Transposon Mutagenesis Identifies Genes Critical for Growth of Pseudomonas nitroreducens TX1 on Octylphenol Polyethoxylates

ABSTRACT

Pseudomonas nitroreducens TX1 is of special interest because of its ability to utilize 0.05% to 20% octylphenol polyethoxylates (OPEO_n) as a sole source of carbon. In this study, a library containing 30,000 Tn5-insertion mutants of the wild-type strain TX1 was constructed and screened for OPEO_n utilization, and 93 mutants were found to be unable to grow on OPEO_n. In total, 42 separate disrupted genes were identified, and the proteins encoded by the genes were then classified into various categories, namely, information storage and processing (14.3%), cellular processes and signaling (28.6%), metabolism (35.7%), and unknown proteins (21.4%). The individual deletion of genes encoding isocitrate lyase (aceA), malate synthase (aceB), and glycolate dehydrogenase (glcE) was carried out, and the requirement for aceA and aceB but not glcE confirmed the role of the glyoxylate cycle in OPEO_n degradation. Furthermore, acetaldehyde dehydrogenase and acetyl-coenzyme A (acetyl-CoA) synthetase activity levels were 13.2- and 2.1-fold higher in TX1 cells grown on OPEO_n than in TX1 cells grown on succinate, respectively. Growth of the various mutants on different carbon sources was tested, and based on these findings, a mechanism involving exoscission to liberate acetaldehyde from the end of the OPEO_n chain during degradation is proposed for the breakdown of OPEO_n.

IMPORTANCE Octylphenol polyethoxylates belong to the alkylphenol polyethoxylate (APEO_n) nonionic surfactant family. Evidence based on the analysis of intermediate metabolites suggested that the primary biodegradation of APEO_n can be achieved by two possible pathways for the stepwise removal of the C₂ ethoxylate units from the end of the chain. However, direct evidence for these hypotheses is still lacking. In this study, we described the use of transposon mutagenesis to identify genes critical to the catabolism of OPEO_n by P. nitroreducens TX1. The exoscission of the ethoxylate chain leading to the liberation of acetaldehyde is proposed. Isocitrate lyase and malate synthase in glyoxylate cycle are required in the catabolism of ethoxylated surfactants. Our findings also provide many gene candidates that may help elucidate the mechanisms in stress responses to ethoxylated surfactants by bacteria.

INTRODUCTION

Octylphenol polyethoxylates (OPEO_n, commercial name Triton X-100) is a nonionic surfactant that belongs to the alkylphenol polyethoxylate (APEO_n) family. These surfactants are used in a range of industrial and household products (1, 2). APEO_n structures typically consist of hydrophilic polyethoxylate chain bound to a hydrophobic moiety, such as alkylphenol or a linear primary/secondary alcohol (3). APEO_n are often discharged into wastewater treatment plants or into the environment, which leads to them ultimately being degraded into shorter ethoxylate (EO) chains. Sometimes, the chains are completely removed to form nonylphenol, octylphenol, and alkylphenol monoethoxylates to triethoxylates (APEO_n, n = 1 to ∼3, respectively) (4, 5). Many studies have shown that these APEO_n metabolites, which have increased hydrophobicity, are more toxic than their parent compounds (6). In fact, these metabolites are able to mimic natural hormones, thus acting as endocrine disruptors when ingested by wildlife, which in turn can affect the environment and ultimately human health (7).

The fate and degradability of APEO_n in the environment have received much attention (8). The biodegradation of APEO_n has been studied using both pure and mixed cultures that grow solely on APEO_n, and several bacterial strains have been reported as being able to degrade the EO chains of APEO_n (1, 8–12). Most such isolates are proteobacteria and are often members of the genus Pseudomonas. Nguyen and Sigoillot (13) isolated four Pseudomonas strains from coastal seawater that grew on OPEO_n, and these generated OPEO with 4 to 5 units of EO chain as the end products. John and White (14) reported a strain of Pseudomonas putida that grew on nonylphenol polyethoxylate (NPEO_n) as the sole source of carbon, and in this case, the final accumulating metabolite was identified by gas chromatography-mass spectroscopy as NPEO₂. Nishio et al. (15) isolated 11 strains of OPEO_n-utilizing bacteria from paddy field soils. One strain, P. putida S-5, was shown to transform OPEO_n to form OPEO₂ and OPEO₃, which were the dominant metabolites accumulating under aerobic conditions. Evidence based on analysis of intermediate metabolites has suggested that the primary biodegradation of APEO_n under aerobic condition is achieved by a stepwise shortening of the EO chains (13, 14, 16). Two possible pathways for the stepwise removal of the C₂ ethoxylate units from the end of the chain have been proposed (Fig. 1). The first is a nonoxidative hydroxyl shift mechanism using ether scission that yields hemiacetal, which then produces acetaldehyde (14, 17). The second is the cleavage of APEO_n by terminal carboxylation, followed by hydrolysis to yield glycolate, which is oxidized to glyoxylate by glycolate dehydrogenase (15, 17). However, clear-cut evidence regarding the operation in organisms of these pathways during OPEO_n degradation is still lacking.

FIG 1.

Proposed pathways for the biodegradation of alkylphenol polyethoxylates from the literature (14, 15, 17).

In our previous studies, P. nitroreducens TX1, which possesses the ability to grow on OPEO_n at a wide range of concentrations, was isolated (1, 9–12, 18). The strain is able to use 0.05% to 20% OPEO_n as a sole source of carbon. The liquid chromatography-mass spectrometry (LC-MS) analysis revealed that the ethoxylate chain was sequentially shortened from the hydroxyl terminal side in this strain (9). To elucidate the biodegradation mechanism of OPEO_n, a library containing 30,000 mutants of strain TX1 was prepared using Tn5 insertion mutagenesis during this study. A total of 93 mutants were identified that were unable to grow on OPEO_n, and these were found to have disrupted 42 individual genes. In-frame deletion of some of the target genes was then performed to confirm the roles of these genes in OPEO_n degradation. The results revealed the important role of the glyoxylate cycle in OPEO_n utilization by strain TX1.

