INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

The antibiotic chloramphenicol blocks translation by interacting with the peptidyl transferase centers of ribosomes (25). Acquired resistance to chloramphenicol in eubacteria is most commonly provided by the enzyme chloramphenicol acetyltransferase (CAT), which is encoded by one of several different types of cat genes (reviewed in reference 27). Some CAT proteins also provide resistance to fusidic acid and crystal violet, but they do so by sequestering these compounds (3, 26). CAT acetylates chloramphenicol once or twice; neither the monoacetate nor the diacetate form of chloramphenicol has been shown to have antibiotic activity (27). Most of the studies of chloramphenicol and CAT have been carried out with either gram-negative or gram-positive bacteria. Little is known about the activities of chloramphenicol and CAT in spirochetes, which are distinct from other bacteria in several characteristics (24, 32).

Spirochaeta aurantia is a pigmented, free-living spirochete found in aquatic environments. In comparison to most other known spirochetes, S. aurantia has simple growth requirements and a fast doubling time (5). These features make it a suitable model organism for genetic studies of spirochetes, and accordingly, we began development of a genetic system for S. aurantia. We had previously shown that a CAT gene of gram-positive bacteria could be expressed in transfected Borrelia burgdorferi (28, 30), and S. aurantia was known to be susceptible to chloramphenicol (4). Thus, there was reason to expect that the CAT gene would provide for positive selection and function as a reporter in S. aurantia as well. However, in preliminary experiments we found that S. aurantia was not only susceptible to chloramphenicol it was also unexpectedly susceptible to diacetyl chloramphenicol, the product of CAT.

In the present study we characterized this phenomenon in more detail and investigated the basis for it. We identified a novel esterase in S. aurantia that is capable of hydrolyzing diacetyl chloramphenicol to chloramphenicol. This esterase appears to be responsible for the unusual susceptibility of this bacterium to diacetyl chloramphenicol. Inasmuch as some actinomycetes have been reported to have diacetyl chloramphenicol esterase activity (23), we compared the activity of S. aurantia with those of selected Streptomyces species.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Media, strains, and culture conditions. The bacteria used were S. aurantia M1 (ATCC 25082), S. aurantia J1 (5), Escherichia coli XL1-Blue MRF' (Stratagene, La Jolla, Calif.), E. coli pGO1 (28), Bacillus subtilis EUR9030 (6), Streptomyces coelicolor A3(2) 2612 (13), Streptomyces lividans 66 TK24 (14), and Streptomyces griseus SKK821 (1). S. aurantia was grown in maltose-peptone-yeast extract (MPY) medium (0.2% [wt/vol] maltose-0.2% peptone-0.1% yeast extract-10 mM potassium phosphate buffer [pH 7.5]) at 22°C. Streptomyces species were grown in yeast extract-malt extract (YEME) broth with 6 mM MgCl₂ (13). E. coli and B. subtilis were grown in Luria-Bertani (LB) broth (Difco Laboratories, Detroit, Mich.) at 37°C with shaking.

Chemicals. Chloramphenicol, chloramphenicol diacetate, o-nitrophenol, p-nitrophenol, -napthol, o-nitrophenol acetate, p-nitrophenol acetate, -napthyl acetate, 4-methylumbelliferyl acetate, and Fast Blue RR were purchased from Sigma Chemical Co. (St. Louis, Mo.). Proteinase K was obtained from Boehringer Mannheim (Indianapolis, Ind.). The esterase reagents were purchased from Sigma.

MICs. Stock solutions were made in dimethyl sulfoxide (DMSO). In a pilot study S. aurantia was inhibited by 3.2% (vol/vol) DMSO, and accordingly, final concentrations of DMSO were below 1% for assays with chloramphenicol or diacetyl chloramphenicol. S. aurantia cells from a logarithmic-growth-phase culture were added to 5 ml of MPY medium for a final density of 10⁶ cells/ml. E. coli was similarly grown in MPY medium or LB broth. Bacteria were counted with a Petroff-Hausser chamber under phase microscopy. The MIC of each compound was the lowest concentration that yielded a cell count of fewer than 10⁷ cells/ml after 72 h of culture for S. aurantia or 24 h of culture for E. coli. Control cultures without antibiotic typically yielded >2 × 10⁸ cells/ml for S. aurantia or 10⁹ cells/ml for E. coli under the same conditions. For determinations of MICs for Streptomyces species, 10 ml of YEME medium was inoculated with spores for a final density of 10⁷/ml; the MIC was the lowest concentration that prevented aggregative growth after 72 h. Each assay was performed in triplicate.

