Science of the Total Environment 323 (2004) 63–69 Ecotoxicological assessment of doxycycline in aged pig manure using multispecies soil systems ´ ´ J. Pro, G. Carbonell, J.V. Tarazona C. Fernandez*, C. Alonso, M.M. Babın, ˜ km 7.5, 28040 Madrid, Spain Laboratory for Ecotoxicology, Department of the Environment, INIA, Ctra de la Coruna Accepted 15 October 2003 Abstract This paper assesses the ecotoxicity of the antibiotic doxycycline in aged spiked pig manure using a multispecies soil system (MS 3) covering plants, earthworms and soil microorganisms. The study reproduced realistic exposure conditions, as well as higher exposure doses covering the uncertainty factors typically employed for covering interspecies variability. MS 3, consisting of columns of natural sieved soil assembled with earthworms and seeds from three plant species, were employed. Pig manure was spiked with doxycycline (75 or 7500 mgyml), aged for 15 days under aerobicyanaerobic conditions and added on top of the soil columns (120 mlycolumn, equivalent to 220 kg Nyha). Water and doxycycline free manure were used as negative controls. Doxycycline (7500 mgyml) solution was used as a positive control. No effects on plants or earthworms were observed. Significant effects on soil phosphatase activity, indicating effects on soil microorganisms, were observed at the highest exposure dose, affecting all soil layers in the doxycycline-solution-treated MS 3 (positive control) but only the top layer in the spiked pig manure system. Chemical analysis confirmed the different behavior of doxycycline in both systems (with and without manure) and those effects were observed in soil with measured concentrations over 1 mgykg soil. The detection of doxycycline in leachates revealed a potential mobility. Leachate concentrations were similar for doxycycline solution and spiked manure treatments. 䊚 2003 Elsevier B.V. All rights reserved. Keywords: Pig manure; Doxycycline; Multispecies soil systems 1. Introduction Veterinary drugs are widely used for intensive animal production, and can appear at significant concentrations in manure. The use of manure as agricultural fertilizer implies the release of drug residues and their metabolites on the environment. The potential environmental effects of these resi*Corresponding author. Tel.: q34-91-347-6866; fax: q3491-357-2293. ´ E-mail address: torija@inia.es (C. Fernandez). dues is of increasing concern (Daughton and Ternes, 1999; Jorgensen and Halling-Sorensen, 2000). Doxycycline is a structural isomer of tetracycline with a broad-spectrum antimicrobial activity (Shaw et al., 1986), which is used for respiratory tract and intestinal tract diseases in farm animals (Aronson, 1980). The available toxicokinetic information indicates low metabolism potential for doxycycline ¨ (Nelis et al., 1981; Bocker, 1983), and after 0048-9697/04/$ - see front matter 䊚 2003 Elsevier B.V. All rights reserved. doi:10.1016/j.scitotenv.2003.10.015 64 ´ C. Fernandez et al. / Science of the Total Environment 323 (2004) 63–69 intravenous or oral administration the drug, mostly unchanged (90–95%), is excreted in urine and faeces (Shaw et al., 1986; Saivin et al., 1988). Doxycycline is used in animal therapy and therefore livestock excreta will contain almost the whole administered dose. Animal manure containing doxycycline will be released in the environment when manure is used as agricultural fertilizer. Direct exposure of terrestrial organisms and indirect exposure of aquatic organisms due to drainage and run-off are expected from this agricultural practice (Halling-Sorensen et al., 1998; Jorgensen and Halling-Sorensen, 2000). The available information on the toxicity of these antibiotics is limited and the standard tests are conducted under unrealistic conditions. Accordingly, the effects of doxycycline on relevant soil taxonomic groups under realistic application conditions were studied using a multispecies soil system (MS 3). Owing to the possible indirect exposure of aquatic organisms, ecotoxicological studies on aquatic invertebrates (Daphnia magna Straus), algae (Chlorella vulgaris Beijerinck) and fish (in vitro toxicity test on fish cell RTG2 and RTl-W1 lines) were also performed. 