Article pubs.acs.org/JAFC
Determination of Hepatotoxic Indospicine in Australian Camel Meat by Ultra-Performance Liquid Chromatography−Tandem Mass Spectrometry Eddie T. T. Tan,†,§ Mary T. Fletcher,*,† Ken W. L. Yong,† Bruce R. D’Arcy,‡ and Rafat Al Jassim‡ †
Queensland Alliance for Agriculture and Food Innovation (QAAFI), The University of Queensland, Health and Food Sciences Precinct, Coopers Plains, QLD 4108 Australia ‡ School of Agriculture and Food Sciences, Faculty of Science, The University of Queensland, Gatton, QLD 4343 Australia § Food Technology Department, Faculty of Applied Sciences, Universiti Teknologi MARA, Shah Alam, Malaysia ABSTRACT: Indospicine is a hepatotoxic amino acid found in Indigofera plant spp. and is unusual in that it is not incorporated into protein but accumulates as the free amino acid in the tissues (including muscle) of animals consuming these plants. Dogs are particularly sensitive to indospicine, and secondary poisoning of dogs has occurred from the ingestion of indospicinecontaminated horse meat and more recently camel meat. In central Australia, feral camels are known to consume native Indigofera species, but the prevalence of indospicine residues in their tissues has not previously been investigated. In this study, a method was developed and validated with the use of ultra-performance liquid chromatography−tandem mass spectrometry (UPLC−MS/MS) to determine the level of indospicine in camel meat samples using isotopically labeled indospicine as an internal standard. UPLC−MS/MS analysis showed that the method is reproducible, with high recovery efficiency and a quantitation limit of 0.1 mg/kg. Camel meat samples from the Simpson Desert were largely contaminated (≈50%) by indospicine with levels up to 3.73 mg/kg (fresh weight) determined. However, the majority of samples (95%) contained less than 1 mg/kg indospicine. KEYWORDS: indospicine, Indigofera, camel, Camelus dromedarius, ultra-performance liquid chromatography−tandem mass spectrometry
■
in the tissues of animals fed Indigofera plant material.9 Such residues have been shown to be highly residual and persist for several months after the cessation of feeding in horses,10 goats,11 cattle,12 and rabbits.13 Indospicine (2) has been shown experimentally to cause liver damage in a range of animals, including mice,2 rats,14 and dogs.15 Dogs are particularly sensitive to indospicine hepatotoxicity, and of concern, from a food safety viewpoint, is that fatalities have been reported when naturally contaminated meat was used as pet food. Secondary poisoning of dogs has occurred from the ingestion of indospicine-contaminated horse meat10 and, more recently, camel meat16 from animals that had grazed in regions where Indigofera species are prevalent. The population of feral camels in Australia has been estimated at 1 million camels, predominantly present in arid inland regions, with a predicted population growth rate of 7− 8% per annum (an increase of approximately 80 000 feral camels per year).17 Investigations of dietary preferences of camels in arid central Australia have provided evidence of the consumption range of Indigofera spp.18 Of 243 plant species observed to be consumed by camels in this region, six Indigofera species were noted, including I. colutea, I. georgei, I. helmsii, and I. hirsuta, which were assessed to be the “main food plant at
INTRODUCTION
