ipso-Hydroxylation and Subsequent Fragmentation: a Novel Microbial Strategy To Eliminate Sulfonamide Antibiotics

Abstract

Sulfonamide antibiotics have a wide application range in human and veterinary medicine. Because they tend to persist in the environment, they pose potential problems with regard to the propagation of antibiotic resistance. Here, we identified metabolites formed during the degradation of sulfamethoxazole and other sulfonamides in Microbacterium sp. strain BR1. Our experiments showed that the degradation proceeded along an unusual pathway initiated by ipso-hydroxylation with subsequent fragmentation of the parent compound. The NADH-dependent hydroxylation of the carbon atom attached to the sulfonyl group resulted in the release of sulfite, 3-amino-5-methylisoxazole, and benzoquinone-imine. The latter was concomitantly transformed to 4-aminophenol. Sulfadiazine, sulfamethizole, sulfamethazine, sulfadimethoxine, 4-amino-N-phenylbenzenesulfonamide, and N-(4-aminophenyl)sulfonylcarbamic acid methyl ester (asulam) were transformed accordingly. Therefore, ipso-hydroxylation with subsequent fragmentation must be considered the underlying mechanism; this could also occur in the same or in a similar way in other studies, where biotransformation of sulfonamides bearing an amino group in the para-position to the sulfonyl substituent was observed to yield products corresponding to the stable metabolites observed by us.

INTRODUCTION

Sulfonamides are widely used as antibiotics, antidiabetics, diuretics, antivirals, and anticancer agents (1–4), and thus, large amounts of these compounds enter the environment every year (5, 6). Contamination with sulfonamides poses environmental concern due to the potential development and dissemination of antibiotic resistances (7). Despite their ubiquity, their microbial metabolism and ultimate fate in the environment thereof are poorly understood.

Several studies showed that sulfamethoxazole (SMX) (Fig. 1a), an important representative of sulfonamide compounds, undergoes partial degradation in wastewater treatment plants under aerobic and anaerobic conditions (8–11). We recently demonstrated that Microbacterium sp. strain BR1, a Gram-positive bacterium isolated from a membrane bioreactor treating effluent contaminated by several pharmaceuticals, was capable of mineralizing the ¹⁴C-labeled aniline moiety of SMX when the latter was supplied as the sole carbon source (12). This was the first unambiguous indication that sulfonamides are subject to growth-linked metabolism in microorganisms.

Fig 1.

Structures of the sulfonamides tested for degradation by resting cells of Microbacterium and the stable transformation products formed from them. (a) Sulfamethoxazole; (b) 3-amino-5-methylisoxazole; (c) sulfadiazine; (d) 2-aminopyrimidine; (e) sulfadimethoxine; (f) 2,6-dimethoxy-4-pyrimidinamine; (g) sulfamethazine; (h) 2,6-dimethyl-4-pyrimidinamine; (i) sulfamethizole; (j) 5-methyl-1,3,4-thiadiazol-2-amine; (k) 4-amino-N-phenylbenzenesulfonamide; (l) aniline; (m) asulam; (n) methylcarbamate; (o) 4-amino-N-cyclohexylbenzenesulfonamide; (p) sulfanilamide.

To our knowledge, Hartig (13) was the first to identify the aminated heteroaromatic side moieties of the sulfonamides SMX and sulfadimethoxine as stable metabolites after biodegradation with activated sludge. This result was recently confirmed by two groups that were able to isolate Microbacterium strains with the ability to degrade the sulfonamides sulfamethazine (SMZ) (14) and sulfadiazine (SDZ) (15). Additionally, both groups identified the aminated heteroaromatic side moieties of the sulfonamide as a stable metabolite after the degradation of the parent compound. Although these stable metabolites were identified, the initial attack of the sulfonamide and the further degradation pathway of the aniline path remained unclear.

Before those findings were reported, only few sulfonamide metabolites (i.e., hydroxylated, acetylated, and glycosylated, etc.) were identified, which were not further degraded (16–18).

