Environ. Sci. Technol. 2005, 39, 3671-3678
Elevated Microbial Tolerance to Metals and Antibiotics in Metal-Contaminated Industrial Environments R A M U N A S S T E P A N A U S K A S , * ,† T R A V I S C . G L E N N , †,‡ CHARLES H. JAGOE,† R. CARY TUCKFIELD,§ ANGELA H. LINDELL,† AND JV. MCARTHUR† Savannah River Ecology Laboratory, University of Georgia, Drawer E, Aiken, South Carolina 29802, Department of Biological Sciences, University of South Carolina, Columbia, South Carolina 29208, and Westinghouse Savannah River Company, Savannah River National Laboratory, Aiken, South Carolina 29808
To test the hypothesis that industrial metal contaminants select for microorganisms tolerant to unrelated agents, such as antibiotics, we analyzed metal and antibiotic tolerance patterns in microbial communities in the intake and discharge of ash settling basins (ASBs) of three coal-fired power plants. High-throughput flow-cytometric analyses using cell viability probes were employed to determine tolerances of entire bacterioplankton communities, avoiding bias toward culturable versus nonculturable bacteria. We found that bacterioplankton collected in ASB discharges were significantly more tolerant to metal and antibiotic exposures than bacterioplankton collected in ASB intakes. Optical properties of microorganisms collected in ASB discharges indicated no defensive physiological adaptations such as formation of resting stages or excessive production of exopolymers. Thus, it is likely that the elevated frequency of metal and antibiotic tolerances in bacterioplankton in ASB discharges were caused by shifts in microbial community composition, resulting from the selective pressure imposed by elevated metal concentrations or organic toxicants present in ASBs.
1. Introduction Toxic metal exposure alters the composition of microbial communities and selects for metal-resistant strains (1, 2). It has also been suggested that metal exposure indirectly selects for bacteria resistant to unrelated toxicants, particularly antibiotics (3, 4). The co-occurrence of resistance to metals and antibiotics has been found in numerous clinical and environmental isolates (3, 4). Examples include Pb and multiple antibiotic resistance in Pseudomonas, Bacillus, Corynebacterium, and Enterobacter (5); Cd and streptomycin resistance in Salmonella (6); Hg and tetracycline resistance in unidentified marine bacteria (1); Pb, Co, Ni, Cu, and * Corresponding author phone: (803)725-2752; fax: (803)725-3309; e-mail:
[email protected]. † University of Georgia. ‡ University of South Carolina. § Westinghouse Savannah River Company. 10.1021/es048468f CCC: $30.25 Published on Web 04/16/2005
2005 American Chemical Society
penicillin resistance in Klebsiella pneumoniae (7); V and multiple antibiotic resistance in Enterobacter cloacae (8); and Cd and erythromycin resistance in S. maltophilia (9). Multiple genes encoding for metal and antibiotic resistance are commonly found on the same plasmids or transposons, for example, transposon Tn21 conferring co-resistance to mercury, aminoglycosides, and sulfonamides (3, 10). In some cases, single enzymes function as efflux pumps for multiple metals and antibiotics; this is defined as cross-resistance (11, 12). In either case, direct selection for metal resistance could indirectly select for organisms or mobile genetic elements conferring antibiotic resistance. Accordingly, elevated frequencies of antibiotic resistance have been observed in the culturable fractions of microbial assemblages in metal-contaminated freshwater streams (13), coastal areas (1, 14), and in digestive tracts of primates with Hg-containing dental amalgams (15). If proven important in real-world environments, metalinduced indirect selection for antibiotic resistance could raise serious public health concerns (3, 4). While environmental concentrations of antibiotic contaminants are typically below levels likely to impact microorganisms, except near animal and fish farms (16), metal contamination is ubiquitous and is still rapidly increasing (17). Most antibiotics are readily degraded in the environment (16), but metals are not, and so can represent a long-term selective pressure. Pathogenic and commensal microorganisms selected for antibiotic resistance in metal-contaminated environments may reach human hosts through work and recreation activities. However, identification of the specific agents that select for resistance has been difficult in previously studied environments, some of which may have been contaminated with antibiotics in addition to metals. Resistance studies of environmental bacteria have also been hampered by methodological limitations. The classic method of determining microbial susceptibility to a toxicant is based on examining the ability of an isolate to reproduce in standardized growth media amended with the toxicant. However, only a small fraction of environmental bacteria, typically below 1%, are culturable in such conditions (18). Thus, resistance patterns inferred from the tiny culturable fraction may not be representative of the entire microbial community. To test the hypothesis that toxic metal exposure indirectly selects for antibiotic-resistant bacteria throughout the entire microbial community, we examined the sensitivity of bacteria from ash settling basins (ASBs) of coal-fired power plants to a diversity of antibiotics. Coal-fired power plants represent a major source of global metal pollution, accounting for 1060% of the anthropogenic emissions of As, Cd, Cr, Cu, Hg, Mn, Mo, Ni, Pb, Se, Sb, and V (19). In the United States, coal-fired power plants generate about 1013 g of metal-rich ash annually, about a third of which is mixed with surface water and deposited to ASBs (20). The overflow of ASBs is usually discharged back to surface waters. The only known inputs to the ASBs investigated in this study were river water and ash from coal combustion. This excluded the risk of ASB contamination with antibiotics in excess of those potentially present in intake water, making these systems ideal for studies of the indirect selection for antibiotic resistance by exposure to industrial metal emissions. To avoid culturability biases, we employed fluorescent, single-cell viability probing and high-throughput (up to 100 samples h-1) flow-cytometric analyses to study the tolerance of ASB intake and discharge bacterioplankton to diverse metals and antibiotics. Because no single physiological viability indicator is universally appropriate, we employed VOL. 39, NO. 10, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
3671
two viability probes in parallel: probing for cell membrane integrity with SYTO-9 and propidium iodide nucleic acid stains (21-23) and probing for esterase activity with carboxyfluorescein diacetate (22, 24). In our preliminary tests, these two probes provided better discrimination of live heterotrophic bacterioplankton from autotrophs and debris compared to alternative methods, such as cell membrane potential probes DiOC2(3) (25) and DiBAC4(3) (26) and probing for respiration activity with CTC (27).
