Environ. Sci. Technol. 2007, 41, 3593-3600
Laccase-Mediated Michael Addition of 15N-Sulfapyridine to a Model Humic Constituent H E I D I M . B I A L K , †,‡ C U R T I S H E D M A N , § ALEX CASTILLO,| AND J O E L A . P E D E R S E N * ,‡,⊥,# Molecular and Environmental Toxicology Center, Departments of Biochemistry and Soil Science, and Environmental Chemistry and Technology Program, University of Wisconsin, Madison, Wisconsin 53706, and Wisconsin State Laboratory of Hygiene, Madison, Wisconsin 53718
Chemical incorporation of sulfonamide antimicrobials into natural organic matter may represent an important process influencing the fate of these synthetic, primarily agents in soil and sediment environments. We previously demonstrated that a fungal peroxidase mediates covalent coupling of sulfonamide antimicrobials to model humic constituents; reactions with the 2,6-dimethoxyphenol syringic acid produced Schiff bases (Bialk et al. Environ. Sci. Technol. 2005, 39, 4436-4473). Here, we show that fungal laccase-mediated reaction of sulfapyridine with the orthodihydroxyphenol protocatechuic acid yields a Michael adduct. We synthesized 15N-enriched sulfapyridine to facilitate determination of the covalent linkage(s) formed between sulfapyridine and protocatechuic acid by NMR spectroscopy. 1H-15N heteronuclear multiple bond correlation experiments and tandem mass spectrometry demonstrated that the sulfapyridine anilinic nitrogen engaged in a Michael addition reaction to oxidized protocatechuic acid to form an anilinoquinone. Michael adducts are more stable than the previously reported imine linkages between sulfonamides and 2,6-dimethoxyphenols. Michael addition to quinone-like structures in soil organic matter is expected to diminish the mobility and biological activity of sulfonamide antimicrobials.
Introduction Sulfonamide antimicrobial agents comprise an important class of human and veterinary pharmaceuticals used to combat bacterial and protozoal infections (1, 2). These sulfanilamide derivatives differ from one another in substitution at the amide nitrogen (for the structure of a representative sulfonamide, see Figure 1). All sulfonamide * Corresponding author phone: (608)263-4971; fax: (608)265-2595; e-mail:
[email protected]. Corresponding author address: Department of Soil Science, University of Wisconsin, 1525 Observatory Drive, Madison, WI 53706-1299. † Present address: Department of Chemistry and Biochemistry, Old Dominion University, Norfolk, VA 23529. ‡ Molecular and Environmental Toxicology Center, University of Wisconsin. § Wisconsin State Laboratory of Hygiene. | Department of Biochemistry, University of Wisconsin. ⊥ Department of Soil Science, University of Wisconsin. # Environmental Chemistry and Technology Program, University of Wisconsin. 10.1021/es0617338 CCC: $37.00 Published on Web 04/12/2007
2007 American Chemical Society
antimicrobials possess the para-NH2 group, a moiety essential for their bacteriostatic activity (1). Sulfonamides competitively inhibit dihydropteroate synthase, thus interfering with folate biosynthesis in microorganisms that must synthesize their own folate (1). Bacteria and other organisms (e.g., mammals) able to utilize preformed folate are insensitive to the biochemical action of sulfonamides (1). Sulfonamide antimicrobials are introduced into soil and water environments via human, animal, and pharmaceutical industry waste disposal (3-6). Dosed individuals excrete unaltered sulfonamides and acetylated metabolites (1). The latter can be reactivated by bacterial deacetylation (7). Sulfonamides are not completely removed by conventional wastewater treatment processes (8, 9) and are released to water and soil environments through wastewater discharges, effluent irrigation, and land application of biosolids (9-12). Sulfonamides used in livestock production enter soil and water environments through runoff and infiltration from manure-amended agricultural fields or confined animal feeding operations (13-15). Once released into the environment, sulfonamides may exert selective pressure for antimicrobial-resistant microorganisms. Sulfonamide concentrations in surface waters (0.02-15 µg‚L-1 (3, 16, 17)), groundwaters (0.046-0.47 µg‚L-1 (15-17)), and soils (0.011 µg‚kg-1 (18)) are generally below inhibitory concentrations (19). Selection of antimicrobial resistant microorganisms and promotion of antibiotic resistance gene transfer by subinhibitory concentrations are poorly understood (19). Pei et al. (20) reported that, compared to pristine sites, river sediments impacted by urban and agricultural activity contained higher sulfonamide concentrations, increased sulfonamide resistance gene diversity, and higher sulfonamide resistance gene copy numbers; however, correlations between sulfonamide concentrations and resistance genes were not significant, and no causal relationship was established. Sulfonamide antimicrobials can induce shifts in bacterial community composition (21), and environmentally relevant concentrations may be sufficient to inhibit microbial activity (22). Characterization of risks posed to human and animal health by the presence of these antimicrobial agents in the environment requires a better understanding of sulfonamide behavior in soil and water ecosystems. A number of studies have examined the equilibrium sorption of sulfonamide antimicrobials to whole soils and specific soil components using short-term batch experiments (23-25). Equilibrium distribution coefficients derived from such studies are useful in predicting sulfonamide mobility and bioavailability to the extent that the interaction of these compounds with environmental particles is governed by noncovalent interactions. Organic compounds containing specific