An Aerobic Hybrid Phthalate Degradation Pathway via Phthaloyl-Coenzyme A in Denitrifying Bacteria

Phthalic acid esters (PAEs) are industrially produced on a million-ton scale per year and are predominantly used as plasticizers. They are classified as environmentally relevant xenobiotics with a number of adverse health effects, including endocrine-disrupting activity. Biodegradation by microorganisms is considered the most effective process to eliminate PAEs from the environment. It is usually initiated by the hydrolysis of PAEs to alcohols and o-phthalic acid. Degradation of o-phthalic acid fundamentally differs in aerobic and anaerobic microorganisms; aerobic phthalate degradation heavily depends on dioxygenase-dependent reactions, whereas anaerobic degradation employs the oxygen-sensitive key enzyme phthaloyl-CoA decarboxylase. We demonstrate that aerobic phthalate degradation in facultatively anaerobic bacteria proceeds via a previously unknown hybrid degradation pathway involving oxygen-sensitive and oxygen-dependent key enzymes. Such a strategy is essential for facultatively anaerobic bacteria that frequently switch between oxic and anoxic environments.

ABSTRACT

The degradation of the xenobiotic phthalic acid esters by microorganisms is initiated by the hydrolysis to the respective alcohols and ortho-phthalate (hereafter, phthalate). In aerobic bacteria and fungi, oxygenases are involved in the conversion of phthalate to protocatechuate, the substrate for ring-cleaving dioxygenases. In contrast, anaerobic bacteria activate phthalate to the extremely unstable phthaloyl-coenzyme A (CoA), which is decarboxylated by oxygen-sensitive UbiD-like phthaloyl-CoA decarboxylase (PCD) to the central benzoyl-CoA intermediate. Here, we demonstrate that the facultatively anaerobic, denitrifying Thauera chlorobenzoica 3CB-1 and Aromatoleum evansii KB740 strains use phthalate as a growth substrate under aerobic and denitrifying conditions. In vitro assays with extracts from cells grown aerobically with phthalate demonstrated the succinyl-CoA-dependent activation of phthalate followed by decarboxylation to benzoyl-CoA. In T. chlorobenzoica 3CB-1, we identified PCD as a highly abundant enzyme in both aerobically and anaerobically grown cells, whereas genes for phthalate dioxygenases are missing in the genome. PCD was highly enriched from aerobically grown T. chlorobenzoica cells and was identified as an identical enzyme produced under denitrifying conditions. These results indicate that the initial steps of aerobic phthalate degradation in denitrifying bacteria are accomplished by the anaerobic enzyme inventory, whereas the benzoyl-CoA oxygenase-dependent pathway is used for further conversion to central intermediates. Such a hybrid pathway requires intracellular oxygen homeostasis at concentrations low enough to prevent PCD inactivation but sufficiently high to supply benzoyl-CoA oxygenase with its cosubstrate.

IMPORTANCE Phthalic acid esters (PAEs) are industrially produced on a million-ton scale per year and are predominantly used as plasticizers. They are classified as environmentally relevant xenobiotics with a number of adverse health effects, including endocrine-disrupting activity. Biodegradation by microorganisms is considered the most effective process to eliminate PAEs from the environment. It is usually initiated by the hydrolysis of PAEs to alcohols and o-phthalic acid. Degradation of o-phthalic acid fundamentally differs in aerobic and anaerobic microorganisms; aerobic phthalate degradation heavily depends on dioxygenase-dependent reactions, whereas anaerobic degradation employs the oxygen-sensitive key enzyme phthaloyl-CoA decarboxylase. We demonstrate that aerobic phthalate degradation in facultatively anaerobic bacteria proceeds via a previously unknown hybrid degradation pathway involving oxygen-sensitive and oxygen-dependent key enzymes. Such a strategy is essential for facultatively anaerobic bacteria that frequently switch between oxic and anoxic environments.

INTRODUCTION

Esters of ortho-phthalic acid (PAE) are industrially produced on the million-ton scale per year and are predominantly used as plasticizers to improve flexibility and other properties of crude-oil-derived plastic materials (1, 2). PAEs are categorized as environmentally relevant xenobiotics that have been massively produced only since the 1960s (3). Among many adverse health effects, the endocrine-disrupting activity of PAEs is of major concern (4, 5). As PAEs are not covalently bound to plastic polymers, they can easily leach out during production, transport, or disposal, causing increasing environmental contamination.