2. Materials and Methods 2.1. Study Sites and Sample Collection. We studied the intake and discharge water of ash-settling basins (ASBs) of three electricity generation facilities located in the southeastern (Plants D-400 and Urquhart) and midwestern United States (Wisconsin plant). Both southeastern plants are located along the Savannah River in Aiken County, South Carolina, with plant Urquhart about 30 kilometers upstream of D-400. Plants D-400 and Urquhart combust bituminous coal from Kentucky (eastern United States), whereas the Wisconsin plant uses low-sulfur bituminous coal from the Powder River basin (western United States). The two eastern plants are similar in size and age (built in 1950s) but differ in their ash handling procedures. Plant D-400 deposits both the bottom ash (the heavy fraction) and the fly ash (the light fraction) into a sequence of three large ash settling basins (ASBs), which have a combined water residence time of about 2 months (28). In contrast, Plant Urquhart uses a series of two small ASBs with combined water residence time of about 16 h, primarily to deposit bottom ash. The Wisconsin plant uses its ASBs to deposit bottom ash only. The surface areas of ash settling basins of plants D-400 and Urquhart are 15 000 and 2500 m2, respectively. Managers of the Wisconsin plant requested that the name, the exact location, and the size of ash settling basins of this plant remain anonymous. Triplicate surface water samples were collected from the intake and discharge of the three electricity generation facilities, resulting in a total of 18 samples. The D-400, Urquhart, and Wisconsin plants were sampled on April 7, July 24, and May 6, 2003, respectively. Surface water samples from the Savannah River (intake to Plants D-400 and Urquhart) and from the Wisconsin River (intake to the Wisconsin plant) were collected immediately upstream of the respective power plant intake. The ASB discharge water was collected from the canals releasing basin overflow. In each case, acid-washed and autoclaved 1-L polycarbonate bottles were filled with sample water by placing the bottles 10-20 cm below the surface and pointing against the current. Before fill, the bottles were prerinsed with sample water three times. At each location, the triplicate samples were collected within a period of five minutes. The samples were processed within 2 h after collection (Savannah River, Plant D-400, and Plant Urquhart) or after overnight refrigeration during shipment (Wisconsin River and plant). 2.2. Water Chemistry. Prior to trace element analyses, sample aliquots were passed through GHP Acrodisc GF 25mm syringe filter with GF/0.45 µm GHP membranes and HPLC certified glass fiber prefilters (Pall Life Sciences, East Hills, NY). Then, the samples were acidified with high-purity HNO3 to a final concentration of 3.5% and stored in polypropylene containers at -20 °C for up to 60 days. Samples were analyzed by ICP-MS (Elan DCR 6100 Plus; Perkin-Elmer, Shelton, CT) according to EPA method 6020 (29). Calibration standards covering a range of 1-500 µg/L were prepared daily by serial dilution of NIST traceable primary standards. Certified reference material (SLRS-4, Riverine water; NRC, Ottawa, Canada), internal standards, and blanks were included in the analysis procedure for quality control purposes. 3672
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 10, 2005
Samples for total dissolved nitrogen, phosphorus, and carbon analyses were filtered (as above) and stored refrigerated in acid-washed and precombusted glass vials with Teflon caps for up to 1 week. The analyses were performed using standard methods at the University of Georgia Chemical Analysis Laboratory. Water pH, specific conductance, and temperature were determined in the field with YSI-3800 water quality logger (YSI, Inc., Yellow Springs, OH). 