functional groups (e.g., phenolic moieties, aromatic amines, primary and secondary amines, amides) can covalently bind to soil and sediment natural organic matter (NOM) (26-29), limiting their mobility and toxicity (28). Chemical incorporation of substituted anilines into NOM can occur via radical-radical coupling and nucleophilic addition reactions (26, 27) and can be mediated by naturally occurring phenoloxidases (e.g., peroxidases, laccases, tyrosinases) and metal oxides (e.g., birnessite) (26-28). The complexity of humic substances renders difficult the elucidation of xenobiotic-NOM coupling mechanisms and the determination of product structure. Substituted phenolic compounds representing structures within humic substances can be profitably employed to provide insight into these reactions (29-31). We previously demonstrated that among the substituted phenols examined, 2,6-dimethoxy phenols VOL. 41, NO. 10, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
3593
(e.g., syringic acid, syringaldehyde) and ortho-dihydroxy phenols (e.g., protocatechuic acid, catechol) were most reactive toward sulfamethazine in incubations containing phenoloxidases or acid birnessite (30). We provided direct spectroscopic evidence that Arthromyces ramosus peroxidase (ARP) mediated the covalent coupling of sulfamethazine with both syringic and protocatechuic acids (30). The ARPmediated reaction of sulfamethazine with the 2,6-dihydroxyphenol syringic acid yielded a Schiff base having an imine linkage via the anilinic nitrogen (30). The nature of the covalent linkage(s) between sulfonamide antimicrobials and ortho-dihydroxyphenols remained to be established. The environmental stability of bound sulfonamide residues is expected to depend on the type of covalent linkage formed with NOM. Substituted anilines may form Michael adducts to quinone-like structures in NOM; such linkages resist hydrolysis and are believed to be not readily reversible under environmental conditions (32). Schiff bases, however, are susceptible to hydrolysis (32). The primary objective of this study was to establish the type(s) of covalent linkage(s) formed between the sulfonamide antimicrobial sulfapyridine (SPD) and the orthodihydroxyphenol protocatechuic acid in reactions mediated by fungal laccase. Based on our previous study (30) and earlier work with substituted anilines (28), we expected covalent cross-coupling to proceed through the SPD anilinic nitrogen. We employed two-dimensional heteronuclear nuclear magnetic resonance (NMR) spectroscopy to determine the nature of cross-coupled products. Because NMR-active 15N nuclei have low natural abundance (0.38%) (33, 34), we synthesized 15N-enriched SPD for use in our experiments. This allowed us to distinguish among possible covalent linkages (e.g., imine, anilinohydroquinone, anilinoquinone). Mass spectrometry was used to obtain corroborating evidence for the structures determined by NMR.
Materials and Methods Chemicals. Protocatechuic acid and DMSO-d6 (∼99% purity) were obtained from Fisher Scientific (Hampton, NH). 15NAcetanilide was purchased from Cambridge Isotope Laboratories (Andover, MA). Dry solvents, 2-aminopyridine, chlorosulfonic acid, and NMR solvents were obtained from Sigma Aldrich (St. Louis, MO) at 99% purity. Starting materials for 15N-SPD synthesis were stored in a vacuum desiccator to minimize water contamination. Chlorosulfonic acid was stored in a fume hood. Glassware contaminated with this acid was thoroughly rinsed with acetone prior to washing with water. Phenoloxidase. Laccases (p-diphenol: oxygen oxidoreductase E.C. 1.10.3.2) mediate one-electron oxidation of phenolic compounds and some aromatic amines to free radicals (35). We assessed the purity of Trametes versicolor laccase (23.7 U‚mg-1) purchased from Novo Nordisk (Copenhagen, Denmark) by sodium dodecyl sulfate-polyacrylamide gel electrophoresis under reducing conditions (4-15% pre-cast gels, Bio-Rad, Hercules, CA) with Coomassie Brilliant Blue G staining. Contaminant proteins were not detected by this method; the enzyme was therefore used as received. Laccase activity was defined as the amount of enzyme needed to achieve a 1.0 min-1 change in absorbance at 468 nm in 3.4 mL of 1 mM 2,6-dimethoxyphenol in citrate phosphate buffer (pH ) 3.8). 15N-Sulfapyridine Synthesis. Because 15N-SPD was not commercially available, we adapted a published SPD synthesis method (36) to produce 15N-SPD from 15N-acetanilide (see Scheme S-1 in the Supporting Information). All reactions were conducted under a fume hood. Preparation of Acetylsulfanilyl Chloride. 15N-Acetanilide (3.0 g) was melted in a round-bottom flask with a gas trap, removed from heat, and allowed to solidify during 10-min 3594
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 10, 2007
cooling in an ice bath. Chlorosulfonic acid (10 mL) was added to the 15N-acetanilide, and the flask was cooled in the ice bath. After cooling, the flask was removed from the ice bath, shaken until the solid was completely dissolved, heated for 10 min in a water bath, and cooled for 10 min in the ice bath. Immediately following addition to a 150-mL beaker half-full with crushed ice and distilled deionized water (ddH2O; Barnstead NANOpure Ultrapure Water System, Dubuque, IA), the acetylsulfanilyl chloride (ASC) product was collected by vacuum filtration. The solid ASC was repeatedly washed with 50-mL portions of cold ddH2O until filtrate pH was ∼7. The solid was then freeze-dried overnight. Reaction of ASC with 2-Aminopyridine and Pyridine. Dry acetone (20 mL) was added to the ASC via a cannula, and the solution was gravity filtered through a 90-mm diameter Whatman cellulose filter. In a separate flask, dry acetone (5 mL) was added to 2.5 g