Biodegradation of PAEs by microorganisms has been identified to be the major process for PAE elimination from the environment (6–9). It is usually initiated by hydrolysis, yielding alcohols and ortho-phthalate (hereafter, phthalate). The degradation of phthalate fundamentally differs in aerobic and anaerobic microorganisms. In aerobic phthalate-degrading bacteria and fungi, two-component Rieske nonheme dioxygenases and cis-diol dehydrogenases convert phthalate to dihydroxyphthalate intermediates that are readily converted to protocatechuate (3,4-dihydroxybenzoate) and CO₂ by cofactor-free decarboxylases (Fig. 1). Protocatechuate then serves as the substrate for ring-cleaving intradiol or extradiol dioxygenases. A variant of such dioxygenase-dependent pathways proceeds via cis-2,3-dihydroxybenzoate (9).

FIG 1.

Microbial degradation of phthalate under aerobic and anaerobic conditions. (A) Aerobic phthalate degradation is, in most cases, initiated by phthalate-3,4-dioxygenases (top pathway) or phthalate-4,5-dioxygenases (bottom pathway), followed by aromatization by cis-diol-dehydrogenases and formation of protocatechuate (3,4-dihydroxybenzoate) by cofactor-free decarboxylases. The latter is then cleaved by intradiol or extradiol oxygenases. (B) Anaerobic phthalate degradation is initiated by thioesterification with CoA either by a CoA transferase (SCPCT, denitrifying bacteria) or by a phthalate CoA ligase (PCL; sulfate-reducing bacteria). Decarboxylation to the central intermediate is catalyzed by phthaloyl-CoA decarboxylase (PCD). The prFMN cofactor of PCD is formed by the prenyltransferase UbiX, with dimethylallyl-monophosphate (DMAP) serving as a prenylating cosubstrate. Enzyme names are presented in italics.

Insights into phthalate degradation in anaerobic bacteria have only recently been obtained in studies with the sulfate-reducing Desulfosarcina cetonia and with denitrifying Thauera and Aromatoleum species (10–12). In denitrifying bacteria, in vitro evidence showed that phthalate is first activated to phthaloyl-coenzyme A (CoA) by a succinyl-CoA-dependent CoA transferase, followed by decarboxylation to benzoyl-CoA, the central intermediate of anaerobic aromatic catabolism (Fig. 1). Benzoyl-CoA is then dearomatized to cyclohexa-1,5-diene-1-carboxyl-CoA by ATP- and ferredoxin-dependent class I benzoyl-CoA reductases (13–15). Further degradation of the product proceeds via the so-called benzoyl-CoA degradation pathway comprising β-oxidation-like reaction sequences and hydrolytic ring cleavage, which finally converts benzoyl-CoA to three acetyl-CoA and CO₂ (16). In the sulfate-reducing bacterium Desulfosarcina cetonia, a similar anaerobic phthalate-degradation pathway was described. Here, the major difference is that phthaloyl-CoA is formed by an ATP-dependent CoA ligase, rather than by a succinyl-CoA-dependent CoA transferase (10). Subsequent dearomatization is then achieved by an ATP-independent class II benzoyl-CoA reductase complex (17, 18).

The formation of the extremely labile phthaloyl-CoA intermediate is a major challenge of anaerobic phthalate degradation. Its planar ring system strongly promotes spontaneous intramolecular substitution of the CoA moiety by the vicinal carboxyl group, yielding free CoA and phthalic anhydride that hydrolyzes to phthalate. Phthaloyl-CoA has a life span of a few minutes and is possibly the most unstable CoA ester in biology (19).

Proteogenomic studies of denitrifying and sulfate-reducing phthalate-degrading bacteria revealed phthalate-induced gene clusters comprising genes putatively encoding (i) tripartite ATP-independent periplasmic (TRAP) transporters system for the uptake of phthalate, (ii) class III CoA transferases (denitrifying bacteria) or AMP-forming CoA ligases (sulfate reducing bacteria), (iii) UbiD-like decarboxylases, and (iv) UbiX-like prenyltransferases that are involved in cofactor maturation of UbiD-like enzymes (10–12) (Fig. 1).