Most commercial sulfonamides are para-substituted aromatic amines. As hydroxyl groups, amino moieties are ortho- and para-directing activators in electrophilic aromatic substitutions, and ortho- and para-substituents of phenolic compounds can be detached by ipso-substitution (19–28). Such biologically mediated ipso-substitutions proceed according to an addition elimination mechanism whereby an electrophilic oxygen species attacks the aromatic ring at the carbon atom bearing the substituent (ipso-position), leading to the formation of a transient hydroxylated cyclohexadienone intermediate. Subsequently, the substituent is eliminated as an anion (type I ipso-substitution) or as a cation (type II ipso-substitution), and para-quinone or quinol, respectively, arises as the product (23, 28). The type of elimination depends on whether the ring-substituent bond electrons in the cyclohexadienone derivative remain with the substituent or with the ring moiety. The reaction, regardless of type, is affected by electronic and steric properties of both the substrate and the attacking agent (29), and the full range of biochemically catalyzed ipso-substitutions remains as yet unexplored.

Here, we show that biologically mediated ipso-substitution leads to a novel bacterial degradation pathway for SMX and several other sulfonamides. From the data, we postulate that the sulfonamides are initially hydroxylated at the ipso-position, forming N-substituted 1-hydroxy-4-imino-cyclohexa-2,5-diene-1-sulfonamide intermediates, which immediately undergo fragmentation to 4-iminoquinone, sulfur dioxide, and the aminated substituent. We discuss the implications of this pathway with regard to both the environmental fate of sulfonamides and the development of resistance to sulfonamide antibiotics.

MATERIALS AND METHODS

Acclimatization of Microbacterium.

Microbacterium cells were acclimatized to SMX by incubating the cells in 25% (vol/vol) Standard I medium (Merck, Grogg Chemie, Stettlen-Deisswil, Switzerland) with 0.5 mM SMX. The cultures were incubated on a rotary shaker (Multitron; Infors HT, Bottmingen, Switzerland) at 130 rpm at 28°C until the optical density at 600 nm (OD₆₀₀) reached 1.2 (after approximately 40 h). Cells used for mineralization or degradation experiments were washed by centrifugation at 4,500 × g at 4°C for 15 min (5804R; Eppendorf, Basel, Switzerland). The supernatant was discarded, and the cells were suspended in the medium required for the subsequent experiment. This washing step was repeated at least three times.

Sulfonamide degradation assay.

The sulfonamide degradation studies were carried out with SMX, sulfadiazine (SDZ), sulfamethazine, sulfamethizole, sulfadimethoxine, 4-amino-N-phenylbenzenesulfonamide, and asulam (Fig. 1). The initial concentration of the sulfonamides was 0.1 mM. A Microbacterium sp. strain BR1 cell suspension was diluted with phosphate-buffered saline (PBS) to an OD₆₀₀ of 0.5, and all experiments were carried out in triplicates. Ten milliliters of the cultures was incubated in 50-ml reaction tubes with screw caps on a rotary shaker at 230 rpm at room temperature (RT). Samples were taken every 30 min for 6 h and finally after 21 h. Abiotic controls consisting of pure PBS and the corresponding sulfonamides were also analyzed in triplicates. The sulfonamides sulfanilamide and 4-amino-N-cyclohexylbenzenesulfonamide (Fig. 1) were incubated at an OD₆₀₀ of 7 as in the experiments described above, as an OD₆₀₀ of 0.5 did not lead to detectable degradation. From these batches, samples were taken after 0 and 1 h of incubation. In addition, a control experiment with cells and SMX as a substrate was carried out to verify the metabolic activity of the Microbacterium sp. strain BR1 batch under the same conditions. After centrifugation of the samples, supernatants were analyzed by high-performance liquid chromatography (HPLC) to monitor the concentrations of the parent compound as well as the corresponding degradation product, except for methylcarbamate (Fig. 1n), which is expected to result from asulam (Fig. 1m) degradation. The respective rates for degraded sulfonamides were determined in the linear degradation range over at least 4 data points (2 h of incubation) and with an r² of >0.98.

Growth assay using SMX as a carbon source.