of 2-aminopyridine. To this flask, 4 mL of pyridine (previously dried over KOH pellets) was added via a syringe. The ASC and pyridine solutions were then combined. After HCl evolution had ceased, the flask was sealed with a rubber septum and agitated to induce 4-acetamido-N-(2-pyridyl)-benzenesulfonamide precipitation. (If necessary, precipitation may be induced by scratching the glassware and cooling in an ice bath.) The flask was allowed to stir overnight. Preparation of 15N-SPD. The precipitate was collected via vacuum filtration and washed with chilled 85% acetone/ 15% ddH2O. After dissolving the solid in 15 mL of 3 M NaOH, the solution was refluxed for 30 min. The solution was allowed to cool to room temperature, and pH was adjusted to ∼6.5 by dropwise addition of 6 M HCl. The 15N-SPD product (collected by vacuum filtration) was recrystallized twice with 85% acetone, collected by vacuum filtration, and washed with ice cold ddH2O. Confirmation of Product Identity. The purity of the final reaction product was determined by chromatographic and spectroscopic methods. Briefly, starting materials (ASC, aminopyridine, pyridine) and final reaction product (15NSPD) were blotted onto a thin layer chromatography plate with a capillary tube, and the plate was placed in a beaker containing 50:50 methanol/chloroform. The retardation factor (Rf, ratio of the distance traveled by a solute to that of the solvent front) for each solute was used to determine the presence of contaminants in the synthesized product. One- and two-dimensional NMR spectroscopy was employed to verify product identify (see below). Laccase-Mediated Reaction of 15N-SPD with Protocatechuic Acid. Protocatechuic acid (0.3 mM) was reacted for 30 min under constant stirring with 0.3 mM 15N-SPD in the presence of ∼1.5 U‚mL-1 laccase in 75 mL of 0.2 M acetate (pH 5.6). Colored products were isolated by solid-phase extraction as previously described (30), freeze-dried, and analyzed by spectroscopy and mass spectrometry. Nuclear Magnetic Resonance Spectroscopy. NMR spectra were obtained on a 500 MHz Varian spectrometer equipped with an inverse Varian HCX probe optimized for a 90° pulse width prior to data collection. Samples for NMR analyses were dissolved in DMSO-d6. The chemical structure of synthesized 15N-SPD (∼0.5 mg‚mL-1) was corroborated by one-dimensional (1H), homonuclear (1H-1H) COrrelation SpectroscopY (COSY), and 1H-15N Heteronuclear Multiple Bond Correlation (HMBC) NMR experiments. HMBC experiments were also performed on cross-coupled products. Onedimensional NMR spectra were collected with 32 scans and a 3-s relaxation delay. HMBC spectra were obtained using gradient pulses for selection, an 8-Hz long-range coupling constant, and a 3-s recycle delay. Scans (128) were collected for 64 increments in the F1 dimension; 2000 data points were collected in the F2 dimension. Both dimensions were processed using
FIGURE 1. 1H-15N HMBC spectrum of synthesized 15N-SPD. The one-bond correlation appears as two cross-peaks due to the two spin states of 15N. The three- (δH∼6.5 ppm) and four- (δH∼7.5 ppm) bond correlations were confirmed by COSY (Figure 2). The spectrum was referenced internally to DMSO-d6 and externally to NH3 (assigned δN ) 0). unshifted sine-squared functions and forward linear prediction using 8 coefficients in the F1 dimension. 1H-15N HMBC spectra were referenced to DMSO-d6, and external NH3 was assigned nitrogen chemical shift, δN ) 0. Two-dimensional Advanced Chemistry Development (ACD) Laboratories Spec Manager and 2D Predictor software were used to predict 1H and 15N chemical shifts (δH and δN) for one reaction product. For COSY experiments, eight scans were collected for each of the 256 increments in the F1 dimension; 2000 data points were collected in the F2 dimension. The F1 dimension was processed using a sine-squared function with 90° phase shifts and a zero filling factor of 2. Mass Spectrometry. Isolated reaction products were separated by high-performance liquid chromatography (HPLC) on an Agilent 1100 system equipped with a photodiode array detector and interfaced to an Applied Biosystems/ MDS SCIEX API 4000 triple quadrupole mass spectrometer (MS) with an electrospray ionization (ESI) source. The Phenomenex Luna C18(2) analytical column (150-mm × 3-mm id, 3-µm pore size) was maintained at 30 °C. Reaction products (10 µL injection volume) were separated by gradient elution at 0.36 mL‚min-1 (solvent A ) 0.1% aqueous formic acid, solvent B ) methanol): 0-5 min, hold at 95%A; 5-15 min, linear gradient from 95%A to 60%A; 15-20 min, linear gradient from 60%A to 28%A; 20-23 min, linear gradient from 28%A to 0%A; 23-25 min, linear gradient from 0%A to 95%A; 25-35 min, hold at 95%A to re-equilibrate column prior to the next injection. UV absorbance (λ ) 254 and 265 nm) was measured to aid peak selection for Q1 scan spectra. Negative ionization mode TurboIonSpray Q1 mass spectral data (50-1000 m/z) were collected with a 1-s scan time. MS detection parameters included the following: curtain gas flow, 20 psi; nebulizer gas flow, 35 psi; heater gas flow, 30 psi.; ion spray voltage, -3000 V; source temperature, 250 °C; declustering potential, -20 V; and entrance potential, -10 V. Candidate reaction product molecular ions observed in the MS scan data were selected for reanalysis in product ion scan mode (other instrument parameters as described above) to examine daughter ions resulting from collision-induced dissociation. The mass of the identified product was confirmed by ESItime-of-flight (ESI-TOF) mass spectrometry (Bruker, Amherst,
MA) in full scan mode at a 0.5 full-width at half-maximum resolution. Cross-coupled product (0.5 mg‚mL-1 in methanol) was analyzed in negative ionization mode.
