- Email: [email protected]
Aerobic degradation of tylosin in cattle, chicken, and swine excreta
ARTICLE IN PRESS
Environmental Research 93 (2003) 45–51
Aerobic degradation of tylosin in cattle, chicken, and swine excreta Jerold Scott Teeter and Roger D. Meyerhoff Lilly Research Laboratories, A Division of Eli Lilly and Company, 2001 West Main Street, Drop Code GL45, Greenfield, IN 46140, USA Received 9 July 2002; received in revised form 14 November 2002; accepted 22 November 2002
Abstract Tylosin, a fermentation-derived macrolide antibiotic, was tested to determine its aerobic degradation rate in cattle, chicken, and swine excreta. For chicken, excreta from a hen administered 14C-tylosin as part of a metabolism study were used. For cattle and swine, 14C-tylosin was added to control excreta. The formation of 14C volatile breakdown products and 14CO2 was not observed throughout the study. Material balance for the cabon-14 label ranged between 94% and 104%. Initial, day-0, concentrations of tylosin-A averaged 119.5274.39, 35.0171.34, and 62.8272.11 mg/g (dry weight basis) for cattle, chicken, and swine excreta samples, respectively. After 30 days, samples averaged 4.1670.69 and 4.1170.69 mg/g tylosin-A in cattle and swine excreta, respectively. No residues of tylosin-A or its factors were apparent in the chicken excreta samples after 30 days of incubation. In each case, tylosin declined to less than 6.5% of the initial level after 30 days. Calculated first-order half-lives under the test conditions were 6.2 days, o7.6 days, and 7.6 days for cattle, chicken, and swine excreta, respectively. The results indicate that tylosin residues degrade rapidly in animal excreta. Therefore, tylosin residues should not persist in the environment. r 2003 Elsevier Science (USA). All rights reserved. Keywords: Tylosin; Degradation; Biodegradation; Environmental chemistry; Environmental fate
1. Introduction Tylosin, CAS Number [1401-69-0], is a fermentationderived macrolide isolated from a strain of Streptomycetes fradiae found in soil from Thailand (McGuire et al., 1961). Tylosin bulk material and formulated products are composed primarily (80–90%) of tylosin factor-A, or tylosin-A, and small amounts of factors B, C, and D (Loke et al., 2000). This natural product exhibits antibacterial activity against numerous gram-positive bacteria and mycoplasma species by binding to ribosomes and inhibiting bacterial protein synthesis (Elanco, 1982). Tylosin has been used worldwide in a variety of food animal species to prevent and treat respiratory, enteric, and other diseases. It has also been used to enhance the efficiency of food utilization. Tylosin may enter the environment through its use in animals. Tylosin, administered either orally or through injection, is excreted from animals primarily through the feces. A considerable amount of tylosin antimicrobial activity is lost through animal metabolism. Approximately 29% of the microbiological activity of an orally
Fax: +1-317-277-5022. E-mail address: [email protected] (J. Scott Teeter).
administered dose of tylosin is recovered in the feces (Elanco, 1982). Tylosin metabolism in animals generally proceeds from factor A to factor B, or from factor A to factor D and then to the metabolite dihydrodesmycosin. Further metabolism and degradation in vivo are expected. Relative microbiological activities for tylosin and its related factors are provided in Fig. 1. Although the majority of tylosin is degraded during transit through the gastrointestinal tracts of various food animal species, further study on the fate of tylosin is of interest due to potential environmental concerns. The excreta of animals that have been treated with tylosin may be used as fertilizer on cropland. Therefore, it is important to evaluate the degradation rate of tylosin in excreted feces. The research presented here provides evidence for the rapid degradation of tylosin residues in excreta.
2. Materials and methods The chemical structure, molecular weight, chemical formula, and position of the 14C label for tylosin-A used in these studies are provided in Fig. 1. Structures of
0013-9351/03/$ - see front matter r 2003 Elsevier Science (USA). All rights reserved. doi:10.1016/S0013-9351(02)00086-5
ARTICLE IN PRESS J. Scott Teeter, R.D. Meyerhoff / Environmental Research 93 (2003) 45–51
46
O CH3
H3C
CH3 O
HO
OCH2 O-R2
* *
O CH3
N(CH3)2 HO
H3C
O
* H3CH2C
CH2-R1
*
O
O
O-R3
CH3
* O
OH
HO
C46 H77 N O17 (tylosin A) MW= 916.12 * denotes position of 14C label
Compound
R1
R2
Tylosin A -CHO -CH3 Tylosin B -CHO -CH3 Tylosin C -CHO -H -CH3 Tylosin D -CH2OH -CH3 Dihydrodesmycosin -CH2OH 1 Relative bioactivity from unpublished sources.
CH3 OH
O
CH3
Mycarose
R3 Mycarose -H Mycarose Mycarose -H
Relative Bioactivity1 1.00 0.83 0.75 0.35 0.31
Fig. 1. Structures and relative bioactivity of tylosin and its known factors. The molecular formula, molecular weight and position of the are noted for tylosin-A.
tylosin factors B, C, D, and dihydrodesmycosin are also provided in Fig. 1. Flasks of animal excreta containing added 14C-tylosin were incubated aerobically for 30 days. At test initiation and after 30 days incubation, samples were collected and assayed for tylosin. Additionally, the study apparatus collected volatile 14C species in order to provide a material balance. The specific details regarding the study design are provided below. 2.1. Preparation of test and control vessels Two hundred grams of control excreta (cattle, chicken, or swine) were mixed with a Brinkmann Polytron homogenizer (Brinkmann Instruments, Westbury, NY, USA) in a plastic beaker. Varying amounts of water was added to facilitate mixing (approximately 20–40 mL). Approximately 50 g was transferred to each control vessel (triplicates of each excreta). An additional 200 g of control excreta (either cattle or swine) were mixed as above and an aliquot of a 14 C-tylosin spiking solution was added. After thorough mixing, triplicate samples were assayed by combustion to measure the homogeneity of the mixture based on the distribution of 14C. After homogeneity was achieved, approximately 50 g were transferred to each test material vessel (triplicates for each excreta).
