Introduction

Octylphenol polyethoxylates (OPEOn, commercial name Triton X-100) is a common nonionic surfactant that is used in industrial and household products1,2. Its structures is made of hydrophilic polyethoxylate chain (on average it has 9.5 ethylene oxide units) bound to a hydrophobic moiety (octylphenol)3. In the environment, OPEOn can be degraded into shorter exothylate chain to form octylphenol or octylphenol mono- to tri-ethoxylates4. These OPEOn metabolites that are known more estrogenic-like than their parent compounds can mimic the natural hormones and thus act as endocrine disruptors in wildlife5,6.

Much attention has been paid to the fate and degradability of OPEOn in the environment7. The biodegradation of OPEOn has been studied using both pure and mixed cultures that grow solely on OPEOn. Several bacterial strains have been reported to degrade the ethoxylate (EO) chain of OPEOn 1,7,8,9,10,11. The oxidation models have been proposed based on biodegradation intermediate metabolites12,13,14,15. In our previous studies, P. nitroreducens TX1 which revealed an ability to grow on 0.05–20% OPEOn as a sole carbon source1,8,9,10,11,16. The formation of carboxylated octyl-moiety from the catabolism of octylphenol polyethoxylates by P. nitroreducens TX1 was described. The LC–MS analysis shows that the ethoxylate chain was shortened and the estrogen-like intermediates were produced. A library containing 30,000 Tn5-insertion mutants of the wild-type strain TX1 was also constructed and screened for OPEOn utilization. The result revealed the role of the glyoxylate cycle in OPEOn degradation15. However, the evidence for the shortening of ethoxylate units is still lacking. In this study, we investigated the enzyme activities involved in OPEOn degradation in strain TX1 and a mechanism involved in the degradation of polyethoxylate chains was reported. The extent of removal for residual OPEOs during the degradation process by a pure recombinant enzyme was also investigated.

Materials and Methods

Chemicals

Triton X-100 was purchased from Merck Chemical Co. (Darmstadt, Germany). The average number of ethoxylate (EO) units for Triton X-100 is 9.5 according to the manufacturer’s information, which corresponds to an average molecular weight of ca. 625. All the reagents were purchased from the Merck Chemical Co. (Darmstadt, Germany) at a purity of 98–99.5%.

Bacterial strains and culture conditions

P. nitroreducens TX1 was isolated from a rice field drainage in Taiwan10. E. coli strains DH5alpha, BL21(DE3) (Invitrogen, Carlsbad, CA) were used in cloning and expression. E. coli HB101 (pRK2013) was used as a helper in triparental mating experiments17. Luria-Bertani (LB) and mineral salts basal (MSB) media were described in the previous studies8,11. For the large-scale cultivation of strain TX1, a stirred bioreactor was used. The inoculum (10% of the working volume) was transferred from the flask of the one-day old culture to the bioreactor, which contained 3 L of the desired medium. Cultivations were conducted in a 5 L stirred bioreactor (BTF-A5L, Bio-Top Inc, Taiwan) at 200 rpm and with aeration of 0.3 vvm (volume air/volume liquid/min) in the MSB medium containing 0.5% OPEOn. The fermentation broth was harvested at late log phase for the enzyme purification.

Preparation of crude cell extract and purification of enzyme involved in OPEOn degradation from strain TX1

Strain TX1 was cultivated in 20 L of MSB-0.5% OPEOn medium for the preparation of crude cell extract. The cells were collected by centrifugation (11,000 g, 10 min, 4 °C), washed with 40 mM Na2HPO4/KH2PO4 buffer (pH 7.0) and suspended in 800 mL of the same buffer. The cell crude extract was prepared as in our previous study18,19 and then treated with 0.3% protamine sulfate. The supernatants were collected at 30,000 g for 30 minutes for enzyme purification. All purification procedures were performed at 4 °C in Na2HPO4/KH2PO4 buffer (pH 7.0) unless otherwise stated. Enzymes were loaded onto a DEAE-Sepharose XK 26 column (155 mL). The column was preequilibrated with buffer and enzymes with active fractions were eluted with 1 M KCl. The active fractions were collected and were fractionated with ammonium sulfate. The precipitate obtained with 0 to 60% saturation of ammonium sulfate was collected and applied to a phenyl Superpose 6 fast flow column (60 mL). The active fractions were eluted with a linear gradient of 1 to 0 M (NH4)2SO4. The concentrated enzyme solution was loaded onto a Sephacryl S-200 (176 mL) for the next purification step. The active fractions were eluted with pH 6.9 buffer. Finally, the active fractions after gel filtration were loaded onto a Mono P HR 5/20 chromatography column (4 mL). The column was preequilibrated with 25 mM diethanolamine-HCl (pH 9.5) and was eluted with 100 mL polybuffer 96 (Pharmacia Fine Chemicals) and titrated with HCl to pH 6.0.

