Genome-wide analysis of the aromatic catabolism pathways

P. stutzeri has recently received particular attention for its metabolic properties, including denitrification, degradation of aromatic compounds, and nitrogen fixation. Since P. stutzeri A1501 was originally isolated from paddy soil and because it contains sets of genes for the β-ketoadipate pathway, it should be able to utilize aromatic compounds. In our study, we observed that this strain can aerobically degrade benzoate and 4-hydroxybenzoate. As the complete genome of P. stutzeri A1501 was sequenced recently [20], we mapped the genes encoding the peripheral pathways for the catabolism of 4-hydroxybenzoate (pob) and benzoate (ben) in the A1501 chromosome (Figure 1A). In many soil bacteria, these peripheral pathway enzymes channel the individual substrates into one of the two branches of the β-ketoadipate pathway, namely the catechol and protocatechuate branches. Sequence comparison indicated that A1501 has genes encoding all of the enzymes involved in the two branches of the β-ketoadipate pathway. The catechol (cat genes) and the protocatechuate branches (pca genes) converge at β-ketoadipate enol-lactone. One set of enzymes, which are encoded by pcaDIJF, completes the conversion of β-ketoadipate enol-lactone to tricarboxylic acid cycle intermediates (Figure 1B).

In the A1501 genome, the cat genes are chromosomally linked with the ben genes and form an 11.5 kb supercluster (PST1666-PST1676). The deduced amino acid sequence of BenR in A1501 shows high similarity (61% identity) to the P. fluorescens Pf-5 BenR protein. However, the catR gene, which positively regulates the catBC and catA operons in other strains [12, 25], is absent in A1501 (Figure 2A). Additionally, the pca genes in P. stutzeri A1501 are contiguous, whereas the pca genes are scattered over several portions of the genome in other Pseudomonas species, such as P. entomophila [21], P. aeruginosa [26], P. fluorescens [27]and P. putida [2] (Figure 2B). PcaR is an Icl family protein and has been reported to regulate most of the pca genes in the protocatechuate branch of the β-ketoadipate pathway in P. putida [12, 28, 29]. In contrast to other Pseudomonas strains, pcaR is located immediately upstream of pcaI in A1501 (Figure 2B). The deduced amino acid sequence of A1501 PcaR shows 85% identity to that of P. putida KT2440. Notably, the pcaK gene, which encodes a 4-hydroxybenzoate permease responsible for 4-hydroxybenzoate transport [30], is absent in A1501 (Figure 2B). The catabolic gene organization in A1501 lacks the catR and pcaK genes, a feature that is not observed in other Pseudomonas strains.

Functional characterization of the β-ketoadipate pathway

A1501 grew well on 4 mM benzoate and reached an OD₆₀₀ of 0.5 after 24 h of incubation, whereas no growth was observed in the presence of 8 mM benzoate. A1501 grow poorly on 0.4 mM 4-hydroxybenzoate, while 4-hydroxybenzoate at concentrations above 0.8 mM completely inhibited bacterial growth (Figure 3). Further investigation of the β-ketoadipate pathway was made by constructing and characterizing three mutants: benR mutant A1601, pcaR mutant A1602 and pcaD mutant A1603 (Table 1). When the wild type and mutants were cultured in media containing lactate, their growth rates were not affected (data not shown). As expected, the benR mutant failed to grow on benzoate, and the pcaR and pcaD mutants failed to grow on 4-hydroxybenzoate as the sole carbon source. Furthermore, both the pcaR and pcaD mutants lost their ability to utilize benzoate as a carbon source. We constructed three complementary plasmids containing the entire pcaD, pcaR and benR genes for further growth complementation assays. Complementation of the three mutants with the corresponding complementary plasmids restored the catabolic activity, and the three corresponding complementary strains grew on benzoate as the sole carbon source (data not shown). Results from gene disruption analyses and genetic complementation tests demonstrate that the three genes are required for the growth of A1501 on benzoate.

High-performance liquid chromatography (HPLC) was used to measure the concentrations of catechol and muconate in the culture supernatants of the wild type A1501 and pcaD mutant A1603 grown on benzoate as the sole carbon source (Figure 4; see Additional file 1). During the initial phase of benzoate catabolism by A1501, small amounts of catechol (~30 μM) and cis, cis-muconate (~500 nM) were detected. After 24 h, benzoate was completely removed from the culture supernatants, and no metabolites could be detected (see Additional file 1). The inability of the pcaD mutant A1603 to grow on benzoate was further confirmed by HPLC analysis of culture supernatants. After 48 h, the concentration of benzoate remained almost unchanged in the culture supernatant of the mutant, while accumulation of catechol and cis, cis-muconate was detected by HPLC (Figure 4). As shown in Figure 1B, inactivation of PcaD completely blocked the conversion of β-ketoadipate enol-lactone to β-ketoadipate, resulting in accumulation of the intermediates catechol and cis, cis-muconate derived from benzoate. These results provide experimental evidence that the two branches of the β-ketoadipate pathway converge at β-ketoadipate enol-lactone and that the products of pcaDIJF complete the conversion of the latter to TCA cycle intermediates in P. stutzeri A1501, as documented in other Pseudomonas strains [2].

