Formation of Toxic 2-Nonyl-p-Benzoquinones from ... - ACS Publications

It has been reported that technical nonylphenol contains, besides α-quaternary nonylphenols, minor amounts of various α-H, α-methyl substituted ter...
0 downloads 0 Views 543KB Size
Article pubs.acs.org/est

Formation of Toxic 2-Nonyl-p-Benzoquinones from α-Tertiary 4-Nonylphenol Isomers during Microbial Metabolism of Technical Nonylphenol Frédéric L. P. Gabriel,†,§ Mauricio Arrieta Mora,†,∥ Boris A. Kolvenbach,‡ Philippe F. X. Corvini,‡ and Hans-Peter E. Kohler†,* †

Eawag, Swiss Federal Institute of Aquatic Science and Technology, CH-8600 Dübendorf Institute for Ecopreneurship, School of Life Sciences, University of Applied Sciences Northwestern Switzerland, Muttenz, Switzerland



S Supporting Information *

ABSTRACT: In many environmental compartments, microbial degradation of α-quaternary nonylphenols proceeds along an ipsosubstitution pathway. It has been reported that technical nonylphenol contains, besides α-quaternary nonylphenols, minor amounts of various α-H, α-methyl substituted tertiary isomers. Here, we show that potentially toxic metabolites of such minor components are formed during ipso-degradation of technical nonylphenol by Sphingobium xenophagum Bayram, a strain isolated from activated sewage sludge. Small but significant amounts of nonylphenols were converted to the corresponding nonylhydroquinones, which in the presence of air oxygen oxidized to the corresponding nonyl-p-benzoquinonesyielding a complex mixture of potentially toxic metabolites. Through reduction with ascorbic acid and subsequent analysis by gas chromatography−mass spectrometry, we were able to characterize this unique metabolic fingerprint and to show that its components originated for the most part from α-tertiary nonylphenol isomers. Furthermore, our results indicate that the metabolites mixture also contained several α, β-dehydrogenated derivatives of nonyl-p-benzoquinones that originated by hydroxylation induced rearrangement, and subsequent ring and side chain oxidation from α-tertiary nonylphenol isomers. We predict that in nonylphenol polluted natural systems, in which microbial ipso-degradation is prominent, 2-alkylquinone metabolites will be produced and will contribute to the overall toxicity of the remaining material.



INTRODUCTION Nonylphenols are environmental contaminants that are toxic to many aquatic organisms and are strictly regulated in Europe.1 Since the beginning of the 1990s, scientific interest has increasingly focused on these compounds, as it became evident that they are able to disturb hormonal homeostasis of vertebrates by mimicking the estrogenic activity of the natural female hormone 17-β-estradiol.2,3 Technical nonylphenol is industrially produced by Friedel−Crafts alkylation of phenol with propylene trimer, a heterogeneous mixture of nonenes formed by acid catalyzed polymerization of propene. Because of this unspecific mode of production, technical nonylphenol is a mixture of more than 150 components that differ in the structure and the position of the alkyl moiety attached to the phenol ring. Depending on the producer, 4-nonylphenols account for ca. 86−94% of the technical mixture, with 2nonylphenols and decylphenols making up ca. 2−9 and 2−5%, respectively.4 Yet, authors of toxicological and environmental monitoring studies generally have treated technical nonylphenol, regardless of its complexity, as a single compound.5,6 However, recent findings show that estrogenic effects and © 2012 American Chemical Society

degradation rates of isomers vary depending on the structure of the alkyl chain.7−12 In the past decade, several Sphingomonads able to utilize nonylphenol as sole carbon and energy source were isolated from activated sludge.13−16 Degradation experiments showed that Sphingobium xenophagum Bayram markedly changes the isomer distribution pattern of technical nonylphenol when growing with this substrate as the sole carbon and energy source.11 In such experiments, degradation stopped after the bacteria consumed about 86% of the 4-nonylphenol. A strong correlation between transformation of individual isomers and their α-substitution pattern was observed. As a rule, isomers with little bulkiness at the α-carbon and those with a main alkyl chain length of 4−6 carbon atoms were degraded most efficiently.11 Received: Revised: Accepted: Published: 5979

January 30, 2012 April 23, 2012 May 4, 2012 May 4, 2012 dx.doi.org/10.1021/es300383m | Environ. Sci. Technol. 2012, 46, 5979−5987

Environmental Science & Technology

Article

Figure 1. Metabolism of α-quaternary (e.g., NP112) (A) and α-tertiary nonylphenol isomers (e.g., NP2) (B) by strain Bayram.15,17,19 The initial reaction for degradation of these substrates is an ipso-hydroxylation, yielding dearomatized intermediates. α-Quaternary 4-alkyl-4-hydroxy-cyclohexa2,5-dien-1-ones spontaneously break down to hydroquinone and a stabilized alkyl cation, which reacts with a water molecule to yield the corresponding nonanol (A). Because of insufficient stabilization by α-substitution, the alkyl moiety of α-tertiary 4-alkyl-4-hydroxy-cyclohexadienones is not released as a cation. These α-tertiary intermediates hence accumulate and are transformed by a side-reaction to the corresponding 2nonylhydroquinones (dienone−phenol rearrangement, NIH-shift).17 2-Nonylhydroquinones are spontaneously oxidized by air oxygen to yield the corresponding p-benzoquinone derivatives (B). Bold and thin reaction arrows symbolize putatively high and medium rates of reaction, respectively.

Through analysis of NIH-shift metabolites 17,25 by gas chormatography−mass spectrometry, we were able to reveal a unique fingerprint of metabolites left behind when technical nonylphenol was undergoing ipso-degradation. Our results indicate that these metabolites are mainly quinonoid compounds derived from α-tertiary nonylphenol isomers.