Antibiotic bioassays. Plate bioassays of antibiotic activity were performed as previously reported (29). In brief, diacetyl chloramphenicol was added to 1 ml of MPY medium for a final concentration of 20 µg/ml with or without 5 × 10⁶ S. aurantia cells. The culture medium was incubated for 14 h at 22°C and then extracted twice with equal volumes of ethyl acetate (Sigma), and the organic fraction was dried in a Speed Vac evaporator (Savant, Farmingdale, N.Y.). The residue or known quantities of chloramphenicol were dissolved in 20 µl of ethyl acetate and applied to a sterile 6-mm-diameter filter paper disk (BBL, Cockeysville, Md.). B. subtilis was grown in LB broth to approximately 10⁹ cells/ml and swabbed onto petri plates containing Mueller-Hinton agar (Difco). The paper disks were dried in air and then placed on the lawn. The plates were incubated for 14 h at 37°C. When 0.5 µg of chloramphenicol was applied onto a disk, the zone of inhibition was 9 mm. There was no detectable inhibition with a disk containing 800 µg of diacetyl chloramphenicol.

CAE assay. Reagents for the fluorescent chloramphenicol acetate esterase (CAE) assay were from the FASTCAT Assay kit (Molecular Probes Inc., Eugene, Oreg.). Cell extracts were prepared from 5 × 10⁸ stationary-phase cells by pelleting and resuspending cells in 400 µl of TE (10 mM Tris HCl [pH 8.0]-1 mM EDTA). These were lysed by the addition of 40 µl of 50 mM Tris HCl (pH 8.0)-100 mM EDTA-100 mM dithiothreitol and a drop of toluene. The suspension was then briefly vortexed and incubated for 30 min at 30°C (28). Streptomyces cell extracts were prepared from stationary-phase cells pelleted and resuspended in TE followed by sonication at 0°C for 15 min. The suspension was then filtered through a 0.2-µm-pore-size filter (Schleicher & Schuell, Keene, N.H.). The protein concentration of the cell extracts was determined with a Bradford reagent kit (Bio-Rad, Richmond, Calif.), and extract volumes were adjusted to give equivalent protein concentrations for each CAE assay. Volumes of 10 µl were incubated with 10 µl of boron dipyromethane difluoride 1-deoxychloramphenicol (BCAM) or 10 µl of boron dipyromethane difluoride 1-deoxychloramphenicol-3-acetate (AcBCAM) (Molecular Probes). The reaction mixtures were incubated at 34°C for 2 h, extracted with cold ethyl acetate, dried, and resuspended in 20 µl of ethyl acetate. After separation by thin-layer chromatography (TLC), the results were analyzed by a fluorescence densitometric assay as previously described (28). The percent conversion of AcBCAM to BCAM was determined by quantitative analysis of digitized images of the fluorescent bands as described previously (29). For studies of cell-associated esterase activity, a 1.5-ml volume of a stationary-phase culture of S. aurantia at 10⁸ cells per ml was centrifuged for 10 min at 10,000 × g. The supernatant was saved, and the cells in pellet form were suspended in 1.0 ml of phosphate-buffered saline solution, pH 7.5 (PBS), centrifuged, and then resuspended in 1.5 ml of PBS.