Strains or plasmids	Relevant characteristic(s)a	Source or reference
Strains
P. stutzeri
A1501	Wild type, Chinese culture CGMCC 0351, Ben++, Cat++, 4HBA+	[22]
A1601	A1501 benR::pK18mob ΔbenR, Ben-, Cat++, 4HBA+	this study
A1602	A1501 pcaR::pK18mob ΔpcaR, Ben-, Cat-, 4HBA-	this study
A1603	A1501 pcaD::pK18mob ΔpcaD, Ben-, Cat-, 4HBA-	this study
A1608	A1601 harboring complement plasmid pLbenR	this study
A1609	A1602 harboring complement plasmid pLpcaR	this study
A1610	A1603 harboring complement plasmid pLpcaD	this study
Plasmids
pK18mob	Kmr; oriColE1 Mob+ lacZα+, used for directed insertional disruption	[45]
pRK2013	Kmr; Tra+, oriColE1	[46]
pLAFR3	Tcr; Tra-, Mob+, cos, RK2 replicon	[47]
pKbenR	Kmr; 293 bp EcoR I-Hind III fragment containing part of benR in pK18mob	this study
pKpcaR	Kmr; 299 bp EcoR I-Hind III fragment containing part of pcaR in pK18mob	this study
pKpcaD	Kmr; 361 bp EcoR I-Hind III fragment containing part of pcaD in pK18mob	this study
pLbenR	Tcr; 1041 bp EcoR I-Hind III fragment containing benR with its native promoter in pLAFR3	this study
pLpcaR	Tcr; 1457 bp EcoR I-Hind III fragment containing pcaR with its native promoter in pLAFR3	this study
pLpcaD	Tcr; 820 bp EcoR I-Hind III fragment containing pcaD with the lac promoter in pLAFR3	this study

^a Ben⁺⁺, good growth on benzoate; Cat⁺⁺, good growth on catechol; 4HBA⁺, weak growth on 4-hydroxybenzoate; Ben^-, no growth on benzoate; Cat^-, no growth on catechol; 4HBA^-, no growth on 4-hydroxybenzoate; Km^r, kanamycin resistant; Tc^r, tetracycline resistant.

As mentioned above, A1501 can grow well on benzoate, but not on 4-hydroxybenzoate, as the sole carbon and energy source. Therefore, we focused on the genetic organization of the A1501 ben-cat region. As shown in Figure 5A, nine ben and cat genes are in the same transcriptional orientation and the lengths of the intergenic regions vary. Gene expression was further studied by amplifying intergenic regions between adjacent genes. Eight pairs of oligonucleotide primers were designed (Table 2). As shown in Figure 5B, in the presence of benzoate, four products of their expected sizes were amplified with the PF/PR primer pairs spanning the borders of benA-benB (456 bp), benB-benC (503 bp), benC-benD (546 bp), and catB-catC (309 bp). No PCR products were observed with the PF/PR primer pairs spanning the borders of benR-benA (782 bp), benD-benK (610 bp), benK-catB (1074 bp), and catC-catA (1030 bp) in the presence or absence of benzoate. These results suggest that nine benzoate metabolic genes are organized in five transcriptional units. In particular, the catBC genes are co-transcribed in the presence of benzoate.

