Chester Clinic

 

Applicability of standard antibiotic toxicity tests to the ambient aquatic environment

F. L. Hellweger et al., Annals of Environmental Science / 2011, Vol 5, 61-66 APPLICABILITY OF STANDARD
1. INTRODUCTION

ANTIBIOTIC TOXICITY TESTS
Antibiotics, used extensively for human medicine and TO THE AMBIENT AQUATIC
agriculture, enter the aquatic environment via ENVIRONMENT
wastewater and other sources, where they have been found at measurable concentrations [1,2]. There may be adverse effects on non-target organisms (i.e. not Ferdi L. Hellweger*, Xiaodan Ruan, bacteria) [3,4]. Also, there is concern that the Elizabeth Cherchia, Sarah Sanchez antibiotics may promote selection of resistant bacteria and thus add to the global reservoir of antibiotic Center for Urban Environmental Studies, Civil & resistance [5,6]. This paper is concerned with the Environmental Engineering Department, Northeastern effect of antibiotics on bacteria. University, Boston, Massachusetts 02115, USA antibiotics typically are far below their effect Received September 28, 2011; in final form December concentration established using the common minimum 12, 2011, accepted December 15, 2011 inhibitory concentration (MIC) laboratory assay. For tetracycline, for example, ambient surface water concentrations are typically ≤0.11 µg/L [1,7], ABSTRACT
although values as high as 6.8 µg/L have been observed in more heavily impacted systems [8]. MIC Antibiotics enter the aquatic environment via values are about 3,000 µg/L for clinical pathogens [9] wastewater and other sources, where they may and 2,000 µg/L for environmental isolates [10]. promote selection of resistant bacteria, and thus add to The conditions in the environment are quite the global reservoir of antibiotic resistance. Ambient different from those of the conventional MIC test. concentrations typically are several orders of Specifically, the DOM concentration in the ambient magnitude below the lowest observed effect aquatic environment typically is about 6 mgC/L, concentration (LOEC) or whereas MIC tests (liquid broth or solid agar) concentration (MIC), which suggests this is unlikely. generally are done in growth medium at DOM However, the dissolved organic matter (DOM) concentration of about 6,000 mgC/L [11]. Toxicity concentration in conventional MIC laboratory assays tests for environmental bacteria also generally add is typically three orders of magnitude higher than in growth media at high concentrations [10,12,13]. the ambient aquatic environment. Partitioning of Antibiotics, including tetracycline, absorb on DOM antibiotics on DOM could affect their bioavailability [14], which may affect bioavailability. making the laboratory MIC values inapplicable to the Accounting for bioavailability is a well- ambient environment. This question was investigated established concept in environmental toxicology. For using laboratory experiments with E. coli, tetracycline hydrophobic organic compounds, it is generally and DOM varied over six orders of magnitude. For the accepted that the truly dissolved (i.e. not bound to DOM concentrations that were able to support solids or DOM) form of the compound is bioavailable significant growth, the calculated MIC endpoint was 1 mg/L. No media effect was observed, which suggests The truly dissolved concentration can be that sorption to MIC test media is insignificant and estimated using a partitioning calculation. Sorption of that the laboratory-determined MIC values are tetracycline on DOM involves a number of applicable to the ambient environment. mechanisms (e.g. cation exchange, [17]), but their quantification requires information on solution chemistry and (more importantly) DOM properties, Keywords: Antibiotic, tetracycline, bioavailability,
which are not available. Therefore, a simple partition coefficient is used here, as was done previously in models of tetracyclines in the aquatic environment [18,19] and soil [16]. The freely dissolved concentration C Corresponding author: Northeastern University, 360 fd (µ g/L) is [20]: Huntington Ave., 400SN, Boston, MA 02115, USA, phone: (617) 373-3992, fax: (617) 373-4419, e-mail: [email protected] www.aes.northeastern.edu, ISSN 1939-2621 F. L. Hellweger et al., Annals of Environmental Science / 2011, Vol 5, 61-66 the ambient environment. This is an important question that needs to be answered to properly address 1 + K d,DOM [DOM] the environmental impacts of antibiotics. Based on our literature review, this issue has not been addressed. where C (µg/L) is the total concentration, Kd,DOM We