2.3. Bacterioplankton Viability Probes and Flow-Cytometric Analyses. To determine the abundance of bacteria with intact cell membranes, aquatic samples were incubated for 10-30 min with the following (all final concentrations): 5 µM SYTO-9 (membrane-permeable, green-fluorescent nucleic acid stain), 30 µM propidium iodide (nonpermeable red-fluorescent nucleic acid stain), 1 mM Na-EDTA (for permeabilization of the G-cell walls), 0.01% Tween-20 (for cell disaggregation), and a known amount (105-106 mL-1) of 2.15-µm fluorescent SkyBlue particles (internal standard for proportiometric cell quantification) (21-23). Bacteria with intact membranes were discriminated flow-cytometrically as well-defined populations of events with intense green fluorescence on green versus red fluorescence biplots (Figure 1A). Total (live and dead) bacterial abundance was determined using the same protocol as above, except that no propidium iodide was added (21-23). Bacteria were discriminated from debris as intensive green fluorescent particles on green fluorescence and light side scatter biplots (30). To quantify bacterioplankton with active esterases, aquatic samples were incubated for 30-60 min with 10 µM CFDA and 1 mM EDTA (22, 24). Tween-20 and microspheres (as above) were added immediately prior to the flow-cytometric analysis. Bacteria with active esterases were discriminated flow-cytometrically on green fluorescence and light side scatter biplots (Figure 1B). The SYTO-9, propidium iodide, and CFDA were purchased from Molecular Probes, Inc (Eugene, OR); the fluorescent SkyBlue particles were purchased from Spherotech, Inc. (Libertyville, IL); and the EDTA and Tween-20 were supplied by Sigma-Aldrich (St. Louis, MO). All flow-cytometric analyses were performed at room temperature using a standard configuration FACSCalibur instrument equipped with a 488nm argon laser and operated with CellQuest software (BD, Franklin Lakes, NJ). FlowJo software (Tree Star, Inc., Ashland, OR) was used for subsequent data analysis. For counts of viable cells used in the tolerance tests (see below), samples were analyzed immediately after incubations. For total in situ counts (Table 1), aquatic samples were preserved with 2% (final concentration) borax-buffered formaldehyde and stored refrigerated for a period of 1 day. For viable in situ counts of Plant Urquhart samples (Table 2), aquatic samples were stored refrigerated for a period of 1 day. 2.4. Bacterioplankton Tolerance Tests. Aliquots of aquatic samples (10 mL in 15 mL tubes) were left as controls or amended with individual metals or antibiotics representing various modes of action, at the following concentrations: 1 mM CdCl2, 1 mM HgCl2, and 1 mM PbCl2 (class B metals); 0.55 mM NiCl2(OH)6 (transition metals); 100 mg L-1 ampicillin (β-lactams); 300 mg L-1 tetracycline (tetracyclines); 50 mg L-1 rifampin (ansamycins); 150 mg L-1 erythromycin (macrolides); 300 mg L-1 chloramphenicol (chloramphenicols); 100 mg L-1 gentamycin, 100 mg L-1 streptomycin, and 300 mg L-1 kanamycin (aminoglycosides); 50 mg L-1 ciprofloxacin and 300 mg L-1 nalidixic acid (quinolones) (31, 32). These concentrations represent the upper end of the range of concentrations used in previous metal and antibiotic resistance studies (1, 5, 13, 14, 32). The tubes were incubated in the dark under slow rotation for 24 h. Then, the abundances of bacteria with intact membranes and with esterase activity were determined as above. Bacterioplankton tolerance to a
FIGURE 1. Representative flow-cytometric dot-plots generated from bacterioplankton samples collected from Wisconsin plant ASB intake and discharge, which were kept as controls or exposed to 100 mg L-1 ampicillin for 24 h and then probed for membrane integrity and esterase activity. Membrane integrity was probed by simultaneous staining with SYTO-9 and propidium iodide. Esterase activity was probed by incubation with CFDA. Regions were drawn around events considered to represent bacteria with intact membranes or with esterase activity according to probe manufacturer’s instructions. Numerals represent the number of events in each region. Arrows indicate the internal standard particles. Equal volumes (19 µL) of each sample were analyzed, and 30% of the events were plotted.
TABLE 1. Water Chemistry and Bacterioplankton Characterization in Aquatic Samples Collected in the Intake and Discharge of Ash Settling Basins of Coal-Fired Power Plantsb plant D-400 intake
discharge
plant Urquhart intake
Wisconsin plant
discharge
pH 6.0 6.4 6.3 5.4 temperature, °C 16 21 20 23 specific conductance, µS cm-1 80 216 58 88 dissolved organic carbon, µM 409 ( 14 127 ( 12 394 ( 9 247 ( 51 dissolved nitrogen, µM 51 ( 1 47 ( 9 30 ( 1 23 ( 1 dissolved phosphorus, µM 3.9 ( 0.4 3.3 ( 0.0