Primer No.	Primer name	Sequence (5'-3')	Amplified fragmenta
1	RT1-5'	AGCGAGAACCAATGGC	782 bp benRA intergenic region
2	RT1-3'	TAGTCGATTCCCAGGG
3	RT2-5'	GCACTGGATCGAGGGAGC	456 bp benAB intergenic region
4	RT2-3'	GTTGTGCGAGGTGCGTGT
5	RT3-5'	GCTTTCGCTACAAGACCG	503 bp benBC intergenic region
6	RT3-3'	CGCACGTTGCTGATGGTC
7	RT4-5'	CGAACCCAAACACCTCAA	546 bp benCD intergenic region
8	RT4-3'	CTCGGCCTCGATCTCATG
9	RT5-5'	TACCAGGAACATGAGAT	610 bp benDK intergenic region
10	RT5-3'	ACGTCTACTTTTCGCATG
11	RT6-5'	GTTCTTCTGTTGCCTG	1074 bp benK and catB intergenic region
12	RT6-3'	TCTTCGATGTCCTTAG
13	RT7-5'	CCTTCGTCACCCTCAACA	309 bp catBC intergenic region
14	RT7-3'	CTTCACGCATCAGGCTCT
15	RT8-5'	GAAGATGATCGTGAAAC	1030 bp catCA intergenic region
16	RT8-3'	TGAAGAAATGAATGTGC
17	benA-5'	CGGCTCGTCCACCTATGTCTAT	186 bp internal fragment
18	benA-3'	AAACCACCGCCCTTCTTGC
19	catB-5'	CCTTCGTCACCCTCAACAG	159 bp internal fragment
20	catB-3'	TCCAGGCTCAGGCCAAGAC
21	pcaD-5'	TTCGCCGAGCATTTCCG	173 bp internal fragment
22	pcaD-3'	CCGATCAGTCCGCCCAT

^aPCR reactions were carried out with the sets of primers indicated to the left.

BenR activates expression of the benABCDoperon in responseto benzoate

In pseudomonads, benzoate catabolism is initiated by the benABCD operon encoding benzoate dioxygenase (BenABC) and 2-hydro-1,2-dihydroxybenzoate dehydrogenase (BenD), whose expression is positively regulated by BenR [9, 31]. In this study, we found that disruption of the benR gene resulted in a simultaneous loss of the ability to utilize benzoate, but this mutant could grow on catechol, suggesting that BenR might be the sole activator of expression of benABC in A1501. To confirm this, we searched the promoter sequence of benA using in silico analysis. The nucleotide sequence upstream of the benABCD operon has the following sequence features: a putative -10/-35-type promoter, a putative BenR-binding region, and a predicted translational start site (Figure 6A). Comparison with the experimentally well-characterized BenR-binding sequences in P. putida [9] indicated a highly conserved BenR site in the promoter region of the A1501 benA gene (Figure 6A). To determine whether benR is required for activation of the P_benA promoter, the expression level of the benABCD operon was tested in the benR mutant A1601. Quantitative real-time PCR results demonstrated that a significant increase in transcription from the P_benApromoter was seen in wild-type A1501 when benzoate was included in the growth medium, whereas the addition of catechol or cis,cis-muconate had a very weak effect (Figure 6B). When BenR was absent, transcription from the P_benA promoter was highly repressed, irrespective of the presence or absence of the inducer (Figure 6B). As reported in P. putida [9], these results led us to conclude that the benABCD operon is under the control of BenR in response to benzoate in A1501.

Benzoate-mediated induction of the catBCoperon in A1501

In P. putida, the catBC operon encodes cis,cis-muconate lactonizing enzyme I (CatB) and muconolactone isomerase (CatC), which catalyze the second and third steps of the catechol branch of the β-ketoadipate pathway, respectively [8]. The transcription of this operon requires CatR and cis,cis-muconate [32]. Additionally, the translational starts of catR and catBC are separated by 136 bp of intervening DNA containing -35/-10-type promoters and the CatR-binding sites in P. putida [13, 33]. However, we found that only 29 nucleotides are present in the noncoding regions between benK and catB in A1501, suggesting that the promoter region of the catBC operon overlaps with the coding region of the benK gene. The promoter region of the catBC operon from A1501 shows very low similarity to those of the three other Pseudomonas strains, notably the lack of the typical binding site for CatR present in the catB promoter region of other Pseudomonas strains (Figure 6C). Although a catR orthologue could not be identified in A1501, quantitative real-time PCR experiments indicated that benzoate has the strongest induction effect on expression of the catBC operon (Figure 6D). Since benzoate induces expression of catB in the benR mutant background and this mutant is unable to metabolize benzoate, we proposed that induction of the catBC expression is not due to the production of benzoate metabolites, such as cis,cis-muconate. As reported in P. putida, induction of the catBC operon requires cis,cis-muconate, an intermediate of benzoate degradation, and CatR, a well-studied activator in the β-ketoadipate pathway [32]. However, benzoate itself has a significant induction effect on expression of the catBC operon in A1501, strongly suggesting the existence of an uncharacterized regulatory mechanism.

Benzoate degradation in A1501 is subject to carbon catabolite repression

In Pseudomonas and Acinetobacter strains, the Crc global regulator controls the expression of genes involved in benzoate degradation when other preferred carbon sources are present in the culture medium [16, 17]. Based on sequence comparison, we found a Crc-like protein in the A1501 genome (Figure 1A). The A1501 Crc-like protein shows highest amino acid identity with P. aeruginosa Crc (86%), whereas relatively low amino acid identity (only 38%) is observed between A1501 and A. baylyi Crc proteins.