MATERIALS AND METHODS

Bacterial strains and culture conditions.

The bacterial strains and plasmids used in this study are listed in Table 1. Strain TX1 was routinely grown at 30°C in Luria-Bertani (LB) or minimal salt basal (MSB) medium with appropriate sources of carbon and energy (12). In liquid culture, cells were grown in a 50-ml culture volume in 250-ml Erlenmeyer flasks with shaking in an incubator at 180 rpm. For plate culture, agar was added at a final concentration of 1.5% (wt/vol). Escherichia coli was grown in LB medium at 37°C and served as the host organism for plasmid retention. Ampicillin (20 μg · ml⁻¹), gentamicin (20 μg · ml⁻¹), and kanamycin (20 μg · ml⁻¹) were used to select the transformed E. coli or TX1 cells.

TABLE 1.

Bacterial strains and plasmids used in this study

Strain or plasmid	Descriptiona	Reference or source
Strains
Pseudomonas nitroreducens
TX1	Wild-type strain TX1, Ampr	Laboratory collection
ΔaceA mutant	Strain TX1 with in-frame deletion of aceA gene, Ampr	This study
ΔaceB mutant	Strain TX1 with in-frame deletion of aceB gene, Ampr	This study
ΔfixC mutant	Strain TX1 with in-frame deletion of fixC gene, Ampr	This study
ΔglcE mutant	Strain TX1 with in-frame deletion of glcE gene, Ampr	This study
ΔaceA(pBaceA) mutant	ΔaceA mutant harboring plasmid pBaceA, Ampr, Gmr	This study
ΔaceB(pBaceB) mutant	ΔaceB mutant harboring plasmid pBaceB, Ampr, Gmr	This study
E. coli
DH5α	F− ϕ80dlacZΔM15 Δ(lacZYA-argF)U169 deoR recA1 endA1 hsdR17(rK− mK+) phoA supE44 λ− thi-1 gyrA96	Laboratory collection
DH5α/λpir	λpir lysogen of DH5α	Laboratory collection
BW20767	Containing pRL27; used as donor for pRL27 conjugations	Larsen et al. (21)
HB101	Containing pRK2013; used as helper for pRL27 conjugations	Figurski and Helinski (22)
Plasmids
pGEM-T Easy	PCR cloning vector, Ampr	Promega
pGaceA	Ampr, pGEM-T Easy with 2.6-kb fragment, including aceA	This study
pGaceB	Ampr, pGEM-T Easy with 3.2-kb fragment, including aceB	This study
pRL27	Kmr, mini-Tn5 transposon (oriR6K) delivery vector	Larsen et al. (21)
pRK2013	Kmr, carrying RK2 transfer genes	Figurski and Helinski (22)
pK18mobsacB	Kmr, oriT(RP4) sacB lacZα Plac Pmbi, mobilization and counterselection	Schafer et al. (24)
pKaceA	pK18mobsacB containing a gene fragment with 98% deletion of internal region of aceA	This study
pKaceB	pK18mobsacB containing a gene fragment with 98% deletion of internal region of aceB	This study
pKfixC	pK18mobsacB containing a gene fragment with 98% deletion of internal region of fixC	This study
pKglcE	pK18mobsacB containing a gene fragment with 98% deletion of internal region of glcE (GenBank accession no. WP_017520229.1)	This study
pBBR1MCS-5	Broad-host-range cloning vector, lacZ, Gmr	Kovach et al. (26)
pBaceA	Gmr, pBBR1-MCS5 with 2.6-kb fragment, including aceA	This study
pBaceB	Gmr, pBBR1-MCS5 with 3.2-kb fragment, including aceB	This study
pBglcE	Gmr, pBBR1-MCS5 with 1.5-kb fragment, including glcE	This study

^aKm^r, Gm^r, and Amp^r indicate resistance to kanamycin, gentamicin, and ampicillin, respectively.

Plasmid and chromosomal DNA isolation.

Total DNA and plasmid DNA were extracted using previously described procedures (19). For the PCR amplifications, a 50-μl PCR mixture, consisting of 0.2 mM each of the four dinucleoside triphosphates (dNTPs), 20 pmol each primer, 10 ng of extracted DNA, and 1.25 U of Taq DNA polymerase (Toyobo Co. Ltd., Japan), with an appropriate amount of reaction buffer, was used. Amplification was performed in a program temperature control system, PC-808 (Astec Co. Ltd.). Other molecular techniques were performed using standard procedures (20) or as recommended by the reagent suppliers. The oligonucleotide primers used in this study are listed in Table 2.

TABLE 2.