Heterodimeric succinyl-CoA:phthalate CoA transferase (SCPCT) from Aromatoleum aromaticum EbN1 was heterologously produced in Escherichia coli by expression of the phthalate-induced sptAB genes (19). The isolated recombinant enzyme predominantly catalyzed the phthalate-dependent hydrolysis of succinyl-CoA, whereas phthaloyl-CoA was formed only at submicromolar concentrations due to its rapid decay. The second step of anaerobic phthalate degradation is catalyzed by phthaloyl-CoA decarboxylase (PCD). It belongs to the UbiD enzyme family of (de)carboxylases that contain a recently identified prenylated flavin mononucleotide (prFMN) cofactor (20, 21). A PCD has so far been isolated and biochemically characterized only from Thauera chlorobenzoica 3CB-1 (22). It specifically catalyzes the decarboxylation of phthaloyl-CoA to benzoyl-CoA, as determined in a coupled assay starting with phthalate and succinyl-CoA in the presence of SCPCT. Hexameric PCD contains prFMN, K⁺, and Fe²⁺ per subunit; the metal cations are proposed to be involved in prFMN binding. In air, the Fe²⁺ was oxidized to Fe³⁺, resulting in a loss of the prFMN cofactor and activity with a half-life of 13.3 ± 3 min. The Fe²⁺ found in PCD substitutes for Mn²⁺ found in oxygen-tolerant UbiD-like enzymes from aerobic organisms (20, 22). PCD was the most abundant enzyme in phthalate-grown cells, and its cellular concentration (≈140 μM) exceeds that of the substrate phthaloyl-CoA by 250-fold. As a result, the labile phthaloyl-CoA produced by SCPCT is immediately captured and decarboxylated by PCD. To minimize phthaloyl-CoA accumulation and its subsequent hydrolysis, optimal PCD:SCPCT activity ratios of 4:1 were identified in vitro, fitting to the observed relative cellular abundance of both enzymes (19). The prFMN cofactor of UbiD-like enzymes, such as PCD, is formed by UbiX-type prenyltransferases that in bacteria use dimethylallyl-monophosphate (DMAP) as a prenylating cosubstrate (23, 24) (Fig. 1). Putative ubiX-like genes are present in all phthalate-induced gene clusters, suggesting that their products act as prenyltransferases for synthesis of the prFMN cofactor of PCDs (9).

Here, we studied aerobic phthalate degradation in T. chlorobenzoica 3CB-1 and Aromatoleum evansii KB740 that were previously reported to use phthalate as a growth substrate under denitrifying conditions. Surprisingly, there was no evidence for the involvement of a phthalate dioxygenase, but instead, T. chlorobenzoica 3CB-1, A. evansii KB740, and probably all other facultatively anaerobic phthalate-degrading bacteria employ a hybrid pathway with reaction sequences from anaerobic phthalate and aerobic benzoyl-CoA oxidation pathways.

RESULTS

Aerobic growth of denitrifying bacteria with phthalate.

The denitrifying Betaproteobacteria T. chlorobenzoica 3CB-1 and Aromatoleum evansii KB740 (the latter was formerly referred to as Azoarcus evansii KB740 [25]) were previously shown to grow with phthalate under denitrifying conditions (11). We tested whether these organisms are also capable of using phthalate as a growth substrate with oxygen as the electron acceptor. Cultivation with 10 mM phthalate was carried out in a 10-liter fermenter where the oxygen concentration was continuously monitored and the flushing rate with air was stepwise increased with increasing optical density (OD) (Fig. 2A and B). Aerobic growth with phthalate was observed with both strains with doubling times of 6.7 (T. chlorobenzoica 3CB-1) and 8 h (A. evansii KB740) and was approximately 1.6-fold lower than during optimal growth with phthalate under denitrifying conditions (11); maximally, 30 to 40 g of cells (wet mass) was obtained within 3 to 4 days.

FIG 2.