Growth experiments with strain Bayram and pure nonylphenol isomers showed that α-quaternary nonylphenol isomers serve as growth substrates, whereas nonylphenols containing α-hydrogens do not. Nevertheless, such isomers are cometabolically transformed to para-hydroxylated metabolites with retained alkyl moieties (Figure 1B).15,17 These findings, together with elaborate 18O labeling experiments, led to elucidation of the degradation pathway. Nonylphenols are hydroxylated at the ipso-position and are thereby transformed to 4-alkyl-4-hydroxy-cyclohexa-2,5-dien-1-ones, from which αquaternary alkyl moieties are able to detach as transient alkyl carbocations.17−19 These carbocations electrophilically react with water molecules to yield the corresponding nonanols (Figure 1A). The detachment of the alkyl moiety results in the formation of hydroquinone, which is further utilized by the bacteria for growth. As the carbocation only forms if sufficiently stabilized by α-alkyl branching, side chains containing αhydrogens are not released (Figure 1B). α-Tertiary 4-alkyl-4hydroxy-cyclohexadienone intermediates hence accumulate and undergo side reactions, for example, rearrangements yielding the corresponding 2-nonylhydroquinones (NIH-shift). In contact with air oxygen, the 2-nonylhydroquinones appear to be oxidized to the corresponding 2-nonyl-p-benzoquinones20 (Figure 1B). A recent analysis by two-dimensional gas chromatography− time-of-flight mass spectrometry reported that technical nonylphenol contains various α-H, α-methyl substituted nonylphenol isomers.4 Because such isomers have the potential to be transformed to 2-nonylhydroquinones in environmental compartments that are dominated by nonylphenol metabolism along an ipso-substitution pathway21,22 and because quinonoid compounds, derived from (substituted) hydroquinones, are agents of oxidative stress and have a high toxic potential,23,24 we investigated the formation of nonylhydroquinones during ipsodegradation of technical nonylphenol in strain Bayram.



EXPERIMENTAL PROCEDURES General. Sphingobium xenophagum Bayram was isolated from activated sludge of a municipal wastewater treatment plant nearby Zürich, Switzerland.15 The abbreviations used for the various 4-nonylphenol isomers are based on the systematic numbering proposed by Guenther et al.26 Technical nonylphenol (>85% as referred to the amount of para-nonylphenol isomers) was purchased from Fluka (Buchs, Switzerland). Purities of chemically synthesized 4-NP10 and 4-NP70 have been described elsewhere. 11 2-(1-ethyl-1,4-dimethylpentyl)hydroquinone was chemically synthesized in an electrophilic substitution reaction involving hydroquinone, 3,6-dimethylheptan-3-ol, and bortrifluoride.27 Minimal medium (Yeast nitrogen base without amino acids, YNB, Difco, Detroit, MI) was prepared following the manufacturer’s recommendations (the pH value was adjusted to 7 with NaOH). Analytical Procedures. For GC-MS analysis, we used a GC 8060 gas chromatograph (Fisons Instruments, Milan, Italy) that was coupled to a MD 800 quadrupole mass spectrometer (Fisons Instruments, Manchester, UK). Helium was used as carrier gas with a flow rate of 1.2 mL/min. The injector was operated in splitless mode at 270 °C. Split and purge flows were adjusted to 10 mL/min. Sample volumes of 2 μL were injected by an A200S autosampler (CTC analytics, Zwingen, Switzerland). Separations were performed on a DB-17 MS capillary column (60 m long, 0.25 mm internal diameter and 0.25 μm film thickness; J&W Scientific, Folsom, California). Oven temperature was increased with 10 °C/min from 60 °C 5980

dx.doi.org/10.1021/es300383m | Environ. Sci. Technol. 2012, 46, 5979−5987

Environmental Science & Technology

Article

Figure 2. GC-MS chromatogram (total ion chromatogram) of the trimethylsilyl derivatives of 2-nonylhydroquinone metabolites produced during ipso-degradation of technical nonylphenol by strain Bayram. Peaks that show minimal interference with neighboring peaks are numbered. Spectra were classified into 6 different groups (A−F) based on the set of characteristic signals displayed.

(0 min) to 200 °C (0 min) and then with 0.5 °C/min from 200 to 213 °C. Interface and source temperatures were set at 250 and 200 °C, respectively. Detection was performed in the electron impact mode at 45 eV. Data were acquired in the full scan mode from 36 to 390 atomic mass units (amu) with 2.9 scans per second. N,O-(Bis(trimethylsilyl)trifluoroacetamide (BSTFA) was used as derivatizing agent. For derivatization, an aliquot solution was concentrated to dryness by flushing with a stream of N2. A surplus of pure BSTFA was added and the mixture was left to stand overnight. The BSTFA was then completely evaporated and the residue was twice dissolved in CH3CN and concentrated. Finally the residue was resuspended in CH2Cl2. For HPLC analysis with ultraviolet (UV) detection, a Gynkotek HPLC system consisting of a M480 gradient pump, a UV/vis UVD3404 diode array detector, and a Gina 50 autosampler (Dionex, Olten, Switzerland) was used. Twenty μL of samples dissolved in 2-propanol were injected. Separation was performed on a reversed-phase column protected with a precolumn (CC 250/4 and CC8/4 Nucleosil 100-5 C18 HD, respectively; Macherey-Nagel, Oensingen, Switzerland). Solvent flow rate was set at 1 mL/min. Prior to injection, the column was equilibrated for 5 min with 100% solvent A (methanol/water 7:3). Then, a linear gradient to 60% A and 40% B (methanol/water 9:3) was applied for 17 min, followed by a linear gradient to 100% B for 8 min. Isocratic conditions (100% B) were then maintained for 12 min. At last, a gradient was applied for 3 min to restore starting conditions (100% A). Degradation Experiments with Resting Cells and Single α-Tertiary 4-Nonylphenol Isomers. Assays were performed in 15 mL vials containing 2.4 mL of 20 mM sodium phosphate buffer (ph 7.4) and 0.3 mL of a suspension of induced cells of strain Bayram (final cell concentration of 6.0