In Australia there are more than 40 species of Indigofera, both native and introduced, with a number of these known to contain hepatotoxin indospicine (2) (Figure 1), including Indigofera circinella, Indigofera colutea, Indigofera hendecaphylla, Indigofera hirsuta, Indigofera linnaei, Indigofera suffruticosa, Indigofera spicata, and Indigofera trita.1−8 The amino acid indospicine (2) is an arginine (1) (Figure 1) analogue and is unusual in that it is not incorporated into proteins but is present as the free amino acid and accumulates in the free form
Received: Revised: Accepted: Published:
Figure 1. Chemical structure of arginine (1); indospicine (2), an analogue of arginine; and D3-L-indospicine (3), an internal standard. © 2014 American Chemical Society
1974
November 21, 2013 January 15, 2014 January 16, 2014 January 16, 2014 dx.doi.org/10.1021/jf4052495 | J. Agric. Food Chem. 2014, 62, 1974−1979
Journal of Agricultural and Food Chemistry
Article
Figure 2. Tandem mass spectrometry fragmentation spectra of (A) indospicine and (B) D3-L-indospicine. Indospicine Extraction. Prior to analysis, Amicon Ultra 0.5 mL, 3K centrifugal filter units (Merck Millipore, Kilsyth, Australia) were prerinsed and centrifuged (10 000 rpm, 20 min) with reverse osmosis water (2 × 300 μL) to remove glycerine and then inverted and spun for 1 min at 1000 rpm. Camel muscle samples were thawed, finely chopped, weighed (0.5 g), and mixed with 0.1% heptafluorobutyric acid (HFBA, 25 mL), follow by homogenization with a Polytron T25 Basic homogenizer (Labtek, Brendale, Australia) for 15 s. The homogenized samples were chilled (4 °C) for 20 min and then centrifuged at 4500 rpm for 20 min at 18 °C. Aliquots of supernatants (1.0 mL) spiked with the internal standard D3-L-indospicine (3) (1 mg/L in 0.1% HFBA, 100 μL) were vortexed for 10 s. A portion (450 μL) was transferred into prerinsed Amicon Ultra 0.5 mL, 3K centrifugal filters, which were centrifuged (10 000 rpm, 20 min), with the filtrates transferred to a limited volume insert for UPLC− MS/MS analysis. UPLC−MS/MS Equipment and Conditions. Sample data were collected by MassLynx 4.1 and processed by TargetLynx application managers. Separation of indospicine was done by a Waters ACQUITY UPLC system liquid chromatography (Waters, Rydalmere, Australia). LC separations were performed using a 100 mm × 2.1 mm i.d., 1.7 μm, BEH C18 column (Waters, Rydalmere, Australia) at 30 °C with flow rate of 0.2 mL/min. The mobile phase was a mixture of (A) H2O with 0.1% HFBA (v/v) (pH 2.15) and (B) acetonitrile with 0.1% HFBA, which were prepared freshly for each analysis. The following gradient was used: 0 min, 99% A; 4 min, 70% A; 7 min, 70% A; 8 min, 99% A; 10 min, 99% A. MS/MS detection was made using a Waters Micromass Quattro Premier triple quadrupole mass spectrometer with an electrospray ionization (ESI) source (Waters, Rydalmere, Australia), operated in the positive mode. Eluted indospicine (2) was quantitated by utilizing selected reaction monitoring (SRM) transitions of m/z 174.2 → 111.0 (verified by transition of m/z 174.2 → 157.1) for indospicine (2) and m/z 177.1 → 114.0 (verified by transition of m/z 177.1 → 113.0) for D3-L-indospicine (3) as internal standard. The capillary voltage was 2.79 kV, cone gas flow was 50 L/h, and desolvation gas flow was 600 L/h. The source and desolvation temperatures were set at 150 and 350 °C, respectively. The argon gas collision energies for indospicine (2) (15 and 12 eV) and D3-L-indospicine (3) (15 and 15 eV) were set with cone voltage at 25 V. Synthesized indospicine (2) (external standard) and D3-Lindospicine (3) (deuterium-labeled internal standard, >99% pure) were used in UPLC−MS/MS analysis. Internal (1 mg/L) and external standard (0.002−2 mg/L) solutions for indospicine (2) were prepared in H2O with 0.1% HFBA. Solutions were frozen for no longer than 1 month.
