Comparison of Lateral Flow Assay, Kidney Inhibition Swab, and

of Agriculture, 1605 Albrecht Boulevard, Fargo, North Dakota 58102-2765, United States ... For a more comprehensive list of citations to this arti...
0 downloads 0 Views 4MB Size
Article pubs.acs.org/JAFC

Comparison of Lateral Flow Assay, Kidney Inhibition Swab, and Liquid Chromatography−Tandem Mass Spectrometry for the Detection of Penicillin G Residues in Sow Urine Weilin L. Shelver,* Shubhashis Chakrabarty,† and David J. Smith Agricultural Research Service, Biosciences Research Laboratory, United States Department of Agriculture, 1605 Albrecht Boulevard, Fargo, North Dakota 58102-2765, United States ABSTRACT: Sows (n = 126) were administered penicillin G; urine, collected at slaughter, was screened by kidney inhibition swab (KIS; 4 h testing time) and then stored at −80 °C (∼1200 days) until analysis by lateral flow assay (LF, ∼5 min testing time) and tandem quadrupole LC-MS/MS (TQ) analysis. The stability of penicillin in urine during storage was verified using TQ analyses. Quantitative results were well-correlated (R2 = 0.98) with only a ∼10% decrease in penicillin concentration during the 3-year storage period. KIS retesting of stored samples returned results consistent with the original analyses. Lateral flow assay results were highly correlated with the KIS and TQ results. A KIS positive sample, which was not confirmed by TQ or LF, was assayed by Triple-TOF LC-MS to determine the cause of the apparent false positive. This study suggests LF can be used to quickly and efficiently screen for penicillin G residues before slaughter. KEYWORDS: Penicillin G, sows, screening test, antibiotics, residues



INTRODUCTION

producers to mitigate risks of marketing penicillin-treated animals having violative residues.5 Lateral flow tests have been often used for on-site screening of drug residues and may return results within 10 min or less when used properly. For example, over 3.5 million screening assays were used to evaluate antibiotic residues in bulk-tank milk during 2015, and a majority of the assays utilized lateral flow strip tests.6 These rapid screening assays have been successfully incorporated into dairy production so that the incidence of drug residues in bulk-tank milk is rare. As an example, out of 3.61 million samples analyzed representing 3.73 million tests, only 579 (∼1 in 6 200 samples) tested positive for drug residues.6 In previous reports,3,5 we have established that urine is an excellent matrix for the prediction of violative penicillin residues in swine kidney when the KIS assay is used as a rapid screening assay. Implementation of the KIS microbial inhibition assay at swine production facilities, however, may be low because specialized equipment is needed, technical training is required, and results are not available for 4−5 h after urine collection. The purpose of this study was to determine the accuracy of simple, rapid, and inexpensive penicillin G lateral flow assays to detect penicillin residues in swine urine using samples collected by Lupton et al.5 during a large (n = 126 sows) residue depletion study. Because the live-phase aspect of the residue depletion trial was conducted in 2012 and the lateral flow assay experiments were carried out in 2015, we also conducted studies to verify penicillin stability under our storage conditions and to determine the repeatability of the KIS assay after prolonged matrix storage. The results obtained from the KIS tests and lateral flow assays were verified using quantitative

Penicillin G procaine is indicated for treatment of a number of bacterial diseases in a variety of food-animal species. When used under label conditions in the United States, approved preslaughter withdrawal periods are 14 days for cattle, 9 days for sheep, and 7 days for swine with kidney tolerances in cattle of 50 ppb.1 Because penicillin G approvals for use in swine were granted using a microbial inhibition assay (which measures total antimicrobial residues) to measure residue depletion,1 the U.S. Food and Drug Administration, Center for Veterinary Medicine, has not established a tolerance for specific penicillin residues in swine tissues.2 Thus, for animals treated “off-label” (i.e., nonlabel dose, alternate route of administration, duration of dosing, etc.) with penicillin G under the auspices of the Animal Medicinal Drug Use Clarification Act, any detectable penicillin G residue in any edible tissue of swine is considered to be violative. Shelver et al.3 described the detection of penicillin G residues in kidney, liver, plasma, urine, skeletal muscle, and injection site of heavy sows using a microbial inhibition assay (KIS) following extra-label penicillin G procaine administration. In that study,3 sows were treated intramuscularly (IM) with 33 000 IU/kg of body weight (bw) (5× label dosage) of penicillin G procaine for three consecutive days and were slaughtered with withdrawal periods extending to 39 days posttreatment. The KIS rapid screening assay, employed by the Food Safety and Inspection Service (FSIS) to screen kidney tissues at slaughter facilities, accurately predicted penicillin G positive kidney samples. The presence of urinary penicillin G residues was an excellent predictor of kidney penicillin G residues greater than the 25 ng/g screening level in use by USDA FSIS.4 Such data suggested that the kidney inhibition swab test, when used with a urine matrix, could be a viable option for the timely ( 176 and m/z 335 > 160 for penicillin G, and m/z 342 > 183 and m/z 342 > 160 for d7-penicillin G. All samples were analyzed in duplicate. Triple TOF LC-MS(/MS) Unknown Identification. A Shimadzu Nexera XR liquid chromatograph (Kyoto, Japan) interfaced with an AB Sciex 5600+ mass spectrometer (Framingham, MA) was used for identification of unknowns. An Acquity BEH C18 column was used for the LC separation of penicillin G, penicillin G metabolites, and other compounds by using a gradient of 5% acetonitrile/0.1% formic acid (solvent A) and acetonitrile/0.1% formic acid (solvent B) from 5% B to 90% B over 18 min at 0.2 mL/min and an injection of 10 μL sample aliquots. Electrospray ionization (DuoSpray Ion Source, AB Sciex) was used for ionization in positive ion mode with capillary voltage of 4.5 kV. A probe temperature of 400 °C was used along with ion source gases I and II and curtain gas, all at 40 psi of nitrogen. The declustering potential used for these experiments was 20 V. For MS/MS, a collision energy of 35 V was used. The data were analyzed by AB Sciex Analyst and Peak View software. Statistical Analysis. The analysis was carried out using Statistical Analysis System by PROBIT procedure with a logistic distribution to fit an S-shaped curve to the data (SAS 9.4, SAS Institute, Inc., Cary, NC). The equation from the fit was used to compute the concentration at which there was a 90% probability of a positive response indicating the presence of penicillin G. The 95% confidence limits were also computed from the data to determine the concentration that would give a positive response 90% of the time.