Colorimetric esterase assays. Assays were performed as described by Morgan et al. (20) and Beaufay et al. (2). Equal amounts of protein in 30-µl volumes were added to 1.0 ml of the following: 100 mM potassium phosphate (pH 7.5) for p-nitrophenylacetate, 20 mM potassium phosphate (pH 7.5)-1 mM EDTA-0.1% Triton X-100 for o-nitrophenylacetate, 50 mM Tris (pH 8.0) for -napthyl acetate, or PBS for 4-methylumbelliferyl acetate. After 10 min, either 10 µl of 100 mM p-nitrophenylacetate, 50 µl of 180 mM o-nitrophenylacetate, or 10 µl of 40 mM -napthyl acetate, all in cold methanol, or 10 µl of 40 mM 4-methylumbelliferyl acetate in DMSO was added, respectively. The reactions proceeded at 22°C, and the absorbance was determined at 10-min intervals with a Spectronic 21D spectrophotometer (Spectronic Instruments Inc., Rochester, N.Y.) at 420 nm for p-nitrophenylacetate (₄₂₀ = 11.614 mM¹ cm¹), 400 nm for o-nitrophenylacetate (₄₀₀ = 2.1525 mM¹ cm¹), 323 nm for -napthyl acetate (₃₂₃ = 1.477 mM¹ cm¹), and 362 nm for 4-methylumbelliferyl acetate (₃₆₂ = 16.72 mM¹ cm¹). A blank with TE instead of the cell extracts was processed similarly, and the values obtained were subtracted as correction for background hydrolysis of the substrates (2, 20).

Gel electrophoresis and staining with esterase substrates. Extracts of S. aurantia or E. coli were subjected to nondenaturing electrophoresis in which the separating gel was 7.5% acrylamide, the stacking gel was 3.5% acrylamide, and the buffer was 38 mM glycine-5 mM Tris HCl (pH 8.3) (31). Approximately 25 µg of each extract was put in each lane; the sample buffer was 50 mM Tris (pH 6.8)-3% bromophenol blue-35% glycerol. Electrophoresis was run at 125 V and at 4°C. The protein standards were phosphorylase b, bovine serum albumin, and ovalbumin. Afterwards, the gels were soaked in 100 mM potassium phosphate (pH 6.5) for 10 min and then incubated in the same buffer with either -napthyl acetate (5 mM) and Fast Blue RR (0.4 mg/ml) for 1 h (20) or 0.02% 4-methylumbelliferyl acetate for 10 min (9). Bands were visualized under white light for hydrolysis of -napthyl acetate or under long-wavelength UV light for hydrolysis of 4-methylumbelliferyl acetate. To determine whether the cell components with esterase activity with these other substrates also had diacetyl chloramphenicol esterase activity, gel slices were excised from unstained gels. The slices were obtained from the same regions of the lanes of S. aurantia or E. coli extracts that had esterase activity by the colorimetric assays. Slices from above and below the identified bands were included as controls. The gel slices were added to 0.5 ml of PBS with 50 µg of diacetyl chloramphenicol/ml, incubated at 37°C for 4 h, and then analyzed by the plate bioassay as described above.

HPLC. High-performance liquid chromatography (HPLC) was performed at Rocky Mountain Instruments Laboratories (Fort Collins, Colo.). The system consisted of the following: a 250- by 4.6-mm C₁₈ column (Yamamura Chemical Company, Kyoto, Japan) with a 120-Å pore size; a Spectra-Physics (San Jose, Calif.) IsoChrom pump and 8780 autosampler with a 20-µl loop; a Hitachi (San Jose, Calif.) 655A UV detector with a 10-µl flow cell, set at 278 nm; and a Waters (Milford, Mass.) Millennium data system with a SATIN module (A/D converter). The mobile phase was 525 ml of HPLC-grade water, 475 ml of acetonitrile, 1 ml of acetic acid, and 500 mg of ammonium acetate (Fisher). This was filtered through a 0.45-µm-pore-size nylon filter (Whatman, Clifton, N.J.), and the flow rate was 1.5 ml/min. Chloramphenicol diacetate and chloramphenicol were the standards. For the assay, 450 µg of protein from an S. aurantia cell extract, which was prepared as described above, was added to 300 µg of diacetyl chloramphenicol in a final reaction volume of 630 µl. At 0, 30, and 60 min after the start of the reactions, 50-µl samples were removed, dried, and resuspended in the mobile phase. In one experiment the extract was boiled for 10 min before the start of the reaction.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

MICs of chloramphenicol and diacetyl chloramphenicol. The MICs of chloramphenicol and diacetyl chloramphenicol for two strains of S. aurantia, one strain of E. coli, and three species of Streptomyces were determined (Table 1). As expected, E. coli was not susceptible to high concentrations of diacetyl chloramphenicol (29). The MIC of chloramphenicol for the streptomycetes was severalfold higher than that for E. coli or S. aurantia, but all three species of Streptomyces were able to grow in high concentrations of diacetyl chloramphenicol. S. aurantia was unique in being nearly as susceptible to diacetyl chloramphenicol as it was to chloramphenicol.