times”, and I. linnaei and I. basedowii, which are“preferred food plants”.18 Short-lived mass developments of forbs were reported to be seasonally utilized by camels, including Indigofera species in summer. Despite the reported dog poisoning incident,16 the extent of indospicine (2) contamination of camel meat in Australia remains largely unknown. Consequently, this study was undertaken to develop and validate a high-throughput ultra-performance liquid chromatography−tandem mass spectrometry (UPLC−MS/MS) method to quantitate indospicine (2) in camel meat tissue. The validated method was then used to determine the concentrations of indospicine (2) in a number of collected camel meat samples obtained from a high camel density area of Australia, the Simpson Desert in the Northern Territory of Australia. Indospicine (2) has previously been measured in plant material by amino acid analysis,1,19,20 HPLC,4,9 and, more recently, LC−MS/MS with phenylisothiocyanate derivitization.21 Analysis of indospicine (2) in meat is more challenging due to the lower levels of toxin and the complex matrix. Indospicine (2) in animal tissues have previously been analyzed by lengthy (nearly 2 h) amino acid analysis1,22 and by phenylisothiocyanate derivatization and UV detection, with resolution from other amino acids being problematic.9 The incorporation of D3-L-indospicine (3) (Figure 1) in this present study overcomes matrix effects noted in our previous LC−MS/ MS analysis of camel tissues.16
■
MATERIALS AND METHODS
Indospicine Standards. Synthesized indospicine (2) and D3-Lindospicine (3) were provided by Dr. James De Voss and Dr. Robert Lang, The University of Queensland. Meat Materials. A total of 99 camels were sampled as part of a cull exercise under the Australian Feral Camel Management Project23 in the Simpson Desert in central Australia. Rump muscle samples were excised from culled camels, placed in specimen containers, frozen, and transported to Health and Food Science Precinct, Brisbane, Australia. All of the samples were stored frozen (−30 °C) prior to extraction and analysis. Camel muscle samples free from indospicine (2) contamination (used for negative control and spike samples) were obtained from a pet food supplier, and these samples had been sourced from 200 km northwest of Rawlinna in Nullabor Plain, West Australia (an area relatively free from Indigofera species),24 and stored frozen. 1975
dx.doi.org/10.1021/jf4052495 | J. Agric. Food Chem. 2014, 62, 1974−1979
Journal of Agricultural and Food Chemistry
Article
Accurate Mass Analysis. Accurate mass analysis of both parent and daughter ions was performed on an AB SCIEX TripleTOF 5600+ system (AB SCIEX, Mt. Waverley, Australia) via direct injection and fragmentation of standard solutions of both indospicine and D3-Lindospicine.
camel meat extracts (spiked samples and Simpson Desert samples) were determined by interpolating the relative intensity with a calibration curve. Calibration curves were constructed for each batch and involved plotting the ratio of the intensity of the positive-ion ESI SRM transition m/z 174 → 111 relative to the internal standard (D3-L-indospicine, 3) SRM transition m/z 177 → 114, versus the indospicine (2) concentrations for standard solutions of 0.002−2.00 mg/L. Seven-point calibration curves with the overlaid chromatograms of internal and external standards are shown in Figure 3.
■
RESULTS AND DISCUSSION UPLC−MS/MS Fragmentation of Indospicine and D3L-indospicine. Fragmentation of the protonated molecular ion of indospicine (2) ([M + H]+ m/z 174) in positive-ion ESI mode produces a number of daughter ions, notably m/z 157 (loss of NH3), m/z 129 (loss of NH3 and CO), m/z 111 (loss of NH3 and HCO2H), and m/z 84 (loss of NH3, CO2H, and HCN), as seen in Figure 2. Accurate mass MS and MS/MS data confirmed the composition of each of the daughter ions, particularly the perhaps unexpected composition of m/z 129 as C6H13N2O (observed 129.1029, calculated 129.1028). We postulate then that the required loss of NH3 and CO occurs via loss of NH3 from the amidine group followed by formation of a cyclic lactone intermediate, from which it is possible to lose CO while retaining the second oxygen atom. In the labeled standard D3-L-indospicine (3), deuterium atoms are incorporated at C-5 (two deuterium) and C-6 (one deuterium), and this is evident in the fragmentation of the protonated molecular ion [M + H]+ m/z 177, which produces corresponding daughter ions of m/z 160/159 (loss of NH3/ NH2D), m/z 132/131 (loss of NH3/NH2D and CO), m/z 114/113 (loss of NH3/NH2D and H2CO2H), and m/z 86/87 (loss of NH3/NH2D, H2CO2H, and HCN), as seen in Figure 2. The fragmentation to almost equal proportion of each of these ion pairs is consistent with the proposed initial loss of NH2 from the terminal amidine group with concomitant loss of either H or D from the monodeuterated C-6 position. Accurate mass MS and MS/MS data again confirmed that the composition of each of the daughter ions is consistent with the listed mass losses. The most dominant daughter ion observed at m/z 111 was chosen as the transition (m/z 174 → 111) for quantitation of indospicine (2) using the selected reaction monitoring (SRM) program on the UPLC−MS/MS, with the m/z 157 daughter ion selected as the transition (m/z 174 → 157) for verification. The efficiency of the ionization and the detection of the precursor and fragments ions were optimized by adjusting collision energy (15 and 12 eV) and cone voltage (25 V). Similarly, the transition m/z 177 → 114 was selected for SRM quantitation of D3-L-indospicine (3), with the transition m/z 177 → 113 for verification with optimized collision energy and cone voltage (15 eV and 25 V, respectively). Development and Validation of a UPLC−MS/MS Method for Indospicine in Camel Meat Samples. The UPLC−MS/MS method reported here is a high-throughput yet sensitive analysis method for indospicine (2) in camel meat tissue, and this is the first reported indospicine analysis in unprocessed camel meat tissue based on the developed and validated UPLC−MS/MS method. The method involves a simple tissue extraction into aqueous solution followed by physical deproteinisation using centrifugal ultrafiltration before UPLC−MS/MS detection. The described UPLC separation on BEH C18 column eluting with water and acetonitrile (0.1% HFBA) gradients provides good peak shape and column retention. Heptafluorobutyric acid (HFBA) is an effective ionpairing agent that is widely used in reverse-phase peptide/ amino acid separations.25 Concentrations of indospicine (2) in
Figure 3. Overlaid SRM chromatograms of indospicine standards (0.002−2.00 mg/L) and internal standard (A) monitored by UPLC− MS/MS with ESI.