Oligonucleotide primers used in this study

Primer name	Sequence (5′–3′)a	Note	Restriction site	Reference or source
tpnRL17-1	AACAAGCCAGGGATGTAACG	For sequencing of transposon insertion		Larsen et al. (21)
tpnRL13-2	CAGCAACACCTTCTTCACGA			Larsen et al. (21)
AceA_FEco	GCGAATTCACGTGCTCGACTAAGCCTTC	For in-frame deletion of aceA in TX1	EcoRI	This study
AceA_RXho	GCCTCGAGTGCGGACATGGCCAATCCTTC		XhoI	This study
AceA_FXho	GCCTCGAGTTCCACTAAGAACTGGCTCAC		XhoI	This study
AceA_RHind	GCAAGCTTCTCTTGATCAGCGGCAGTTT		HindIII	This study
AceB_FEco	GCGAATTCGAAGCGAAGAACGGGTACAG	For in-frame deletion of aceB in TX1	EcoRI	This study
AceB_RXho	GCCTCGAGTTCAGTCATTGCTTGCCTCAC		XhoI	This study
AceB_FXho	GCCTCGAGGGTCTGTAAGCACCTCGCGCC		XhoI	This study
AceB_RHind	GCAAGCTTAACCGTCTGCAGTCTTTCTGA		HindIII	This study
FixC_FEco	GCGAATTCCAGTGATCCGTCGAAGGTG	For in-frame deletion of fixC in TX1	EcoRI	This study
FixC_RSma	GCCCCGGGGGCGGGCATCAGAAGGGCTCC		SmaI	This study
FixC_FSma	GCCCCGGGGCCAGCTGAATAGGACAAGG		SmaI	This study
FixC_RHind	GCAAGCTTCAGTTCAGCTCCACCTCGAT		HindIII	This study
GlcE_FEco	GCGAATTCAGTTTCGATTCGGTGGAAAA	For in-frame deletion of glcE in TX1	EcoRI	This study
GlcE_RXho	GCCTCGAGATCGGCCATCAGAAACGCTCC		SmaI	This study
GlcE_FXho	GCCTCGAGGAGCTCTGATGCAAACCAACC		SmaI	This study
GlcE_RHind	GCAAGCTTCGTCTGCACTATGGCTTCG		HindIII	This study
aceA(f)_F	GCGAATTCACGTGCTCGACTAAGCCTTC	For construction of pBaceA	EcoRI	This study
aceA(f)_R	GCAAGCTTCTCTTGATCAGCGGCAGTTT		HindIII	This study
aceB(f)_F	GCGAATTCGAAGCGAAGAACGGGTACAG	For construction of pBaceB	EcoRI	This study
aceB(f)_R	GCAAGCTTAACCGTCTGCAGTCTTTCTGA		HindIII	This study
glcE(f)_F	TGCCCGGGGAGCTGGAACGCGCCG	For construction of pBglcE	SmaI	This study
glcE(f)_R	GCACAAGCTTGATCAGGTAGATGCGCC		HindIII	This study

^aRestriction enzyme sites are underlined.

Transposon mutagenesis.

The plasmid pRL27, which carries the transposon Tn5 with a kanamycin (Km) resistance gene, was chosen as the vector for transferring the transposon to P. nitroreducens TX1. The introduction and subsequent transposition of the modified mini-Tn5 transposon into the P. nitroreducens TX1 genome were carried out as previously described by mating TX1 and E. coli BW20767, which carries the transposon delivery vector pRL27 (21, 22). The donor BW20767(pRL27), the helper E. coli HB101(pRK2013), and the recipient strain TX1 were mixed at a ratio of 1:1:1, spotted on an LB agar plate, and cultured at 30°C for 12 h to allow conjugational transfer of the pRL27 plasmid into the recipient cells. Transconjugants were initially selected on MSB agar medium containing 0.5% succinate, ampicillin, and kanamycin. They were then subjected to OPEO_n utilization screening as follows. First, single colonies were streaked on MSB agar containing 0.5% OPEO_n. The plates were then incubated in an inverted position for several days at 30°C. Mutant strains that did not show visible growth on the plate were designated OPEO_n-negative mutants and were then subjected to further study.

Plasmid rescue and recovery of interrupted genes.

Larsen et al. have shown that one-step cloning of the transposon with its associated flanking DNA can be accomplished after mutation by pRL27 (21). The selected OPEO_n-negative mutants were grown overnight in LB plus kanamycin and ampicillin, and then 1.5 ml of each culture was transferred to microcentrifuge tubes. Next, the cells were harvested by centrifugation at 10,000 × g for 1 min. Chromosomal DNA was isolated from the pelleted cells as described above and digested with the restriction enzyme BamHI, which does not cut within the transposon sequence of pRL27. The digested DNA was cleaned, which was followed by self-ligation using the T4 ligase at 20°C for 1 h. Material from ligated mixtures was transformed into E. coli DH5α λpir, as previously described, and plated onto LB agar-kanamycin plates in order to select for cells transformed with the ligated pRL27 plasmid that also contained the flanking P. nitroreducens TX1 DNA (23). Next, the transposon with its flanking DNA was isolated using the Midi Plus ultrapure plasmid extraction system (Viogene, Taipei, Taiwan). Primers tpnRL17-1 and tpnRL13-2, which anneal to positions within the transposon sequence in pRL27 and read outwards into flanking DNA regions, were used for sequencing (21). Sequencing of the DNA interrupted by the transposon was carried out on an automatic genetic analyzer (Applied Biosystems). The genes disrupted by the transposon were identified using the BioEdit software by a local BLAST search against the TX1 draft genome sequence and also against the GenBank database.

In-frame deletion mutagenesis.

Four genes (aceA, aceB, fixC, and glcE, which encode isocitrate lyase, malate synthase, dehydrogenase, and glycolate dehydrogenase, respectively) were inactivated by in-frame deletion to avoid any polar effect. The in-frame aceA deletion mutant of strain TX1 was constructed by allelic replacement, as previously described (24). A gene fragment containing about a 98% deletion of the internal region of aceA was created by overlap extension PCR, as described previously (25). The primers used were Ace_FEco, Ace_RXho, Ace_FXho, and Ace_RHind. The internally deleted gene fragment was cloned into a suicide vector (pK18mobsacB), which was named pKaceA. The aceA deletion mutant of TX1 was created by triple mating between strains TX1, E. coli DH5α(pKaceA), and E. coli HB101(pRK2013). The TX1 ΔaceA strain was screened as described previously (24) and confirmed by PCR. The same procedure was applied to construct the TX1 ΔaceB, TX1 ΔfixC, and TX1 ΔglcE mutants using the primers that are described in Table 2.

Complementation of the TX1 ΔaceA, TX1 ΔaceB, and TX1 ΔglcE mutants.