Bioassay for conversion to chloramphenicol. S. aurantia may have been inhibited in its growth by diacetyl chloramphenicol itself, but this phenomenon has not been reported with other bacteria. An alternative explanation was the conversion of diacetyl chloramphenicol to chloramphenicol by the cells or the medium. To investigate this possibility, a microbiological bioassay for chloramphenicol was carried out. For the assay, different combinations of medium, S. aurantia cells, and diacetyl chloramphenicol were first incubated together. The medium was then extracted, concentrated, and applied to a disk. The disk was placed on a lawn of B. subtilis, which is susceptible to chloramphenicol but not to diacetyl chloramphenicol (29). The combination of S. aurantia, diacetyl chloramphenicol, and medium produced a zone of inhibition of 25 mm for both strains (data not shown). There was no detectable inhibition of the growth of the B. subtilis indicator with extracts of MPY medium alone, medium with diacetyl chloramphenicol, or medium with S. aurantia in the absence of diacetyl chloramphenicol. (We had previously shown that extracts of non-serum-containing medium, diacetyl chloramphenicol, and E. coli or B. burgdorferi cells did not produce inhibition [29].)

CAE assay. For a more direct assessment of the production of chloramphenicol from diacetyl chloramphenicol, cell extracts from S. aurantia M1 were prepared and incubated with a fluorescent chloramphenicol acetate compound (AcBCAM). Two control reactions were carried out: (i) BCAM and the S. aurantia extract and (ii) AcBCAM with the MPY medium. The reaction products were separated by TLC. Results of the experiment are shown in Fig. 1A.

View larger version (37K): [in this window] [in a new window]   FIG. 1.   CAE assay of cell extracts of S. aurantia under different conditions. Each sample was incubated with fluorescent chloramphenicol acetate (AcBCAM) or fluorescent chloramphenicol (BCAM) and separated by TLC, and the substrate and product were visualized under UV light. (A) S. aurantia cell extract with AcBCAM or BCAM; MPY medium with AcBCAM. (B) AcBCAM incubated with the following, from left to right: untreated S. aurantia cell extracts, S. aurantia cell extracts pretreated with heat (70°C) or proteinase K, washed S. aurantia cells (cell pellet), or the spent medium of the culture.

HPLC assay. The compound or compounds produced from diacetyl chloramphenicol by S. aurantia M1 cells were identified by HPLC that included diacetyl chloramphenicol and chloramphenicol as standards (Fig. 2A and B, respectively). At the start of the reaction, the single large peak at a retention time of 7.72 min represented the diacetyl chloramphenicol (Fig. 2C). The HPLC peak at a retention time of 3.22 min indicated the presence of chloramphenicol (Fig. 2B). After 1 h of incubation of the extract with diacetyl chloramphenicol, the findings were reversed: the diacetyl chloramphenicol peak had disappeared, and the largest peak was that of chloramphenicol (Fig. 2D). If the S. aurantia cell extract was boiled first, there was no detectable conversion of diacetyl chloramphenicol to chloramphenicol, even after 2 h of incubation (data not shown).

View larger version (11K): [in this window] [in a new window]   FIG. 2.   HPLC of diacetyl chloramphenicol (A), chloramphenicol (B), and diacetyl chloramphenicol and an S. aurantia extract at 0 min (C) and after 60 min of incubation (D). The x axis of each graph is the retention time in minutes. The y axis is the readout of absorbance of the column outflow at 278 nm. The small peaks in both samples containing the S. aurantia extract (C and D) were at 2.29 min from injection.

Range of esterase activity. To assess the ability of the extract to hydrolyze other esters, we used four different colorimetric substrates for carboxylesterases: o-nitrophenylacetate, p-nitrophenylacetate, -napthyl acetate, and 4-methylumbelliferyl acetate (2, 16, 20). Table 2 shows the results of incubation of extracts of S. aurantia or E. coli with these substrates. The S. aurantia cell extract hydrolyzed each of the esters to a greater extent than the E. coli cell extract did.