The limit of quantitation (LOQ) determined for this work was at least 9:1 for the peak-to-peak signal-to noise ratio for the quantitation transition (m/z 174 → 111). Both quantitation and qualification ions were monitored for each analysis. The average ratio of the quantitation peak to that of the qualification peak for spiked samples (n = 21) was 4.49. Confirmation of indospicine (2) was also carried out by examining the quantitation/qualification ion ratio, with the requirement that the ratio be within ±20%. Hence, only samples that have the ratio of 3.59 to 5.39 are considered to contain actual indospicine (2). The calibration curves (n = 8) demonstrated good linearity with R2 values of 0.9960 ± 0.004 throughout the analysis. Recovery efficiency and reproducibility of this analytical method were demonstrated by replicate analyses (n = 7) of the matrix spiked at three concentration levels: 2.5, 5.0, and 50.0 mg/kg indospicine-free camel meat. The spiked matrices were then extracted by the method described in indospicine (2) extraction. These spiking studies had recoveries of 106, 109, and 102%, respectively. The mean concentrations (standard deviation (SD), relative standard deviation (RSD)) of seven analyses for each spiked sample were 2.65 mg/kg (SD 0.24, RSD 9.06%), 5.43 mg/kg (SD 0.42, RSD 7.82%), and 51.01 mg/kg (SD 5.02, RSD 9.85%), respectively. The relative standard deviation was in all cases below 10% for three spiked samples, and this indicated that the method is reproducible with good precision. A significant clinical challenge with UPLC−MS/MS is the potential for matrix effects that cause interferences or impact ionization efficiency. The matrices of 1976
dx.doi.org/10.1021/jf4052495 | J. Agric. Food Chem. 2014, 62, 1974−1979
Journal of Agricultural and Food Chemistry
Article
Figure 4. UPLC−MS/MS analysis of (A) individual and (B) overlaid SRM chromatograms for indospicine: (a) 0.05 mg/L standard, (b) 2.5 mg/L spiked sample, (c) Simpson Desert camel meat sample (SD_79) (1.41 mg/L), and (d) blank sample.