To determine whether the OPEO_n utilization deficiency was due to an inactivated gene, the TX1 ΔaceA, TX1 ΔaceB, and TX1 ΔglcE mutants were complemented with appropriate wild-type genes expressed in a broad-host-range vector pBBR1MCS5 (26). For example, the aceA expression vector was constructed as follows: a PCR fragment (2.6 kb) obtained by amplifying the chromosomal DNA of TX1 with the primers aceA(f)_F and aceA(f)_R was ligated into a pGEM-T Easy vector (Promega Co.). The resulting vector (pGaceA) was then amplified in E. coli DH5α, purified, and digested with EcoRI and HindIII to generate the PCR product. Next, the retrieved PCR-amplified fragment was ligated into pBBR1MCS-5, which had been cut with the same restriction enzymes. The resulting plasmid (pBaceA) was then introduced into TX ΔaceA by conjugation using the helper strain E. coli HB101(pRK2013). The presence of the intact aceA gene was confirmed by DNA sequencing. The recombinant TX1 ΔaceA(pBaceA) strain was selected on MSB agar containing 0.5% OPEO_n, ampicillin, and gentamicin. The same procedure was applied for complementation of the TX1 ΔaceB and TX1 ΔglcE mutants with pBaceB and pBglcE, respectively, which were constructed using the primers listed in Table 2.

Preparation of crude cell extract from strain TX1.

A 0.5-liter culture of TX1 grown on either 0.5% OPEO_n-MSB or 0.5% succinate-MSB was used for the preparation of crude cell extract. The cells were collected by centrifugation (11,000 × g, 10 min, 4°C), washed with 10 mM potassium phosphate buffer (pH 7.0), and suspended in 5 ml of the same buffer. The cell suspension was subjected to sonication for 3 min with 50% pulse on ice using the Sonicator 4000 (Misonix, Farmingdale, NY) to disrupt the cells (27). In the process of cell lysis, 0.15 mM protease inhibitor (phenylmethylsulfonyl fluoride) was added. After the removal of unbroken cells and cell debris by centrifugation (11,000 × g, 10 min, 4°C), the supernatant was collected by ultracentrifugation (200,000 × g, 1 h, 4°C) and used as the crude cell extract. Protein concentrations were determined using the Bradford protein assay with bovine serum albumin as the standard.

Enzyme assays.

The acetaldehyde dehydrogenase activity assay was performed by measuring the rate of appearance of NADH at 340 nm in 1-cm-path cuvette at 25°C with a Beckman DU640 spectrophotometer (Beckman Coulter, Krefeld, Germany). The incubation mixtures contained the following constituents in a final volume of 1 ml: 10 mM potassium phosphate buffer (pH 7.0), 10 mM 2-mercaptoethanol, 2 mM NAD⁺, 0.5 mg of cell crude extract, and 1 mM acetaldehyde. The reaction was started by the addition of acetaldehyde. For isocitrate lyase activity, a spectrophotometric assay was used to measure isocitrate-dependent formation of glyoxylate. The standard reaction mixture (0.5 ml) contained 10 mM potassium phosphate buffer (pH 7.0), 5 mM MgCl₂, 2 mM dithioerythritol, 3.5 mM phenylhydrazine, and 0.5 mg of cell crude extract. The reactions were started by the addition of 2 mM isocitrate, and the formation of the glyoxylate phenylhydrazone derivative was monitored at 324 nm. Malate synthase activity in the crude cell extracts was measured by the method of Srere et al. by tracking the glyoxylate- and acetyl-CoA-dependent release of CoA (28). Acetyl-CoA synthetase activity in the crude cell extracts was measured in the presence of ATP, acetate, and CoA, as described in a previous study (29).

RESULTS AND DISCUSSION

Transposon mutagenesis of P. nitroreducens TX1.

The availability of the draft genome sequence of P. nitroreducens TX1 provides an opportunity for investigating genes that play significant roles in OPEO_n utilization (18). In order to identify genes involved in OPEO_n catabolism, the Tn5 vector pRL27 (21), which has been used widely in Pseudomonas genetics, was used to construct a mutant library of strain TX1, as described in Materials and Methods. In total, 30,000 Tn5 insertion mutants of TX1 were successfully obtained. After 48 h of incubation at 30°C, these mutants were screened, and 93 (0.31%) of them failed to grow on 0.5% OPEO_n-MSB but still grew on 0.5% succinate-MSB.

A total of 6,341 open reading frames (ORFs) have been putatively identified in the TX1 draft genome (18), and therefore, the Tn5 mutant library would seem to have 4.7× ORF coverage of the TX1 genome. The stability of the Tn5 transposon in the transformants was also tested using two of the mutants (C94-4 and A16-7). After 10 generations of the single-colony propagation by subculture on LB plates, the initial and final colonies were found to still retain the kanamycin resistance and, in addition, the Tn5 transposon had remained in the same genomic position on the genome, with this being determined by PCR. These findings confirm that the transposon insertion events present in the transformants from this library are stable.

The transposon insertion points of the 93 mutants defective in OPEO_n utilization were identified by the plasmid rescue, followed by nucleotide sequencing using a transposon-derived primer set (21). The insertion loci harboring Tn5 were then mapped to 42 independent coding genes, three noncoding sequences, and one 23S rRNA. These results are summarized in Table 3. Of these 42 genes, 11 had been mutated multiple times to give a number of different mutant strains for a given specific gene. These were the dedA gene, which had 12 mutants; the hit gene, with 11 mutants; the aceB gene, with seven mutants; the aceA and rfaG genes, with six mutants; the rfe gene, with four mutants; the coq7 gene and a gene encoding a hypothetical protein, with three mutants; and the yfgC and fixC genes and a gene encoding a transcriptional regulator, with two mutants (Table 3). These genes with multiple mutations were used for further investigation. Among these 42 disrupted genes, 39 genes could be located within the draft genome of TX1 (accession no. AMZB01000000), while the other three genes, namely, those of the K11-8, C2-42, and T43-26 mutants, were found to be similar to genes present in the genomes of other strains, including Mycobacterium abscessus, Comamonas testosteroni, and Pseudomonas knackmussii (accession numbers WP_049232543, WP_003060809, and WP_043252236.1, respectively), but not within the draft genome of TX1 (18).

TABLE 3.