View larger version (76K): [in this window] [in a new window]   FIG. 3.   Nondenaturing 7.5% polyacrylamide gel electrophoresis and colorimetric analysis for esterase activity in S. aurantia M1 and J1 and in E. coli. Carboxylesterase activity was detected in situ in the gels with -napthyl acetate and Fast Blue RR (A) or with 4-methylumbelliferyl acetate (B). The bands in the gel shown in panel B were visualized under fluorescent light. Molecular sizes for the markers phosphorylase b (97 kDa), bovine serum albumin (68 kDa), and ovalbumin (43 kDa) are shown to the left of the gels.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

We show here that S. aurantia has a cell-associated esterase activity that converts diacetyl chloramphenicol to chloramphenicol. The product of the reaction was identified as chloramphenicol by three different methods: (i) microbiological assays with chloramphenicol-sensitive and chloramphenicol-resistant bacteria, (ii) TLC with a specific substrate and standards, and (iii) HPLC analysis of the products of the reaction. While it is still possible that S. aurantia is susceptible to low concentrations of diacetyl chloramphenicol itself, this seems unlikely.

The diacetyl chloramphenicol esterase activity was not influenced by prior incubation with sub-MIC concentrations of the substrate or product. The activity was eliminated or much reduced by boiling, heating to 70°C, and protease treatment. In other experiments, changing the temperature, amount of oxygen, amount of light, or carbon sources for growth had no effect on the amount of the esterase activity of S. aurantia (C. D. Sohaskey, unpublished data).

S. aurantia also had carboxylesterase activity for the following other esters: o-nitrophenylacetate, p-nitrophenylacetate, -napthyl acetate, and 4-methylumbelliferyl acetate. Until the putative enzyme is purified or a knockout mutation is produced, we cannot conclude that a single enzyme is responsible for the hydrolysis of all compounds tested here. However, the single band of S. aurantia that had esterase activity for -napthyl acetate and 4-methylumbelliferyl acetate also was able to convert diacetyl chloramphenicol to chloramphenicol by bioassay. This finding indicates that a single protein or an oligomer (16) had activity for both diacetyl chloramphenicol and the other esters.

The function of the S. aurantia esterase or esterases in nature is not known. It may serve this microorganism in its mud, marsh, and pond environments by breaking down complex organic compounds for nutrition and by inactivating toxic substances. Carboxylesterases of mammals have a wide range of substrates and are thought to hydrolyze many exogenous compounds (15, 19). Other bacterial carboxylesterases are responsible for detoxification or for the hydrolysis of diacylglycerides (12, 16, 18). Esterases that hydrolyzed diacetyl chloramphenicol had previously been noted only in the bacterial genera Streptomyces and Corynebacterium (23). We confirmed that three Streptomyces species had diacetyl chloramphenicol esterase activities, but these were at lower levels than those we observed with S. aurantia. This lower activity in the streptomycetes may explain why CAT has provided resistance in some Streptomyces (11).

The streptomycetes, like S. aurantia, inhabit environments containing toxins as well as complex nutrients. S. aurantia might prove to be an important source for biologically unique enzymes, as are the Streptomycetes. This spirochete is relatively easy to culture and can be grown in either defined or rich media (4). The main factor limiting progress in this field is the lack of genetic tools. Indeed, the original impetus for this study was the application of a recombinant cat as a selectable marker for transformation of S. aurantia. But since this spirochete produces an enzyme that would counteract CAT activity, it is questionable whether a transformed CAT gene would provide resistance to chloramphenicol. The esterase activity could be reduced by heating before in vitro CAT assays (28). But treatment at 70°C would not be possible for selection and maintenance of viable transformants.

Moreover, the S. aurantia carboxylesterase may also counteract the effects of other resistance gene products, such as the acetyltransferases that inactivate aminoglycosides, thus abrogating other selection mechanisms. For S. aurantia and other bacteria with potent esterase activity, selection using resistance mechanisms that are not based on acetylation may be necessary. In the case of streptomycetes and chloramphenicol, these other mechanisms include hydrolases (17, 22), phosphorylases (21), and extrusion pumps (10).