Results and Discussion Characterization of the Synthesized 15N-SPD. Synthesized 15N-SPD was characterized by 1H-15N-HMBC and COSY experiments. In the HMBC spectrum (Figure 1), one-, three-, and four-bond correlations arose from J-couplings between 1H nuclei attached to phenyl carbons and the 15N-labeled aniline nitrogen. Coupling assignments were confirmed by COSY (Figure 2). In COSY, protons belonging to the same spin system yield one-bond coupling patterns. The phenyl protons producing three- and four-bond correlations in the HMBC spectrum are also coupled to one another in the COSY spectrum. The Supporting Information further details interpretation of the HMBC and COSY spectra. Laccase-Mediated Coupling of 15N-Sulfapyridine to Protocatechuic Acid. Fungal laccase mediates sulfonamide antimicrobial transformation in the presence of orthodihydroxyphenols (30). Chemical incorporation of aromatic amines into NOM can proceed via radical-radical coupling or by nucleophilic addition to electron-poor sites in humic material (e.g., quinones) (26, 37). Radical-radical coupling does not appear to be an important mechanism in laccasemediated coupling of sulfonamide antimicrobials to substituted phenols (30). ARP-mediated coupling of sulfamethazine to the 2,6-dimethoxyphenol syringic acid appears to proceed via oxidative decarboxylation of syringic acid to generate a quinone that serves as a site for nucleophilic attack by the sulfonamide anilinic nitrogen (30). Both ARP and laccase oxidize ortho-diphenols (such as protocatechuic acid) to ortho-quinones (35, 38) creating sites favorable for Michael addition reactions (39) with aromatic amines (26). We therefore anticipated laccase-mediated formation of Michael adducts between 15N-SPD and protocatechuic acid. Michael adducts are considerably more resistant to hydrolysis than are Schiff bases (e.g., the adducts formed between sulfonamides and syringic acid (30)) (32). Interpretation of the 1H-15N HMBC Spectrum. The 1H-15N HMBC spectrum of products of laccase-mediated SPDVOL. 41, NO. 10, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
3595
FIGURE 2. COSY spectrum of synthesized 15N-SPD. The proton producing δH ∼ 6.5 ppm on the F1 scale is coupled to that producing δH ∼ 7.5 ppm on the F2 scale verifying the four-bond correlation from the protons attached to carbons meta to the aniline nitrogen (∼7.5 ppm) of 15N-SPD (Figure 1). The one-bond correlation appearing at δH ∼ 5.7 and 6.0 ppm on both scales is not coupled to other protons as expected due to the absence of a hydrogen on the carbon atom to which the aniline nitrogen is attached. Proton H did not appear in this spectrum but was assigned based on one-dimensional 1H experiments (Figure S-2). The sample was referenced to DMSO-d6. protocatechuic cross-coupling exhibited cross-peaks from both unreacted and reacted 15N-SPD (Figure 3). The anilinic nitrogen of unreacted 15N-SPD produced a cross-peak at δN ∼ 70 ppm. Three- and four-bond correlations with protons ortho (δH ∼ 6.5 ppm) and meta (δH ∼ 7.5 ppm) to the anilinic nitrogen were assigned based on COSY data (Figure 2). Weak detection of a four-bond correlation (observed as a doublet) for unreacted 15N-SPD was likely due to low signal-to-noise (33, 34). Laccase-mediated reaction of 15N-SPD with protocatechuic acid produced new cross-peaks at two different δN (compare Figures 1 and 3) indicating covalent bond formation. The new cross-peaks resulted from three-bond correlations; cross-peaks from single-bond correlations were not observed in this experiment as indicated by the absence of cross-peaks from protons attached to the anilinic nitrogen of unreacted 15N-SPD. The cross-peak at δN ∼ 142 ppm correlated to aromatic protons on the anilinic ring ortho to the labeled nitrogen (doublet at δH ∼ 7.2 ppm, split due to the meta protons) and a proton associated with a double bond or alkene (δH ∼ 5.3 ppm, due to a proton associated with protocatechuic acid linked to SPD through the anilinic nitrogen) (33, 34). The δN ∼ 118 ppm cross-peak appeared as a weak doublet and correlated to only an aromatic proton (δH ∼ 6.8 ppm) ortho to the SPD anilinic nitrogen. The HMBC experiment 3596
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 10, 2007
detected no additional protons within three bonds of the reacted aniline nitrogen. In addition to reacting with protocatechuic acid monomers, 15N-SPD likely also reacted with self-coupled protocatechuic acid oligomers. The most probable explanation of the absence of additional 1H connectivities associated with δN ∼ 118 ppm is 15N-SPD nucleophilic addition to self-coupled protocatechuic acid products lacking protons within three bonds of the linkage. Thorn et al. (26) reported that peroxidase-mediated crosscoupling of aniline to humic substances led to anilinoquinone formation based on a tentative peak assignment in the δN range of 100-122 ppm. Cross-peaks in Figure 3 associated with cross-coupled products exhibited δN (∼118 and 142 ppm) similar to those for anilinoquinones (26). Other functional groups with δN between 100 and 150 ppm include enamine, enaminone, quinolone, and indole moieties (26). Enamines and enaminones are formed by reaction of β-unsaturated alkane with primary amines. Primary amines also undergo coupling reactions to form heterocyclic rings such as quinolones and indoles (26). The δN for cross-coupled 15N-SPD are most consistent with the scenario in which the SPD anilinic nitrogen engages in nucleophilic attack on an oxidized form of protocatechuic acid while retaining a proton. Proposed Reaction Pathway. Scheme 1 presents a proposed reaction pathway for laccase-mediated Michael addition of
FIGURE 3. 1H-15N HMBC spectrum of unreacted 15N-SPD and reaction products resulting from the incubation of 15N-SPD with protocatechuic acid and laccase. The appearance of new cross-peaks corresponding to nitrogen reaction products have two different δN (∼118 and 142 ppm). The cross-peak at δN ∼ 118 ppm is correlated to an aromatic proton between δH ) 6.8 and 6.9 ppm. The cross-peak at δN ∼ 142 ppm is correlated to aromatic (δH ∼ 7.2 ppm) and alkene (δH ∼ 5.3 ppm) protons. These chemical shifts are consistent with the anilinoquinone linkage shown in Scheme 1. The spectrum was referenced to DMSO-d6 and NH3 (δN ) 0).