Characterization of the OPEO_n-negative mutants of P. nitroreducens TX1 obtained by Tn5 insertional mutagenesis

Mutanta	Gene	Functionb	Protein accession no.	Function group (COGs)c
C94-4, C100-12, N8-29, N10-35, N36-50, P30-12	aceA	Isocitrate lyase	WP_017517106.1	COG2224C
A16-7, C87-31, C90-8, K1-02, N38-2, P10-33, T16-15	aceB	Malate synthase	WP_017520204.1	COG2225C
B1-31, B1-33	fixC	Dehydrogenase (flavoprotein)	WP_017520493.1	COG0644C
A2-42	hisP	ABC-type histidine transport system	WP_017517948.1	COG4598E
C39-17	gdh2	NAD-specific glutamate dehydrogenase	WP_017519332.1	COG2902E
C96-34	rhtB	Putative threonine efflux protein	WP_017521173.1	COG1280E
T8-50	ilvB	Thiamine pyrophosphate-requiring enzyme	WP_017516325.1	COG0028EH
N43-34	putP	Na+/proline symporter	WP_017518917.1	COG0591ER
C48-18, C77-4, C95-33, P19-3, P19-32, P38-47, P42-36, T10-24, T15-48, T57-1, T58-5	hit	Diadenosine tetraphosphate hydrolase and other HIT family hydrolases	WP_017517344.1	COG0537FGR
B3-49	bglX	Beta-glucosidase-related glycosidase	WP_017517232.1	COG1472G
P40-18	zwf	Glucose-6-phosphate 1-dehydrogenase	WP_017520335.1	COG0364G
K32-8	gshA	Gamma-glutamylcysteine synthetase	WP_017520855.1	COG2918H
A21-39	ubiE	Methylase involved in ubiquinone biosynthesis	WP_017521165.1	COG2226H
C73-29, C97-46, P24-25	coq7	Demethoxyubiquinone hydroxylase	WP_017517343.1	COG2941H
B3-29	fadB	3-Hydroxyacyl-CoA dehydrogenase	WP_017519287.1	COG1250I
A11-1	yafY	Predicted transcriptional regulator	WP_017518996.1	COG2378K
C4-30	araC	AraC-type DNA-binding domain-containing protein	WP_017521913.1	COG2207K
B5-50	aRO8	Transcriptional regulator	WP_017518869.1	COG1167KE
C15-19	hepA	Superfamily II DNA/RNA helicase	WP_017522209.1	COG0553KL
K11-8	dob10	DEAD/DEAH box helicase	WP_049232543	COG4581L
C2-42	Tn3	Transposase	WP_003060809	COG1961L
C82-3, T44-10, T45-21, T67-19, T78-50, T77-39	rfaG	Glycosyltransferase	WP_017520449.1	COG0438 M
C85-47, T41-40, T66-4, T66-19	rfe	UDP-N-acetylglucosamine-1-phosphate transferase	WP_017520347.1	COG0472 M
N52-26	gt2	Glycosyltransferases	WP_017520448.1	COG1216 M
P33-4	nlpD	Membrane protein related to metalloendopeptidase	WP_017517352.1	COG0739 M
P55-13	wcaA	Glycosyltransferase involved in cell wall biogenesis	WP_017517134.1	COG0463 M
T43-26	galU	UTP-glucose-1-phosphate uridylyltransferase	WP_043252236.1	COG1210 M
B7-14	wcaG	Nucleoside diphosphate-sugar epimerase	WP_017520346.1	COG0451MG
B6-16	tar	Methyl-accepting chemotaxis protein	WP_017516259.1	COG0840NT
N24-8	anmK	Predicted molecular chaperone	WP_017517351.1	COG2377O
A21-12	COG4590	ABC-type uncharacterized transport system	WP_017520681.1	COG4590R
A7-47	ybcL	Membrane-bound metal-dependent hydrolase	WP_017520412.1	COG1988R
B2-48, B7-43	yfgC	Putative Zn-dependent protease	WP_017519734.1	COG4783R
K17-10	roxA	Ribosomal protein L16 Arg81 hydroxylase	WP_017517102.1	COG2850S
K21-12	tauE	Cytochrome biogenesis protein	WP_017516824.1	COG2836S
C3-45, C5-33, C9-50, C61-29, C63-28, C64-38, P32-21, P33-22, T11-1, T15-44, T17-21, T46-48	dedA	Membrane protein	WP_017521718.1	COG0586S
N30-8	atoC	Response regulator	WP_017518918.1	COG2204T
P36-32	GGDEF	GGDEF domain, diguanylate cyclase	WP_017517199.1	COG2199T
C12-38	norM	Na+-driven multidrug efflux pump	WP_017518643.1	COG0534V
B3-46, B4-25		Transcriptional regulator	WP_017516487.1	NC
C11-3, T56-25, T57-21		Hypothetical protein	WP_017520115.1	NC
A28-42		Hypothetical protein	WP_017520049.1	NC
T12-28		23S rRNA		NC
A6-49		Noncoding sequence		NC
B2-47		Noncoding sequence		NC
P34-24		Noncoding sequence		NC

^aThe representative strains (in bold) were selected for further investigation, as described in the text.

^bHIT, histidine triad.

^cThe proteins were classified into functional categories according to the clusters of orthologous groups (COG). The functional categories are information storage and processing, including COG categories K (transcription) and L (replication, recombination, and repair); cellular processes and signaling, including O (posttranslational modification, protein turnover, and chaperones), M (cell wall/membrane/envelope biogenesis), N (cell motility), T (signal transduction mechanisms), and V (defense mechanisms); metabolism, including C (energy production and conversion), G (carbohydrate transport and metabolism), E (amino acid transport and metabolism), F (nucleotide transport and metabolism), H (coenzyme transport and metabolism), and I (lipid transport and metabolism); poorly characterized, including R (general function prediction only) and S (function unknown). NC, proteins not classified into a COG.

Functional characterization of the mutant genes.