mg fresh weight/mL assay mixture). Preparation of the cell suspension has been described elsewhere.19 Experiments were started by adding 14 μL of a 60 mM methanolic solution of 4NP10 or 4-NP70. Assay mixtures were stirred at room temperature with a magnet (8 × 3 mm) and aerated for 1 h with a flow of air flushing onto the surface of the liquid. Assay mixtures were then shock frozen at −80 °C (storage at −20 °C) and extracted 3 x with 2 mL of CH2Cl2. After evaporation of the solvent, the extract was dissolved in an appropriate volume of CH2Cl2 and analyzed by GC−MS. Degradation Experiments with Growing Cells and Technical Nonylphenol. Sterile 300 mL cotton plugged Erlenmeyer vials containing each 50 mL of minimal medium and 50 mg technical nonylphenol (1 mg/mL) were inoculated each with 1 mL of a 9 day old minimal medium liquid culture (optical density of ca. 0.4 at 546 nm, 1 mg technical nonylphenol/ml as sole carbon source) and incubated at 25 °C on a rotary shaker (240 rpm). Experimental incubations were stopped at days 7 and 9 and appropriate control incubations at days 0 and 9 by shock freezing at −80 °C. The vials were stored at −20 °C until they were analyzed. Thawed assay mixtures were extracted 3× in the culture vials with 20 mL of CH2Cl2. Organic phases were completely evaporated on a rotary evaporator. Residues were taken up in CH2Cl2 by repeated application of 1 mL samples of solvent. The pooled solvent was transferred to a small vial and the total solvent volume was determined by weighing. Aliquots were taken for GC-MS and HPLC analysis. To prepare HPLC samples, CH2Cl2 was exchanged with 2-propanol. Characterization of 2-nonylhydroquinone metabolites was performed with the day-7 extract. To reduce the 2-nonyl-p-benzoquinones to the corresponding hydroquinone derivatives for GC-MS analysis, 500 μL of an 5981

dx.doi.org/10.1021/es300383m | Environ. Sci. Technol. 2012, 46, 5979−5987

Environmental Science & Technology

Article

Figure 3. Typical mass spectra of peaks in the GC-MS chromatogram of the trimethylsilyl derivatives of nonylhydroquinones formed during ipsodegradation of technical nonylphenol. Spectra were classified into 6 different groups (Figure 4): Group A (the spectrum of peak 8 is shown), B (peak 11), C (peak 6), D (peak 9), E (peak 1), F (peak 4). Group A spectra represent isomers with an α-H, α-methyl substitution and the same is most likely true for Group B spectra. Group E spectra are probably produced by isomers with an α-H, α-ethyl substitution (considering the NIH-shift mechanism, an α-dimethyl substitution seems unlikely). Spectra of Groups C, D, and F most likely represent mixtures of coeluting isomers (see SI Figure S5). The 2 spectra at the bottom are those of the standards NHQ112 and 4-NP112.

were added to an aliquot of day-7 extract (500 μL), which had been resuspended in 500 μL CH3OH. Group-Wise Separation of the 2-Nonylhydroquinones by Semipreparative HPLC. The 2-nonylhydroquinones (UVmax ca. 290 nm) obtained by reduction of the 2-nonyl-pbenzoquinones (UVmax ca. 250 nm) contained in a 500 μL aliquot of the day-7 extract (see above) were separated from remaining nonylphenols (UVmax ca. 280 nm) by semipreparative HPLC-UV. The pooled nonylhydroquinone

aqueous solution of ascorbic acid (1 mg/mL) was added to 500 μL aliquot of a CH2Cl2 extract and the mixture was vortexed 5 times for ca. 2 min with intervals of 15 min. The CH2Cl2 phase was used for derivatization with BSTFA and GC-MS analysis. To reduce the aliquot destined for HPLC analysis, 10 μL of a concentrated 2-propanol solution of ascorbic acid (10 mg/mL) were added to 100 μL of HPLC sample and the mixture was incubated for 1 h at room temperature. For semipreparative purposes, 50 μL methanolic ascorbic acid solution (10 mg/mL) 5982

dx.doi.org/10.1021/es300383m | Environ. Sci. Technol. 2012, 46, 5979−5987

Environmental Science & Technology

Article

Figure 4. Classification of typical mass spectra obtained from chromatographic peaks in Figure 2. Spectra were grouped on the basis of the intensities of the peaks at m/z 281 and 378, and the characteristics of the range m/z 281 − 378.