Aerobic growth of T. chlorobenzoica 3CB-1 and A. evansii KB740 with phthalate. (A and B) Representative growth curves for aerobic cultivation in a 10-liter fermenter with 10 mM phthalate of T. chlorobenzoica 3CB-1 (A) and A. evansii KB740 (B). ○, OD₅₇₈; ■, phthalate. (C and D) Dependence of the doubling time (t_D) on the agitation rate during aerobic growth with 10 mM phthalate (□) and benzoate (■) in 100-ml Erlenmeyer flasks on a rotary shaker. (C) T. chlorobenzoica 3CB-1. (D) A. evansii KB740. Error bars indicate mean deviation of three biological replicates.

To test whether the oxygen concentration had an effect on the doubling time during aerobic growth with phthalate or benzoate, we cultivated T. chlorobenzoica 3CB-1 and A. evansii KB740 in 100 Erlenmeyer flasks on a rotary shaker at various agitation rates (0 to 400 rpm) (Fig. 2C and D). As expected, a slightly lower doubling time was observed in agitated (>100 rpm) versus nonagitated cultures (Fig. 2C and D). However, even at the highest agitation rate (400 rpm), the doubling time did not increase. Doubling time was 2-fold (T. chlorobenzoica 3CB-1) or 1.6-fold (A. evansii KB740) lower in benzoate versus phthalate-grown cells.

Differentially abundant proteins in T. chlorobenzoica 3CB-1.

To identify prominent proteins that were differentially abundant during aerobic and anaerobic growth with phthalate, we analyzed the soluble protein fractions of T. chlorobenzoica 3CB-1 after growth with phthalate and benzoate under aerobic and anaerobic conditions. After one-dimensional SDS-PAGE analysis (10% acrylamide), a prominent 60-kDa band was observed that was present in cells grown with phthalate/nitrate and phthalate/air (Fig. 3, band B). The 60-kDa bands obtained from aerobically and anaerobically grown cells were excised from the gel, tryptically digested, and analyzed by ultraperformance liquid chromatography (UPLC) coupled to quadrupole time of flight (Q-TOF) mass spectrometry (MS) after electrospray ionization (ESI). For details of MS results, see Table S1 in the supplemental material. In extracts from aerobically and anaerobically grown cells, the proteins were identified as PCD, which had recently been purified from wild-type cells grown with phthalate under denitrifying conditions (22) (UniProt accession no. A0A193DUB4, 68.5% sequence coverage) (see Table S1 in the supplemental material). A band apparently comigrating with PCD in cells grown anaerobically with benzoate (Fig. 3, band E) was identified as a benzoate CoA ligase/nitrite reductase component, whereas no tryptic peptides from PCD were detected in this band. The PCD band was reproducibly 10% to 20% more intense in cells grown with phthalate in air than in cells grown with phthalate/nitrate. Another protein band (band F) was identified as UbiX-like prenyltransferase in cells grown with phthalate but not in cells grown with benzoate (NCBI accession no. WP_075149175.1). The BoxAB components of benzoyl-CoA monooxygenase, the key enzyme of the aerobic box-degradation pathway in denitrifying bacteria, together with the BoxC benzoyl-CoA dihydrodiol lyase component were identified in cells grown aerobically, with either phthalate or benzoate as the carbon source (Fig. 3, bands A, C, D; Table S1). Among these components, BoxC had a higher abundance in aerobically versus anaerobically grown cells. Note that under anaerobic conditions, additional proteins of the benzoyl-CoA degradation pathway are induced with similar molecular masses of the aerobic BoxA and BoxB components.

FIG 3.

SDS gel of extracts from T. chlorobenzoica 3CB-1 grown aerobically/anaerobically with phthalate/benzoate (10% acrylamide). Lane 1, phthalate/O₂; lane 2, phthalate/nitrate; lane 3, benzoate/O₂; lane 4, benzoate/nitrate (15 μg protein per lane); M, molecular mass standards. The arrow labels refer to proteins bands that were excised, tryptically digested, and analyzed by UPLC-ESI-Q-TOF mass spectrometry. The results of these analyses are summarized in Table S1 in the supplemental material. The assigned proteins are the following: benzoyl-CoA dihydrodiol lyase (component BoxC) (A), PCD (B), benzoyl-CoA monooxygenase component BoxB (C), benzoyl-CoA monooxygenase component BoxA (D), benzoate CoA ligase/NirS nitrite reductase component (E), and UbiX-like prenyltransferase (F).