Benzoate degradation by A1501 involves the oxidation of benzoate into catechol in a two-step process catalyzed by BenABC and BenD, two peripheral pathway enzymes of the catechol pathway. The catechol aromatic ring is converted by the action of CatA, CatB and CatC to cis,cis-muconate, and then to β-ketoadipate-enol-lactone, which is transformed into acetyl-CoA and succinyl-CoA by PcaD, PcaIJ, and PcaF from the β-ketoadipate pathway. Therefore, the benA, catB, and pcaD genes were selected for further analysis. In the presence of the inducer benzoate, highly significant differences in expression were observed, depending on the nature of the non-inducing carbon source (Figure 7). The expression of the three selected genes was most efficiently induced by benzoate when cells were grown on lactate and succinate alone, but was decreased significantly when the carbon source was glucose or acetate (Figure 8). When cells grew on lactate, the expression of benA and catB was efficiently induced by benzoate, respectively; when glucose plus succinate was used as the carbon source, induction was significantly lower (Figure 7A, B). The benA and catB genes showed a similar repression pattern to the pcaD gene, with the slight difference being that acetate was an intermediate-repressing carbon source. Using glucose or succinate as individual carbon sources led to a strong decreasing or increasing effect on expression of the pcaD gene, respectively, whereas growth on a combination of glucose plus succinate and inducer resulted in high induction (Figure 7C). These results suggest that benzoate degradation in A1501 is subject to carbon catabolite repression. Our experimental evidence, combined with the identification of the Crc-like protein in A1501, may be indicative of distinct activities of Crc at different genes or in various bacteria, as previously shown in A. baylyi and P. putida [34, 35]. Further experiments are required to construct an A1501 mutant lacking the Crc-like protein and to investigate role of this protein in carbon catabolite repression.

4-hydroxybenzoate enhances the ability of A1501 to degrade benzoate

A study reported that high concentrations of aromatic hydrocarbons are harmful to cells because they disrupt membrane components [36]. In the plate assay, A1501 grew extremely poorly on 4-hydroxybenzoate as the sole carbon source with colonies of less than 1.0 mm in diameter after 3 days, whereas it produced normal-sized colonies (> 5 mm) on benzoate alone in the same period. These results indicate that 4-hydroxybenzoate itself directly inhibits A1501 growth, which is likely caused by the toxicity of 4-hydroxybenzoate. It is unclear whether the lack of pcaK results in the loss of 4-hydroxybenzoate transport, leaving A1501 unable to metabolize 4-hydroxybenzoate efficiently. In subsequent experiments, growth of A1501 was examined in a mixture of 4 mM benzoate and 0.4 mM 4-hydroxybenzoate. A1501 showed a shorter lag phase and a higher growth rate when cells were grown on the mixture than when benzoate was supplied alone (Figure 8A). Furthermore, under the latter growth conditions, the culture gradually became dark brown in color because of autoxidation of the accumulated catechol (data not shown). However, when the 4-hydroxybenzoate concentration increased to 0.8 mM, growth of A1501 was completely inhibited (Figure 8A). These results indicate that 4-hydroxybenzoate at low concentrations can enhance the ability of A1501 to grow on benzoate.

We then evaluated the effect of 4-hydroxybenzoate on the metabolism of benzoate using HPLC. When 4 mM benzoate alone was provided to the culture, it was completely consumed within 26 h, and metabolic intermediates were present. When 4 mM benzoate and 0.4 mM 4-hydroxybenzoate were provided together as growth substrates, benzoate was completely consumed within 18 h, while no discernible loss of 4-hydroxybenzoate was detected (Figure 8B). Additionally, analysis of the intracellular metabolites by HPLC revealed accumulation of catechol derived from benzoate both in the presence and absence of 4-hydroxybenzoate in the growth medium. The concentration of catechol reached 0.28 mM when A1501 grew on benzoate alone, whereas the concentration of catechol reached approximately 0.12 mM when both benzoate and 4-hydroxybenzoate were in the growth medium (Figure 8C). Collectively, these results suggest that 4-hydroxybenzoate can significantly enhance the ability of A1501 not only to degrade benzoate, but also to remove the catechol accumulated from benzoate.