Mass spectra of the silylated nonylhydroquinones showed an intensive molecular peak at m/z 380,28 which together with other peaks was assigned to the “m/z 380 sequence”, stretching from m/z 380 to m/z 267 and containing m/z terms spaced by intervals of m/z 14 (first interval of m/z 15) (m/z 380, m/z 365, m/z 351, m/z 337, ..., m/z 267) (Figure 3, Supporting Information (SI) Table S1). Fragment peaks belonging to the m/z 380 sequence were most likely produced by ions that lost one or several α-substituents and thus correspond to peaks of the m/z 220 sequence in spectra of underivatized nonylphenols (m/z 220, m/z 205, m/z 191, m/z 177, ..., m/z 107) (Figure 3).11,29 However, spectra belonging to certain chromatographic peaks contained m/z signals associated with a “m/z 378 sequence” stretching from m/z 378 to m/z 265 (m/z 378, m/z 363, m/z 349, ..., m/z 265) (Figure 3). Mass spectra of the selected chromatographic peaks were classified into 6 groups (A, B, C, D, E, and F) (Figures 2−4, SI Table S1), notwithstanding some uncertainties concerning spectral purities. Spectra of group A isomers (GC peaks 3, 7, 8, and 10, Figure 2) were characterized by the absence of signals between the intense molecular peak at m/z 380 and the base peak at m/z 281 ([M·+ − C7H15·]). This clearly indicates that these isomers bear an α-C7H15 substituent, which is easily cleaved (Figure 3 and SI Figure S3), a view that is strongly corroborated by the fact that well-characterized isomers with this type of substitution, NHQ2, NHQ10, and NHQ70, (derived from 4-NP2, 4-NP10 and 4-NP70, respectively) produce typical Group A spectra (see ref 19 and the paragraph “Degradation experiments with resting cells and single α-tertiary 4-nonylphenol isomers” in the Experimental Procedures). The presence of an α-C7H15 substituent implies that Group A isomers also have α-H and α-CH3 substituents (Figure 3). The base peaks at m/z 281 in the mass spectra of Groups B, C, and D strongly indicated the presence of isomers with an αmethyl substituent. But in contrast to group A spectra, Group B spectra showed between m/z 380 and 281 several small to medium sized peaks associated with m/z 380 and m/z 378 sequences (Figure 3, SI Table S1). In Group B spectra, however (GC signals 11 and 12 in Figure 2), these peaks were very small (5−7%) and most likely represented contaminating signals and, therefore, we concluded that Group B isomers had the same α-substitution as Group A isomers (α-H, α-CH3) (Figure 3, SI Table S1). Spectra of Groups C (GC peaks 5 and 6) and D (GC peak 9) contained characteristic m/z 378sequence peaks at m/z 307 and m/z 321, respectively (Figure 3, SI Table S1). These spectra showed an m/z 378 peak with an

fractions (methanol−water solution) were completely concentrated on a rotary evaporator. Traces of water were eliminated by flushing with N2. The residue was dissolved in CH3CN and transferred to a small vial for derivatization with BSTFA (see above). The CH2Cl2 solution containing the derivatized nonylhydroquinone was concentrated to a volume of 30 μL to obtain a GC-MS chromatogram with a sharp baseline and intense peaks (Figure 2). Proportion of Degraded Technical Nonylphenol. Remaining amounts of technical nonylphenol were determined by analyzing reduced aliquots of culture extracts (see above) and technical nonylphenol standards with HPLC-UV (external calibration). Recovery of technical nonylphenol in the nonincubated control (day-0 extract) was 93.1 ± 5.2% (mean value derived from two HPLC-runs ± standard deviation) and was defined as 100% for calculating the proportion of nonylphenol remaining after day 7 (13.0 ± 0.7%) and day 9 (25.2 ± 1.2%) (the sample sacrificed at day 9 showed lower degradation efficiency).



RESULTS

In degradation experiments with strain Bayram and technical nonylphenol as sole carbon source, ca. 81% (average value of two incubations) of the nonylphenols (UVmax ca. 280 nm, retention time ≥25 min) were metabolized within a week of incubation. When analyzing the culture extracts by means of HPLC-UV, we could detect only very small amounts of 2nonylhydroquinone metabolites (UVmax ca. 290 nm, retention time ca. 10 min). However, the presence of lipophilic products (retention time ca. 22.5 min) with UV-spectra highly similar to that of p-benzoquinone (UVmax at 250 nm) strongly suggested that the bulk of the 2-nonylhydroquinone metabolites had been oxidized to 2-nonyl-p-benzoquinones. We were not able to isolate the nonyl-p-benzoquinones by HPLC fractionation, since they quickly decomposed upon evaporating the solvent. Nonetheless, reduction with ascorbic acid efficiently converted the alkyl-benzoquinones back to the corresponding hydroquinone derivatives, which in contrast were stable and could be group-wise separated from the remaining nonylphenol substrates by means of HPLC-UV. The nonylhydroquinone fraction was subjected to derivatization with a silylating reagent and analyzed by GC-MS with an optimized elution program, resulting in a chromatogram that showed an isomer fingerprint. Only selected peaks that were judged to represent pure compounds were further analyzed (Figure 2). 5983

dx.doi.org/10.1021/es300383m | Environ. Sci. Technol. 2012, 46, 5979−5987

Environmental Science & Technology

Article

Figure 5. Proposed pathways for formation of nonylhydroquinones from nonylphenols. α-Tertiary 4-alkyl-4-hydroxy-cyclohexa-2,5-dien-1-one intermediates undergo a phenol-dienone rearrangement (NIH shift), in which the alkyl moiety formally moves as an anion to the vicinal position (attack of a nucleophile on the 4-position of a Michael system).17 It is conceivable that a small proportion of α-quaternary intermediates also undergo this rearrangement27 (A). Detachment of the α-quaternary alkyl moiety as a cation and electrophilic attack of the cation on the vicinal carbon atom of the rearomatized ring has been proposed as an alternative pathway18 (B). Certain nonylhydroquinone metabolites formed during ipso-degradation of technical nonylphenol may be derived from ortho-substituted nonylphenol components (C).