the extract can significantly (p ≤ 0.05) suppress the indospicine and isotope-labeled internal standard responses. However, the use of stable isotope-labeled internal standards in this study to compensate for matrix effects and to increase accuracy resulted in the quantitation of matrix and matrix-free extracts having no substantial differences. Reagent spike recovery efficiency was not done due to this internal standard compensation principle. A blank matrix sample, spiked matrix samples, and spiked check samples were analyzed for each batch of analyses as part of the quality control for the validity of the analyses. The validity of indospicine (2) present in camel meat samples was ensured by spiked check samples being reported within 70% ≤ spike ≤130%. The study showed that the developed and validated method is reproducible with good recovery. Figure 4 shows UPLC−MS/MS analysis of individual and overlaid SRM chromatograms corresponding to indospicine (2), standard, spiked, real, and blank samples. Trace amounts of glycerine present on the Amicon Ultra 0.5 mL, 3K centrifugal filter unit used in the extraction could interfere with the analysis by enhancing UPLC−MS/MS indospicine (2) responses. Hence, Amicon Ultra 0.5 mL, 3K centrifugal filter units are prerinsed twice to remove the interference. The response of indospicine (2) was then not significantly affected by the ultrafiltration step. To further minimize the effect of the Amicon Ultra 0.5 mL, 3K centrifugal filter, the internal standard was added prior to filtration in the sample preparation steps. While the use of an internal standard can correct for the effects of matrix on measured signals, as discussed earlier, an internal standard can also compensate for centrifugal filtration effects that occur when it cofilters with indospicine. In the method development process, recovery efficiency referrals to centrifugal filtered and noncentrifugal filtered external standards were carried out, and there was no significant difference between them. Therefore, the preparation of external standards for each batch of quantitation did not involve the ultrafiltration step. The indospicine (2) concentrations of the matrix spike samples were analyzed after 2 months of frozen storage and were still within standard deviation ranges. Thus, indospicine (2) is stable under cold
storage conditions with low storage losses. This is in line with an indospicine stability study reported by Simpson and Hegarty.22 The current developed method is less time-consuming than the amino acid analyzer method,2,19,20 more accurate and specific than the HPLC method,9 and less tedious through the absence of derivatization in the sample preparation.21 The indospicine (2) method reported here and the method developed by Gardner and Riet-Correa21 are generally alike in their ease at achieving indospicine resolution from other components. Indospicine in Camel Meat. To demonstrate the applicability of the developed and validated method, an analysis of a large number of meat samples from feral camel sampled as part of a cull exercise in the Simpson Desert was carried out. UPLC−MS/MS analysis of the feral camel meat samples revealed the presence of indospicine (2) at concentrations of 0−3.73 mg/kg. Figure 5 shows the number of contaminated camel meat samples by range of indospicine (2) concentrations. The indospicine (2) concentrations have a skewed distribution toward a zero value. These results show that indospicine (2) does contaminate camel meat in central Australia and are consistent with the previous detection of indospicine (2) in minced camel meat.16 However, more than 50% of the samples analyzed in this study have concentrations that are below the LOQ (0.1 mg/kg), while less than 5% of the samples have concentrations ≥1 mg/kg. Even so, several samples contained concentrations as high as 3.73 mg/kg. These results suggest caution, since a clinical investigation found that dietary exposure to an indospicine-contaminated pet food diet, of sweet potato/camel meat (2:1) containing indospicine at 2.1 mg/kg, could cause severe hepatic disease in dogs when consumed over extended time periods.16 The camel meat used to make this diet would have contained 3 times this level (or 6.3 mg/kg), almost twice the maximal levels seen in the present study. The low prevalence of indospicine (2) residues in camels in this study may, however, have been affected by the below average and very much below average rainfall in central Australia for the past 12 months (April 2012−May 2013)26 1977
dx.doi.org/10.1021/jf4052495 | J. Agric. Food Chem. 2014, 62, 1974−1979
Journal of Agricultural and Food Chemistry
Article