performed a number of growth experiments (L/kgC) is the DOM partition coefficient and [DOM] with E. coli and tetracycline at various DOM (kgC/L) is the DOM concentration. Partitioning of concentrations to determine the effect of DOM tetracycline to growth media has not been concentration. At DOM concentrations high enough to investigated, but Kd,DOM for various environmental support significant growth (1.2 - 6,000 mgC/L, almost DOMs (e.g. humic acid, natural organic matter) are four orders of magnitude), no media effect was available (Table 1). Using the above equation with observed, suggesting that sorption to MIC test media Kd,DOM = 104.2 L/kgC (Table 1), a total concentration is negligible and that MIC values are applicable to the of C = 1,000 µg/L corresponds to a freely dissolved ambient aquatic environment. concentration of Cfd = 10 µg/L in the MIC test ([DOM] = 6,000 mgC/L) and 900 µg/L in the ambient aquatic environment ([DOM] = 6 mgC/L). This 2. MATERIALS AND METHODS
suggests that the potency of tetracycline may be increased by two orders of magnitude in the ambient Escherichia coli K-12 MG 1655 was cultured using aquatic environment. standard methods (details in [11]). Experiments with Limited studies have explored the effect of exponential and stationary phase cells were performed environmental factors on the toxicity of tetracycline. to cover a range of conditions. Bacterial colony- Garrett and Miller [21] observe no significant effect of forming units (CFU) were 4.0×108 CFU/100mL and 1.5×1011 CFU/100mL after 2 hr (exponential phase) concentrations were only varied by a factor of two and and 22 hr (stationary phase) of incubation, other parameters (salt) were different as well. Chander respectively. For growth medium, Luria-Bertani broth et al. [22] found reduced toxicity in a soil-water (TEKNOVA) and Mueller-Hinton II cation-adjusted mixture with higher tetracycline affinity. These broth (TEKNOVA) were used. The organic carbon experiments do not cover the 1,000-fold difference in concentration of LB medium was measured to be DOM concentration but they are generally consistent approximately 0.3 gC/gLB [11]. Thus, 20g LB/L with the partitioning mechanism. corresponds to 6,000 mgC/L, and 22 gMHB/L We were concerned that the bioavailability in the corresponds to 6,600 mgC/L, and concentrations of MIC test and ambient environment are very different, and that the MIC toxicity values may not be Tetracycline hydrate (99%) (Aldrich) was dissolved applicable to the ambient environment. Specifically, and diluted in deionized water in 15 mL Falcon tubes we hypothesized that the high DOM concentration in wrapped in aluminum foil to prevent light significantly reduce the environmental potency of antibiotics is much higher in

Table 1 Tetracycline partitioning to dissolved organic matter (DOM)
log Kd,DOM (L/kgC) Elliot soil humic acid (ESHA)(a) Gu and Karthikeyan [24] Sithole and Guy [25] River and wetland NOM Verma et al. [26] (a) 0.01 M I, estimated from data in reference. (b) sorption and desportion, 0.01 M NaCl, estimated from data in reference. (c) Fit to linear portion below Ceq = 5 µM, assumed foc = 0.34. www.aes.northeastern.edu, ISSN 1939-2621 F. L. Hellweger et al., Annals of Environmental Science / 2011, Vol 5, 61-66 For the toxicity experiments, LB or MHB increasing DOM is consistent with a reduction in medium and tetracycline were added to phosphate bioavailability due to partitioning. buffer solution (PBS, [23]) in 250 mL Pyrex wide- Visual examination of the data in Fig. 1 suggests mouth flasks. The flasks were wrapped with growth inhibition for the 1,000 and 10,000 µg/L aluminum foil, covered with cotton swabs, stirred at tetracycline concentrations for all experiments, except 400 rpm, and kept at 20 °C. Experiments with and those with the lowest DOM concentration (panel A4). without equilibration were performed to cover a range For this set of experiments, no significant differences of conditions. For equilibration, bacteria were added are evident and densities for all tetracycline after 24 hours. For no equilibration, bacteria were concentrations are relatively close (note y-axis scale). added after 10 min. Cell densities were counted using We attribute this to the low DOM available for membrane filtration [11]. A sample was filtered growth. For the lower tetracycline concentrations (0, through 0.45 µm filters, put on LB agar plates, 1, 10 and 100 µg/L) there is no consistent pattern for incubated at 37°C for 24 hours, and colonies were the effect of tetracycline on growth. For example, the counted