Protein quantification, in-gel digestion and protein identification

A 12% SDS-PAGE gel was used for determination of molecular weight of the purified enzyme and enzyme purity. After separation in SDS-PAGE gels, the proteins were visualized by staining using mass compatible Comassie blue. Excised gel pieces were washed with deionized water twice, then destained with 200 μL ammonium bicarbonate (ABC, NH4HCO3)/50% v/v acetonitrile (ACN, CH3CN) for 15 min and dehydrated by incubation with 100 μL 100% ACN for 5 min. The process was repeated until gels were destained completely. The gel pieces were further dried by vacuum concentration for 10 min. Prior to the tryptic digestion, the sample was diluted with 25 mM ABC to give a final urea concentration of less than 0.6 M. Trypsin was added with weight ratio of 20 (protein): 1 (trypsin) and the sample was incubated in a waterbath at 37 °C for 12–16 h. The reaction was stopped with 10 μL 0.1% formic acid. The digested peptides were extracted by ultrasonication for 5 min and stopped for 5 min (repeat three times) and the peptides were kept in the supernatant in the new tube. The sonication procedure was repeated by adding 0.1% formic acid/50% ACN. The peptides were concentrated by vacuum concentration. The resulting peptide mixture was then subjected to the CapLC system (Waters, Milford, MA, USA) utilizing a capillary column (75 µm i.d., 10 cm in length) with a linear gradient from 5 to 50% ACN containing 0.1% formic acid over 46 min. The separated peptides were on-line analyzed under positive survey scan mode on a nano-ESI-Q-TOF instrument (Micromass, Manchester, UK). The scan range was from m/z 400 to1600 for MS and m/z 50 to 2000 for MS/MS. The raw data were acquired and processed using MassLynx V 4.1 software (Micromass) and were converted to PKL files by the ProteinLynx 2.2.5 software. The PKL files were analyzed using the MASCOT search engine (http://www.matrixscience.com). The proteins with scores above the significant threshold (P < 0.05) are shown as identified proteins.

Enzyme assay and hydrogen peroxide assay

Enzyme activity was assayed by oxygen uptake rate in a 1.5 mL reaction mixture containing 50 mM KH2PO4/Na2HPO4 (pH 8.0), 10 μM MgSO4, 5 μM Co(NO3)2, 100 μM FeSO4, 200 μM ZnSO4, 10 μM CuSO4, 60 μM MnSO4, 40 μM Na2EDTA, 0.05% OPEOn, 144 µg enzyme and 0.5 mM NADH at 30 °C by an oxygen monitor (Biological Oxygen Monitor, Yellow springs co. Ohio, USA). To estimate the kinetic parameters of the enzyme, the pure enzyme activity was measured in the forward reactions according to the literature20. Protein concentrations were determined using the Bradford protein assay with bovine serum albumin as the standard21. Specific enzyme activity is reported as nmole/min/mg. The hydrogen peroxide assay was based on the detection of H2O2 using the amplex red fluorescent dye22. In the presence of horseradish peroxidase, the amplex red reagent reacts with H2O2 with a 1:1 stoichiometry producing highly fluorescent resorufin. 1.5 µg pure enzyme were added into 0.25 mL reaction mixture containing 50 mM KH2PO4/Na2HPO4 (pH 8.0), 0.05 mM NADH, 0.2 mM ZnCl2 at 30 °C for 1 h and then horseradish peroxidase (0.2 U/ml) and amplex red reagent (1 µM) were added. Production of resorufin was followed by an increasing absorbance at 571 nm. Concentrations of H2O2 were calculated by comparing absorbance of samples to a series of H2O2 standards (0.2~1 µM) treated with the amplex red mixture.