2. Materials and methods 2.1. Chemicals Doxycycline hyclate (CAS 564-25-0), batch number 22497, was supplied by Hipra Laboratories S.A. (17170 Amer, Girona, Spain) as a microcrystalline yellow powder. Pig manure was supplied from a conventional pig farm where doxycycline was not used. Soil was obtained from an agricultural site (calcisol, pH 7.9, 0.43% organic matter) with no pesticide or manure applications in the previous 5 years. Dechlorinated tap water was used for irrigation. 2.2. Ageing of spiked pig manure Doxycycline was added to pig manure at concentrations of 75 and 7500 mgyl. The lowest concentration corresponds to a dose of 12 mgykgy day, mean animal weight 90 kg, volume of manure 4.75 lyanimalyday, manure dilution factor 3 and non-metabolised doxycycline excretion 100%. The highest concentration corresponded to 100 times the lowest one. Spiked manure, initial concentrations 75 and 7500 mgyl, was kept in the dark, at 22"2 8C, and sporadically shaken for 15 days. After this period, doxycycline concentration was determined by HPCL. The remaining doxycycline contents were 45.3 and 3772 mgyl, respectively. Taking into account the maximum application rate of 210 kg Nyha (Directive CEE 91y676) and the total nitrogen content in the manure (3.08 gy l), 120 ml of fortified manure was applied on each soil column (surface 165 cm2 MS 3 systems). 2.3. Multispecies soil systems MS 3 consist of columns of natural sieved soil where soil dwelling macroorganisms (plants, invertebrates) are incorporated. In this system, a lighty dark period and daily irrigation produces a gradient of conditions from the top to the bottom of the column. For this reason, the soil–air interface, water transport and degradationysorption kinetics are reproduced in a better way than that in the standard soil bioassays, while the use of an homogeneous sieved soil and laboratory-cultured macroorganisms facilitated the reproducibility of the results. The system allows a realistic incorporation of the chemical, resembling the expected agricultural practices for the inclusion of manure. A leachate collection device (funnel and bottle) was integrated. The following treatments were studied in triplicate: doxycycline 15 days aged doxycycline spiked pig manure (45 and 3772 mgyl, low and high concentration APM (aged pig manure), respectively). Doxycycline water solution (7500 mgyl, as doxycycline base), a water control and a blank control (doxycycline free APM). All treatments were added in a total volume of 120 ml per MS 3. Columns were prepared with 9 kg of sandysoil mixture (2:8) and irrigated with 2 l of water. After 48 h, the different treatments were applied on the soil surface, 10 Eisenia fetida Savigny adults, 10 Triticum aestivum Linnaeus seeds and 10 Vicia sativa Linnaeus seeds were introduced in each ´ C. Fernandez et al. / Science of the Total Environment 323 (2004) 63–69 65 column. One hour later, columns were irrigated with 120 ml of water. A fixed volume of 120 ml of water was added daily, leachates were collected weekly and doxycycline was analyzed by HPLC. Lightydark period 16y8 h, light intensity (800 lux) and temperature (22"2 8C) were kept constant. After 21 days, earthworms and germinated seeds were counted, biomass of the aerial parts of developed plants was determined and three microbial endpoints were measured. Soil enzymatic activities (phosphatase and dehydrogenase) were quantified according to Freeman et al. (1995) and Rogers and Li (1985), respectively. Glucose-induced soil respiration was determined on triplicates of 8 g of soil at 30 8C monitoring the CO2 production for 24 h, using the m-Trac 4200 system (SY-LAB, GmbH P.O. Box 47, A-3002 Pukersdorf, Austria), based on the variation of conductivity of a KOH 0.2% water solution. Respiration rates were calculated in the linear phase of the respiration curves. Statistical comparisons were done by one-way analysis of variance test and least significant differences multiple range test (Statgraphics Plus Ver. 4.1), comparing doxycycline water solution with water control and doxycycline APM with blank control. followed by RPHPLC-DAD (Reverse Phase High Performance