ml at 40°C.

Analytical methods.

The UV-VIS spectra was observed by a UV2450 spectrophotometer (Shimadzu). The PA, 6HPA, 3,6DHPA, and 2,5DHP concentrations were determined by HPLC analysis on a Shimadzu AD20 system equipped with a Phecda C₁₈ reversed-phase column (250 mm by 4.60 mm, 5 μm). The concentrations of the compounds were calculated using standard samples. The mobile phase consisted of methanol, water, and formic acid (12.5:87.5:0.2 [vol/vol/vol]) at a flow rate of 0.6 ml/min, at 30°C. LC/TOF-MS analysis was performed in a TripleTOF 5600 (AB SCIEX) mass spectrometer as described previously (44). The Zn²⁺ concentration of the PicC protein was analyzed using the ICP-OES method as described in a previous study (45).

Data availability.

The PicC sequence and the complete genome sequence of A. faecalis JQ135 have been deposited in the GenBank database under accession numbers ARS01287 and CP021641, respectively.