These results suggest that PCD and UbiX-like prenyltransferase are employed for both aerobic and anaerobic phthalate degradation via phthaloyl-CoA, whereas further benzoyl-CoA degradation involves benzoyl-CoA monooxygenases under aerobic conditions. Aerobic phthalate degradation via phthaloyl-CoA has not been reported before. Instead, phthalate-specific two-component Rieske nonheme iron dioxygenases are employed. Using BLAST, we searched in the genomes of T. chlorobenzoica 3CB-1 and A. aromaticum EbN1 for genes encoding such dioxygenases using phthalate 3,4-dioxygenase from Arthrobacter keyseri (UniProt accession no. Q9AGK3), phthalate 4,5-dioxygenase from Pseudomonas putida (UniProt accession no. Q05182), and phthalate 2,3-dioxygenase from Pseudomonas sp. PTH10 (NCBI accession no. BBB03253) as query sequences. The only sequence with an expect value threshold below e⁻⁵⁰ was a gene product annotated as vanillate O-demethylase (VanB), which shares a sequence identity of 41% with phthalate 4,5-dioxygenase but is expected to catalyze a methyl transfer reaction.

In vitro assays for enzymes involved in aerobic phthalate degradation.

To substantiate the expected PCD-, UbiX-, and SCPCT-dependent aerobic phthalate degradation pathway in T. chlorobenzoica 3CB-1 and A. evansii KB740, in vitro assays were carried out. Using extracts from cells grown aerobically with phthalate, the succinyl-CoA (0.4 mM) and phthalate (0.5 mM)-dependent formation of benzoyl-CoA was observed with a specific activity of 42.8 ± 8 nmol min⁻¹mg⁻¹ (T. chlorobenzoica 3CB-1) and 34 ± 5 nmol min⁻¹mg⁻¹ (A. evansii KB740) using the UPLC-based benzoyl-CoA monitoring assay. In agreement with studies of anaerobic phthalate degradation (11), acetyl-CoA could not substitute for succinyl-CoA in these assays. Moreover, no succinyl-CoA and phthalate-dependent formation of benzoyl-CoA was observed in extracts from cells grown aerobically with benzoate. The activity was reproducibly 1.1- to 1.2-fold higher than in extracts of cells grown with phthalate/nitrate, which matches the higher abundance of PCD in aerobically versus anaerobically grown cells. Similar to what was seen in the in vitro assays with extracts from anaerobically grown cells, virtually no phthaloyl-CoA intermediate was detectable during the succinyl-CoA dependent conversion of phthalate to benzoyl-CoA plus CO₂ (<2 μM). The amount of free CoA formed was approximately 10% of the succinyl-CoA added; this value is comparable to previous observations (11, 22).

The oxygen sensitivity of phthalate- and succinyl-CoA-dependent benzoyl-CoA formation in extracts from aerobically grown cells was investigated by incubating anaerobically prepared extracts under aerobic and anaerobic conditions in the presence of 50 mM KCl. Notably, K⁺ ions were recently reported to have a protective effect on PCD toward oxidative damage (11). The anaerobically incubated enzyme showed, in contrast to previous studies (11), a slow oxygen-independent background instability over time. Under aerobic conditions, PCD activity in extracts from anaerobically grown cells decreased time dependently with a half-life of 50 ± 8 min (mean value ± standard deviation of three biological replicates) (Fig. 4). This oxygen sensitivity of PCD was reproducibly lower in extracts from aerobically grown cells than extracts from cells grown under denitrifying conditions (half-life, 32 ± 7 min). As the induction of enzymes involved in detoxification of reactive oxygen species is specifically expected in aerobically grown cells, the apparent lower sensitivity of PCD is likely explained by the action of oxygen-scavenging enzymes. Notably, isolated PCD is even more sensitive in air, with a half-life of 13.3 min (22). In summary, the results indicate that levels of oxygen sensitivity of PCD in cell extracts were similar during aerobic and anaerobic growth with phthalate.

FIG 4.