visually. highest growth rate in panels A2 and A3 are for 0 and The minimum inhibitory concentration (MIC), the 10 µg/L tetracycline, respectively. These differences lowest observed effect concentration (LOEC), is are likely due to experimental variability, as illustrated calculated as follows. For each experiment, the growth by the experiments with 0 µg/L tetracycline in panel rate is calculated as the regression slope of the natural A1 (identical experimental conditions). The calculated logarithms of the cell densities vs. time. The MIC for MIC (see Methods section) is 1,000 µg/L for all each set of experiments is taken as the minimum panels in Fig. 1 (except panel A4). concentration for which the growth rate is For the experiments with sufficient media significantly lower (α = 1%) than that of the concentration to support growth, the MIC end point is corresponding experiments with lower tetracycline experimental conditions, including type of media (LB vs. MHB), growth phase of cells (exponential vs. stationary) and sorption equilibration (equilibration vs. 3. RESULTS AND DISCUSSION
no equilibration). There is no correlation between MIC and DOM. The experiments covered DOM The results are presented in Figure 1, which shows the concentrations from those of pure growth media (6 time course of cell densities for experiments with gC/L) down to those far below those typical of surface different tetracycline and DOM concentrations. The waters (12 µgC/L). No media effect is observed, growth media serves as substrate for growth and it which suggests no significant sorption to MIC test may (as hypothesized) serve as a partitioning medium. media, and laboratory-derived MIC values should be Therefore, the growth rate is expected to be higher at applicable to the ambient aquatic environment. higher DOM concentration due to two factors: higher These results are for tetracycline, and the situation nutrients and less bioavailable tetracycline. For may be different for other antibiotics. Of course, the example, in the experiments with MHB media, the bioavailability in the aquatic environment will also be growth rate for the 1,000 µg/L tetracycline treatment affected by partition to natural DOM and solids, is higher at 6,600 than at 120 mgC/L DOM (yellow which has to be considered [16]. line in panels B1 and B2). Is this due to higher We previously presented a model of tetracycline substrate or less available tetracycline? The effect of in the Poudre River [19]. In that study, we used a DOM on the bioavailability of tetracycline cannot be partitioning coefficients based on environmental judged by differences in the growth rates of DOMs to calculate partitioning to MIC test media (see experiments with different DOM concentration. Introduction section), which resulted in a significant To interpret the results, we first examine increase in potency. This led to selection of antibiotic differences between experiments resistant bacteria in the river. The results presented tetracycline concentrations for a given DOM type and here suggest that there is no significant sorption to concentration (i.e. within each panel in Figure 1). MIC test media, and therefore the presence of Specifically, we determine the minimum inhibitory tetracycline-resistant bacteria in the Poudre River concentration (MIC) for each panel by comparing the cannot be explained by the effect of the antibiotic (i.e. growth rates. Then we compare the MIC for different see Models 3B2 and 3C2, Fig. S4, ref. [19]). DOM concentrations. An increase in MIC with www.aes.northeastern.edu, ISSN 1939-2621 F. L. Hellweger et al., Annals of Environmental Science / 2011, Vol 5, 61-66 (A) Luria-Bertani (B) Mueller-Hinton 6,000 mgC/L 6,600 mgC/L Figure 1 Time
course of bacteria densities in water Concentration (µg L-1): tetracycline and dissolved organic Log10 of density in Luria-Bertani broth, stationary phase cells. B: Mueller- exponential phase cells. Number in italics shows DOM concentration. Circles (squares) are experiments 0.012 mgC/L www.aes.northeastern.edu, ISSN 1939-2621 F. L. Hellweger et al., Annals of Environmental Science / 2011, Vol 5, 61-66 4. ACKNOWLEDGEMENTS