The proteins encoded by the 42 unique genes identified in this study (OPEO_n growth-associated proteins) were grouped into functional classes using the clusters of orthologous groups (COG), as shown in Table 3. When this was done, six proteins (14.3%) were classified into the information storage- and processing-related categories (K and L). Furthermore, the cellular process- and signaling-related categories (O, M, N, T, and V) consisted of 12 proteins (28.6%). Fifteen proteins (35.7%) were present in the metabolism-related categories (C, G, E, F, H, and I). Two poorly characterized COG groups (R and S) were found to contain six proteins (14.3%). In addition, for the category “no related COGs” (the protein is not predicted to belong to any of the currently defined COGs), there were found to be three proteins (7.1%).

Overall, 12 (28.5%) genes were predicted to encode catabolic enzymes. Three proteins (isocitrate lyase, malate synthase, and FixC dehydrogenase) are known to be involved in energy production and conversion, whereas nine proteins (diadenosine tetraphosphate hydrolase, NAD⁺-specific glutamate dehydrogenase, thiamine pyrophosphate-requiring enzyme, β-glucosidase-related glycosidase, glucose-6-phosphate 1-dehydrogenase, γ-glutamylcysteine synthetase, UbiE methylase involved in ubiquinone biosynthesis, demethoxyubiquinone hydroxylase, and 3-hydroxyacyl-CoA dehydrogenase) are known to be involved in the metabolism of amino acids, carbohydrates, coenzymes, and lipids. Of these, diadenosine tetraphosphate hydrolase (hit) is a key enzyme controlling the in vivo concentration of the dinucleotide diadenosine tetraphosphate that has been proposed to play a range of roles in processes, such as control of DNA replication and repair, signaling in stress response, and apoptosis (30).

OPEO_n is one of the most widely used nonionic surfactants; it is used in biology to lyse cells to allow the extraction of protein and other cellular organelles. Previous studies have suggested that OPEO_n affects the cell membrane by disrupting its structural integrity and functioning, which results in increased membrane fluidity and permeabilization (31). In our previous work, we found that P. nitroreducens TX1 is able to tolerate up to 20% OPEO_n in MSB medium. Therefore, the membrane of TX1 might have special structural features that allow OPEO_n to be tolerated. In this context, it should be noted that seven (16.6%) out of the 42 identified proteins, namely, three glycosyltransferases (rfaG, gt2, and wcaA), a UDP-N-acetylglucosamine-1-phosphate transferase (rfe), a membrane protein related to metalloendopeptidase (nlpD), a UTP-glucose-1-phosphate uridylyltransferase (galU), and a nucleoside-diphosphate-sugar epimerase (wcaG), are known to be linked to cell wall/membrane/envelope biogenesis. For example, the deletion of ssg, which encodes a glycosyltransferase in Pseudomonas alkylphenolia KL28, has been shown to cause the loss of lipopolysaccharide O antigen, which alters the composition of the exopolysaccharide. Furthermore, this mutant strain was found to have reduced surface spreading, reduced pellicle formation, and reduced biofilm formation; these were probably due to the cumulative effects of lipopolysaccharide truncation and alterations to the cell's exopolysaccharide composition (23). In some bacterial species, UDP-N-acetylglucosamine-1-phosphate transferase has been shown to play an important role in the biosynthesis of various polymers within the bacterial cell wall, such as common antigen and lipopolysaccharide O antigen (32). UTP-glucose-1-phosphate uridylyltransferase is an enzyme associated with glycogenesis. This enzyme synthesizes UDP-glucose from glucose-1-phosphate. UDP-glucose has been reported to be involved in galactose utilization, in glycogen synthesis, and in the synthesis of various carbohydrate moieties, such as glycolipids, glycoproteins, and proteoglycans (33). Interestingly, yfgC codes for a putative Zn-dependent protease, and although its function is unknown, the chemical and genetic data suggest that it also plays a role in outer membrane protein biogenesis (34). Furthermore, five (11.9%) out of the 42 genes were predicted to encode transport proteins, among which two ABC transporters were identified. Such transport proteins were likely to play important roles in the transport of OPEO_n into the cell. For the functions of the DedA membrane protein, the recent genetic approaches have revealed important roles in membrane homeostasis. Bacterial DedA family mutants display such intriguing phenotypes as cell division defects, temperature sensitivity, altered membrane lipid composition, elevated envelope-related stress responses, and loss of proton motive force (35). Finally, there are a significant number of mutants that fall within the hypothetical, unknown, and unclassified gene categories, which suggests that there is still a large number of unknown genes across the P. nitroreducens genome potentially involved in OPEO_n metabolism that remain to be explored, including the three not found in the TX1 draft genome. Further studies on these genes using complementation and other strategies will improve our understanding of the strain TX1 and its tolerance and degradation of OPEO_n.

Role of the glyoxylate cycle in OPEO_n degradation by TX1.

Isocitrate lyase and malate synthase, which are both key enzymes in the glyoxylate cycle, were frequently detected during screening of mutants unable to grow on 0.5% OPEO_n. In the glyoxylate cycle, two molecules of acetyl-CoA are converted into oxaloacetate, thus bypassing the reactions of the tricarboxylic acid (TCA) cycle in which CO₂ is released. The glyoxylate cycle is essential when cells are growing on C₂ compounds, because this pathway allows the synthesis of all cellular compounds with three or more carbon atoms; examples include the biosynthesis of amino acids and nucleotides. To confirm the role of these two enzymes in OPEO_n utilization, in-frame deletion mutants lacking the aceA and aceB genes were created from strain TX1. As expected from the transposon mutagenesis results, the TX1 ΔaceA and TX1 ΔaceB mutants were unable to grow on OPEO_n. The in-frame deletion mutants were then used for complementation analysis. Two plasmids, pBaceA and pBaceB, which carry the aceA and aceB genes, respectively, were constructed and transferred independently from E. coli into the TX1 ΔaceA and TX1 ΔaceB mutant strains by conjugation. In both cases, the introduction of the wild-type copy of the gene resulted in a return to the wild-type growth pattern, confirming the role of the aceA and aceB in OPEO_n utilization. In addition, isocitrate lyase and malate synthase activities in cell extract of wild-type cells grown on OPEO_n were 277.6 ± 5 (mean ± standard deviation) and 15.1 ± 5 nmol · min⁻¹ · mg⁻¹, respectively. These activities were downregulated 8.8- and 1.7-fold in cells grown on succinate (31.3 ± 6.5 and 8.7 ± 2 nmol · min⁻¹ · mg⁻¹, respectively). These results clearly reveal that the glyoxylate cycle plays a critical role in OPEO_n degradation by strain TX1.