DISCUSSION In the past few years, all the major 4-nonylphenol isomers in technical nonylphenol have been structurally elucidated12,29,30 (see ref 11 for an overview). But despite the continuing importance of technical nonylphenol as a major bulk chemical in many parts of the world, its exact composition is still far from being completely known. With the emergence of twodimensional gas chromatography/time-of-flight mass spectrometry (GC×GC-ToFMS), scientific interest in elucidating the structures of the minor isomers has intensified,4,31 because certain of these unidentified components may substantially contribute to the overall endocrine activity of the technical mixture.11 We show here that potentially toxic metabolites were formed from minor components during the microbial degradation of technical nonylphenol and that such metabolites, therefore, contribute to the toxicity potential of nonylphenol mixtures. Our experiments clearly indicated that many of the reactive alkylquinone metabolites accumulating in the degradation experiments were derived from unidentified minor isomers that contain tertiary α-carbons. The presence of αtertiary nonylphenols in technical mixtures is remarkable since all the well-characterized components of technical mixtures have quaternary α-carbons.11,12,29,30 However, our biodegradation experiments corroborate recent findings of an analysis by two-dimensional gas chromatography−time-of-flight mass spectrometry reporting that technical nonylphenol contains a number of α-H, α-methyl substituted nonylphenol isomers, one of them eluting shortly after 4-NP194 (4-(1,3-dimethyl-1propylbutyl)phenol).4 Analyzing a technical mixture by means of gas chromatography-tandem mass spectrometry combined with cluster analysis, Moeder et al. tentatively identified one αH, α-methyl and 13 otherwise substituted α-tertiary 4nonylphenols among the isomers eluting before the bulk of 4-nonylphenols ((5%-phenyl)-methylpolysiloxane as separation phase).32 Here, we could clearly identify seven α-H, α-methyl substituted nonylphenols and tentatively detect five such structures among the early eluting isomers of a technical mixture (SI Figure S1).

intensity more than half that of the m/z 380 peak. In the case of peak 5, however, width and retention time of the signal on the m/z 307 trace did not match well with those of signals on other mass traces, indicating that the spectrum obtained at this retention time represented an overlap of spectra produced by different, coeluting compounds (SI Table S1). Mass spectra of C isomers had a certain resemblance with that of the α-methyl, α-ethyl substituted reference compound NHQ112 (2-(1,4dimethyl-1-ethylpentyl)hydroquinone). There were nevertheless important differences between these spectra. The spectrum of NHQ112 contained the most distinctive peak at m/z 309 instead of m/z 307, did not show a peak at m/z 378, and displayed a small but significant m/z 351 peak ([M·+ − ·C2H5]) (Figure 3). Moreover, the m/z 267 peak in the NHQ112 spectrum was more intensive than the m/z 281 peak. In view of these substantial differences, we conclude that Group C isomers did not have an α-methyl, α-ethyl substitution. Group E (GC peaks 1 and 2) and F (GC peak 4) spectra showed intense peaks at m/z 295 (65 − 68%, 380 series) and 293 (100%), respectively. The m/z 293 peak in the group F spectrum belonged to the m/z 378 sequence; accordingly this spectrum contained a prominent m/z 378 peak, which was about twice as high as the m/z 380 peak. However, in group E spectra the m/z 380 peak represented the base peak, and the intensity of the m/z 378 peak was smaller than 4%. In contrast to the spectra of Groups A, B, C, and D with base peaks at m/z 281, spectra of Group E and F displayed much smaller m/z 281 peaks (12−20%, Figure 4) (In the group F spectrum, the m/z 281 peak was most likely produced by a coeluting compound, Figure 3, SI Table S1). Whereas Group E spectra presumably correspond to nonylhydroquinones containing an α-H, α-ethyl substitution (however, α-dimethyl substitution cannot be definitely excluded), spectra of Groups C, D, and F with prominent m/z 378 sequence peaks did not allow assignments to specific structures. 5984

dx.doi.org/10.1021/es300383m | Environ. Sci. Technol. 2012, 46, 5979−5987

Environmental Science & Technology

Article

this intrinsic activity for their defense against potential enemies. Well-known examples are methyl- and ethyl-substituted pbenzoquinones in the secretions of many arthropods that serve as an active defense principle, which is strongly repellent to many predators.33,34 Biological effects of quinones can be explained by two inherent properties: electrophilicity and oxidative strength.23 Acting as electrophiles, p-benzoquinones and naphtoquinones tend to deplete cellular glutathione reserves and react with amine and thiol groups of macromolecules (in humans they behave as haptens and induce various forms of contact dermatitis35,36). The formation of nonyl-p-benzoquinones may explain why during ipso-degradation of technical nonylphenol undissolved, reddish-brown, putatively polymeric material accumulates,13 whereas with 4-t-octylphenol (a single compound with a quaternary α-carbon) as substrate, culture media remained less sullied (unpublished results). Accumulation of toxic nonyl-p-benzoquinones might be one of the factors contributing to the incomplete degradation of technical nonylphenol observed during incubations with growing cells of strains Bayram and TTNP3.11,28,37 As oxidants, quinones undergo redox cycling and may be reduced by one-electron reductases to yield the corresponding semiquinones,38 which are able to donate the exceeding electron to molecular oxygen. The resulting superoxide anion may undergo further reactions to generate hydroxy radicals, an even more reactive oxygen species able to abstract hydrogen atoms from nonactivated sites in macromolecules.23,24 Considering the lipophilic and oxidative properties of nonyl-pbenzoquinones it is tempting to suggest that these compounds might also disturb the electron transport by interfering with ubiquinones in cell membranes (SI Figure S4). ipso-Degradation of technical nonylphenol leaves behind undegraded material enriched with minor isomers.4,11 We show here that during ipso-degradation of the technical mixture small but significant amounts of nonylphenols are converted to the corresponding nonylhydroquinones and nonyl-p-benzoquinonesa complex mixture of potentially toxic metabolites that originate for the most part from α-tertiary compounds. Most recently, Deng et al.39 showed that quinone methide intermediates were formed during incubation of nonylphenol (Pestanal, Sigma-Aldrich) with human liver microsomes, but unfortunately, did not specify that only isomers bearing αhydrogens, representing minor components of technical mixtures, can actually be transformed to these reactive species. We predict that in nonylphenol polluted natural systems, in which microbial ipso-degradation is prominent, 2-alkylquinone metabolites will be produced. It is conceivable that these reactive compounds will contribute to the overall toxicity of the remaining material, but because of their chemical reactivity, it will be difficult to detect these metabolites in such systems. In the laboratory, microbial ipso-degradation provides an elegant means to enrich specific minor components of technical mixtures or to filter them out by transforming them into compounds with retained alkyl structure. In the future, ipsodegradation “coupled” with GC×GC-ToFMS or GC-ion trap MS might enable us to better identify and characterize these minor isomers also directly in the environment.