Degradation products to environmentally relevant bacteria, including selected This work was supported by funding from the Charles tetracycline-resistant bacteria. Arch Environ River Conservancy, and the Cabot Foundation. The Cont Toxicol, 2002, 42: 263-271. National Science Foundation (NSF) sponsored the [11] Bucci V, Vulić M, Ruan X, Hellweger FL. Northeast Louis Stokes Alliance for Minority Population Dynamics of Escherichia coli in Participation (NE-LSAMP) and the Bernard M. Surface Water. J Am Water Res Assoc, 2011, 1- Gordon Center for Subsurface Sensing and Imaging 9. DOI: 10.1111/j.1752-1688.2011.00528.x (Gordon-CenSSIS). [12] Halling-Sørensen B, Sengeløv G, Ingerslev F, reviewers provided constructive criticism. Jensen LB. Reduced antimicrobial potencies of oxytetracycline, tylosin, sulfadiazin, streptomycin, ciprofloxacin, and olaquindox due 5. REFERENCES
to environmental processes. Arch Environ Cont Toxicol, 2003, 44:7-16. Kolpin DW, Furlong ET, Meyer MT, Thurman [13] Kümmerer K, Alexy R, Huttig J, Scholl A. EM, Zaugg SD, Barber LB, Buxton HT. Standardized tests fail to assess the effects of Pharmaceuticals, hormones, and other organic antibiotics on environmental bacteria. Water wastewater contaminants in U.S. streams, Res, 2004, 38: 2111-2116. 1999−2000: A national reconnaissance. Environ [14] Tolls J. Sorption of veterinary pharmaceuticals Sci Tech, 2002, 36:1202-1211. in soils: A review. Environ Sci Technol, 2001, Yang S, Carlson K. Evolution of antibiotic occurrence in a river through pristine, urban and [15] Di Toro, DM, Zarba CS, Hansen DJ, Berry WJ, agricultural landscapes. Water Res, 2003, Swartz RC, Cowan CE, Pavlou SP, Allen HE, Thomas NA, Paquin PR. Technical basis for Boxall ABA, Kolpin DW, Halling-Sørensen B, establishing sediment quality criteria for Tolls J. Are veterinary medicines causing nonionic organic chemicals using equilibrium environmental risks? Environ Sci Technol, partitioning. Environ Toxicol Chem, 1991, 2003, 37: 286A-294A. Brain RA, Johnson DJ, Richards SM, Sanderson [16] McClellan K, Halden RU. Pharmaceuticals and H, Sibley PK, Solomon KR. Effects of 25 personal care products in archived U.S. pharmaceutical compounds to Lemna gibba biosolids from the 2001 EPA National Sewage using a seven-day static-renewal test. Environ Sludge Survey. Water Res, 2010, 44:658-68. Toxicol Chem, 2004, 23:371–382. [17] Gu C, Karthikeyan KG, Sibley SD, Pedersen Kümmerer K. Resistance in the environment. J JA. Complexation of the antibiotic tetracycline Antimicrob Chemother, 2004 , 54: 311-320. with humic acid. Chemosphere, 2007, 66:1494- Pruden A, Pei R, Storteboom H, Carlson KH. Antibiotic resistance genes as emerging [18] Rose PE, Pedersen JA. Fate of oxytetracycline contaminants: Studies in Northern Colorado. in streams receiving aquaculture discharges: Environ Sci Tech, 2006, 40:7445-7450. Model simulations. Environ Toxicol Chem, Hirsch R, Ternes T, Haberer K, Kratz KL. Occurrence of antibiotics in the aquatic [19] Hellweger FL, Ruan X, Sanchez S. A simple environment. Sci Tot Environ, 1999, 225: 109- model of tetracycline antibiotic resistance in the aquatic environment (with application to the Liu H, Zhang G, Liu CQ, Li L, Xiang M. The Poudre River). Int J Environ Res Public Health, occurrence of chloramphenicol and tetracyclines 2011, 8:480-497. in municipal sewage and the Nanming River, [20] Chapra SC. Surface Water-Quality Modeling. Guiyang City, China. J Environ Monitor, 2009, Boston: McGraw-Hill, 1997. [21] Garrett ER, Miller GH. Kinetics and Bryskier A, ed. Antimicrobial Agents: mechanisms of action of antibiotics on Antibacterials and Antifungals. Washington: microorganisms III. Inhibitory action of ASM Press, 2005. tetracycline and chloramphenicol on [10] Halling-Sørensen B, Sengeløv G, Tjørnelund J. Escherichia coli established by total and viable Toxicity of Tetracyclines and Tetracycline counts. J Pharm Sci, 1965, 54: 427-431. www.aes.northeastern.edu, ISSN 1939-2621 F. L. Hellweger et al., Annals of Environmental Science / 2011, Vol 5, 61-66 [22] Chander Y, Kumar K, Goyal SM, Gupta SC. [25] Sithole BB, Guy RD. Models for tetracycline in Antibacterial activity of soil-bound antibiotics. J aquatic environments. Water Air Soil Pollut, Environ Qual, 2005, 34:1952-1957. 1987, 32: 303−314. [23] U.S. Environmental Protection Agency. Method [26] Verma B, Headley JV, Robarts RD. Behaviour 1603: Escherichia coli (E. coli) in Water by and fate of tetracycline in river and wetland Membrane Filtration Using Modified waters on the Canadian Northern Great Plains. J Membrane-Thermotolerant Escherichia coli Environ Sci Health A, 2007, 42:109-117. Agar (Modified MTEC). EPA-821-R-02-023. Washington: EPA, 2002. [24] Gu C, Karthikeyan KG. Sorption of the antibiotic tetracycline to humic-mineral Northeastern University, 2011 complexes. J Environ Qual, 2008, 37:704-711. www.aes.northeastern.edu, ISSN 1939-2621

Source: http://openjournals.neu.edu/aes/journal/article/download/v5art8/v5p61-66