Growth of the mutants on different carbon sources.

Eleven mutants (TX1 ΔaceA, TX1 ΔaceB, TX1 ΔfixC, C48-18, C73-29, C82-3, C85-47, B2-48, C3-45, B3-46, and C11-3), which are representatives of the 11 genes present in the Tn5 library as multiple events, TX1 ΔglcE, and the three complemented strains [TX1 ΔaceA(pBaceA), TX1 ΔaceB(pBaceB), and TX1 ΔglcE(pBglcE)] were tested for their growth on 0.1% OPEO_n, 0.1% NPEO_n, 0.1% dodecyl octaethoxylate (AEO₈), 0.1% ethanol, 0.1% acetate, 0.1% glycolate, or 0.1% pyruvate as the sole source of carbon (Table 4). The glcE gene was deleted to test the cleavage of OPEO_n by terminal carboxylation, followed by hydrolysis to yield glycolate in strain TX1 (Fig. 1, oxidation). Wild-type TX1 is unable to grow on polyethylene glycol 400 (PEG 400), polyethylene glycol 1000 (PEG 1000), acetaldehyde, or glyoxylate; therefore, these compounds were not used for growth testing of the mutants. The generation time of wild-type TX1 was much longer when grown on glycolate (10.8 to 11.8 h) than on OPEO_n, NPEO_n, AEO₈, ethanol, acetate, or pyruvate (1.5 to 4.8 h) (Table 4). All mutants were able to grow on pyruvate, with generation times ranging from 3.6 to 4 h. When NPEO_n or AEO₈ was used as the sole source of carbon, five mutants, TX1 ΔfixC, TX1 ΔglcE, TX1 ΔaceA(pBaceA), TX1 ΔaceB(pBaceB), and TX1 ΔglcE(pBglcE), showed growth on these compounds, with generation times ranging from 1.5 to 3.7 h, whereas the rest of mutants failed to grow at all. When ethanol or acetate was used as the sole source of carbon, the two glyoxylate cycle mutants (TX1 ΔaceA and TX1 ΔaceB) failed to grow, but others were able to grow on ethanol or acetate, with generation times ranging from 2.7 to 4.8 h. In addition, the two mutants TX1 ΔaceB and TX1 ΔglcE failed to grow on 0.1% glycolate, but the other mutants were able to grow on it, with generation times ranging from 10.8 to 11.8 h.

TABLE 4.

Growth of TX1 mutants on various carbon sources

Straina	Description	Generation time (mean ± SD) (h)b
		OPEOn	NPEOn	AEO8	Ethanol	Acetate	Pyruvate	Glycolate
TX1	Wild type	1.5 ± 0.1	3.2 ± 0.1	3.7 ± 0.2	4.8 ± 0.2	2.7 ± 0.2	3.6 ± 0.4	11.3 ± 3.1
TX1 ΔaceA	Deletion of isocitrate lyase gene	−	−	−	−	−	3.6 ± 0.4	11.5 ± 2.5
TX1 ΔaceB	Deletion of malate synthase gene	−	−	−	−	−	3.6 ± 0.4	−
TX1 ΔfixC	Deletion of dehydrogenase gene	1.5 ± 0.1	3.1 ± 0.1	3.7 ± 0.1	4.8 ± 0.2	2.7 ± 0.2	3.6 ± 0.4	10.8 ± 2.2
TX1 ΔglcE	Deletion of glycolate dehydrogenase gene	1.5 ± 0.1	3.2 ± 0.1	3.7 ± 0.1	4.7 ± 0.1	2.9 ± 0.3	4 ± 0.2	−
C48-18	Mutation at diadenosine tetraphosphate hydrolase gene	−	−	−	4.7 ± 0.2	2.7 ± 0.2	4 ± 0.2	11.2 ± 4.2
C73-29	Mutation at demethoxyubiquinone hydroxylase gene	−	−	−	4.7 ± 0.2	2.7 ± 0.2	4 ± 0.2	11.3 ± 3.3
C82-3	Mutation at glycosyltransferase gene	−	−	−	4.7 ± 0.2	2.9 ± 0.3	4 ± 0.2	11.5 ± 2.8
C85-47	Mutation at UDP-N-acetylglucosamine-1-phosphate transferase gene	−	−	−	4.7 ± 0.1	2.7 ± 0.2	4 ± 0.2	11.7 ± 2.5
B2-48	Mutation at putative Zn-dependent protease gene	−	−	−	4.8 ± 0.2	2.9 ± 0.3	4 ± 0.2	10.9 ± 3.2
C3-45	Mutation at gene encoding membrane protein	−	−	−	4.7 ± 0.2	2.9 ± 0.3	4 ± 0.2	11.6 ± 2.7
B3-46	Mutation at gene encoding transcriptional regulator	−	−	−	4.8 ± 0.1	2.9 ± 0.3	4 ± 0.2	11.8 ± 2.8
C11-3	Mutation at gene encoding unknown protein	−	−	−	4.7 ± 0.1	2.9 ± 0.3	4 ± 0.2	11.7 ± 3.1
TX1 ΔaceA(pBaceA)	Complementation of isocitrate lyase gene	1.5 ± 0.1	3.2 ± 0.1	3.6 ± 0.2	4.7 ± 0.1	2.9 ± 0.3	4 ± 0.2	11.1 ± 3.2
TX1 ΔaceB(pBaceB)	Complementation of malate synthase gene	1.5 ± 0.1	3.1 ± 0.1	3.7 ± 0.1	4.7 ± 0.2	2.9 ± 0.3	4 ± 0.2	10.8 ± 2.6
TX1 ΔglcE(pBglcE)	Complementation of glycolate dehydrogenase gene	1.5 ± 0.1	3.1 ± 0.1	3.7 ± 0.1	4.8 ± 0.2	2.9 ± 0.3	4 ± 0.2	11.2 ± 2.8

^aEach strain was cultured in 5 ml of MSB plus 0.1% of the relevant carbon source.