A major challenge in our study was to interpret mass spectra of derivatized metabolites containing both m/z 380 and m/z 378 sequence peaks. One should note that pairs of signals with an interval of m/z 2 are also present in electron impact mass spectra of underivatized nonylphenols, although this particular feature has not attracted the attention of researchers so far. The spectrum of the α-methyl, α-ethyl substituted isomer 4-NP112 (4-(1-ethyl-1,4-dimethylpentyl)phenol) for instance contain small, but well-defined signals at m/z 147, 119, and 105 (“m/z 218 sequence”) beside intense signals at m/z 149, 121, and 107 (“m/z 220 sequence”), respectively (Figure 3; SI Figure S2 shows a tentative proposal to explain the formation of ions with m/z 147, 119, and 105). However, we deemed mechanistic proposals aimed to explain the loss of H2 from m/z 380 ions implausible and rejected them (not shown). To account for the “m/z 378 sequence” peaks in the spectra of the Groups C, D, and F, we propose that these spectra result from overlapping of spectra belonging to different compounds (Figure 3). This hypothesis was corroborated by carefully comparing the mass traces relevant to a given spectrum (see SI Table S1 and Figure S5). The scenario we propose is that certain nonylhydroquinones coelute with non-2-en-2-yl- and non-3-en-3-yl-hydroquinones, respectively, thus accounting for signals of the m/z 380 and 378 series. Speculations about the mechanism of formation of these α−β-dehydrogenated metabolites comprise action of some oxidizing agent such as nonyl-p-benzoquinones themselves or, more likely, reactive oxygen species formed, directly or indirectly, by reaction of nonylhydroquinones with air oxygen (see below). Also note that all structures assigned to metabolites in Figure 3 either represent α-tertiary nonylhydroquinones or (in the case of the α−β dehydrogenated substances) are possibly derived from such compounds. This accords well with the notion that α-quaternary and α-tertiary nonylphenols undergo different pathways after the initial ipso-hydroxylation (Figure 1A and B).15 However, we cannot exclude the possibility that minute fractions of α-quaternary cyclohexadienone intermediates underwent an NIH-shift. Indeed, small amounts of a particular α-quaternary nonylhydroquinone, NHQ111 (2-(1ethyl-1,3-dimethylpentyl)hydroquinone), have been isolated,27 though formation of this α-quaternary metabolite might alternatively be explained by a mechanism different from the NIH-shiftrelease of the alkyl moiety from the cyclohexadienone intermediate as a cation and subsequent recombination of the cation with the rearomatized carbon ring at a vicinal position (Figure 5).18 It is important to note that this hypothetic mechanism cannot apply to α-tertiary compounds because, as judged from experimental evidence, αtertiary 4-alkyl-4-hydroxy-cyclohexadienones do not release the side chain (Figure 1).15,17 Finally, nonylhydroquinones might in principle also be formed by para-hydroxylation of orthononyphenols (Figure 5C), minor components of technical mixtures that elute before the group of para-substituted isomers on non polar capillary columns.4,30 The low retention times of the newly detected α-H, α-methyl nonylphenols resemble indeed the presumed retention times of ortho-nonylphenols (SI Figure S1). Moreover, Bayram seems to be able to partially degrade early eluting components of technical mixtures.11 We propose that in the presence of air oxygen, the 2nonylhydroquinones formed during ipso-degradation of technical nonylphenol oxidized to the corresponding p-benzoquinone derivatives, inherently reactive compounds, which are potentially toxic to cells.23 Numerous living organisms exploit



ASSOCIATED CONTENT

S Supporting Information *

Table showing mass spectra associated with the relevant peaks of the GC−MS chromatogram of trimethylsilyl derivatized 5985

dx.doi.org/10.1021/es300383m | Environ. Sci. Technol. 2012, 46, 5979−5987

Environmental Science & Technology

Article

nonylhydroquinone metabolites and a figure showing the corresponding m/z traces, total ion chromatogram and relevant m/z traces of technical nonylphenol, fragmentation pathways explaining mass spectra, additional information on the chemodiversity of alkyl-p-benzoquinones. Mass spectra of specific α-tertiary nonylhydroquinones, HPLC-UV chromatograms illustrating the reduction of nonyl-p-quinones. This material is free of charge via the Internet at http://pubs.acs.org.