The growth of Microbacterium sp. strain BR1 was tested in a separate experiment. Cultures were grown in 250-ml Erlenmeyer flasks filled with 100 ml of MMO medium (45) amended with yeast extract (0.5 mg liter⁻¹) and vitamins (0.5 ml liter⁻¹) (46). SMX (0.5 mM) was added as an electron donor and carbon source. A total of 0.05% of a culture (OD₆₀₀ = 1) grown on Standard I medium served as the inoculum. The cells were washed twice and resuspended in MMO medium as described above, before adding them to the minimal medium. Cells with SMX, treatments inoculated with autoclaved inoculum, and treatments inoculated with cells but without SMX were set up in triplicate. Cultures were incubated at 28°C on a rotary shaking incubator (130 rpm). One milliliter of sample was centrifuged for 30 min at 16,000 × g at 4°C, the supernatant was discarded, and the pellet was stored at −80°C until DNA extraction was performed. The growth of strain BR1 was measured by quantitative PCR (qPCR), as described in the supplemental material.

4-Aminophenol degradation.

In the degradation assays, the initial 4-aminophenol concentration was 75 μM, and the OD₆₀₀ of Microbacterium sp. strain BR1 cells was 7.0. Samples were taken for 4-aminophenol analysis at 0, 7.5, 15, 30, 45, and 60 min after the start of incubation. The samples were directly mixed with ice-cold methanol (20% [vol/vol] final concentration) and stored in an ice bath containing water and ethanol in the dark to stop both biotic and abiotic transformation of 4-aminophenol. All degradation experiments were set up in triplicates. The samples were then centrifuged at 32,000 × g at 4°C for 5 min. 4-Aminophenol was detected in the supernatants by a colorimetric method adapted from methods described previously by Van Bocxlaer et al. (30). One hundred microliters of Na₂HPO₄ (0.5 M in H₂O [pH 12]) was mixed with 10 μl MnCl₂ (1 mM in H₂O) and 10 μl resorcinol (48 mM in H₂O) before 100 μl of the sample was added. After a 5-min reaction time, the absorption was measured at 550 nm on a Synergy 2 multimode microplate reader (Biotek, Luzern, Switzerland). Due to the abiotic oxidation of 4-aminophenol, the starting concentration of the abiotic sample was measured directly after setting up the experiment. For the calculation of the 4-aminophenol concentration from the absorption value, a standard was freshly prepared.

Measurements of oxygen consumption rates.

Oxygen consumption rate (OCR) measurements were performed by using an XF96 extracellular flux analyzer (Seahorse Bioscience, USA) based on fluorimetric O₂ detection. This system allows the measurement of 96 different oxygen-consuming reactions in parallel and the separated injection of 4 different substrates (ports A to D on the sensor cartridge). This system was proven to be suitable for assays of cultured cells (31) as well as oxygen-consuming enzymes (32). The hydration of the CFA96 sensor cartridge was carried out overnight with 200 μl calibration solution per well. Just before the calibration of the sensor cartridge, port A was loaded with 25 μl 50 mM PBS (pH 7), and port B was loaded with 25 μl of freshly prepared 400 μM substrate solutions in PBS (final reaction mixture concentration, 50 μM). The reaction microplate was loaded with 150 μl acclimatized and nonacclimatized Microbacterium. sp. strain BR1 cells (final OD₆₀₀ of 2) for the biotic assays or with PBS for abiotic controls. Every setup was carried out in quadruplicates at a constant temperature of 32°C. The protocol for the measurements was as follows: mixing for 30 s, measurement for 20 min, injection of port A, mixing for 30 s, measurement for 10 min, injection of port B, mixing for 30 s, and measurement for 4 h. The oxygen consumption rate was calculated in the linear range (r² > 0.99) of oxygen consumption after injection of the substrate.

Cell disruption.

For the ¹⁸O experiment, a frozen Microbacterium stock with an OD₆₀₀ of 6.7 in PBS was thawed at RT, and 1.5 mg ml⁻¹ lysozyme (100,000 units mg⁻¹; Fluka, Buchs, Switzerland) was added. The mixture was incubated on a rotary shaker (Swip; Edmund Bühler, Hechingen, Germany) at RT for 1 h. After centrifugation at 15,000 × g at 4°C for 15 min, the supernatant was discarded, and the cell pellet was suspended to the initial volume by adding ice-cold PBS. The cells were disrupted with the OmniLyse HL kit (ClaremontBio Solutions, Upland, CA, USA). The OmniLyse bead chamber was prewashed with PBS in advance, as described in the OmniLyse protocol, and each lysozyme-treated cell aliquot was disrupted separately by withdrawing and infusing it 10× through the bead chamber. The disrupted-cell suspension was finally centrifuged at 16,000 × g at 4°C for 10 min. The supernatants were pooled and used for the cell extract degradation experiments.