SCHEME 1. Proposed Reaction Pathway for Laccase-Mediated Michael Addition of 15N-SPD to Protocatechuic Acida
a Protocatechuic acid is first oxidized by laccase to a quinone (A). The sulfapyridine aniline nitrogen attacks the carbon at position 3 on the quinone resulting in Michael adduct (B). Isomerization of this product (C) followed by reoxidation (D) results in a quinone. This product is consistent with the NMR correlations shown in Figure 3 and the m/z ratio determined by ESI-TOF-MS (Figure 4).
15N-SPD to protocatechuic acid. Initially, laccase oxidizes protocatechuic acid to an ortho-quinone (Scheme 1A). Michael addition of 15N-SPD to the carbon at position 3 (Scheme 1B) is followed by tautomerization (Scheme 1C) and reoxidization by laccase to an ortho-quinone (Scheme 1D). The resulting adduct bears a single proton associated with a double-bonded carbon within three bonds of the linkage with SPD, consistent with the cross-peak at δH ∼ 5.3 ppm (alkene region, Figure 3). If the cross-coupled product was not reoxidized, the proton correlation would appear in the aromatic region rather than the alkene region on the F2 axis (Figure 3). The aromatic proton correlated to the same δN results from hydrogens on the anilinic ring ortho to the labeled nitrogen. Michael addition of 15N-SPD to position 5 of oxidized protocatechuic acid would not produce a correlation from a proton within three bonds as observed in the 1H-15N HMBC spectrum (Figure 3). Furthermore, the electron distribution
in oxidized protocatechuic acid disfavors nucleophilic addition of SPD to the carbon at position 5. The observed δH were consistent with those predicted by ACD Proton Prediction software for the proposed cross-coupled product in Scheme 1 (data not shown). Differences between the observed δN for the reacted aniline nitrogen and that predicted by ACD Nitrogen Prediction software for the proposed product were noted and most likely resulted from the lack of aromatic structures shown in Scheme 1 in the software database. Mass Spectrometry. The proposed cross-coupled product corresponding to the δN ∼ 142 ppm cross-peaks (Figure 3) had a predicted molecular mass of 400.37 u. ESI-TOF-MS indicated the presence of a reaction product with m/z ) 400.06 (Figure 4), consistent with the final reaction product in Scheme 1. These data support the proposed structural assignment for the cross-coupled product. Ions with this m/z ratio were not detected in control reactions of protocatechuic acid and laccase. VOL. 41, NO. 10, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
3597
FIGURE 4. ESI-TOF mass spectrum of a product formed during incubation of 15N-SPD with protocatechuic acid and laccase. The m/z 400.06 ion is consistent with the final product shown in Scheme 1. The data were obtained in negative ionization mode. The mass labels corresponding to the noise have been removed. Further confirmation of the cross-coupled product structure was obtained by HPLC-ESI-MS/MS. Fragmentation of the m/z 400.06 ion produced daughter ions (Figure 5) consistent with the proposed Michael adduct corresponding to δN ∼ 142 ppm cross-peaks in the HMBC spectrum (Figure 3). Although the m/z 140.3 and 168.4 ions could also arise from fragmentation of unreacted 15N-SPD, the presence of the m/z 123.1 ion suggests Michael adduct formation. A second reaction product with m/z 645.1, presumably corresponding to that producing the cross-peak at δN ∼ 118 ppm (Figure 3), was also subjected to collision-induced dissociation. The resulting mass spectrum contained a number of high molecular mass daughter ions, consistent with cross-coupling of 15N-SPD to self-coupled protocatechuic trimers (data not shown). Self-coupling of substituted phenolic compounds produces complex polymeric structures (40, 41). The 1H-15N HMBC and MS/MS data did not allow structural elucidation of this reaction product beyond confirming that the covalent linkage formed between 15NSPD and altered protocatechuic acid was consistent with an anilinoquinone. Environmental Implications. Chemical incorporation of sulfonamide antimicrobials into NOM is expected to decrease their mobility and bioaccessibility, and provides an explanation for previous reports of apparent nonextractable residue formation (7, 42). Our results suggest that sulfonamide antimicrobials may form Michael adducts with quinone-like moieties in NOM in addition to the Schiff bases we previously reported (30). Bound residues formed by Michael addition reactions (especially after the oxidation step C in Scheme 1) are more likely to persist in soil than those formed by Schiff 3598
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 10, 2007