Construction of bacterial strains

To inactivate the lipoamide dehydrogenase (lpd) gene in wild-type TX1, a gene fragment containing about 400 bp of the internal region of lpd gene (accession number WP_017518066.1) was amplified and cloned into pK18mobsacB plasmid. The primers used were F_lpd: 5′-GCGAATTCGAAGACCCTGACCAAGCAAG (EcoRI, underlined); R_lpd: 5′-GCAAGCTTATTTCCGGGTGGGTGTAGAT (HindIII, underlined). The gene fragments were cloned at the EcoRI/HindIII site into the pK18mobsacB plasmid. The resulting plasmid was named pKlpd. The lpd-internal mutation of TX1 was created by triparental mating between strains TX1, E. coli DH5α (pKlpd) and E. coli HB101 (pRK2013) as previously described23. Transconjugants were screened on LB containing both ampicillin and kanamycin, and finally confirmed by PCR.

To clone the lpd gene into the pET28a vector, a method termed “sticky-end PCR”, which generates PCR product bearing cohesive ends compatible with any intended restriction sites, was used24. Briefly, two pairs of PCR primers differing only in the 5′ ends were designed to amplify lpd. PCR was performed in two separate tubes using primer set 1 (F1: 5′-CATGAGCCAGAAATTCGACGTG-3′ and R1: 5′-GGCGCTTCTTGCGGTTGGC-3′) and set 2 (F2: 5′-AATTCATGAGCCAGAAATTCGACGTG-3′ and R2: 5′-TCGAGGCGCTTCTTGCGGTTGGC-3′), respectively. The differences in the 5′ termini of the primers matched the overhangs generated by the EcoRI and XhoI cleavage are underlined in parentheses. The two PCR products were then mixed, followed by denaturation and annealing, which resulted in half of the mixed products bearing cohesive ends. The mixed DNA fragment was ligated with an EcoRI/XhoI-digested pET28a plasmid. The resulting plasmid was named pElpd and then transformed into E. coli BL21 (DE3).

Purification of Lpd in recombinant E. coli

The cell crude extract from E. coli BL21 was filtered through a 0.22 µm filter and loaded onto HisTrap affinity column (5 mL; GE Healthcare). The unabsorbed and loosely bound proteins were eluted from the column by washing with 5 column volumes of 30 mM imidazole in 20 mM KH2PO4/Na2HPO4 buffer (pH 7.0). His-tag Lpd protein was eluted with 500 mM imidazole in the same buffer. All the fractions containing Lpd activity were combined. To further purify the enzyme, gel filtration was performed with a Superose 6 column that was preequilibrated with 50 mM KH2PO4/Na2HPO4 at pH 7.0. The proteins were eluted with the same buffer and all the active fractions were combined. All procedures were done at 4 °C.

Identification of degradation products from OPEOn

To carry out the analysis of degradation products, each 4 mL enzyme reaction (enzyme activity was performed in a reaction mixture containing 50 mM KH2PO4/Na2HPO4 (pH 8.0), 10 μM MgSO4, 5 μM Co(NO3)2, 100 μM FeSO4, 200 μM ZnSO4, 10 μM CuSO4, 60 μM MnSO4, 40 μM Na2EDTA, 0.05% OPEOn, 96 µg/mL pure Lpd enzyme and 0.5 mM NADH at 30 °C) at its particular time point was mixed vigorously with 10 mL of 72.4% of MgSO4 and 0.2 mL of 2.5 M H2SO4 followed by extraction with 25 mL of chloroform (CHCl3) three times14. A portion of the organic phase was collected and dried immediately using a rotary evaporator. Acetonitrile (2 mL) was then added to dissolve the dry residue for liquid chromatography– mass spectrometry (LC–MS) analysis. OPEOn and its degradation products were analyzed by a high-performance liquid chromatography (HPLC) system (Waters Alliance 2690, Milford, MA, USA) equipped with an electrospray ionization– mass spectrometer (Platform LC; Micromass, Manchester, UK). The injection volume was 20 µL and the flow rate was set at 0.5 mL/min. A 5 µm C18 column (Waters µBondapak, 3.9 × 150 mm) was used for the separation. The OPEOn was analyzed in the positive mode. In the positive mode, a mobile phase containing 0.1% aqueous formic acid and acetonitrile (3:7) was applied to analyze OPEOn and the products. The potentials of the electrospray ionization source were set at 3.5 V for the capillary voltage and 50 V for the cone voltage. The source temperature was 100 °C and the flow rate of nitrogen gas was controlled at 300 L h−1.