For all other experiments with cell extracts carried out on larger scales, frozen Microbacterium stocks with an OD₆₀₀ of 30 were thawed and treated with lysozyme as described above. After centrifugation at 15,000 × g at 4°C for 15 min, the cells were disrupted by sonication (80% amplitude, 0.5 cycle, for 20 min) in an ice bath containing water and ethanol at −3°C, after adding 10% (wt/wt) glass beads to the cell suspension. The suspension was then centrifuged at 60,000 × g at 4°C for 20 min. The protein content of the crude cell extract was measured with the Pierce bicinchoninic acid (BCA) protein assay kit (Thermo Scientific, Olten, Switzerland).

Cell extract degradation assays.

The cell extract degradation assays were carried out with the undiluted crude cell extract in duplicates. SMX was added in PBS to a final concentration of 100 μM. Cofactor stocks were prepared freshly in PBS and added to the assay mixtures at a final concentration of 1 mM (NADH and NADPH, respectively). Samples were taken, and the proteins contained within were precipitated by adding 25% (vol/vol) 1 M HCl to the sample. The mixture was incubated for 15 min at RT before centrifugation at 16,000 × g at 4°C for 10 min. The supernatant was analyzed by HPLC.

For the cofactor dependence evaluation and the assays to determine sulfite formation from SMX in cell extracts, 300 μl of the sample was transferred into ultrafiltration tubes (Amicon Ultra 10K device; Millipore, Germany) and centrifuged at 14,000 × g at 4°C for 15 min to retain proteins. The flowthrough was subjected to HPLC and ion chromatography.

Incubations of cell extracts under an ¹⁸O₂ atmosphere.

Twenty-milliliter headspace gas chromatography-mass spectrometry (GC-MS) vials (Agilent Technologies, Germany) sealed with butyl-rubber stoppers were flushed with nitrogen. Subsequently, 4 ml of ¹⁸O₂ (isotopic purity, 97%; Sigma-Aldrich, Switzerland) was drawn by syringe from a vessel held upside down under water before being filled with ¹⁸O₂. The oxygen was quickly injected into the vials after inserting a second needle for pressure equalization. Subsequently, a mixture of SMX and cell extract containing 1.2 mg protein ml⁻¹, prepared as described above, was added to the vials without cofactors; finally, 50 μl of aqueous NADH solution was added to a final concentration of 1 mM to start the reaction, while the final concentration of SMX was 0.1 mM in a 1-ml total volume. After 30 min of incubation, the vials were opened, and 500 μl of the reaction mixture was subjected to ultrafiltration in 0.5-ml centrifugal filters (Amicon Ultra 10K device; Millipore, Germany) and centrifuged at 14,000 × g at 4°C for 15 min to retain proteins. Filtrates were then transferred into HPLC vials for liquid chromatography-mass spectrometry (LC-MS) analysis.

RESULTS

Growth of Microbacterium sp. strain BR1 on mineral medium with sulfamethoxazole as the sole carbon source and identification of 3A5MI as a dead-end metabolite.

Microbacterium sp. strain BR1 was able to grow on 0.5 mM (126.5 mg liter⁻¹) SMX as the principal source of carbon and energy when supplied with vitamins and trace amounts (0.5 mg liter⁻¹) of yeast extract (see Fig. S1 in the supplemental material). Growth was determined by real-time PCR quantifying the 16S rRNA gene copy number. To identify SMX degradation metabolites in growing cultures, we analyzed the supernatants by means of a high-performance liquid chromatograph coupled to a diode array detector (HPLC-DAD) and a high-performance liquid chromatograph coupled to a mass spectrometer (LC-MS). We detected a UV-absorbing metabolite (measurable absorbance below 260 nm) that was absent in controls. The metabolite was identified as 3-amino-methylisoxazole (3A5MI) (Fig. 1b) by comparison of retention times, UV-visible (UV-Vis) absorption spectra, and electron spray ionization (ESI) mass spectra (see Fig. S2 in the supplemental material) with the commercially available 3A5MI standard. Further experiments performed with resting cells of Microbacterium sp. strain BR1 revealed that the decrease in the concentration of SMX coincided well with the increase in the concentration of 3A5MI (see Fig. 2c and Fig. S3 in the supplemental material for HPLC-DAD chromatograms).