Each test was read independently by two scorers blinded with respect to sample identification. Sample Preparation and TQ LC-MS/MS Analysis. Fortyfive percent of urine samples (25 of 56) from the original study,5 having penicillin residues greater than the assay limit of quantitation (4.1 ng/mL, as described by Lupton et al.5) but chi square < 0.0001 and a log likelihood of −19.88, indicating a reasonable data fit. Using the fit parameters, a 90% positive detection limit was calculated to be 3.6 ng/mL with a 95% confidence limit of 3.3 to 4.1 ng/mL. When accounting for dilution, this would represent a nominal detection limit for the lateral flow assay in undiluted urine of 12 ng/mL. The qualitative results we obtained from use of the lateral flow assay with urine were highly consistent with the KIS test results with few exceptions (Table 3). For instance, a single urine sample (withdrawal day 15, trial 2, treatment 1) was scored as positive by the KIS test whereas the lateral flow test returned negative results using both a visual analysis and the automated reader (data highlighted in blue; Table 3). For this sample, no quantifiable penicillin G was obtained by LC-MS/ MS, nor was the presence of penicillin G detected by the lateral flow assay, yet both the 2012 and 2015 KIS assays returned positive results. Such an outcome suggests the presence of antibacterial activity in the absence of the penicillin G molecule. Unidentified microbial inhibitor (UMI) results from KIS assays do occur. For example, FSIS reported 6 UMIs in the 2013 residue sample results10 and 4 UMIs in the 2014 residue sample results11 using KIS assays. Further investigation using triple-TOF LC-MS(/MS) to search for unknown compounds in the urine sample did not identify known antibiotics but confirmed the presence of penilloic acid (Figure 3), indicative that this particular sow had received penicillin G. Penicillin G Stability in Urine. Stability of penicillin during storage has been reported to be temperature- and matrixdependent. For example, Boison et al.12 showed that benzylpenicillin was relatively stable in incurred plasma samples stored at −20 °C for 2 months, with 10−20% of the residue

Figure 4. Concentrations of penicillin G residues in sow urine samples quantified when collected, in 2012 (x-axis), and after 3 years of storage at −80 °C, in 2015 (y-axis).

residues of penicillin G were relatively stable after storage for ∼3 years at −80 °C. To our knowledge, this is the first stability study with urine and the longest stability test of penicillin G stored under ultracold conditions that has been reported. In addition to the highly consistent results between the 2012 and 2015 KIS analyses, results obtained after TQ LC-MS/MS analysis of urine samples in 2012 and 2015 using the same instrument and analytical method (Figure 4) were highly correlated (R2 = 0.98). The slope of the correlation between the quantitative results obtained in 2012 and 2015 was 0.9, suggesting a maximum degradation of 10%. E

DOI: 10.1021/acs.jafc.6b05049 J. Agric. Food Chem. XXXX, XXX, XXX−XXX

Article

Journal of Agricultural and Food Chemistry Payne et al.14 suggested that screening animals for penicillin residue prior to marketing would help to ensure that violative residues would not occur at harvest. Lupton et al.5 showed that penicillin residues are substantially more concentrated in urine than in kidneys at the time of slaughter. Thus, the presence of penicillin G in urine is an excellent predictor that the antibiotic has not been completely depleted from tissues. Previously, we have shown that a bacterial inhibition swab test (KIS test), employed by the FSIS to screen kidney tissues at slaughter facilities, accurately predicted KIS-positive kidney samples when used with urine. That is, positive urinary KIS results were an excellent predictor of kidney penicillin residues greater than the 25 ng/g screening level in use by USDA FSIS.4 In the current study, we have established that, with the use of a lateral flow assay, it is possible to rapidly (∼5 min) and accurately determine the presence of penicillin G in urine. We believe that the use of such an assay at production sites could substantially reduce marketing of penicillin-treated animals that harbor violative kidney penicillin residues.