Oxygen sensitivity of PCD-catalyzed benzoyl-CoA formation from phthalate and succinyl-CoA in extracts of T. chlorobenzoica 3CB-1. Anaerobically prepared extracts of cells grown with phthalate/O₂ (A) and anaerobically with phthalate/nitrate (B). ■, incubation in air; □, anaerobic incubation in the glove box (95:5 N₂:H₂, by volume).

Enrichment, activity, and cofactors of PCD from T. chlorobenzoica 3CB-1 cells grown aerobically with phthalate.

To verify whether the PCDs involved in aerobic and anaerobic phthalate degradation in T. chlorobenzoica 3CB-1 are indeed identical enzymes, PCD was purified from extracts of cells grown aerobically with phthalate. The benzoyl-CoA-forming activity in extracts was enriched via anion-exchange chromatography on DEAE-Sepharose and gel filtration in an anaerobic glove box (95% N₂/5% H₂, by volume). During enrichment, the PCD activity monitoring assays contained purified SCPCT to form the substrate of PCD from succinyl-CoA and phthalate. A 60-kDa band was 3.4-fold enriched with a yield of 26% (Table 1; Fig. 5). The specific activity (mean ± standard deviation) was 145 ± 32 nmol min⁻¹mg⁻¹, which was similar to the enzyme enriched from anaerobically grown cells (140 ± 5 nmol min⁻¹mg⁻¹) (22). The 60-kDa band was excised and tryptically digested. UPLC-ESI-Q-TOF analysis identified the protein as PCD (NCBI accession no. WP_075149174.1). The enzyme contained 1.7 ± 0.22 mol Fe and 0.98 ± 0.13 mol prFMN per mol subunit of PCD, indicating that PCD contains the same cofactors regardless of whether it is produced under aerobic or denitrifying conditions.

TABLE 1.

Enrichment of PCD from extracts of T. chlorobenzoica 3CB-1 grown with phthalate/O₂^a

Enzyme or enrichment step	Total protein (mg)	Total activity (nmol min–1)	Specific activity (nmol min–1 mg–1) (mean ± SD)	Enrichment (fold)	Recovery (%)
100,000 × g supernatant	356	15,240	42.8 ± 8	1	100
DEAE-Sepharose	85	7,760	91.2 ± 20	2.1	51
Gel filtration	27.5	3,980	145 ± 32	3.4	26

^aPCD was purified from 4.25 g (wet mass) of Thauera chlorobenzoica cells.

FIG 5.

Enrichment of PCD from soluble extracts of T. chlorobenzoica 3CB-1 grown aerobically with phthalate. M, molecular mass standard (numbers on left refer to molecular masses in kDa); CE, cell extract; DEAE, PCD activity containing fraction obtained after DEAE chromatography; SE, PCD activity containing fraction after size exclusion chromatography. The arrow points to the enriched protein band identified as PCD from T. chlorobenzoica 3CB-1 (NCBI accession no. WP_075149174.1).

DISCUSSION

In this work, we provided evidence that T. chlorobenzoica 3CB-1 and A. evansii KB 740 use an identical enzyme inventory for the conversion of phthalate and succinyl-CoA to benzoyl-CoA and CO₂ during aerobic and anaerobic growth with phthalate (Fig. 6). This finding was not expected, as aerobic phthalate degradation via protocatechuate or 2,3-dihydroxybenzoate has so far always been associated with the use of dioxygenases (6–8). We propose that all denitrifying phthalate-degrading bacteria, so far all belonging to the genera Aromatoleum and Thauera, degrade phthalate via phthaloyl-CoA under both oxic and anoxic conditions. Under aerobic conditions, the benzoyl-CoA formed is subsequently degraded in the so-called benzoyl-CoA oxidation (box) pathway that involves only a single oxygen-dependent enzyme, a class I diiron monooxygenase referred to as benzoyl-CoA-2,3-epoxidase (16, 26, 27). Notably, this enzyme can itself be considered a cytoplasmic oxygen-scavenging system to protect PCD. The obvious adaptation to low-oxygen concentrations also requires high-affinity respiratory oxidases. Indeed, genes for cytochrome cbb₃ oxidase subunits, which are abundant in organisms living under low oxygen concentrations (28), are present in all phthalate-degrading Thauera and Aromatoleum species (e.g., TChl_RS162240/45 in T. chlorobenzoica 3CB-1).

FIG 6.