Liquid Chromatography-Diode Array ¨ Detection, Waters 2695-2696) (Bocker, 1980; Riond et al., 1989; Alonso et al., 2002). An aliquot sample (50-ml leachate or 10-g soil) was buffered with 25-ml buffer pH 5 (400 ml 0.2 M Na2HPO4 0.2 Mq250 ml 0.1 M citric acid 0.1 Mq1.5 g EDTA sodium salt, to pH 5 with NaOH 40%), sonicated for 15 min and centrifuged (soil sample). The solution was loaded on an SPE OASIS HLB 3 cc (Waters) cartridge previously conditioned with 5-ml methanol, 5-ml MilliQ water. The cartridge was vacuum dried and eluted with 1-ml methanol, 2-ml HPLC mobile phase. The extract was injected and chromatographed on a SYMMETRY C18, 5 mm, 4.6=25 mm2 (WAT054275) column, at 40 8C temperature, flow 1.2 mlymin, mobile phase buffer pH 6.5 (1 l 0.1 N NaH2PO4q 5.7, ml NN9-dimethyloctylamine):methanol:acetonitrile (53y35y12). Diode array was set to scan l 200–500 nm for doxycycline identification and the extracted l 357 chromatogram was used for doxycycline quantitation by the means of an external standard calibration curve. 2.4. Ecotoxicological tests The ecotoxicological profile of doxycycline raw material (Table 1) showed Chlorella vulgaris (IC50, 48 hs15.2 mgyl) as the most sensitive tested organism; other aquatic organisms and earthworms seemed to be less sensitive to this compound. For fish cell lines, the most sensitive endpoint was EROD activity, followed by bgalactosidase. Both in vitro endpoints were more sensitive than the in vivo acute lethality. The observed effects of fortified pig manures on soil microorganisms using the MS 3 systems were assessed through phosphatase activity (Fig. 1), dehydrogenase activity (Fig. 2) and glucoseinduced respiration (Fig. 3). For statistical analysis water control was compared with the doxycycline water solution and blank control with high and low doxycycline concentrations in APM. No effects for dehydrogenase activity were observed. However, reductions of 40–50% were observed for phosphatase activity at the highest doxycycline dose. These effects were observed The ecotoxicological profile of doxycycline raw material was studied using the Organisation for Economic Co-operation and Development (OECD) standardised methods: Daphnia magna LC50 (48 h) (method 202), Eisenia fetida LC50 (method 207) and Danio rerio Hamilton LC50 (96 h) (method 203). Chlorella vulgaris IC50 (48 h) was performed on 96 well cell culture plates (Environment Canada, 1992; Ramos et al., 1996). In vitro toxicity test (EROD, b-galactosidase and Neutral Red incorporation) on fish cells (RTG2 and RTl-W1 lines) were carried out according to ´ et al. (2001) and Borenfreund and Puerner Babın (1985), respectively. 2.5. Doxycycline quantitative analysis Quantitation of doxycycline in leachate and soil samples was done by SPE (Solid Phase Extraction) 3. Results and discussion ´ C. Fernandez et al. / Science of the Total Environment 323 (2004) 63–69 66 Table 1 Ecotoxicological profile of doxycycline raw material Taxonomic group Test specie L(I)C50 (95% Conf. Int.) mgyl Aquatic invertebrates Aquatic plants Terrestrial invertebrates Aquatic vertebrates Daphnia magna Chlorella vulgaris Eisenia foetida Danio rerio 156.14 (136.6–178.4). 48 h 15.2 (13.7–16.8). 48 h )1000. 14 days )50. 96 h Fish cells RTL-W1 RTG-2 EROD mgyl b-Gal mgyl N.R. mgyl 51.57 (28.44–93.48) N.D. 60.67 (31.6–116.3) )200 )200 )200 EROD, ethoxy-resorufin O-deethylase; b-Gal, beta-galactosidase; N.R., Neutral Red incorporation; N.D., not determined. Fig. 1. Phosphatase activity (mg methyl-umbelliferone (MU)yg dried soil). APMsaged pig manure. Fig. 3. Respiration rate (mg CO2y100 g dried soil). APMs aged pig manure. Fig. 2. Dehydrogenase activity (mg 1,3,5-triphenylformazanyg dried soil). APMsaged pig manure. only in the top soil layer in the MS 3 treated with spiked pig manure, but at all soil layers when doxycycline was applied in water. No effects of doxycycline treatments were observed on plants and earthworms (Fig. 4). Baguer et al. (2001) observed that the antimicrobials oxytetracycline and tylosin have very low toxicity on soil invertebrates including earthworms, enchytraeids and collembola. Batchelder (1982) and Jjemba (2002) observed effects of