(11) Gabriel, F. L. P.; Routledge, E. J.; Heidlberger, A.; Rentsch, D.; Guenther, K.; Giger, W.; Sumpter, J. P.; Kohler, H.-P. E. Isomerspecific degradation and endocrine disrupting activity of nonylphenols. Environ. Sci. Technol. 2008, 42, 6399−6408. (12) Katase, T.; Okuda, K.; Kim, Y.-S.; Eun, H.; Takada, H.; Uchiyama, T.; Saito, H.; Makino, M.; Fujimoto, Y. Estrogen equivalent concentration of 13 branched para-nonylphenols in three technical mixtures by isomer-specific determination using their synthetic standards in SIM mode with GC−MS and two new diastereomeric isomers. Chemosphere 2008, 70, 1961−1972. (13) Tanghe, T.; Dhooge, W.; Verstraete, W. Isolation of a bacterial strain able to degrade branched nonylphenol. Appl. Environ. Microbiol. 1999, 65, 746−751. (14) Fujii, K.; Urano, N.; Ushio, H.; Satomi, M.; Kimura, S. Sphingomonas cloacae sp. nov., a nonylphenol-degrading bacterium isolated from wastewater of a sewage-treatment plant in Tokyo. Int. J. Syst. Evol. Microbiol. 2001, 51, 603−610. (15) Gabriel, F. L. P.; Giger, W.; Guenther, K.; Kohler, H.-P. E. Differential degradation of nonylphenol isomers by Sphingomonas xenophaga Bayram. Appl. Environ. Microbiol. 2005, 71, 1123−1129. (16) Shan, J.; Jiang, B. Q.; Yu, B.; Li, C. L.; Sun, Y. Y.; Guo, H. Y.; Wu, J. C.; Klumpp, E.; Schaffer, A.; Ji, R. Isomer-specific degradation of branched and linear 4-nonylphenol isomers in an oxic soil. Environ. Sci. Technol. 2011, 45, 8283−8289. (17) Gabriel, F. L. P.; Heidlberger, A.; Rentsch, D.; Giger, W.; Guenther, K.; Kohler, H.-P. E. A novel metabolic pathway for degradation of 4-nonylphenol environmental contaminants by Sphingomonas xenophaga Bayram. ipso-Hydroxylation and intramolecular rearrangement. J. Biol. Chem. 2005, 280, 15526−15533. (18) Corvini, P. F. X.; Hollender, J.; Ji, R.; Schumacher, S.; Prell, J.; Hommes, G.; Priefer, U.; Vinken, R.; Schäffer, A. The degradation of α-quaternary nonylphenol isomers by Sphingomonas sp. strain TTNP3 involves a type II ipso-substitution mechanism. Appl. Microbiol. Biotechnol. 2006, 70, 114−122. (19) Gabriel, F. L. P.; Cyris, M.; Jonkers, N.; Giger, W.; Guenter, K.; Kohler, H.-P. E. Elucidation of the ipso-substitution mechanism for side chain cleavage of α-quaternary 4-nonylphenols and 4-tbutoxyphenol in Sphingobium xenophagum Bayram. Appl. Environ. Microbiol. 2007, 73, 3320−3326. (20) Kohler, H.-P. E.; Gabriel, F. L. P.; Giger, W. ipso-Substitution − A novel pathway for microbial metabolism of endocrine-disrupting 4nonylphenols, 4-alkoxyphenols, and bisphenol A. Chimia 2008, 62, 358−363. (21) Giger, W.; Gabriel, F. L. P.; Jonkers, N.; Wettstein, F. E.; Kohler, H. P. E. Environmental fate of phenolic endocrine disruptors: Field and laboratory studies. Phil. Trans. R. Soc. 2009, 367, 3941−3963. (22) Montgomery-Brown, J.; Li, Y.; Dang, W.-H.; Mong, G. M.; Campbell, J. A.; Reinhard, M. NP1EC Degradation pathways under oxic and microoxic conditions. Environ. Sci. Technol. 2008, 42, 6409− 6414. (23) Monks, T. J.; Hanzlik, R. P.; Cohen, G. M.; Ross, D.; Graham, D. G. Quinone Chemistry and Toxicity. Toxicol. Appl. Pharmacol. 1992, 112, 2−6. (24) Kappus, H. Overview of enzyme systems involved in bioreduction of drugs and in redox cycling. Biochem. Pharmacol. 1986, 35, 1−6. (25) Guroff, G.; Daly, J. W.; Jerina, D. M.; Renson, J.; Witkop, B.; Udenfriend, S. Hydroxylation-induced migration: The NIH shift. Science 1967, 157, 1524−1530. (26) Guenther, K.; Kleist, E.; Thiele, B. Estrogen-active nonylphenols from an isomer-specific viewpoint: A systematic numbering system and future trends. Anal. Bioanal. Chem. 2006, 384, 542−546. (27) Corvini, P. F. X.; Meesters, R. J. W.; Schäffer, A.; Schröder, H. F.; Vinken, R.; Hollender, J. Degradation of a nonylphenol single isomer by Sphingomonas sp. strain TTNP3 leads to a hydroxylationinduced migration product. Appl. Environ. Microbiol. 2004, 70, 6897− 6900. (28) Corvini, P. F. X.; Meesters, R.; Mundt, M.; Schäffer, A.; Schmidt, B.; Schröder, H.-F.; Verstraete, W.; Vinken, R.; Hollender, J.

AUTHOR INFORMATION

Corresponding Author

*Phone: +41 44 823 55 21; e-mail: [email protected]. Present Addresses §

Institute of Clinical Chemistry and Laboratory Medicine, University of Rostock, Rostock, Germany ∥ Apartado postal 1812−7050, Cartago, Costa Rica. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS This work was financially supported by the Swiss National Science Foundation (Projects No. SNF-200021-120574 and SNF-200021-130319).