^b−, no growth; OPEO_n, Triton X-100; NPEO_n, Triton N-101; AEO₈, dodecyl octaethoxylate. Data are the results from three independent experiments.

Proposed mechanism for the degradation of OPEO_n by P. nitroreducens TX1.

The biodegradation of long-chain APEO_n has been studied using isolated bacterial strains that grow solely on OPEO_n or NPEO_n. Most isolates are from the genus Pseudomonas, which belongs to the Gammaproteobacteria (9, 13, 14). Based on the results of metabolite analyses, the initial degradation reaction of APEO_n has been found to be a shortening of the ethoxylate chain by exoscission of the EO chains from the hydroxyl terminal side (13, 14, 16) (Fig. 1). Our group isolated Pseudomonas strain TX1, which has the ability to grow on OPEO_n at a wide range of concentrations (9). LC-MS analysis of the intermediate metabolites revealed the presence of a gradual shortening process that affected the EO chains during OPEO_n utilization. The results suggested that the same exo-type biodegradation process is occurring in TX1 as in other isolates.

The terminal oxidation model for the biodegradation of PEG under aerobic conditions has been reviewed previously (36, 37). In this model, the biodegradation of the PEG molecules is initiated through the oxidation of the terminal EO unit to a carboxyl group, which is followed by the liberation of glycolate. The carboxyl is converted to glyoxylate by glycolate dehydrogenase (GlcE). Tasaki et al. (38) isolated the adh1 gene from P. putida S-5 and expressed this gene in E. coli. By measuring the reduction of 2,6-dichlorophenolindophenol (DCPIP) spectrophotometrically at 600 nm using a cell crude extract, alcohol dehydrogenase (Adh1) was shown to have activity against OPEO_n of various EO chain lengths (38). However, its relevance to the ether cleavage step in OPEO_n biodegradation remains unknown. The use of redox dye-based assay with crude extracts can be misleading, because crude cell extract may contain multiple activities that are able to utilize DCPIP as an electron acceptor (17). Furthermore, the detection of glycolate and an increase in GlcE activity have not been reported; such findings would support the terminal oxidation model. In our study, a fixC gene encoding a dehydrogenase was found among the mutants. However, an in-frame deletion mutant of the fixC gene retained the ability to grow on OPEO_n in a manner similar to the wild type. These results suggest that the transposon insertion into fixC has a polar effect, and the fixC gene product itself may not play an important role in OPEO_n utilization. In addition, TX1 grows only slowly on glycolate but does not grow at all on glyoxylate, PEG 400, or PEG 1000 as the sole source of carbon. In order to determine whether glycolic acid is involved in the metabolism of OPEO_n by strain TX1, a glcE in-frame deletion mutant was constructed. As shown in Table 4, TX1 ΔglcE grew on OPEO_n to a level comparable to that of the wild type. Interestingly, TX1 ΔglcE and TX1 ΔaceB were unable to grow on glycolate, whereas TX1 ΔaceA was able to grow. Interestingly, one isocitrate lyase gene was found in the TX1 genome (18). In addition, the isocitrate lyase activity in the cell extract of TX1 ΔaceA cells grown on glycolate or succinate was not detectable. These results suggest that the degradation of glycolate does not involve the glyoxylate cycle, because the isocitrate lyase mutant was able to grow on glycolate in a manner similar to the wild-type TX1 (Fig. 2). Considering the results, we conclude that the degradation of OPEO_n by strain TX1 does not demonstrate an oxidative pathway involving the liberation of glycolate.

FIG 2.

Proposed pathway and mechanism (black lines) for the degradation of OPEO_n by P. nitroreducens TX1.

In the nonoxidative biodegradation model, biodegradation of the EO chain proceeds via the shift of a hydroxyl group, followed by the liberation of acetaldehyde, which is then quickly transformed to acetate by dehydrogenases that are ubiquitously distributed throughout the cell (39). The pathway for acetate degradation is well known and involves the conversion of acetate into acetyl-CoA, which then enters in the central carbon metabolism of the cell (39). In the glyoxylate cycle, the input of two molecules of acetyl-CoA results in a net synthesis of one molecule of succinate, which is then available for biosynthetic purposes (40). As shown in Table 4, TX1 was able to grow on either ethanol or acetate as the sole source of carbon, with generation times ranging from 2.7 to 4.8 h, but TX1 ΔaceA and TX1 ΔaceB were unable to grow on these carbon sources. These results suggest that the glyoxylate cycle plays a critical role in ethanol or acetate catabolism. The ethanol oxidation product acetate must first be activated to acetyl-CoA so it can be used as a carbon source and for the replenishment of intermediates within the TCA cycle; this would occur via the glyoxylate bypass in strain TX1. Our results suggest that nonoxidative biodegradation with the liberation of acetaldehyde is the most likely pathway for OPEO_n utilization in strain TX1 (Fig. 2). In addition, acetaldehyde dehydrogenase and acetyl-CoA synthetase activities were upregulated 13.1- and 2.1-fold in cells grown on OPEO_n (211 ± 14 and 68.8 ± 1.1 nmol · min⁻¹ · mg⁻¹, respectively) versus TX1 grown on succinate (16 ± 12 and 32.6 ± 1.2 nmol · min⁻¹ · mg⁻¹, respectively). Further studies on the mechanism used for the liberation of acetaldehyde will be needed in order to improve our understanding of the degradation of OPEO_n by strain TX1.