Bacterial strains, plasmids and growth conditions

The bacterial strains and plasmids used in this work are listed in Table 1. Bacterial strains were grown in Luria-Bertani (LB) and minimal lactate-containing medium (medium K), as previously described [43]. When required, carbon sources were supplemented at the following final concentrations: 4 mM glucose, 4 mM succinate, 4 mM lactate, 4 mM acetate, 4 mM benzoate, 0.4 mM catechol and 0.4 mM 4-hydroxybenzoate. The following antibiotics were added as required at the indicated final concentrations: 10 μg/ml tetracycline (Tc) and 50 μg/ml kanamycin (Km).

Construction of nonpolar mutants

We constructed a nonpolar insertion into the benR, pcaR, and pcaD genes, respectively, by homologous suicide plasmid integration, as described previously [44], using pK18mob as the vector [45]. DNA fragments (~300 bp) were amplified using the total DNA of A1501 as the template and appropriate oligonucleotide primers. Oligonucleotide primers were designed to generate amplicons for the creation of nonpolar mutations enabling transcription of downstream genes. The amplicons were ligated into the vector pK18mob and the resulting plasmids were introduced into P. stutzeri A1501 from Escherichia coli JM109 by triparental conjugation using pRK2013 [46] as the helper plasmid. The nonpolar mutant strains A1601, A1602, and A1603 were generated in which benR, pcaR, and pcaD, respectively, were disrupted without blocking the transcription of downstream genes. Correct recombination was confirmed by PCR analysis. For further growth complementation assays, we used the broad host vector pLAFR3 to construct three complementary plasmids, pLbenR, pLpcaD and pLpcaR, as described previously [47]. Three complementary plasmids and the corresponding complementary strains are listed in Table 1.

RT-PCR and Quantitative real-time PCR

Total RNA was isolated with an SV Total RNA Isolation System (Promega, Madison, WI, USA) and treated with RNase-free DNase I (Promega). The integrity of RNA was analyzed by agarose gel electrophoresis. To check for DNA contamination, samples were analyzed with PCR using primers for benA. First-strand cDNAs were synthesized from 1 μg of total RNA in a 20 μl reaction volume using the Protoscript First-Strand cDNA Synthesis Kit (New England Biolabs, Ipswich, MA, USA).

For quantitative real-time PCR (Q-PCR) experiments, primer pairs, as shown in Table 2, were designed based on the published reference genome sequence of P. stutzeri A1501 using the Primer 4 server. Amplicons (100 to 200 bp) and reaction specificity were confirmed by agarose gel electrophoresis and product dissociation curves. Q-PCR reactions contained 1 μl of cDNA, 10 μl of 2× QuantiTect SYBR Green PCR Master Mix (Qiagen, Hilden, Germany), 0.5 μl of each primer (20 μM stock), and 8 μl of RNase-free water. Amplifications were conducted on an ABI PRISM 7000 Real Time PCR System (Applied Biosystems, Foster City, CA, USA) under the following conditions: 10 min at 95°C, followed by 40 cycles of 15 s at 95°C, 31 s at 55°C, and 31 s at 72°C, followed by a melting-curve program (55°C to 99°C, with a 5-s hold at each temperature). Q-PCR data were analyzed using the ABI PRISM 7000 Sequence Detection System Software (Applied Biosystems). All cDNA samples were run in triplicate. The expression of l6S rRNA was used as an internal control and the signal was used to normalize variations due to different reverse transcription efficiencies. The comparative CT (threshold cycle) method was used to determine the average fold induction of mRNA by comparing the CT of the target gene to that of the reference gene, as described previously [48]. The average fold change and standard deviation from three independent RNA samples are reported for each point tested.

High-performance liquid chromatography (HPLC) analysis

To monitor metabolism, the pcaD mutant and wild-type strains were grown in minimal medium supplemented with benzoate or a mixture of benzoate and 4-hydroxybenzoate. One-milliliter culture samples were centrifuged to pellet cells. Any cells remaining in the supernatant were removed by passage through a low-protein-binding, 0.22 μm pore size, syringe filter (MSI, Westborough, MA, USA). HPLC analysis was performed using an Agilent Technologies (Santa Clara, CA, USA) 1200 series chromatography system. A 20-μl sample of the filtrate was analyzed on a C18 reverse-phase HPLC column (Agilent Technologies). Elution at a rate of 0.8 ml/min was carried out with 30% acetonitrile and 0.1% phosphoric acid, and the eluant was detected at 254 nm. Under these conditions, the retention times for benzoate, catechol, cis, cis-muconate, and 4-hydroxybenzoate standards were 6.071, 2.388, 3.358, and 2.770 min, respectively. Peak areas corresponding to standard and experimental samples were integrated using the manufacturer's software package (Agilent Technologies).