Fig 2.

Degradation of different sulfonamides by resting cells of Microbacterium sp. strain BR1 and concomitant formation of metabolites (set 1). Shown are the concentrations of the parent compound in the culture supernatant (circles) and the abiotic control (squares) and the concentrations of the corresponding metabolite in the culture supernatant (triangles) and the abiotic control (inverted triangles). (a) Sulfadiazine and 2-aminopyrimidine; (b) sulfamethizole and 5-methyl-1,3,4-thiadiazol-2-amine; (c) sulfamethoxazole and 3-amino-5-methylisoxazole; (d) sulfadimethoxine and 2,6-dimethoxy-4-pyrimidinamine; (e) sulfamethazine and 2,6-dimethyl-4-pyrimidinamine.

Identification of 4-aminophenol in assays with cell extracts of Microbacterium sp. strain BR1.

Additional experiments were carried out in order to gain further insights into the degradation pathway of SMX. Resting-cell experiments with [¹⁴C]SMX, labeled at its aniline moiety, showed the formation of additional transient and polar metabolites, albeit they did not accumulate to concentrations allowing reliable mass spectrometric analysis (see Fig. S4 in the supplemental material). Therefore, we prepared crude cell extracts of Microbacterium sp. strain BR1 and incubated them with 100 μM SMX along with the cofactors NADH and NADPH and without cofactors. 3A5MI was detected only when NADH was used as the cofactor (45.5 μM [standard deviation {SD} = 3.6 μM]; n = 3). Concomitant with the formation of 3A5MI, we observed the formation of an additional metabolite in such incubation mixtures. The new metabolite was identified as 4-aminophenol by comparison of its chromatographic behavior and its mass spectrum (LC-MS) with those of an authentic standard (HPLC mass spectra in Fig. 3a and b and LC-MS extracted ion chromatograms in Fig. S5a and S5c in the supplemental material). Incubations of resting cells of Microbacterium sp. strain BR1 with 75 μM 4-aminophenol as a substrate led to a distinct degradation of this compound compared to controls (see Fig. S6 in the supplemental material). Additionally, we confirmed 4-aminophenol degradation by Microbacterium sp. strain BR1 by means of oxygen consumption rate measurements. In such experiments, we demonstrated that SMX-acclimatized cells had a significantly higher oxygen consumption rate in the presence of 4-aminophenol than did nonamended controls (see Table S1 in the supplemental material).

Fig 3.

HPLC mass spectra of 4-aminophenol formed in supernatants after incubations of Microbacterium sp. strain BR1 cell extracts with NADH and SMX. (a) Authentic standard; (b) incubation under a ¹⁶O₂ atmosphere; (c) incubation under an ¹⁸O₂ atmosphere.

Degradation of SMX by cell extracts under an ¹⁸O₂ atmosphere.

In order to determine the type of enzyme activity involved in the initial attack of SMX and the origin of the hydroxyl group of 4-aminophenol, we incubated cell extracts of Microbacterium sp. strain BR1 with SMX and NADH under an ¹⁸O₂ atmosphere. These assays led to the formation of 4-aminophenol giving a molecular ion with a mass-to-charge ratio of 112 when analyzed by LC-MS in the positive-ionization mode. This ratio corresponds to a mass shift of the molecular ion by 2 atomic mass units in comparison to that of controls incubated under a ¹⁶O₂ atmosphere (HPLC mass spectra in Fig. 3b and c and LC-MS extracted ion chromatograms in Fig. S5c, S5d, S5e, and S5f in the supplemental material) and confirmed that the hydroxyl group introduced into the aniline moiety originated from molecular oxygen.

Degradation of [¹⁴C]SDZ by resting cells of Microbacterium sp. strain BR1.