Results

Existence of metabolic enzymes related to OPEOn degradation in strain TX1

P. nitroreducens TX1 is of special interest because of its capability to use 0.05–20% OPEOn as a sole carbon source1,8,9,10,11,16. In this study, oxygen consumption activity was used to evaluate the involvement of O2 in the degradation of OPEOn in a whole cell system. The result revealed that OPEOn-dependent oxygen consumption activity induced in TX1 cells prepared from MSB containing 0.5% OPEOn (33.8 nmole/min) was 3.7 fold higher than that grown on 0.5% succinate (Table 1). These finding suggested the presence of metabolic enzymes in the TX1 cell which are related to the growth on OPEOn as sole carbon source. Using oxygen consumption activity as an enzyme assay, we attempted to purify the metabolic enzymes involved.

Table 1 Oxygen uptake in Pseudomonas nitroreducens TX1 grown on different concentrations of OPEOn or succinate.
Full size table

Purification of enzyme involved in OPEOn degradation from strain TX1

The cell-free extract of strain TX1 was prepared from 20 L culture. The purification steps for the enzyme are summarized in Table 2. Enzyme activity was assayed by oxygen consumption as described in “Materials and Methods”. The enzyme was purified 18-fold with a 0.4% yield that appeared to be highly purified on an SDS-PAGE gel (Fig. 1). The specific activity of the purified enzyme was 41.3 nmol/min/mg. On SDS-PAGE, the molecular mass of a monomeric protein was estimated to be 52 kDa. By gel filtration, the active fraction was eluted at a retention time corresponding to approximately 98 kDa. The finding suggested that the enzyme exists as a dimer in the native state. The optimum temperature and pH were found to be 30 °C and 8.0, respectively. The effects of various metal ions on purified enzyme activity were examined (Table 3). Some metal ions (Mn2+, Mg2+, Zn2+, and Cu2+) enhanced enzyme activity. Excess of EDTA was checked and revealed a similar activity as the “no metals” control.

Table 2 Purification of OPEOn - oxygen consuming enzyme from P. nitroreducens TX1.
Full size table
Figure 1
figure 1

(A) 12% SDS-PAGE of the purification steps from P. nitroreducens TX1. Lane 1: Crude cell extract of strain TX1; lane 2: protamine sulfate treated; lane 3: DEAE-Sepharose; lane 4: 25–60% ammonium sulfate; lane 5: Phenyl-Sepharose; lane 6: Sephacryl S-200; lane 7: Mono P HR 5/20; lane M: molecular mass markers. (B) 12% SDS-PAGE of the purification steps from recombinant E. coli BL21 (DE3). Lane 1: Crude cell extract of recombinant cells; lane 2: His-trap; lane 3: Superose 6. The protein sample contains 5 μg in each well. The gel was stained by Coomassie Brilliant blue.

Full size image
Table 3 Metal effect on purified enzyme analyzed by oxygen consumption activity.
Full size table

Analysis of degradation products from pure enzyme

The products of OPEOn degraded by the enzyme in the presence of metals and an excess of NADH were analyzed by HPLC-MS. Figure 2 shows mass spectra of the OPEOn at zero time (A) and that after the incubation for 2 h (B). In both mass spectra, a series of intense ion peaks (■) of OPEOn molecules with a sodium cation attached [R-(CH2CH2O)n-H + Na+] are clearly observed. However, compared with the mass spectrum of the original sample, the ion peak distribution of mass spectra obviously shifts to lower mass with the elapse of the incubation time. After 2 hours of reaction, the products obtained from the degradation experiment revealed a prominent ion peak at m/z = 493.3, corresponding to OPEO6. When comparing OPEOn without the presence of the enzyme, the products were mainly OPEO9 (m/z = 625.3), the dominant form in the original substrate. Mass spectra of the reaction showed a characteristic pattern due to loss of individual EO [CH2-CH2-OH] units of m/z = 44.0 from the parental OPEOn. The shortest detectable compound in our condition is OPEO3 (m/z = 371.3). However, the other product with smaller molecular mass is not detected in this study. The reasons might be that small EO units are (1) lost in extraction steps; (2) co-eluted with initial solvent peak in LC. The degradation products detected in this study indicate that the degradation mechanism for OPEOn by a pure enzyme from strain TX1 involves the shortening of ethoxylate units in the polyethoxylate chains.