14
C-label
The chicken test material was obtained from a hen treated with 14C-tylosin as part of a tylosin metabolism study. Excreta from this animal were collected and blended 1:1 (w:w) with water. The excreta were stored frozen prior to use. For use in this study, the material was thawed at room temperature and mixed in approximate ratio of 1:1 by weight with freshly obtained control chicken excreta. The material was blended with a Brinkmann Polytron homogenizer and tested as before for homogeneity. Prior to adding the test and control material, the gross tare weights of the test vessels and caps were measured and recorded. This permitted the determination of moisture lost throughout the incubation period based on gravimetric measurements. The vessels were connected to a volatile compound trapping system and a vacuum system. Ambient air was drawn through the vessels to provide oxygen and sweep volatile compounds evolved into the trapping system (Fig. 2). 2.2. Incubation After preparation, and throughout the study, the test vessels were covered with foil to exclude light. The vessels were incubated in a room maintained at 20711C. Periodically, at intervals of no more than 7 days, the vessels were weighed and water was added to the flasks
ARTICLE IN PRESS J. Scott Teeter, R.D. Meyerhoff / Environmental Research 93 (2003) 45–51
(a)
47
Backflow Trap (plastic scintillation vial)
Inflow To vacuum pump
14 CO -Trap (graduated cylinder) 2 Volatile Organic Trap (graduated cylinder)
(b)
Backflow Traps
To vacuum pump
14 CO2 Traps Volatile Organic Traps (scintillation vials)
Fig. 2. Aerobic incubation and volatile compound trapping apparatus (similar to Eirkson et al., 1987) used to determine the degradation of tylosin in cattle, chicken, and swine excreta. (a) Volatile trapping system used for chicken excreta. Back-flow traps were empty scintillation vials. (b) Volatile organic traps contained Scintisafe Scintillation Cocktail (Fisher Scientific, Fair Lawn, NJ, USA) and the 14CO2 traps contained Harvey Carbon-14 Cocktail (R.J. Harvey Instrument Corporation, Hillsdale, NJ, USA).
to approximate the day-0 weight, except on day 30 when no water was added. 2.3. Moisture content determination At study initiation and termination, moisture contents were determined for the contents of each vessel. Moisture contents were determined gravimetrically after drying the samples at approximately 1031C for about 24 h. Moisture contents were expressed as the percent of the original wet weight. 2.4. Sampling and analysis for volatile compounds and 14CO2 production
14
C-organic
Gas washing bottles were used to trap volatile 14Cproducts from the cattle and swine excreta. On days 7, 14, 21, and 30, the flow of air through the cattle and swine vessels was stopped and the traps were removed. The volume of solution (Harvey Carbon-14 Cocktail) in each trap was recorded. One-milliliter aliquots from each trap were counted by liquid scintillation counting (LSC) for 14C. Scintillation cocktail was added to the empty back-flow traps and the entire solution was
counted by LSC. The traps were replaced at each sampling interval. All LSC data was obtained using a Beckman LS6000 Liquid Scintillation Counter (Beckman-Coulter, Fullerton, CA, USA). For the study with chicken excreta, small 20-mL scintillation vials (Zinsser Analytic Polyvials, Frankfurt, Germany) were used as traps. The traps were changed frequently to keep them from drying up. On days 3, 6, 8, 10, 13, 15, 17, 20, 22, 24, 27, and 30, the traps from the chicken systems were removed and additional liquid scintillation cocktail was added as necessary to the vials (to ensure at least 10 mL of solution in the vials as required for scintillation counting). The trap samples were counted by LSC. The traps were replaced with new vials and solution at each sampling interval. 2.5. Collection of samples from vessels After loading the test vessels with the appropriate excreta, duplicate samples were removed, extracted, and assayed by HPLC for tylosin-A. Duplicate samples were also taken to assay by combustion. Additional prepared control material was used for the extraction recovery samples (spikes). On day 30, the bulk of the vessel
ARTICLE IN PRESS 48
J. Scott Teeter, R.D. Meyerhoff / Environmental Research 93 (2003) 45–51
contents was transferred to individual plastic cups and homogenized. Duplicate samples were taken for combustion and another duplicate set of samples were taken for extraction and HPLC analysis. Triplicate samples were taken from each vessel for moisture content determination. On day 30, material from control vessels was used for recovery determination. 2.6. Extraction of tylosin from excreta Excreta samples were extracted on days 0 and 30. Ten milliliters of refrigerated (41C) extractant solution (methanol:acetonitrile:0.1 M ascorbic acid, 45:45:10, v:v:v) was added to 3–5 g of solid sample in a 50-mL plastic centrifuge tube. The tube was capped, shaken, and the contents homogenized with a Brinkmann Polytron homogenizer for approximately 15–30 s. The tube was centrifuged at 1400g for 10 min and the clear supernatant decanted to a plastic beaker. A second 10mL aliquot of extractant was added to the tube and the process was repeated. Eighty milliliters of Milli-Q water (Millipore, a division of Waters Corporation, Milford, MA, USA) was added to the pooled supernatant and the solution was rested at room temperature for 10 min. A 1.00-mL aliquot of each raw extract was assayed by LSC after addition of 12 mL of Scintisafe Scintillation Cocktail (Fisher Scientific, Fair Lawn, NJ, USA). The remainders of the raw extracts were loaded on Sep-Pak C18 Plus SPE cartridges (Waters Corporation, Milford, MA, USA) that were preconditioned with 5 mL of methanol followed by 5 mL Milli-Q water applied under vacuum. The extracts were loaded at a flow rate of less than 6 mL/min. After loading, the SPE column was washed with 10 mL Milli-Q water followed by 10 mL of 20% acetonitrile in water (v:v). The columns were dried under vacuum for at least 3 min at a minimum of 127 mmHg. The columns were eluted with 2.5 mL of 5% acetic acid in methanol and the eluates dried under a gentle stream of air at 351C. Dried residues were redissolved in 2.00 mL of sample diluent (methanol:water, 70:30, v:v) with sonication and brief vortexing. The final extracts were filtered through 0.45-mm PTFE filters (Gelman Sciences, Ann Arbor, MI, USA) prior to HPLC analysis. Extracts were diluted as necessary to be within the calibration range of the analytical detector and in light of the anticipated sample levels. Calibration standards were prepared between 0.25 and 2.5 mg/mL. Aliquots (0.10 mL) of each final eluate were assayed by LSC. Samples of the solid residues from the extractions were combusted to determine radioactive material balance.