In order to obtain further evidence with regard to the fate of the heterocyclic side chain and the sulfonyl moiety of sulfonamides, a resting-cell experiment was carried out with [¹⁴C]SDZ (Fig. 1c), a different sulfonamide compound. SDZ was ¹⁴C labeled at its heterocyclic moiety, allowing radioactive tracing of its fate during degradation. During incubation, we observed only two ¹⁴C peaks upon analysis by means of HPLC coupled to a liquid scintillation radiodetector (LSRD) and HPLC-DAD (see Fig. S7 in the supplemental material). We identified these peaks as 2-aminopyrimidine and the parent compound SDZ (Fig. 1d and c, respectively) by means of HPLC-DAD. These results showed that SDZ was also degraded by Microbacterium sp. strain BR1 and that here, analogously to the formation of 3A5MI in incubation mixtures with SMX, 2-aminopyrimidine was formed as the sole metabolite that contained the ¹⁴C label of the original side group. Therefore, the data rule out the transient formation of side-chain metabolites containing the sulfonyl moiety.

Sulfite formation during SMX degradation.

In a further experiment with cell extracts of Microbacterium sp. strain BR1, we were able to show that sulfite but not sulfate was formed concomitantly with the degradation of SMX and the formation of 3A5MI. The concentrations of SMX and its metabolite 3A5MI, as well as those of sulfite and sulfate, were monitored. SMX was degraded by the cell extract at a rate of 1.85 μM min⁻¹ (SD = 0.22 μM min⁻¹), whereas 3A5MI and sulfite were formed at rates of 2.01 μM min⁻¹ (SD = 0.05 μM min⁻¹) and 1.92 μM min⁻¹ (SD = 0.05 μM min⁻¹), respectively (Fig. 4). In such incubation mixtures, we did not observe the biotic formation of sulfate.

Fig 4.

Formation of 3-amino-5-methylisoxazole and sulfite during the degradation of SMX by cell extracts of Microbacterium sp. strain BR1. Shown are the concentrations of SMX (circles), sulfite (triangles), net sulfate (inverted triangles), and 3-amino-5-methylisoxazole (squares) over time.

Degradation of different sulfonamides by SMX-adapted resting cells of Microbacterium sp. strain BR1.

In two incubation experiments, we tested whether additional sulfonamides could be degraded by Microbacterium sp. strain BR1 cells grown on SMX. In a first set of experiments, SMX, SDZ, sulfadimethoxine, sulfamethazine, or sulfamethizole was added at a final concentration of 100 μM to resting-cell suspensions with an OD₆₀₀ of 0.5. All compounds were degraded, and metabolites corresponding to the aminated heteroatomic side group (3A5MI for SMX, 2-aminopyrimidine for SDZ, 2,6-dimethoxypyrimidin-4-amine for sulfadimethoxine [Fig. 1f and e, respectively], 2,6-dimethyl-4-pyrimidinamine for sulfamethazine [Fig. 1h and g, respectively], and 5-methyl-1,3,4-thiadiazol-2-amine for sulfamethizole [Fig. 1j and i, respectively]) accumulated in the assay mixtures (Fig. 2). They could be identified by comparing their respective retention times and absorption spectra obtained by HPLC-DAD to those of the authentic standards. The degradation rates of the sulfonamides corresponded well to the formation rates of the respective metabolites, with almost stoichiometric turnover. The highest degradation rate was calculated for SDZ, at 2.51 μM min⁻¹ g⁻¹ (dry weight) (SD = 0.02 μM min⁻¹ g⁻¹ [dry weight]; n = 3), followed by sulfamethizole, at 2.26 μM min⁻¹ g⁻¹ (dry weight) (SD = 0.03 μM min⁻¹ g⁻¹ [dry weight]; n = 3). SMX was degraded at a rate of 2.09 μM min⁻¹ g⁻¹ (dry weight) (SD = 0.02 μM min⁻¹ g⁻¹ [dry weight]; n = 3), and sulfadimethoxine and sulfamethazine were degraded at lower rates of 1.64 μM min⁻¹ g⁻¹ (dry weight) (SD = 0.07 μM min⁻¹ g⁻¹ [dry weight]; n = 3) and 1.53 μM min⁻¹ g⁻¹ (dry weight) (SD = 0.01 μM min⁻¹ g⁻¹ [dry weight]; n = 3), respectively.