base reactions (e.g., imines); the latter are more easily hydrolyzed in aqueous solution (32). Reports of nonextractable sulfonamide residues in soil (7, 42) may therefore reflect Michael addition to quinone moieties in soil NOM. While our studies of sulfonamide antimicrobial coupling with model humic constituents shed light on the types of covalent linkages possible with NOM, determination of the relative importance of Michael and Schiff adducts requires investigation of coupling reactions with humic substances. Chemical incorporation of sulfonamide antimicrobials into soil organic matter is expected to diminish their biological activity and therefore reduce the selective pressure of these compounds for antimicrobial resistant microorganisms. Sulfonamide antimicrobials exert their bacteriostatic effect by binding to intracellular dihydropteroate synthase (1). Covalent coupling to NOM may limit the entry of sulfonamides into microbial cells thereby denying access to intracellular targets. Furthermore, the free amino group (anilinic nitrogen) is required for competitive inhibition of dihydropteroate synthase (1). Stable adduct formation via the anilinic nitrogen would drastically diminish antibacterial potency (2). The extent of covalent coupling of sulfonamide antimicrobials to NOM is expected to be closely related to the availability of suitable reaction sites. Covalent coupling of sulfonamides to model humic constituents appears to proceed via nucleophilic addition to quinone-like moieties. Quinones appear to be important constituents of NOM (43, 44), although they may comprise a relatively small fraction of oxygenated functional groups (44). Phenoloxidases secreted by saprophytic fungi can catalyze the formation of
FIGURE 5. Daughter ions resulting from MS/MS analysis of the Michael adduct corresponding to the cross-peaks at δN ∼ 142 ppm in Figure 3. The proposed structure of each fragment is placed above the corresponding m/z. Ionization sites resulting in each fragment ion are noted. The arrow associated with the structure above ion cluster 123.1 represents delocalization of the nitrogen double bond. The m/z of 91.3 is a background signal. electron-poor sites favorable for nucleophilic attack by aromatic amines (29, 30); manganese oxides can similarly oxidize phenolic moieties to produce electron-deficient sites. Catechols and quinones synthesized de novo by brown rot fungus (46) may participate in cross-coupling reactions with contaminants. De novo quinone synthesis may be important in environments with low extracellular phenoloxidase activity due to enzyme sorption and inactivation processes and/or pH extremes. Soil solution pH is expected to influence the cross-coupling reaction kinetics for sulfonamide antimicrobials. As the pH approaches the anilinic nitrogen pKa (2.3 for SPY (47)), this moiety becomes increasingly protonated, reducing its electron density and nucleophilicity. The redox chemistry of soil and subsurface environments is expected to impact the stability of bound sulfonamide residues (32). In covalent coupling reactions with substituted phenols, the aromatic amine of the sulfonamide antimicrobial serves as the nucleophile. Other compounds possessing primary and secondary aromatic amines (e.g., reduced TNT metabolites, chlorinated anilines with various substitution patterns, some herbicides) can engage in coupling reactions to substituted phenols and NOM (27-29). A variety of pharmaceuticals contains aromatic amines (e.g., fluoroquinolones, trimethoprim, some antineoplastic agents, some diuretics) and may also engage in addition reactions to NOM. Terminal and secondary peptidic amines form Michael adducts with quinone-like structures in humic material, a mechanism that may preserve peptides in soils and sediments (48). Synthesis of 15N-SPD facilitated identification of covalent linkages formed in laccase-mediated coupling reactions with protocatechuic acid. We anticipate that stable isotope-labeled sulfonamide antimicrobials will prove useful in future mechanistic studies of sulfonamide interaction with soils and sediments.
Acknowledgments We thank Charlie Fry and Monika Ivancic (University of WisconsinsMadison NMR facility) for helpful advice and Eric Bialk, Ieva Reich and Glen Hinckley for fruitful discussions. The UWsMadison NMR facility is supported in part by National Science Foundation grant CHE-9629688. This research was funded by USDA CSREES Project WIS04621 and the Wisconsin Small System Waste Management Project. We gratefully acknowledge the constructive comments of three anonymous reviewers.
Supporting Information Available Text and figures addressing synthesis and characterization of 15N-sulfapyridine. This material is available free of charge via the Internet at http://pubs.acs.org.