Table 1 Gradient HPLC conditions for tylosin assay Column Detector Flow rate Injection volume Column temperature Run time Mobile phase A Mobile phase B Gradient conditions:
250 mm 4.6 mm i.d. Spherisorb Phenyl, 5-mm particle size uv 280 nm 1.5 mL/min 100 mL 301C 30 min 1:1 (v:v) acetonitrile:water 0.02 M dibutyl ammonium phosphate Time %A %B 0.00 100 0 6.00 100 0 6.01 38 62 25.00 58 42 25.01 100 0 30.00 100 0
(HPLC). Details of the gradient separation and detection are listed in Table 1. Reference standard material was used for the quantitative determination of tylosin-A. Tylosin concentrations were determined by comparison of sample peak areas against a multipoint standard curve. Extraction and clean-up efficiency were assessed with control excreta (approximately 5 g) spiked with 150 mg tylosinA. The limit of detection for tylosin-A in extract solution was 0.08 mg/mL. The limit of quantitation for tylosin-A in extract solution was 0.25 mg/mL. Reference sample material was used for the qualitative identification of tylosin factors B, C, and D. Qualitative identification was based on coincident retention time. When injected as individual qualitative standards and assayed by HPLC, tylosin factors A, B, C, and D eluted from the analytical column between 11 and 15 min. Area and relative concentration (quantitative) determination for tylosin factors was based on the peak area response of tylosin-A reference standard material. 2.8. Combustion analysis Samples for total 14C were weighed onto quartz weigh boats and combusted for 4 min in a Harvey Biological Oxidizer (R.J. Harvey Instrument Corporation, Hillsdale, NJ, USA). The released 14CO2 was trapped in Harvey Carbon-14 Cocktail (R. J. Harvey Instrument Corporation, Hillsdale, NJ, USA) and measured by LSC.
3. Results
2.7. Analytical method (HPLC methods and extraction)
3.1. Moisture content
Analysis for tylosin was performed using reverse phase high performance liquid chromatography
The initial moisture content of the homogenized control and test excreta are provided in Table 2.
ARTICLE IN PRESS J. Scott Teeter, R.D. Meyerhoff / Environmental Research 93 (2003) 45–51 Table 2 Moisture content, 14C in volatile and 14CO2 traps at study conclusion, HPLC analytical values at days 0 and 30, total percent of initial tylosin degraded in 30 days
49
14
C material balance and
Matrix
Initial moisture Final moisture Percent of initial Day 30 HPLC Percent of initial Total material Day 0 HPLC 14 content (%) content (%) C found in traps balance (%) mean concentration mean concentration tylosin degraded of tylosin7std. in 30 days at study end (%) of tylosin7std. dev. (mg/g)a dev. (mg/g)a
Cattle Cattle control Chicken Chicken control Swine Swine control
82.3 82.2 75.0 65.3 75.9 77.0
85.1–85.8 83.0–85.8 84.4–85.5 77.6–80.9 76.7–77.7 59.3–72.0
o0.8 o0.006 o0.1 o0.01 o0.8 o0.005
104 na 94 na 101 na
119.5274.39 nd 35.0171.34 nd 62.8272.11 o13c
4.1670.69 nd o2.23b nd 4.1170.69 o1
96.5 na 493.6 na 93.5 na
Notes: nd— not detected; na—not analyzed. a n=6. b Calculated based on analytical limit of detection and dilution factor. c Most swine in the USA receive tylosin prophylactically. Therefore, tylosin residues in swine excreta are to be expected.
Changes in the final moisture content, primarily an increase in moisture content, relative to time zero is attributable to dried material deposited on the sides of the flasks. Moisture contents were kept relatively constant throughout the studies. 3.2.
14
C material balance
The test material vessels evolved 14C-volatile compounds corresponding to less than 0.8% of the initial 14 C contained in the test systems. Combustion analysis data for 14C in the solid samples on day 30 summed with the 14C evolved in the traps provided overall material balance figures of 104%, 94%, and 101% for cattle, chicken, and swine, respectively (Table 2). 3.3. HPLC analytical results Initial, day 0, concentrations of tylosin-A averaged 119.5274.39, 35.0171.34, and 62.8272.11 mg/g (n ¼ 6) on a dry weight basis for cattle, chicken, and swine samples, respectively (Table 2). On day 30, samples averaged 4.1670.69 and 4.1170.69 mg/g tylosin-A (n ¼ 6) for cattle and swine, respectively. No residues of tylosin or its factors were apparent in the chicken excreta samples after 30 days of incubation (with a limit of detection of 2.23 mg/g, Fig. 3). However, trace residues of possible tylosin factors were evident in the HPLC chromatograms on day 30 for the swine and cattle samples (Figs. 4 and 5) although the levels were very low. Calculated first-order degradation half-lives of tylosin-A in cattle, chicken, and swine excreta were 6.2 days, o7.6 days, and 7.6 days, respectively. Approximately 96.5%, 93.5%, and more than 93.6% of the initial tylosin-A was degraded in 30 days for the cattle, swine, and chicken excreta, respectively (Table 2).
Fig. 3. Representative HPLC chromatograms for tylosin in chicken excreta: (a) control day-0, dilution factor (DF) 1 , (b) day-0 test sample, DF 50 , and (c) day-30 test sample, DF 40 .
ARTICLE IN PRESS J. Scott Teeter, R.D. Meyerhoff / Environmental Research 93 (2003) 45–51
50
14
Unextractable bound residues increased slightly as measured by combustion at 30 days. As shown in Table 3, day-0 extraction recoveries for 14C averaged 79% for the cattle excreta and 76% for the swine excreta as compared to the total sample combustion values. Day30 extraction recoveries for 14C compounds from the excreta were largely efficient, ranging between average values of 63% and 94% as compared to the total sample combustion values (Table 3). Day-0 residual 14C concentrations in the extracted solid ranged from 19.0% to 21.6% as compared to the combustion values for the cattle and swine excreta, respectively. Day-30
residual 14C concentrations in the extracted solid averaged 53.2%77.0%, 16.4%71.0%, and 36.4%71.0% for the cattle, chicken, and swine excreta samples, respectively. When compared to day 0, these data suggest that the solid retains more of the 14C material over time. This may be in part due to incorporation of breakdown products into the matrix or extraction inefficiency as sorption occurs over the duration of the study. On day 0, 80.1%72.6% and 81.3%71.2% of the 14C that was in the bulk extract was recovered in the final extract after SPE clean up for the cattle and swine excreta, respectively. After 30 days of incubation, the values were 51.2%79.5% and 15.5%74.0% for the
Fig. 4. Representative HPLC chromatograms for tylosin in swine excreta: (a) control day-0, dilution factor (DF) 30 , (b) day-0 test sample, DF 30 , and (c) day-30 test sample, DF 2 .