Figure 2
figure 2

Mass spectra of OPEOn degradation by the purified enzyme in the present of metals and NADH. (A) At zero time, the mass spectrum of reaction shows a characteristic pattern of the OPEOn parental ion pattern. The most abundant ethoxymer is OPEO9 (m/z = 625.3). (B) After 2 hours reaction, the profile shifts to lower EO chain products with OPEO6 (m/z = 493.3) as most abundant ethoxymer. The reaction mixture (4 mL) contains 50 mM KH2PO4/Na2HPO4 (pH 8.0), 10 μM MgSO4, 5 μM Co(NO3)2, 100 μM FeSO4, 200 μM ZnSO4, 10 μM CuSO4, 60 μM MnSO4, 40 μM Na2EDTA, 0.05% OPEOn, 96 µg/mL pure Lpd enzyme and 0.5 mM NADH at 30 °C.

Full size image

Protein identification

The identity of the protein band was determined by partial amino acid sequence via ESI-MS/MS. The major protein in the 52-kDa band matched to dihydrolipoamide dehydrogenase from P. nitroreducens TX1 (accession number WP_017518066), which had the highest Mowse score and the largest number of peptide matches (31 peptide matches). These peptide matches cover 47% of the protein and are spread throughout the protein (Supplemental Fig. S1). The calculated molecular weight of Lpd (50 kDa) is in close agreement with that observed in our SDS-PAGE analysis. The absorption spectrum of the purified enzyme shows maximal absorbance at 375 and 458 nm. As the enzyme had high homology with Lpds, the Lpd activity was confirmed according to the literature20. When the concentrations of NAD+ and dihydrolipoamide were varied, the initial velocities represented a series of parallel lines, suggestive of a ping-pong reaction mechanism25. The result revealed the kinetic parameters of Lpd (KM, NAD+ = 0.52 mM; Vmax = 0.09 µmole/min) (Fig. S2). The specific activity of Lpd in this experiment was 108.2 µmole/min/mg which was higher than that from P. fluorescens (35.7 µmole/min/mg)26 and P. putida (68 µmole/min/mg)27.

Analysis of the genome fragment containing the lpd gene and insertion mutagenesis

According to the draft genome sequence from strain TX116, the lpd gene locates in contig099 (accession number NZ_AMZB01000099). A 7.9 kb fragment located in the 57613–65519 position of contig099 was analyzed (Fig. S3). The succinate dehydrogenase, 2-oxoglutarate dehydrogenase, and dihydrolipoamide succinyltransferase genes were located upstream of lpd, whereas the malate-CoA ligase gene was located downstream of lpd. All genes are involved in the TCA cycle. The lpd consists of 1,437 bp (Fig. S4), encoding 478 amino acids. A search of a nucleotide sequence database (http://www.ncbi.nlm.nih.gov/BLAST) was performed and revealed that lpd had the highest similarity (94%) to the dihydrolipoamide dehydrogenase gene from P. denitrificans (WP_015477517.1). For different genera, the highest similarity (83%) to the dihydrolipoamide dehydrogenase gene from Azotobacter vinelandii (3LAD) was observed. The sequence comparison was carried out and revealed that the FAD-binding domain (Gly-11, 13 and 16), the NAD-binding domain (Gly-186, 188, 191 and 202) and the residues involved in the dimerization are completely conserved in Lpd (TX1), as compared to others (Fig. 3). A redox active center containing two cysteine residues (Cys-49 and Cys-54) are directly engaged in thiol-disulfide exchange reactions during catalysis. To further confirm the role of lpd gene on the growth of TX1 with OPEOn, efforts in inactivate the lpd gene were made, but no mutant was found, suggesting that Lpd is essential to cell.