Literature Cited (1) National Research Council. The Use of Drugs in Food Animals: Benefits and Risks; National Academy Press: Washington, DC, 1999. (2) Goodman & Gilman’s The Pharmacological Basis of Therapeutics, 10th ed.; Hardman, J. G., Limbird, L. E., Gilman, A. G., Eds.; McGraw Hill: New York, 2001. (3) Kolpin, D. W.; Furlong, E. T.; Meyer, M. T.; Thurman, E. M.; Zaugg, S. D.; Barber, L. B.; Buxton, H. T. Pharmaceuticals, hormones, and other organic wastewater contaminants in U.S. Streams, 1999-2000: A national reconnaissance. Environ. Sci. Technol. 2002, 36, 1202-1211. (4) Karthikeyan, K. G.; Meyer, M. T. Occurrence of antibiotics in wastewater treatment facilities Wisconsin, U.S.A. Sci. Total Environ. 2005, 361, 196-207. (5) Kinney, C. A.; Furlong, E. T.; Werner, S. L.; Cahill, J. D. Presence and distribution of wastewater-derived pharmaceuticals in soil irrigated with reclaimed water. Environ. Toxicol. Chem. 2006, 25, 317-326. (6) Holm, J. V.; Rugge, K.; Bjerg, P. L.; Christensen, T. H. Occurrence and distribution of pharmaceutical organic compounds in the VOL. 41, NO. 10, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
3599
(7) (8)
(9)
(10)
(11)
(12)
(13) (14) (15) (16) (17)
(18) (19) (20)
(21)
(22) (23)
(24) (25) (26)
(27)
groundwater downgradient of a landfill (Grindsted, Denmark). Environ. Sci. Technol. 1995, 29, 1415-1420. Berger, K. V.; Petersen, B.; Bunning-Pfaue, H. Persistenz von Gu ¨lle-Arzneistoffe in der Nahrungskette. Arch. Lebensmittelhyg. 1986, 37, 85-108. Pe´rez, S.; Eichhorn, P.; Aga, D. S. Evaluating the biodegradability of sulfamethazine, sulfamethoxazole, sulfathiazole, and trimethoprim at different stages of sewage treatment. Environ. Toxicol. Chem. 2005, 24, 1361-1367. Go¨bel, A.; Thomsen, A.; McArdell, C. S.; Joss, A.; Giger, W. Occurrence and sorption behavior of sulfonamides, macrolides, and trimethoprim in activated sludge treatment. Environ. Sci. Technol. 2005, 39, 3981-3989. Pedersen, J. A.; Soliman, M. A.; Suffet, I. H. Human pharmaceuticals, hormones and personal care product ingredients in runoff from agricultural fields irrigated with treated wastewater. J. Agric. Food Chem. 2005, 53, 1625-1632. Kinney, C. A.; Furlong, E. T.; Werner, S. L.; Cahill, J. D. Presence and distribution of wastewater-derived pharmaceuticals in soil irrigated with reclaimed water. Environ. Toxicol. Chem. 2006, 25, 317-326. Kinney, C. A.; Furlong, E. T.; Zaugg, S. D.; Burkhardt, M. R.; Werner, S. L.; Cahill, J. D.; Jorgensen, G. R. Survey of organic wastewater contaminants in biosolids destined for land application. Environ. Sci. Technol. 2006, 40, 7207-7215. Burkhardt, M.; Stamm, C.; Waul, C.; Singer, H.; Muller, S. Surface runoff and transport of sulfonamide antibiotics and tracers on manured grassland. J. Environ. Qual. 2005, 34, 1363-1371. Kay, P.; Blackwell, P. A.; Boxall, A. B. A. Fate of veterinary antibiotics in a macroporous tile drained clay soil. Environ. Toxicol. Chem. 2004, 23, 1136-1144. Batt, A. L.; Snow, D. D.; Aga, D. S. Occurrence of sulfonamide antimicrobials in private water wells in Washington County, Idaho, U.S.A. Chemosphere 2006, 64, 1963-1971. Hirsch, R.; Ternes, T. A.; Haberer, K.; Mehlich, A.; Ballwanz, F.; Kratz, K.-L. Occurrence of antibiotics in the environment. Sci. Total Environ. 1999, 225, 109-118. Lindsey, M. E.; Meyer, M. T.; Thurman, E. M. Analysis of trace levels of sulfonamide and tetracycline antimicrobials in groundwater and surface water using solid-phase microextraction and liquid chromatography/mass spectrometry. Anal. Chem. 2001, 73, 4640-4646. Ho¨per, H.; Kues, J.; Nau, H.; Hamscher, G. Eintrag und Verbleib von Tierarztneimittelwirkstoffen in Bo¨den. Bodenschutz 2002, 4, 141-148. Ku ¨ mmerer, K. Resistance in the environment. J. Antimicrob. Chemother. 2004, 54, 311-320. Pei, R.; Kim, S.-C.; Carlson, K. H.; Pruden, A. Effect of river landscape on the sediment concentrations of antibiotics and corresponding antibiotic resistance genes. Water Res. 2006, 40, 2427-2435. Schmitt, H.; Beelen, P. V.; Tolls, J.; Leeuwen, C. L. V. Pollutioninduced community tolerance of soil microbial communities caused by the antibiotic sulfachloropyridazine. Environ. Sci. Technol. 2004, 38, 1148-1153. Thiele-Bruhn, S.; Beck, I. C. Effects of sulfonamide and tetracycline antibiotics on soil microbial activity and microbial biomass. Chemosphere 2005, 59, 457-465. Thiele-Bruhn, S.; Seibicke, T.; Schulten, H.-R.; Leinweber, P. Sorption of sulfonamide pharmaceutical antibiotics on whole soils and particle-size fractions. J. Environ. Qual. 2004, 33, 13311342. Boxall, A. B. A.; Blackwell, P.; Cavallo, R.; Kay, P.; Tolls, J. The sorption and transport of a sulphonamide antibiotic in soil systems. Toxicol. Lett. 2002, 131, 19-28. Gao, J.; Pedersen, J. A. Adsorption of sulfonamide antimicrobial agents to clay minerals. Environ. Sci. Technol. 2005, 39, 95099516. Thorn, K. A.; Pettigrew, J. P.; Goldenberg, W. S. Covalent binding of aniline to humic substances. 2. 15N NMR studies of nucleophilic addition reactions. Environ. Sci. Technol. 1996, 30, 27642775. Thorn, K. A.; Kennedy, K. R. 15N NMR investigation of the covalent binding of reduced TNT amines to soil humic acid, model compounds, and lignocellulose. Environ. Sci. Technol. 2002, 36, 3787-3796.