Fig. 5. Representative HPLC chromatograms for tylosin in cattle excreta: (a) control day-0, dilution factor (DF) 30 , (b) day-0 test sample, DF 30 , and (c) day-30 test sample, DF 2 .
3.4. Extract and final sample
Table 3 Percentage of
C residues
14
C extracted and measured by LSC as compared to time-zero Cattle
Average Std.dev. %RSD
14
C as determined by combustion
Swine
Chicken
Day 0
Day 30
Day 0
Day 30
Day 0
Day 30
79.05 2.27 2.87
63.12 4.04 6.40
75.55 2.63 3.47
71.18 2.75 3.86
na
93.88 6.48 6.90
Notes: na—not applicable, samples were not assayed to provide these values; n ¼ 6 for all values.
ARTICLE IN PRESS J. Scott Teeter, R.D. Meyerhoff / Environmental Research 93 (2003) 45–51
cattle and swine excreta samples, respectively. Although not measured on day 0, after 30 days of incubation, the value was 2.2%70.1% for the chicken excreta. This data indicates that initially, a high level of 14C material was attributable to tylosin and was recovered in the final extracts. With aging, the 14C residues, no longer being structurally or chemically similar to tylosin, are isolated out of the day 30 final extracts by the SPE procedure. This result parallels the analytical values for tylosin as measured by HPLC.
51
the transient nature of tylosin in the environment by analyzing 104 surface water samples from sites where tylosin might be found. Only 13.5% of the samples assayed contained tylosin, and those only had concentrations ranging from 0.04 to 0.28 mg/L. While the studies presented here were conducted under controlled laboratory conditions, seasonal and temporal variations in the field may occur and effect the rate of degradation. However, microflora induction effects from treated animals may also provide for more rapid degradation than was observed in these studies.
4. Discussion/conclusions Veterinary drugs, including tylosin, may enter the soil through the feces of treated animals (van Gool et al., 1993). Baguer et al. (2000) investigated the effects of tylosin on soil fauna (earthworms, springtails, and enchytraeids) and concluded that no effects were observed at levels as high as 3000 mg/kg. In another study, designed to measure the biodegradability of tylosin in soil–manure slurries, the authors concluded that tylosin and its degradation products disappeared rapidly (Ingerslev and Halling-Sorensen, 2001). Degradation half-lives for tylosin in soil–manure slurries ranged between 4.1 and 8.1 days. The authors also concluded that mineralization of tylosin and its degradation products must occur rapidly since their study methodologies were designed to cover and detect a wide range of possible metabolites. Loke et al. (2000) observed similar results in both aerobic and methanogenic manure systems, concluding that the degradation half-lives of tylosin in the aqueous phase were less than 2 days. Similarly, tylosin degraded rapidly in a composting trial (Vogtmann et al., 1984). In compost containing swine and chicken litter, tylosin at initial levels of 22.4 and 224.0 ppm was not detected after 75 days. The results presented here further indicate that tylosin residues in cattle, chicken, and swine excreta degrade rapidly. Measured half-lives for aerobic degradation are 6.2 days, o7.6 days, and 7.6 days for cattle, chicken, and swine excreta, respectively. These rapid half-lives indicate that tylosin residues will not be persistent in the environment. This is also confirmed by Gavalchin and Katz (1994), who conducted soil degradation studies at 41C, 201C, and 301C and reported that no tylosin was found after 30 days incubation at 201C and 301C, and only 40% of the initial tylosin was recovered at 41C. Kolpin et al. (2002) recently confirmed
Acknowledgment Special thanks to Alison N. Perkins for her detailed review and editorial comments on this manuscript.
References Baguer, A.J., Jensen, J., Krogh, P.H., 2000. Effects of the antibiotics oxytetracycline and tylosin on soil fauna. Chemosphere 40, 751–757. Elanco, C., 1982. Tylan Injection Veterinarian’s Technical Handbook. Elanco Products Company Indianapolis, IN. Eirkson, C., Harrass, M.C., Osborne, C.M., Sayre, P.G., Zeeman, M., 1987. Aerobic biodegradation in water. In: Environmental Assessment Technical Assistance Handbook. US FDA, Washington, 3.11. PB87–175345. Gavalchin, J., Katz, S.E., 1994. The persistence of fecal-borne antibiotics in soil. J. Assoc. Anal. Commun. Int. 77 (2), 481–485. Ingerslev, F., Halling-Sorensen, B., 2001. Biodegradability of metronidazole, olaquindox, and tylosin and formation of tylosin degradation products in aerobic soil-manure slurries. Ecotoxicol. Environ. Saf. 48, 311–320. Kolpin, D.W., Furlong, E.T., Meyer, M.T., Thurman, E.M., Zaugg, S.D., Barber, L.B., Buxton, H.T., 2002. Pharmaceuticals, hormones, and other organic wastewater contaminants in us streams 1999–2000: a national reconnaissance. Environ. Sci. Technol. 36, 1202–1211. Loke, M.L., Ingerslev, F., Halling-Sorensen, B., Tjornelund, J., 2000. Stability of tylosin a in manure containing test systems determined by high performance liquid chromatography. Chemosphere 40, 759–765. McGuire, J.M., Boniece, W.S., Higgens, C.E., Hoehn, M.M., Stark, W.M., Westhead, J., Wolfe, R.N., 1961. Tylosin, a new antibiotic, I. Microbiological studies. Antibiot. Chemother. 11 (5), 320–327. van Gool, S., 1993. Possible environmental effects of antibiotic residues in animal manure. Tijdschr. Diergeneeeskd. 8–10 (Dutch English Summary). Vogtmann, H., Fragstein, P.V., Draeger, P., 1984. The degradation of agrochemicals during composting. Peat Agriculture Horticulture Proceedings Second International Symposium, pp. 357–378.