Figure 3
figure 3

The amino acid sequence alignment of dihydrolipoamide dehydrogenase from Pseudomonas nitroreducens TX1 with these from Pseudomonas denitrificans (GenBank accession number WP_015477517.1), Azotobacter Vinelandii (3LAD) and Escherichia coli (P00391). Identical amino acid residues are highlighted in gray. ▲, the residues involved in the dimerization; ■, the residues involved in FAD binding.

Full size image

Characterization of recombinant Lpd

Because of the low yield (0.4%) of purified protein from wild type TX1 and the long procedure in purification steps, the construction of recombinant Lpd was performed. BL21 (DE3) cells carrying the pElpd plasmid were grown at 20 °C for 20 h to maximize the protein yield. Lpd was purified using a HisTrap Chelating column with a recovery of about 99%. His-tagged proteins were further purified using a Superose 6 gel filtration column. The enzyme was purified 60-fold with a 51% yield. The molecular weight (58 kD), including a 6xHis-tag of the purified protein was in good agreement with the calculated molecular weight of Lpd. By gel filtration, the active fraction was eluted at a retention time corresponding to approximately 127 kDa (data not shown). The result re-confirmed that Lpd is a dimer in its native state.

Degradation of each ethoxymer by pure recombinant Lpd

The extent of removal for residual OPEOs during the degradation process by pure recombinant Lpd was also investigated. Figure 4 shows the disappearance of each ethoxymer of the major distribution. The amount of OPEOn decreases as the enzyme-substrate reaction time increases. After 4 hours, the treatment of purified Lpd reduces OPEO15 level by approximately 90%. In this figure, we only use the ethoxymers with odd numbers of the ethoxylate units to represent the results. The disappearance rates for OPEOn homologs were OPEO15 > OPEO13 > OPEO11 > OPEO9 > OPEO7 > OPEO5, meaning the longer the ethoxylate chain, the higher chance they were attacked by the reaction generated by this enzyme.

Figure 4
figure 4

Degradation of OPEOn by purified recombinant Lpd in the presence of metals and NADH. The reaction mixture (4 mL) contained 50 mM KH2PO4/Na2HPO4 (pH 8.0), 10 μM MgSO4, 5 μM Co(NO3)2, 100 μM FeSO4, 200 μM ZnSO4, 10 μM CuSO4, 60 μM MnSO4, 40 μM Na2EDTA, 0.05% OPEOn, 96 µg/mL pure Lpd enzyme and 0.5 mM NADH at 30 °C. Total ion count of representative OPEOn (n = 5, 7, 9, 11, 13, 15) is expressed as percentage of the initial ion count of the individual ethoxymer according to the function of reaction time. The ethoxymer containing odd numbers of ethoxylate units represents the degradation of each monomer in this figure. Data points and error bars represent the means and standard deviations of triplicate measurements.

Full size image

Discussion

In this study, we investigated the enzyme activities involved in OPEOn degradation in strain TX1. Since the degradation of OPEOn required O2 consumption, the oxygen-consuming enzymes were monitored and purified from strain TX1. The pure enzyme was identified as dihydrolipoamide dehydrogenase by its mass and amino acid sequences. The absorption spectrum of the purified Lpd was also studied and showed maximal absorbance at 375 and 458 nm, indicating the presence of FAD as a cofactor in Lpd. The results were consistent with the previous studies, in which Lpd is a homo-dimer25,28. Efforts to inactivate the lpd gene in strain TX1 was made, however no mutant was obtained. The reason might be due to Lpd serving as a component of some important complex enzyme systems such as the pyruvate dehydrogenase complex and the branched chain ketoacid dehydrogenase complex29.