some antimicrobials on terrestrial plants but only at doses much higher than those employed in this study. Therefore, the results observed for doxycy- ´ C. Fernandez et al. / Science of the Total Environment 323 (2004) 63–69 cline are in agreement with those reported for other antimicrobials. The chemical analysis of soil (Fig. 5) explained the apparent discrepancy observed for phosphatase activity. When applied with aged manure, doxycycline remained mostly in the top soil layer, which reached concentrations three orders of magnitude higher than those observed in deeper layers. This effect is not observed when doxycycline was applied with water, where a homogeneous distribution, giving similar concentrations through the soil column, was observed. As expected, the effects are not related to the dose, but to the reached soil concentration. Soil concentrations above 1000 mgykg dry soil lead to inhibition in soil phosphatase activity. Doxycycline was detected in all leachate samples (Fig. 6). As expected, the concentrations were higher for the highest doses, but no differences were observed for the applications of spiked aged manure and the direct application in water. The leaching only represents a very low percentage of the total amount of chemical applied to the MS 3. As expected, soil microorganisms represent the most sensitive group for doxycycline. The MS 3, reproducing standard or realistic exposure conditions through the application of aged spiked pig manure, indicates that no effects are observed at concentrations below 1 mgykg soil d.w. The acute 67 Fig. 5. Doxycycline concentration measured in MS 3 soil (mgykg dried). APMsaged pig manure. toxicity of doxycycline to the standard aquatic species is also low. Green algae, with a 48 h IC50 of 15.2 mgyl, appears as the most sensitive group. The European risk assessment protocol for veterinary medicines (EMEAyCVMP, 1997) recommends the use on an application factor of 100 to the most sensitive concentration. Therefore, the Predicted No Effect Concentration for doxycycline in water should be 0.15 mgyl. However, it has been described (Jorgensen and Halling-Sorensen, 2000) that blue-green algae can be two to three orders of magnitude more toxic for antimicrobials than green algae, therefore the application of factor of 100 might not cover this sensitive group. Fig. 4. Length of aerial part of plants (mm) and percentage of survival earthworms. APMsaged pig manure. 68 ´ C. Fernandez et al. / Science of the Total Environment 323 (2004) 63–69 Fig. 6. Cumulative quantity of doxycycline measured in leachates (mgycolumn). APMsaged pig manure. An additional interesting result is the sensitivity ˜ et of the in vitro tests with fish cell lines. Castano al. (1996) demonstrated a correlation between in vivo and in vitro data for three cellular endpoints, ATP, neutral red viability and detachment measured through the total cellular mass (Kenacyd Blue protein assay; Knox et al., 1986), but the sensitivity of the cellular parameters was clearly lower than in vivo lethality. The enzymatic endpoints measured in this study are not only more sensitive than standard cellular endpoints such as neutral red viability but also more sensitive than in vivo lethality. The use of fish testing for routine screening of antimicrobials, which are already tested in terrestrial vertebrates and expected to be of low toxicity to fish, is controversial. The in vitro method employed in this study could be used as an alternative to animal testing. The data presented in this paper provide the ecotoxicological information required for assessing the environmental risk of doxycycline. No information on this antimicrobial has been reported in databases such as those published by HallingSorensen et al. (1998) or Servos et al. (2002). The results are in agreement with those observed for other antimicrobials for soil invertebrates (Baguer et al., 2001; Batchelder, 1982) and for terrestrial plants (Jjemba, 2002). The results also indicate the potential of innovative approaches, such as the use of the MS 3 and in vitro toxicity test on fish cell lines, for the effects assessment of pharmaceuticals. Acknowledgments This work