Environmental Research 93 (2003) 45–51
Aerobic degradation of tylosin in cattle, chicken, and swine excreta Jerold Scott Teeter and Roger D. Meyerhoff Lilly Research Laboratories, A Division of Eli Lilly and Company, 2001 West Main Street, Drop Code GL45, Greenfield, IN 46140, USA Received 9 July 2002; received in revised form 14 November 2002; accepted 22 November 2002
Abstract Tylosin, a fermentation-derived macrolide antibiotic, was tested to determine its aerobic degradation rate in cattle, chicken, and swine excreta. For chicken, excreta from a hen administered 14C-tylosin as part of a metabolism study were used. For cattle and swine, 14C-tylosin was added to control excreta. The formation of 14C volatile breakdown products and 14CO2 was not observed throughout the study. Material balance for the cabon-14 label ranged between 94% and 104%. Initial, day-0, concentrations of tylosin-A averaged 119.5274.39, 35.0171.34, and 62.8272.11 mg/g (dry weight basis) for cattle, chicken, and swine excreta samples, respectively. After 30 days, samples averaged 4.1670.69 and 4.1170.69 mg/g tylosin-A in cattle and swine excreta, respectively. No residues of tylosin-A or its factors were apparent in the chicken excreta samples after 30 days of incubation. In each case, tylosin declined to less than 6.5% of the initial level after 30 days. Calculated first-order half-lives under the test conditions were 6.2 days, o7.6 days, and 7.6 days for cattle, chicken, and swine excreta, respectively. The results indicate that tylosin residues degrade rapidly in animal excreta. Therefore, tylosin residues should not persist in the environment. r 2003 Elsevier Science (USA). All rights reserved. Keywords: Tylosin; Degradation; Biodegradation; Environmental chemistry; Environmental fate
1. Introduction Tylosin, CAS Number [1401-69-0], is a fermentationderived macrolide isolated from a strain of Streptomycetes fradiae found in soil from Thailand (McGuire et al., 1961). Tylosin bulk material and formulated products are composed primarily (80–90%) of tylosin factor-A, or tylosin-A, and small amounts of factors B, C, and D (Loke et al., 2000). This natural product exhibits antibacterial activity against numerous gram-positive bacteria and mycoplasma species by binding to ribosomes and inhibiting bacterial protein synthesis (Elanco, 1982). Tylosin has been used worldwide in a variety of food animal species to prevent and treat respiratory, enteric, and other diseases. It has also been used to enhance the efficiency of food utilization. Tylosin may enter the environment through its use in animals. Tylosin, administered either orally or through injection, is excreted from animals primarily through the feces. A considerable amount of tylosin antimicrobial activity is lost through animal metabolism. Approximately 29% of the microbiological activity of an orally
Fax: +1-317-277-5022. E-mail address: [email protected] (J. Scott Teeter).
administered dose of tylosin is recovered in the feces (Elanco, 1982). Tylosin metabolism in animals generally proceeds from factor A to factor B, or from factor A to factor D and then to the metabolite dihydrodesmycosin. Further metabolism and degradation in vivo are expected. Relative microbiological activities for tylosin and its related factors are provided in Fig. 1. Although the majority of tylosin is degraded during transit through the gastrointestinal tracts of various food animal species, further study on the fate of tylosin is of interest due to potential environmental concerns. The excreta of animals that have been treated with tylosin may be used as fertilizer on cropland. Therefore, it is important to evaluate the degradation rate of tylosin in excreted feces. The research presented here provides evidence for the rapid degradation of tylosin residues in excreta.
2. Materials and methods The chemical structure, molecular weight, chemical formula, and position of the 14C label for tylosin-A used in these studies are provided in Fig. 1. Structures of
0013-9351/03/$ - see front matter r 2003 Elsevier Science (USA). All rights reserved. doi:10.1016/S0013-9351(02)00086-5
ARTICLE IN PRESS J. Scott Teeter, R.D. Meyerhoff / Environmental Research 93 (2003) 45–51
46
O CH3
H3C
CH3 O
HO
OCH2 O-R2
* *
O CH3
N(CH3)2 HO
H3C
O
* H3CH2C
CH2-R1
*
O
O
O-R3
CH3
* O
OH
HO
C46 H77 N O17 (tylosin A) MW= 916.12 * denotes position of 14C label
Compound
R1
R2
Tylosin A -CHO -CH3 Tylosin B -CHO -CH3 Tylosin C -CHO -H -CH3 Tylosin D -CH2OH -CH3 Dihydrodesmycosin -CH2OH 1 Relative bioactivity from unpublished sources.
CH3 OH
O
CH3
Mycarose
R3 Mycarose -H Mycarose Mycarose -H
Relative Bioactivity1 1.00 0.83 0.75 0.35 0.31
Fig. 1. Structures and relative bioactivity of tylosin and its known factors. The molecular formula, molecular weight and position of the are noted for tylosin-A.
tylosin factors B, C, D, and dihydrodesmycosin are also provided in Fig. 1. Flasks of animal excreta containing added 14C-tylosin were incubated aerobically for 30 days. At test initiation and after 30 days incubation, samples were collected and assayed for tylosin. Additionally, the study apparatus collected volatile 14C species in order to provide a material balance. The specific details regarding the study design are provided below. 2.1. Preparation of test and control vessels Two hundred grams of control excreta (cattle, chicken, or swine) were mixed with a Brinkmann Polytron homogenizer (Brinkmann Instruments, Westbury, NY, USA) in a plastic beaker. Varying amounts of water was added to facilitate mixing (approximately 20–40 mL). Approximately 50 g was transferred to each control vessel (triplicates of each excreta). An additional 200 g of control excreta (either cattle or swine) were mixed as above and an aliquot of a 14 C-tylosin spiking solution was added. After thorough mixing, triplicate samples were assayed by combustion to measure the homogeneity of the mixture based on the distribution of 14C. After homogeneity was achieved, approximately 50 g were transferred to each test material vessel (triplicates for each excreta).