In another experimental set, degradation of 100 μM asulam and 4-amino-N-phenylbenzenesulfonamide was tested by using resting cells with an OD₆₀₀ of 0.5 and compared to degradation of SMX (Fig. 5). At a rate of 1.42 μM min⁻¹ g⁻¹ (dry weight) (SD = 0.03 μM min⁻¹ g⁻¹ [dry weight]; n = 3), SMX was degraded slower than in the first set, while asulam was degraded at a rate of 2.38 μM min⁻¹ g⁻¹ (dry weight) (SD = 0.07 μM min⁻¹ g⁻¹ [dry weight]; n = 3), and 4-amino-N-phenylbenzenesulfonamide was degraded at a rate of 0.90 μM min⁻¹ g⁻¹ (dry weight) (SD = 0.02 μM min⁻¹ g⁻¹ [dry weight]; n = 3). Although lower degradation rates than those in the previous experiments were observed in this case, all compounds were completely degraded. As expected, 4-amino-N-phenylbenzenesulfonamide led to the formation of aniline, which was identified by comparing its retention time and absorption spectrum obtained by HPLC-DAD to those of the authentic standard, which eventually was also degraded. Asulam was degraded when incubated with resting cells of strain BR1, but the expected metabolite methylcarbamate was never detected. Nevertheless, additional incubation experiments with asulam and a crude cell extract of Microbacterium sp. strain BR1 clearly showed that sulfite was released concomitant to the degradation of the parent sulfonamide compound (see Fig. S8 in the supplemental material). In a third experiment, resting cells with an OD₆₀₀ of 7 were incubated with 100 μM SMX, sulfanilamide, and 4-amino-N-cyclohexylbenzenesulfonamide, respectively. The higher cell density was chosen, as preliminary tests did not lead to the degradation of the latter two compounds. While SMX was degraded as before (73.7 μM; SD = 0.9 μM), no degradation was observed for sulfanilamide and 4-amino-N-cyclohexylbenzenesulfonamide.

Fig 5.

Degradation of different sulfonamides by resting cells of Microbacterium sp. strain BR1 and concomitant formation of metabolites (set 2). Shown are the concentrations of the parent compound in the culture supernatant (circles) and the abiotic control (squares) and the concentrations of the corresponding metabolite in the culture supernatant (triangles) and the abiotic control (inverted triangles). (a) SMX and 3-amino-5-methylisoxazole; (b) 4-amino-N-phenylbenzenesulfonamide and aniline; (c) asulam.

DISCUSSION

In this study, we show that Microbacterium sp. strain BR1 employs a novel pathway based on a type I ipso-substitution mechanism to metabolize sulfonamide antibiotics.

We demonstrated that upon incubation with SMX, Microbacterium sp. strain BR1 produced 4-aminophenol as a transient metabolite, which was subsequently further degraded and concomitantly released sulfite and a dead-end metabolite, 3A5MI, in equimolar amounts. From these data, we conclude that SMX was initially hydroxylated at the ipso-position, forming 1-hydroxy-4-imino-N-(5-methylisoxazol-3-yl)cyclohexa-2,5-diene-1-sulfonamide as an intermediate (compound b) (Fig. 6), which then underwent fragmentation to 4-iminocyclohexa-2,5-dienone (4-iminoquinone) (compound c) (Fig. 6), sulfur dioxide (compound d) (Fig. 6), and 3A5MI (compound h) (Fig. 6).

Fig 6.

Proposed mechanism for the degradation of SMX by Microbacterium sp. strain BR1. SMX is hydroxylated. Note that [FeIII-O-OH] represents the hydroperoxo-iron of a P450-dependent monooxygenase, which has been chosen arbitrarily as an electron donor (compound a), followed by electron rearrangement (compound b). This results in a concerted cleavage, giving rise to the formation of benzoquinone-imine (compound c), sulfur dioxide (compound d), and a 3-imino-5-methylisoxazole intermediate (compound e). Reduction of compound c yields 4-aminophenol (compound f), while compound d is hydrated to form sulfite (compound g), and compound e is transformed to 3-amino-5-methylisoxazole by accepting a proton.