has been co-funded by a research agreement between INIA and HIPRA laboratories and the EU Project ERAVMIS CT1999-00003EVK. References ´ ´ C, Garcıa ´ P, Carbonell G, Alonso C, Fernandez C, del Rıo Tarazona JV. Degradation kinetic of doxycycline in pig manure. In: Abstracts book of SETAC Europe 12th Annual Meeting, Challenges in Environmental Risk Assessment and Modelling, 12–16 May 2002, Vienna, Austria, p. 163. Aronson AL. Pharmacotherapeutics of the newer tetracyclines. JAVMA 1980;10(2):1061 –1068. ´ MM, Garcıa ´ P, Fernandez ´ Babın C, Alonso C, Cabonell G, Tarazona JV. Toxicological characterisation of sludge from sewage treatment plants using toxicity identification evaluation protocols based on in vitro toxicity tests. Toxicol in Vitro 2001;15:519 –524. Baguer AJ, Jensen J, Krogh PH. Effects of the antibiotics oxytetracycline and tylosin on soil fauna. Chemosphere 2001;40:751 –757. Batchelder AR. Chlortetracycline and oxytetracycline effects on plant growth and development in soil systems (cattle manure). J Environ Quality 1982;11(4):675 –678. ´ C. Fernandez et al. / Science of the Total Environment 323 (2004) 63–69 ¨ Bocker R. Rapid analysis of doxycycline from biological samples by high-performance liquid chromatography. J Chromatogr 1980;187:439 –441. ¨ Bocker R. Analysis and quantitation of a metabolite of doxycycline in mice, rats, and humans by high-performance liquid chromatography. J Chromatogr 1983;274:255 –262. Borenfreund E, Puerner JA. Toxicity determined in vitro by morphological alterations and neutral red absorption. Toxicol Lett 1985;24:119 –124. ˜ A, Cantarino MJ, Castillo P, Tarazona JV. Correlations Castano between the RTG-2 cytotoxicity test EC50 and in vivo LC50 rainbow trout bioassay. Chemosphere 1996;32:2141 – 2157. Daughton CG, Ternes T. Pharmaceuticals and personal care products in the environment: agents of subtle change? Environ Health Persp 1999;107(Suppl. 6):907 –938. Biological test method: growth inhibition test using the freshwater alga Selenastrum capricornutum. In: Environmental Protection Series Report EPS EPS 1yRMy25; 1992. EMEAyCVMPy055y96-FINAL. Environmental risk assessment for veterinary medicinal products other than GMOcontaining and immunological products. The European Agency for the Evaluation of Medicinal Products; 1997. Freeman C, Liska G, Ostle NJ, Jones SE, Lock MA. The use of fluorogenic substrates for measuring enzyme activity in peatlands. Plant Soil 1995;175:147 –152. Halling-Sorensen B, Nielsen SN, Lanzky PF, Ingerslev F, ¨ Lutzhoft HCH, Jorgensen SE. Occurrence, fate and effects of pharmaceutical substances in the environment—a review. Chemosphere 1998;36(2):357 –393. 69 Jjemba PK. The effect of chloroquine, quinacrine and metronidazole on both soybean plants and soil microbiota. Chemosphere 2002;46:1019 –1025. Jorgensen SE, Halling-Sorensen B. Drugs in the environment. Chemosphere 2000;40(7):691 –699. Knox P, Uphill PF, Fry JR, Benford J, Balls M. The FRAME multicenter project on in vitro cytotoxicology. Food Chem Toxicol 1986;24:457 –463. Nelis HJCF, DeLeenheer AP. Evidence for metabolic inertness of doxycycline. J Pharm Sci 1981;70:226 –228. ˜ MJ. Desarrollo Ramos C, de la Torre AI, Tarazona JV, Munoz ´ de Chlorella vulgaris, utilizando de un ensayo de inhibicion ´ 1996;13:97 – un test en microplacas. Revista de Toxicologıa 100. Riond J-L, Hedeen KM, Tyczkowska K, Riviere JE. Determination of doxycycline in bovine tissues and body fluids by high-performance liquid chromatography using photodiode array ultraviolet-visible detection. J Pharm Sci 1989;78(1):44 –47. Rogers JE, Li SW. Effect of metals and other inorganic ions on soil microbial activity: soil dehydrogenase assay as a simple toxicity test. Bull Environ Contam Toxicol 1985;34:858 –865. Saivin S, Houin G. Clinical pharmacokinetics of doxycycline and minocycline. Clin Pharmacokinet 1988;15:355 –366. Servos MR, Bennie DT, Starodub ME, Orr JC. Pharmaceuticals and personal care products in the environment: a summary of published literature. National Water Research Institute Contribution No. 02-309. Environment Canada; 2002. Shaw DH, Rubin SI. Pharmacologic activity of doxycycline. JAVMA 1986;189(7):808 –810.