(6) GLH Incorporated. National milk drug residue data base. Fiscal year 2015 annual report: October 1, 2014−September 30, 2015. www. kandc-sbcc.com/nmdrd/fy-14.pdf (accessed 9/15/2016). (7) Salter, R. S.; Douglas, D.; McRobbie, L.; Quintana, J.; Legg, D.; Schwartz, J.; Conaway, D.; McPhee, C.; Saul, S.; Markovsky, R. Validation of the charm 3 SL3 beta-lactam test for screening raw milk in compliance with the U.S. pasteurized milk ordinance. Performance Tested Method 071002. J. AOAC Int. 2011, 94, 348−357. (8) Apley, M.; Coetzee, H.; Gehring, R.; Karriker, L. Pharmacokinetics and tissue residues of procaine penicillin G in sows after administration of 33,000 IU/kg intramuscularly and by needle-free injection in the hip. National Pork Board Research Report NPB no. 07-234; 2009. (9) US FSIS. Inhibition screen test for antimicrobial drugs. CLG R41. http://www.fsis.usda.gov/wps/wcm/connect/74e06adb-9c8e-477f8ecf-9e76bd03f4bc/CLG-ADD3.pdf?MOD=AJPERES (accessed 3/ 21/2016). (10) US FSIS. U.S. National Residue Program: 2013 Residue Sample Results. http://www.fsis.usda.gov/wps/wcm/connect/0de80b0c05d5-48f6-b494-8a361d2de0c4/2013-Red-Book.pdf?MOD=AJPERES (accessed 9/15/2016). (11) US FSIS. U.S. National Residue Program: 2014 Residue Sample Results. http://www.fsis.usda.gov/wps/wcm/connect/2428086b-f8ec46ed-8531-a45d10bfef6f/2014-Red-Book.pdf?MOD=AJPERES (accessed 9/15/2016). (12) Boison, J. O.; Korsrud, G. O.; MacNeil, J. D.; Yates, W. D. G.; Papich, M. G. Effect of cold-temperature storage on stability of benzylpenicillin residues in plasma and tissues of food-producing animals. J. AOAC Int. 1992, 75, 974−978. (13) Berti, M. A.; Maccari, M. Stability of frozen rat plasma containing different antibiotics. Antimicrob. Agents Chemother. 1975, 8, 633−637. (14) Payne, M. A.; Craigmill, A.; Riviere, J. E.; Webb, A. I. Extralabel use of penicillin in food animals. J. Am. Vet. Med. Assoc. 2006, 229, 1401−1403.

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Tel.: 701-239-1425. ORCID

Weilin L. Shelver: 0000-0002-4721-9945 Present Address †

ORISE Post-Doctoral Research Fellow at the Animal Metabolism-Agricultural Chemicals Research Unit, USDA, ARS, Biosciences Research Laboratory, Fargo, ND, 58102 (S.C.).

Notes

Names are necessary to report factually on available data; however, the USDA neither guarantees nor warrants the standard of the product, and the use of the name by USDA implies no approval of the product to the exclusion of others that may also be suitable. USDA is an equal opportunity provider and employer. The authors declare no competing financial interest.



ACKNOWLEDGMENTS The authors wish to thank Dr. Mark West for statistical test consultation. The skillful technological support provided by Kira Rahn, Amy McGarvey, and Jason Holthusen is greatly appreciated.



REFERENCES

(1) Freedom of information summary, supplemental new animal drug application, Norocillin, Penicillin G procaine, injectable suspension for cattle, sheep, swine, and horses; NADA 065-010; Approval date: April 23, 2010. (2) 21 CFR Section 556.510(b). (3) Shelver, W. L.; Lupton, S. J.; Newman, D. J.; Larsen, S.; Smith, D. J. Depletion of penicillin G residues in heavy sows after intramuscular injection. Part II: application of kidney inhibition swab tests. J. Agric. Food Chem. 2014, 62, 7586−7592. (4) US FSIS. Screening and confirmation of animal drug residues by UHPLC-MS-MS. CLG R45. http://www.fsis.usda.gov/wps/wcm/ connect/b9d45c8b-74d4-4e99-8eda-5453812eb237/CLG-MRM1. pdf?MOD=AJPERES (accessed 9/15/2016). (5) Lupton, S. J.; Shelver, W. L.; Newman, D. J.; Larsen, S.; Smith, D. J. Depletion of penicillin G residues in heavy sows after intramuscular injection. Part I: tissue residue depletion. J. Agric. Food Chem. 2014, 62, 7577−7585. F

DOI: 10.1021/acs.jafc.6b05049 J. Agric. Food Chem. XXXX, XXX, XXX−XXX