14
C-label
The chicken test material was obtained from a hen treated with 14C-tylosin as part of a tylosin metabolism study. Excreta from this animal were collected and blended 1:1 (w:w) with water. The excreta were stored frozen prior to use. For use in this study, the material was thawed at room temperature and mixed in approximate ratio of 1:1 by weight with freshly obtained control chicken excreta. The material was blended with a Brinkmann Polytron homogenizer and tested as before for homogeneity. Prior to adding the test and control material, the gross tare weights of the test vessels and caps were measured and recorded. This permitted the determination of moisture lost throughout the incubation period based on gravimetric measurements. The vessels were connected to a volatile compound trapping system and a vacuum system. Ambient air was drawn through the vessels to provide oxygen and sweep volatile compounds evolved into the trapping system (Fig. 2). 2.2. Incubation After preparation, and throughout the study, the test vessels were covered with foil to exclude light. The vessels were incubated in a room maintained at 20711C. Periodically, at intervals of no more than 7 days, the vessels were weighed and water was added to the flasks
ARTICLE IN PRESS J. Scott Teeter, R.D. Meyerhoff / Environmental Research 93 (2003) 45–51
(a)
47
Backflow Trap (plastic scintillation vial)
Inflow To vacuum pump
14 CO -Trap (graduated cylinder) 2 Volatile Organic Trap (graduated cylinder)
(b)
Backflow Traps
To vacuum pump
14 CO2 Traps Volatile Organic Traps (scintillation vials)
Fig. 2. Aerobic incubation and volatile compound trapping apparatus (similar to Eirkson et al., 1987) used to determine the degradation of tylosin in cattle, chicken, and swine excreta. (a) Volatile trapping system used for chicken excreta. Back-flow traps were empty scintillation vials. (b) Volatile organic traps contained Scintisafe Scintillation Cocktail (Fisher Scientific, Fair Lawn, NJ, USA) and the 14CO2 traps contained Harvey Carbon-14 Cocktail (R.J. Harvey Instrument Corporation, Hillsdale, NJ, USA).
to approximate the day-0 weight, except on day 30 when no water was added. 2.3. Moisture content determination At study initiation and termination, moisture contents were determined for the contents of each vessel. Moisture contents were determined gravimetrically after drying the samples at approximately 1031C for about 24 h. Moisture contents were expressed as the percent of the original wet weight. 2.4. Sampling and analysis for volatile compounds and 14CO2 production
14
C-organic
Gas washing bottles were used to trap volatile 14Cproducts from the cattle and swine excreta. On days 7, 14, 21, and 30, the flow of air through the cattle and swine vessels was stopped and the traps were removed. The volume of solution (Harvey Carbon-14 Cocktail) in each trap was recorded. One-milliliter aliquots from each trap were counted by liquid scintillation counting (LSC) for 14C. Scintillation cocktail was added to the empty back-flow traps and the entire solution was
counted by LSC. The traps were replaced at each sampling interval. All LSC data was obtained using a Beckman LS6000 Liquid Scintillation Counter (Beckman-Coulter, Fullerton, CA, USA). For the study with chicken excreta, small 20-mL scintillation vials (Zinsser Analytic Polyvials, Frankfurt, Germany) were used as traps. The traps were changed frequently to keep them from drying up. On days 3, 6, 8, 10, 13, 15, 17, 20, 22, 24, 27, and 30, the traps from the chicken systems were removed and additional liquid scintillation cocktail was added as necessary to the vials (to ensure at least 10 mL of solution in the vials as required for scintillation counting). The trap samples were counted by LSC. The traps were replaced with new vials and solution at each sampling interval. 2.5. Collection of samples from vessels After loading the test vessels with the appropriate excreta, duplicate samples were removed, extracted, and assayed by HPLC for tylosin-A. Duplicate samples were also taken to assay by combustion. Additional prepared control material was used for the extraction recovery samples (spikes). On day 30, the bulk of the vessel
ARTICLE IN PRESS 48
J. Scott Teeter, R.D. Meyerhoff / Environmental Research 93 (2003) 45–51
contents was transferred to individual plastic cups and homogenized. Duplicate samples were taken for combustion and another duplicate set of samples were taken for extraction and HPLC analysis. Triplicate samples were taken from each vessel for moisture content determination. On day 30, material from control vessels was used for recovery determination. 2.6. Extraction of tylosin from excreta Excreta samples were extracted on days 0 and 30. Ten milliliters of refrigerated (41C) extractant solution (methanol:acetonitrile:0.1 M ascorbic acid, 45:45:10, v:v:v) was added to 3–5 g of solid sample in a 50-mL plastic centrifuge tube. The tube was capped, shaken, and the contents homogenized with a Brinkmann Polytron homogenizer for approximately 15–30 s. The tube was centrifuged at 1400g for 10 min and the clear supernatant decanted to a plastic beaker. A second 10mL aliquot of extractant was added to the tube and the process was repeated. Eighty milliliters of Milli-Q water (Millipore, a division of Waters Corporation, Milford, MA, USA) was added to the pooled supernatant and the solution was rested at room temperature for 10 min. A 1.00-mL aliquot of each raw extract was assayed by LSC after addition of 12 mL of Scintisafe Scintillation Cocktail (Fisher Scientific, Fair Lawn, NJ, USA). The remainders of the raw extracts were loaded on Sep-Pak C18 Plus SPE cartridges (Waters Corporation, Milford, MA, USA) that were preconditioned with 5 mL of methanol followed by 5 mL Milli-Q water applied under vacuum. The extracts were loaded at a flow rate of less than 6 mL/min. After loading, the SPE column was washed with 10 mL Milli-Q water followed by 10 mL of 20% acetonitrile in water (v:v). The columns were dried under vacuum for at least 3 min at a minimum of 127 mmHg. The columns were eluted with 2.5 mL of 5% acetic acid in methanol and the eluates dried under a gentle stream of air at 351C. Dried residues were redissolved in 2.00 mL of sample diluent (methanol:water, 70:30, v:v) with sonication and brief vortexing. The final extracts were filtered through 0.45-mm PTFE filters (Gelman Sciences, Ann Arbor, MI, USA) prior to HPLC analysis. Extracts were diluted as necessary to be within the calibration range of the analytical detector and in light of the anticipated sample levels. Calibration standards were prepared between 0.25 and 2.5 mg/mL. Aliquots (0.10 mL) of each final eluate were assayed by LSC. Samples of the solid residues from the extractions were combusted to determine radioactive material balance.