The function of Lpd from TX1 was demonstrated in vitro for the degradation of ethoxylate chain containing nonionic surfactants such as Triton X-100. Based on the analysis of OPEOn degraded by Lpd in the presence of metals and an excess of NADH using HPLC-mass spectrometry, the profile of the OPEOn parental ion pattern (the most abundant component is OPEO9 (m/z = 625.3)) shifts to lower EO chain products with OPEO6 (m/z = 493.3) as the most abundant component after 2 hours reaction. The products detected in this study indicate that the biodegradation mechanism for OPEOn by Lpd involves the cleavage of ethoxylate units in polyethoxylate chains of the surfactant structure. For the shortening of ethoxylate units in polyethoxylate chains during OPEOn degradation, three models – nonoxidative biodegradation, oxidative biodegradation and a mechanism involving the attack by •OH radicals – have been proposed previously30,31. Production of hydrogen peroxide and superoxide radical catalyzed by lipoamide dehydrogenase was first reported in 1955 and then confirmed in 196932,33,34. In addition, when Lpd oxidizes NADH, it also has the ability to reduce ferrous ion presence in the reaction mixture and occurs in a linear proportion manner35. The mixture of ferrous ions and H2O2, namely Fenton’s reagents, will generate hydroxyl radicals according to the reaction: Fe2+ + H2O2 → Fe3+ + −OH + •OH36,37,38,39,40. The formation of H2O2 from purified Lpd (TX1) in the presence of metal and an excess of NAD(P)H was investigated in this study. To test the possibility that H2O2 derived from the Lpd catalyzes the reaction, we measured the production of H2O2 by using the Amplex Red fluorescent dye. H2O2 generation was dependent on the presence of Lpd and NADH (Fig. S5). The effect of Lpd concentration on hydrogen peroxide formation was measured by increasing the amount of purified Lpd (0~14 µg). A linear relation between H2O2 and Lpd was obtained (Fig. S5A). In addition, the effect of NADH on the production of H2O2 by Lpd was also analyzed by changing the amount of NADH (0~2 mM). A relation between NADH and the production of H2O2 was directly proportional (Fig. S5B). In addition, the longer the ethoxylate chain of the ethoxymers, the higher was the degradation rate observed (Fig. 4). Therefore, we propose the mechanism for the shortening of the ethoxylated chain of OPEOn involves the attack by •OH radicals that are formed through the Lpd reaction (Fig. 5). In this reaction, Lpd might perform three functions: (1) the oxidation of NAD(P)H35,41; the consumption of O2 to subsequently generate H2O2 32,33,34; and the reduction of metals35,41. With the combination of the three reactions (Lpd: NAD(P)H + M2+ + O2 + H+ → NAD(P)+ + M3+ + •OH + OH−), •OH is produced and then attacks the surfactant. The attack by •OH radicals to OPEOn was studied previously31. According to Brand et al. (1998), three different sites of OPEOn are proposed to be attacked by •OH radicals: (1) CH2 and CH3 groups of the alkyl chain, (2) the aromatic ring, and (3) CH2 groups of the ethoxylate chain. The attack on the CH2 groups of the ethoxy chain is highly favored. The attack of the •OH radicals on OPEOn leads to the formation of OPEOn with a smaller number of ethoxy units and low molecular mass units. These low molecular mass units might convert into acetyl-CoA that goes into the glyoxylate cycle (Fig. 5). Thus, this study demonstrates the bacteria can respond to environmental stress by using existing housekeeping enzymes. It might also be applied to the bacterial degradation of other widely used ethoxylated compounds such as polyethylene glycol, Tween and ethoxylated glycerol types of nonionic surfactants.

Figure 5
figure 5

The proposed mechanism for the shortening of the ethoxylated chain of OPEOn involves the attack by OH• radicals that are formed in the presence of lipoamide dehydrogenase, NADH and an excess of metals. The low molecular mass EO units might convert into acetyl-CoA that goes into the glyoxylate cycle to serve as carbon source. M, metals; EO, ethoxylate.

Full size image

Conclusions

Microbes in the environments play a major role in the degradation of widely used surfactants. In this study, Pseudomonas nitroreducens TX1 is of special interest because of its capability to use a wide range (0.05–20%) of nonionic ethoxylated surfactants such as octylphenol polyethoxylate (OPEOn) as a sole carbon source. We demonstrate a novel catalytic reaction carried out by dihydrolipoamide dehydrogenase involved in the radical formation to attack ethoxylated surfactants which serve as a sole carbon source. It also suggests that a diverse and novel strategy by which bacteria can catabolize a sole carbon source present in the environments.