REFERENCES

(1) Anonymous, Directive 2003/53/EC of the European Parliament and of the Council of 18 June 2003 amending for the 26th time Council Directive 76/769/EEC relating to restrictions on the marketing and use of certain dangerous substances and preparations (nonylphenol, nonylphenol ethoxylate and cement). Off. J. Eur. Union 2003, L 178, 24-27. (2) Soto, A. M.; Justicia, H.; Wray, J. W.; Sonnenschein, C. p-Nonylphenol: An estrogenic xenobiotic released from ″modified″ polystyrene. Environ. Health Perspect. 1991, 92, 167−173. (3) Servos, M. R. Review of the aquatic toxicity, estrogenic responses and bioaccumulation of alkylphenols and alkylphenol polyethoxylates. Water Qual. Res. J. Can. 1999, 34, 123−177. (4) Eganhouse, R.; Pontolillo, J.; Gaines, R.; Frysinger, G.; Gabriel, F. L. P.; Kohler, H.-P. E.; Giger, W.; Barber, L. Isomer-specific determination of 4-nonylphenols using comprehensive two-dimensional gas chromatography/time-of-flight mass spectrometry. Environ. Sci. Technol. 2009, 43, 9306−9313. (5) Brooke, L. T.; Thursby, G., Aquatic Life Ambient Water Quality CriteriaNonyphenol, EPA-822-R-05-005; Office of Water: Washington DC, 2005. (6) Hansen, B. G.; Munn, S. J.; De Bruijn, J.; Pakalin, S.; Luotamo, M.; Berthault, F.; Vegro, S.; Heidorn, C. J. A.; Pellegrini, G.; Vormann, K.; Allanou, R.; Scheer, S. European Commission. EUR 20387 EN European Union Risk Assessment Report. 4-Nonylphenol (Branched) and Nonylphenol; Office for Official Publications of the European Communities: Luxembourg, 2002; Vol. 10. (7) Routledge, E. J.; Sumpter, J. P. Structural features of alkylphenolic chemicals associated with estrogenic activity. J. Biol. Chem. 1997, 272, 3280−3288. (8) Kim, Y.-S.; Katase, T.; Horii, Y.; Yamashita, N.; Makino, M.; Uchiyama, T.; Fujimoto, Y.; Inoue, T. Estrogen equivalent concentration of individual isomer-specific 4-nonylphenol in Ariake sea water. Mar. Pollut. Bull. 2005, 51, 850−856. (9) Preuss, T. G.; Gehrhardt, J.; Schirmer, K.; Coors, A.; Rubach, M.; Russ, A.; Jones, P. D.; Giesy, J. P.; Ratte, H. T. Nonylphenol isomers differ in estrogenic activity. Environ. Sci. Technol. 2006, 40, 5147−5153. (10) Shioji, H.; Tsunoi, S.; Kobayashi, Y.; Shigemori, T.; Ike, M.; Fujita, M.; Miyaji, Y.; Tanaka, M. Estrogenic activity of branched 4nonylphenol isomers examined by yeast two-hybrid assay. J. Health Sci. 2006, 52, 132−141. 5986

dx.doi.org/10.1021/es300383m | Environ. Sci. Technol. 2012, 46, 5979−5987

Environmental Science & Technology

Article

Contribution to the detection and identification of oxidation metabolites of nonylphenol in Sphingomonas sp. strain TTNP3. Biodegradation 2007, 18, 233−245. (29) Thiele, B.; Heinke, V.; Kleist, E.; Guenther, K. Contribution to the structural elucidation of 10 isomers of technical p-nonylphenol. Environ. Sci. Technol. 2004, 38, 3405−3411. (30) Wheeler, T. F.; Heim, J. R.; LaTorre, M. R.; Janes, A. B. Mass spectral characterization of p-nonylphenol isomers using highresolution capillary GC-MS. J. Chromatogr. Sci. 1997, 35, 19−30. (31) Ieda, T.; Horii, Y.; Petrick, G.; Yamashita, N.; Ochiai, N.; Kannan, K. Analysis of nonylphenol isomers in a technical mixture and in water by comprehensive two-dimensional gas chromatography-mass spectrometry. Environ. Sci. Technol. 2005, 39, 7202−7207. (32) Moeder, M.; Martin, C.; Harynuk, J.; Górecki, T.; Vinken, R.; Corvini, P. F. X. Identification of isomeric 4-nonylphenol structures by gas chromatography-tandem mass spectrometry combined with cluster analysis. J. Chromatogr., A 2005, 1102, 245−255. (33) Eisner, T.; Meinwald, J. Defensive secretions of arthropods. Science 1966, 153, 1341−1350. (34) Aneshansley, D. J.; Eisner, T. Biochemistry at 100°C: Explosive secretory discharge of Bombardier beetles (Brachinus). Science 1969, 165, 61−63. (35) Krebs, M.; Christensen, L. P. 2-methoxy-6-pentyl-1,4-dihydroxybenzene (miconidin) from Primula obconica: A possible allergen? Contact Dermatitis 1995, 33, 90−93. (36) Reynolds, G. W.; Epstein, W. L. Unusual contact allergens from plants in the family Hydrophyllaceae. Contact Dermatitis 1986, 14, 39− 44. (37) Corvini, P. F. X.; Vinken, R.; Hommes, G.; Schmidt, B.; Dohmann, M. Degradation of the radioactive and non-labelled branched 4(3′,5′-dimethyl-3′-heptyl)-phenol nonylphenol isomer by Sphingomonas TTNP3. Biodegradation 2004, 15, 9−18. (38) Iyanagi, T.; Yamazaki, I. One-electron-transfer reactions in biochemical systems. V. Difference in the mechanism of quinone reduction by the NADH dehydrogenase and the NAD(P)H dehydrogenase (DT-diaphorase). Biochim. Biophys. Acta 1970, 216, 282−294. (39) Deng, P.; Zhong, D.; Nan, F.; Liu, S.; Li, D.; Yuan, T.; Chen, X.; Zheng, J. Evidence for the bioactivation of 4-nonylphenol to quinone methide and ortho-benzoquinone metabolites in human liver microsomes. Chem. Res. Toxicol. 2010, 23, 1617−1628.

5987

dx.doi.org/10.1021/es300383m | Environ. Sci. Technol. 2012, 46, 5979−5987