Table 1 Gradient HPLC conditions for tylosin assay Column Detector Flow rate Injection volume Column temperature Run time Mobile phase A Mobile phase B Gradient conditions:
250 mm 4.6 mm i.d. Spherisorb Phenyl, 5-mm particle size uv 280 nm 1.5 mL/min 100 mL 301C 30 min 1:1 (v:v) acetonitrile:water 0.02 M dibutyl ammonium phosphate Time %A %B 0.00 100 0 6.00 100 0 6.01 38 62 25.00 58 42 25.01 100 0 30.00 100 0
(HPLC). Details of the gradient separation and detection are listed in Table 1. Reference standard material was used for the quantitative determination of tylosin-A. Tylosin concentrations were determined by comparison of sample peak areas against a multipoint standard curve. Extraction and clean-up efficiency were assessed with control excreta (approximately 5 g) spiked with 150 mg tylosinA. The limit of detection for tylosin-A in extract solution was 0.08 mg/mL. The limit of quantitation for tylosin-A in extract solution was 0.25 mg/mL. Reference sample material was used for the qualitative identification of tylosin factors B, C, and D. Qualitative identification was based on coincident retention time. When injected as individual qualitative standards and assayed by HPLC, tylosin factors A, B, C, and D eluted from the analytical column between 11 and 15 min. Area and relative concentration (quantitative) determination for tylosin factors was based on the peak area response of tylosin-A reference standard material. 2.8. Combustion analysis Samples for total 14C were weighed onto quartz weigh boats and combusted for 4 min in a Harvey Biological Oxidizer (R.J. Harvey Instrument Corporation, Hillsdale, NJ, USA). The released 14CO2 was trapped in Harvey Carbon-14 Cocktail (R. J. Harvey Instrument Corporation, Hillsdale, NJ, USA) and measured by LSC.
3. Results
2.7. Analytical method (HPLC methods and extraction)
3.1. Moisture content
Analysis for tylosin was performed using reverse phase high performance liquid chromatography
The initial moisture content of the homogenized control and test excreta are provided in Table 2.
ARTICLE IN PRESS J. Scott Teeter, R.D. Meyerhoff / Environmental Research 93 (2003) 45–51 Table 2 Moisture content, 14C in volatile and 14CO2 traps at study conclusion, HPLC analytical values at days 0 and 30, total percent of initial tylosin degraded in 30 days
49
14
C material balance and
Matrix
Initial moisture Final moisture Percent of initial Day 30 HPLC Percent of initial Total material Day 0 HPLC 14 content (%) content (%) C found in traps balance (%) mean concentration mean concentration tylosin degraded of tylosin7std. in 30 days at study end (%) of tylosin7std. dev. (mg/g)a dev. (mg/g)a
Cattle Cattle control Chicken Chicken control Swine Swine control
82.3 82.2 75.0 65.3 75.9 77.0
85.1–85.8 83.0–85.8 84.4–85.5 77.6–80.9 76.7–77.7 59.3–72.0
o0.8 o0.006 o0.1 o0.01 o0.8 o0.005
104 na 94 na 101 na
119.5274.39 nd 35.0171.34 nd 62.8272.11 o13c
4.1670.69 nd o2.23b nd 4.1170.69 o1
96.5 na 493.6 na 93.5 na
Notes: nd— not detected; na—not analyzed. a n=6. b Calculated based on analytical limit of detection and dilution factor. c Most swine in the USA receive tylosin prophylactically. Therefore, tylosin residues in swine excreta are to be expected.
Changes in the final moisture content, primarily an increase in moisture content, relative to time zero is attributable to dried material deposited on the sides of the flasks. Moisture contents were kept relatively constant throughout the studies. 3.2.
14
C material balance
The test material vessels evolved 14C-volatile compounds corresponding to less than 0.8% of the initial 14 C contained in the test systems. Combustion analysis data for 14C in the solid samples on day 30 summed with the 14C evolved in the traps provided overall material balance figures of 104%, 94%, and 101% for cattle, chicken, and swine, respectively (Table 2). 3.3. HPLC analytical results Initial, day 0, concentrations of tylosin-A averaged 119.5274.39, 35.0171.34, and 62.8272.11 mg/g (n ¼ 6) on a dry weight basis for cattle, chicken, and swine samples, respectively (Table 2). On day 30, samples averaged 4.1670.69 and 4.1170.69 mg/g tylosin-A (n ¼ 6) for cattle and swine, respectively. No residues of tylosin or its factors were apparent in the chicken excreta samples after 30 days of incubation (with a limit of detection of 2.23 mg/g, Fig. 3). However, trace residues of possible tylosin factors were evident in the HPLC chromatograms on day 30 for the swine and cattle samples (Figs. 4 and 5) although the levels were very low. Calculated first-order degradation half-lives of tylosin-A in cattle, chicken, and swine excreta were 6.2 days, o7.6 days, and 7.6 days, respectively. Approximately 96.5%, 93.5%, and more than 93.6% of the initial tylosin-A was degraded in 30 days for the cattle, swine, and chicken excreta, respectively (Table 2).
Fig. 3. Representative HPLC chromatograms for tylosin in chicken excreta: (a) control day-0, dilution factor (DF) 1 , (b) day-0 test sample, DF 50 , and (c) day-30 test sample, DF 40 .
ARTICLE IN PRESS J. Scott Teeter, R.D. Meyerhoff / Environmental Research 93 (2003) 45–51
50
14
Unextractable bound residues increased slightly as measured by combustion at 30 days. As shown in Table 3, day-0 extraction recoveries for 14C averaged 79% for the cattle excreta and 76% for the swine excreta as compared to the total sample combustion values. Day30 extraction recoveries for 14C compounds from the excreta were largely efficient, ranging between average values of 63% and 94% as compared to the total sample combustion values (Table 3). Day-0 residual 14C concentrations in the extracted solid ranged from 19.0% to 21.6% as compared to the combustion values for the cattle and swine excreta, respectively. Day-30
residual 14C concentrations in the extracted solid averaged 53.2%77.0%, 16.4%71.0%, and 36.4%71.0% for the cattle, chicken, and swine excreta samples, respectively. When compared to day 0, these data suggest that the solid retains more of the 14C material over time. This may be in part due to incorporation of breakdown products into the matrix or extraction inefficiency as sorption occurs over the duration of the study. On day 0, 80.1%72.6% and 81.3%71.2% of the 14C that was in the bulk extract was recovered in the final extract after SPE clean up for the cattle and swine excreta, respectively. After 30 days of incubation, the values were 51.2%79.5% and 15.5%74.0% for the
Fig. 4. Representative HPLC chromatograms for tylosin in swine excreta: (a) control day-0, dilution factor (DF) 30 , (b) day-0 test sample, DF 30 , and (c) day-30 test sample, DF 2 .