3600
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 10, 2007
(28) Dec, J.; Bollag, J.-M. Phenoloxidase-mediated interactions of phenols and anilines with humic materials. J. Environ. Qual. 2000, 29, 665-676. (29) Kim, J. E.; Fernandes, E.; Bollag, J.-M. Enzymatic coupling of the herbicide bentazon with humus monomers and characterization of reaction products. Environ. Sci. Technol. 1997, 31, 2392-2398. (30) Bialk, H. M.; Simpson, A. J.; Pedersen, J. A. Cross-coupling of sulfonamide antimicrobial agents with model humic constituents. Environ. Sci. Technol. 2005, 39, 4463-4473. (31) Tatsumi, K.; Feyer, A.; Minard, R. D.; Bollag, J.-M. Enzymemediated coupling of 3,4-dichloroaniline and ferulic acid - A model for pollutant binding to humic materials. Environ. Sci. Technol. 1994, 28, 210-215. (32) Parris, G. E. Covalent binding of aromatic amines to humates. 1. Reactions with carbonyls and quinones. Environ. Sci. Technol. 1980, 14, 1099-1106. (33) Friebolin, H. Basic One- and Two-Dimensional NMR Spectroscopy; John Wiley & Sons: New York, 1998; pp 246-248. (34) Claridge, T.-D. W. High-Resolution NMR Techniques in Organic Chemistry; Pergamon: New York, 1999; Tetrahedron Organic Chemistry Series Vol. 19. (35) Heinzkill, M.; Bech, L.; Halkier, T.; Schneider, P.; Anke, T. Characterization of laccases and peroxidases from wood-rotting fungi (Family Coprinaceae). Appl. Environ. Microbiol. 1998, 64, 1601-1606. (36) Lehman, J. W. Operational Organic Chemistry; Allyn and Bacon: Boston, MA, 1981; pp 310-320. (37) Bollag, J.-M.; Dec, J. Phenoloxidase-mediated interactions of phenols and anilines with humic materials. J. Environ. Qual. 2000, 29, 665-676. (38) Dec, J.; Haider, K.; Bollag, J.-M. Decarboxylation and demethoxylation of naturally occurring phenols during coupling reactions and polymerization. Soil. Sci. 2001, 166, 660-671. (39) Michael, A. U ¨ ber die Addition von Natriumacetessig- und Natriummalonsa¨urea¨thern zu den Aethern ungesa¨ttigter Sa¨uren. J. Prakt. Chem. 1887, 35, 349-356. (40) Dec, J.; Haider, K.; Bollag, J.-M. Decarboxylation and demethoxylation of naturally occurring phenols during coupling reactions and polymerization. Soil. Sci. 2001, 166, 660-671. (41) Park, J.-W.; Dec, J.; Kim, J.-E.; Bollag, J.-M. Effect of humic constituents on the transformation of chlorinated phenols and anilines in the presence of oxidoreductive enzymes or birnessite. Environ. Sci. Technol. 1999, 33, 2028-2034. (42) Langhammer, J.-P.; Fu ¨ hr, F.; Bu ¨ nning-Pfaue, H. Verbleib von Sulfonamid-Ru ¨ ckstanden aus der Gu ¨ lle im Boden und Nutzpflanzen. Lebensmittelchemie 1990, 44, 93. (43) Thorn, K. A.; Arterburn, J. B.; Mikita, M. A. 15N and 13C NMR investigation of hydroxylamine-derivatized humic substances. Environ. Sci. Technol. 1992, 26, 107-116. (44) Cory, R. M.; McKnight, D. M. Fluorescence spectroscopy reveals ubiquitous presence of oxidized and reduced quinones in dissolved organic matter. Environ. Sci. Technol. 2005, 39, 81428149. (45) Sutton, R.; Sposito, G. Molecular structure of soil humic substances: The new view. Environ. Sci. Technol. 2005, 39, 9009-9015. (46) Paszczynski, A.; Crawfford, R.; Funk, D.; Goodell, B. De novo synthesis of 4,5-dimethoxycatechol and 2,5-dimethoxyhydroquinone by the brown rot fungus Gloeophyllum trabeum. Appl. Environ. Microbiol. 1999, 65, 674-679. (47) Lin, C.-E.; Chang, C.-C.; Lin, W.-C. Migration behavior and separation of sulfonamides in capillary zone electrophoresis II. Positively charged species at low pH. J. Chromatogr., A 1997, 759, 203-209. (48) Pang-Hung, H.; Hatcher, P. New evidence for covalent coupling of peptides to humic acids based on 2D NMR spectroscopy: A means for preservation. Geochim. Cosmochim. Acta 2005, 69, 4521-4533.
Received for review July 20, 2006. Revised manuscript received January 8, 2007. Accepted February 9, 2007. ES0617338