Proposed aerobic phthalate degradation pathway in denitrifying bacteria. Left (light red background), enzymatic steps involved in the conversion of phthalate to benzoyl-CoA known from anaerobic phthalate degradation pathways; right (light blue background), enzymatic steps of the aerobic benzoyl-CoA oxidation (box) pathway known from facultatively anaerobic bacteria (16).

The supply of cytoplasmic benzoyl-CoA-2,3-epoxidase with its cosubstrate and the sufficient protection of PCD from oxygen damage depend on a number of factors that may vary in different organisms: (i) the oxygen-sensitivity of PCD, (ii) the affinity of benzoyl-CoA-2,3-epoxidase to oxygen, (iii) the presence and efficiency of cytoplasmic redox buffers and/or oxygen detoxification systems, and (iv) the efficiency of respiratory oxygen consumption. The doubling time of T. chlorobenzoica 3CB-1 with phthalate was unaffected at high oxygen concentrations in the medium (Fig. 2C). This finding suggests that the intracellular oxygen concentration is maintained at an optimal level over a wide range of extracellular oxygen concentrations. Future work will be required to understand the full interplay of adaptations in facultatively anaerobic, aromatic compound-degrading bacteria to cope with frequently changing oxygen concentrations.

The oxygen sensitivity of PCD is assigned to the oxidation of the Fe²⁺ involved in prFMN binding to Fe³⁺ that causes a loss of prFMN (22). Although the standard redox potential of the PCD-bound Fe^3+/2+ couple is unknown, it is expected to be in the range of 0 mV, as it is typical for protein-coordinated Fe^3+/2+ cofactors, e.g., present in rubredoxins (29). In E. coli, intracellular glutathione concentration is around 5 mM and the ratio of the reduced to the oxidized form in aerobically growing cells ranges from 50:1 to 200:1, giving an estimated cellular redox potential between –0.24 and –0.26 V (30). Assuming a similar scenario in denitrifying bacteria growing aerobically with phthalate, the Fe cofactor of PCD will remain in the reduced +2 state in whole cells, even at low intracellular oxygen concentrations.

So far, the phthalate degradation pathway via phthaloyl-CoA has exclusively been found to be connected to the anaerobic benzoyl-CoA degradation pathway in facultatively and obligately anaerobic bacteria (9–12), whereas aerobic phthalate degradation was generally associated with the use of ring-hydroxylating dioxygenases (6–8). In the pathway described in this work, elements that were restricted previously to anaerobic (phthaloyl-CoA pathway) and aerobic (box-pathway) degradation pathways are combined; therefore, we refer to it here as a hybrid pathway (Fig. 6).

MATERIALS AND METHODS

Cultivation of bacteria and preparation of cell extracts.

T. chlorobenzoica strain 3CB-1 (DSMZ 18012) and A. evansii KB740 (DSMZ 6898) were grown aerobically at 30°C in 100-ml Erlenmeyer flasks on a rotary shaker (if not otherwise stated at a 150-rpm agitation rate). The mineral salt medium contained 10 mM phthalate and was as described for growth under denitrifying conditions (11), with the exception that nitrate was omitted. Growth on the 10-liter scale was carried out in a bioreactor (BioFlo/CelliGen, Eppendorf), that was operated in a continuous fed-batch mode using 0.5 M phthalate under increasing flushing with air (20 liter h⁻¹ to 150 liter h⁻¹). The pH was monitored by an internal pH electrode and automatically adjusted to pH 7.8 using 2 M HCl and 2 M NaOH. Cells were harvested in the exponential growth phase to an optical density at 578 nm (OD₅₇₈) and were kept frozen in liquid nitrogen until use. Cell extracts were prepared under anaerobic conditions as described previously (11).

LC/MS and LC/diode array analyses of metabolites and proteins.

Metabolites were analyzed by LC/MS using a Waters Acquity I-class UPLC with a Waters C₁₈ high-strength silica (HSS) T3 column (2.1 mm by 100 mm, 1.8-μm particle size) coupled to either a Waters Synapt G2-Si high-definition MS (HDMS) ESI/Q-TOF system or a Waters Acquity photodiode array detector. The programs used for CoA thioester analyses and prFMN detection were as described previously (22). Mass spectra were recorded in the positive mode with parameters described previously (22). Evaluation of LC/MS metabolite data was performed using the MassLynx software (Waters); for evaluation of LC-UV/visible data, MassLynx or Empower (Waters) was used.