Fig. 5. Representative HPLC chromatograms for tylosin in cattle excreta: (a) control day-0, dilution factor (DF) 30 , (b) day-0 test sample, DF 30 , and (c) day-30 test sample, DF 2 .
3.4. Extract and final sample
Table 3 Percentage of
C residues
14
C extracted and measured by LSC as compared to time-zero Cattle
Average Std.dev. %RSD
14
C as determined by combustion
Swine
Chicken
Day 0
Day 30
Day 0
Day 30
Day 0
Day 30
79.05 2.27 2.87
63.12 4.04 6.40
75.55 2.63 3.47
71.18 2.75 3.86
na
93.88 6.48 6.90
Notes: na—not applicable, samples were not assayed to provide these values; n ¼ 6 for all values.
ARTICLE IN PRESS J. Scott Teeter, R.D. Meyerhoff / Environmental Research 93 (2003) 45–51
cattle and swine excreta samples, respectively. Although not measured on day 0, after 30 days of incubation, the value was 2.2%70.1% for the chicken excreta. This data indicates that initially, a high level of 14C material was attributable to tylosin and was recovered in the final extracts. With aging, the 14C residues, no longer being structurally or chemically similar to tylosin, are isolated out of the day 30 final extracts by the SPE procedure. This result parallels the analytical values for tylosin as measured by HPLC.
51
the transient nature of tylosin in the environment by analyzing 104 surface water samples from sites where tylosin might be found. Only 13.5% of the samples assayed contained tylosin, and those only had concentrations ranging from 0.04 to 0.28 mg/L. While the studies presented here were conducted under controlled laboratory conditions, seasonal and temporal variations in the field may occur and effect the rate of degradation. However, microflora induction effects from treated animals may also provide for more rapid degradation than was observed in these studies.
4. Discussion/conclusions Veterinary drugs, including tylosin, may enter the soil through the feces of treated animals (van Gool et al., 1993). Baguer et al. (2000) investigated the effects of tylosin on soil fauna (earthworms, springtails, and enchytraeids) and concluded that no effects were observed at levels as high as 3000 mg/kg. In another study, designed to measure the biodegradability of tylosin in soil–manure slurries, the authors concluded that tylosin and its degradation products disappeared rapidly (Ingerslev and Halling-Sorensen, 2001). Degradation half-lives for tylosin in soil–manure slurries ranged between 4.1 and 8.1 days. The authors also concluded that mineralization of tylosin and its degradation products must occur rapidly since their study methodologies were designed to cover and detect a wide range of possible metabolites. Loke et al. (2000) observed similar results in both aerobic and methanogenic manure systems, concluding that the degradation half-lives of tylosin in the aqueous phase were less than 2 days. Similarly, tylosin degraded rapidly in a composting trial (Vogtmann et al., 1984). In compost containing swine and chicken litter, tylosin at initial levels of 22.4 and 224.0 ppm was not detected after 75 days. The results presented here further indicate that tylosin residues in cattle, chicken, and swine excreta degrade rapidly. Measured half-lives for aerobic degradation are 6.2 days, o7.6 days, and 7.6 days for cattle, chicken, and swine excreta, respectively. These rapid half-lives indicate that tylosin residues will not be persistent in the environment. This is also confirmed by Gavalchin and Katz (1994), who conducted soil degradation studies at 41C, 201C, and 301C and reported that no tylosin was found after 30 days incubation at 201C and 301C, and only 40% of the initial tylosin was recovered at 41C. Kolpin et al. (2002) recently confirmed
Acknowledgment Special thanks to Alison N. Perkins for her detailed review and editorial comments on this manuscript.
References Baguer, A.J., Jensen, J., Krogh, P.H., 2000. Effects of the antibiotics oxytetracycline and tylosin on soil fauna. Chemosphere 40, 751–757. Elanco, C., 1982. Tylan Injection Veterinarian’s Technical Handbook. Elanco Products Company Indianapolis, IN. Eirkson, C., Harrass, M.C., Osborne, C.M., Sayre, P.G., Zeeman, M., 1987. Aerobic biodegradation in water. In: Environmental Assessment Technical Assistance Handbook. US FDA, Washington, 3.11. PB87–175345. Gavalchin, J., Katz, S.E., 1994. The persistence of fecal-borne antibiotics in soil. J. Assoc. Anal. Commun. Int. 77 (2), 481–485. Ingerslev, F., Halling-Sorensen, B., 2001. Biodegradability of metronidazole, olaquindox, and tylosin and formation of tylosin degradation products in aerobic soil-manure slurries. Ecotoxicol. Environ. Saf. 48, 311–320. Kolpin, D.W., Furlong, E.T., Meyer, M.T., Thurman, E.M., Zaugg, S.D., Barber, L.B., Buxton, H.T., 2002. Pharmaceuticals, hormones, and other organic wastewater contaminants in us streams 1999–2000: a national reconnaissance. Environ. Sci. Technol. 36, 1202–1211. Loke, M.L., Ingerslev, F., Halling-Sorensen, B., Tjornelund, J., 2000. Stability of tylosin a in manure containing test systems determined by high performance liquid chromatography. Chemosphere 40, 759–765. McGuire, J.M., Boniece, W.S., Higgens, C.E., Hoehn, M.M., Stark, W.M., Westhead, J., Wolfe, R.N., 1961. Tylosin, a new antibiotic, I. Microbiological studies. Antibiot. Chemother. 11 (5), 320–327. van Gool, S., 1993. Possible environmental effects of antibiotic residues in animal manure. Tijdschr. Diergeneeeskd. 8–10 (Dutch English Summary). Vogtmann, H., Fragstein, P.V., Draeger, P., 1984. The degradation of agrochemicals during composting. Peat Agriculture Horticulture Proceedings Second International Symposium, pp. 357–378.