Besides SMX, various other sulfonamides were metabolized by resting cells of Microbacterium sp. strain BR1 pregrown on SMX (Fig. 2 and 5). Noteworthy, only those sulfonamides were metabolized for which the aminated side-chain fragments could delocalize the pair of electrons coming from the heterocyclic cleavage of the amide bond and, therefore, were able to act as moderate leaving groups (shown for 3A5MI in Fig. 6). Sulfanilamide and 4-amino-N-cyclohexylbenzenesulfonamide were not metabolized, in accordance with the notion that NH₂⁻ and cyclohexyl-NH⁻ are very poor leaving groups.

At the moment, it remains unclear whether the substituent spontaneously undergoes fragmentation once hydroxylation occurs or whether the reaction is enzymatically catalyzed. In any case, this mechanism seems to be an effective strategy for microorganisms to metabolize sulfonamides to innocuous compounds. So far, only biologically mediated ipso-hydroxylations have been described for which the subsequently eliminated leaving groups maintained their molecular skeleton, and to that extent, the strategy elucidated here is novel to the best of our knowledge.

The hydroxyl group of 4-aminophenol originating from molecular oxygen and the NADH-dependent enzyme activity are strong indications that the hydroxylating enzyme activity was due to a monooxygenase. Flavoprotein monooxygenases (33, 34) and cytochromes P450 (25) are known to catalyze electrophilic hydroxylations of substituted aromatic compounds, whereby hydroperoxo-flavin (35) and hydroperoxo-iron (44) species, respectively, act as the donor of electrophilic oxygen. It remains to be elucidated whether the hydroxylation activity of the Microbacterium sp. strain BR1 enzyme can be categorized as either one.

Sulfonamide antibiotics are used worldwide in large amounts (5, 6) and can be detected in the μg liter⁻¹ range in various environmental matrices (36–38). Such concentrations, well below the inhibitory concentration, might still drive selection for utilization as a carbon and energy source. Previous reports described the appearance of the aminated side chains as stable metabolites after sulfonamide degradation, either with activated sludge (13) or, as recently shown, with isolated bacteria (14, 15). A Microbacterium sp., which has 99% identity to Microbacterium sp. strain BR1 based on 16S rRNA genes, was able to degrade sulfamethazine under the concomitant release of 2,6-dimethyl-4-pyrimidinamine (14). Furthermore, Microbacterium lacus strain SDZm4 (99% identity to Microbacterium sp. strain BR1) was able to degrade SDZ under the concomitant release of 2-aminopyrimidine (15). These findings, together with the results presented here, lead to the assumption that the isolates have a common degradation mechanism for sulfonamides, which may be transferable to other sludge or soil communities able to degrade sulfonamides.

This is the first publication examining this common degradation mechanism and testing different sulfonamides with one isolated strain. The novel pathway described here is based on ipso-hydroxylation and subsequent fragmentation for the metabolism of sulfonamides. A better understanding of such reactions will help in appraising the risks underlying the increased appearance of drug resistance in microorganisms (39). For instance, Mycobacterium tuberculosis shows rather limited growth inhibition by SMX because it is able to disrupt the antibiotic activity of SMX through N-acetylation (40). Along these lines, it can safely be assumed that microorganisms able to metabolize sulfonamides by the mechanism proposed here will also show decreased susceptibility to sulfonamide antibiotics. On the one hand, evolution and propagation of novel microbial degradation pathways for pollutants will contribute to the elimination of such compounds from the environment and to the development of successful bioremediation strategies. To this extent, our findings can have an impact on the design of sulfonamide antibiotics, which are less persistent in the environment. On the other hand, if the target pollutants happen to be antimicrobials, there will also be a certain risk that such novel metabolic traits may contribute to drug resistance. As of now, sulfonamide resistance has often been attributed to the presence of sul genes, coding for an alternative dihydropteroate synthase to alleviate inhibition of the folic acid biosynthesis pathway targeted by sulfonamides (41–43). However, the catabolism of sulfonamides may present a parallel resistance mechanism that has so far been overlooked. Further research will have to address more intensely the complex relationship between microbial metabolism of antimicrobial compounds at low concentrations and the development and propagation of resistance mechanisms derived from such metabolic traits.