For protein identification from SDS gels, differentially abundant protein bands were excised and cysteine residues were reduced using dithiothreitol and alkylated by treatment with iodoacetamide. After in-gel digestion with trypsin (Sigma-Aldrich), the resulting peptides were separated and analyzed as described previously (22). The resulting spectra were analyzed with ProteinLynx Global Server (Waters) by matching with the UniProt database of Thauera chlorobenzoica (minimal fragment ion matches per peptide, 3; minimal fragment ion matches per protein, 3; minimal peptide matches per protein, 7; false discovery rate, 3%).

Enzyme assays.

PCD activity in cell extracts or enriched enzyme samples was measured in a coupled spectroscopic assay in a glove box (95% N₂, 5% H₂) at 30°C in 100 mM sodium phosphate (pH 7.5) supplemented by 50 mM KCl, 0.5 mM phthalate, and 0.4 mM succinyl-CoA. In control assays, succinyl-CoA was replaced by 0.4 mM acetyl-CoA. Enzymatic reactions were started by the addition of cell-free extracts or purified enzyme. Enzymatic reactions were stopped by precipitation in 0.9 M HCl/10% acetonitrile (vol/vol). Product formation was quantified based on calibration curves of standards. The oxygen sensitivity of PCD activity in extracts from aerobically and anaerobically grown cells was determined in the standard buffer at pH 7.5; constant pH values were always controlled. Extracts were desalted using PD-10 columns (GE Healthcare). A total of 30 μl per time point was incubated on ice in a rubber, closed glass vial. For aerobic incubation, 240-μl extract was exposed to oxygen on ice in a 5-ml glass vial. Samples taken at intervals were tested for benzoyl-CoA formation using the standard assay.

Enrichment of PCD from aerobically grown T. chlorobenzoica 3CB-1.

All steps were carried out under anoxic conditions (95% N₂ and 5% H₂). Around 4 g of phthalate-grown cells (wet mass) was suspended in 14 ml 50 mM morpholinepropanesulfonic acid (MOPS)/KOH, 10 mM KCl, and 1 mM dithioerythritol (DTE; pH 7.5) (buffer A) and lysed using a French pressure cell. After ultracentrifugation, 13 ml of the supernatant was applied to a 50-ml DEAE-Sepharose column (GE Healthcare) at a flow rate of 3 ml⁻¹min⁻¹. Fractions were eluted with 100 mM, 200 mM, and 1 M KCl in buffer A. Size exclusion chromatography was conducted on a 320-ml Superdex HiLoad 26/60 gel filtration column (GE Healthcare) in buffer A containing 100 mM KCl at a flow rate of 2.5 ml min⁻¹. The column was calibrated under the same conditions with standards of known molecular mass (thyroglobulin, ferritin, alcohol dehydrogenase, carboanhydrase, and cytochrome c). During the enrichment of PCD, the assays for determining PCD activity always contained enriched SCPCT, as previously described for the enrichment of PCD from anaerobically grown cells from T. chlorobenzoica strain 3CB-1 (19).

Determination of prFMN and Fe.

The cofactors of enriched PCD, prFMN and Fe²⁺, were analyzed as described previously (22).

Densitometric abundance analyses of PCD on SDS gels.

To estimate the relative abundance of PCD in extracts of cells grown aerobically or anaerobically with phthalate, 20 μg of soluble protein was applied per lane on 9% to 10% polyacrylamide gels. After Coomassie staining, relative abundancies were estimated using the ChemDoc XRS+ visualizing system and the ImageLab software (Bio-Rad).

BLAST analyses.

To search for potential genes encoding phthalate-3,4- or phthalate-4,5 dioxygenases, the phthalate 3,4-dioxygenase from Arthrobacter keyseri (UniProt accession no. Q9AGK3) and phthalate 4,5-dioxygenase from Pseudomonas putida (NCBI accession no. Q05182) were used as query sequences. The threshold criteria for positive hits were E values of ≤e⁻⁵⁰ and sequence identity of ≥50%.