(2) Hegarty, M. P.; Pound, A. W. Indospicine, a hepatotoxic amino acid from Indigofera spicata: Isolation, structure, and biological studies. Aust. J. Biol. Sci. 1970, 23, 831−42. (3) Charlwood, B. V.; Morris, G. S.; Grenham, M. J., A chemical database for the Leguminosae. In Database in Systematics, Alkin, R., Bisby, F. A., Eds.; Academic Press: London, 1984; pp 201−208. (4) Ossedryver, S. M.; Baldwin, G. I.; Stone, B. M.; McKenzie, R. A.; van Eps, A. W.; Murray, S.; Fletcher, M. T. Indigofera spicata (creeping indigo) poisoning of three ponies. Aust. Vet. J. 2013, 91, 143−149. (5) Morton, J. F. Creeping indigo (Indigofera spicata Forsk.) (Fabaceae): A hazard to herbivores in Florida. Econ. Bot. 1989, 43, 314−327. (6) Wilson, P. G.; Rowe, R. A revision of the Indigofereae (Fabaceae) in Australia. 1. Indigastrum and the simple or unifoliolate species of Indigofera. Telopea 2004, 10, 651−682. (7) Wilson, P. G.; Rowe, R. Three new species of Indigofera (Fabaceae: Faboideae) from Cape York Peninsula. Telopea 2008, 12, 285−292. (8) Wilson, P. G.; Rowe, R. A revision of the Indigofereae (Fabaceae) in Australia. 2. Indigofera species with trifoliolate and alternately pinnate leaves. Telopea 2008, 12, 293−307. (9) Pollitt, S.; Hegarty, M. P.; Pass, M. A. Analysis of the amino acid indospicine in biological samples by high performance liquid chromatography. Nat. Toxins 1999, 7, 233−240. (10) Hegarty, M. P., Non-metallic chemical residues in toxic plants with potential importance to animal and human health. In Vet Update ’92; Osborne, H. G., Ed.; University of Queensland. Continuing Professional Education: Brisbane, 1992; pp 323−332. (11) Young, M. P. Investigation of the toxicity of horsemeat due to contamination by indospicine. Ph.D. Thesis, University of Queensland, Brisbane, Australia, 1992. (12) AZRI, Arid Zone Research Institute. Indospicine in beef. In Northern Territory Department of Primary Industries and Fisheries Technical Annual Report 1987−1988; NTDPIF: Alice Springs, Australia, 1989; p 39. (13) Pollitt, S. Residue implications of indospicine, a toxic, nonprotein amino acid. Ph.D. Thesis, University of Queensland, Brisbane, Australia, 2001. (14) Christie, G. S.; Wilson, M.; Hegarty, M. P. Effects on the liver in the rat of ingestion of Indigofera spicata, a legume containing an inhibitor of arginine metabolism. J. Pathol. 1975, 117, 195−205. (15) Kelly, W. R.; Young, M. P.; Hegarty, M. P.; Simpson, G. D. The hepatotoxicity of indospicine in dogs. In Poisonous Plants; James, L. F., Keeler, R. F., Bailey, E. M., Cheeke, P. R., Hegarty, M. P., Eds.; Iowa State University Press: Ames, IA, 1992; pp 126−130. (16) FitzGerald, L. M.; Fletcher, M. T.; Paul, A. E. H.; Mansfield, C. S.; O’Hara, A. J. Hepatotoxicosis in dogs consuming a diet of camel meat contaminated with indospicine. Aust. Vet. J. 2011, 89, 95−100. (17) Edwards, G.; Zeng, B.; Saalfeld, W.; Vaarzon-Morel, P.; McGregor, M. Managing the Impacts of Feral Camels in Australia: A New Way of Doing Business. DKCRC Report 47; Desert Knowledge Cooperative Research Centre: Alice Springs, Australia, 2008. (18) Dörges, B.; Heucke, J.; Dance, R. The palatability of Central Australian plant species to camels. Technote No. 116; Department of Primary Industry, Fisheries and Resources: Northern Territory Government, Alice Springs, Australia (www.nt.gov.au/d/Content/ File/p/Technote/TN116.pdf), July 11, 2013. (19) Aylward, J. H.; Court, R. D.; Haydock, K. P.; Strickland, R. W.; Hegarty, M. P. Indigofera species with agronomic potential in the tropics. Rat toxicity studies. Aust. J. Agric. Res. 1987, 38, 177−86. (20) Miller, R. W.; Smith, C. R., Jr. Seeds of Indigofera species. Their content of amino acids that may be deleterious. J. Agric. Food Chem. 1973, 21, 909−12. (21) Gardner, D. R.; Riet-Correa, F. Analysis of the toxic amino acid indospicine by liquid chromatography−tandem mass spectrometry. IJPPR 2011, 1, 20−27. (22) Simpson, G. D.; Hegarty, M. P. Determination of indospicine in biological material. Tropical Agronomy Technical Memorandum 57;
Figure 5. Number of camel meat samples contaminated by a range of indospicine concentrations (mg/kg).
before camel meat samples were being collected from Simpson Desert. Indigofera plants are known to be seasonally abundant, with the greatest density occurring after rain, which may lead to sporadic instances of higher Indigofera consumption by camels and potentially much higher indospicine (2) residues than found in the present study. In conclusion, this is the first nonderivatized UPLC−MS/MS method developed and validated for the quantitation of indospicine (2) in camel meat samples, with a limit of detection down to 0.1 mg/kg. It was used for the definitive identification of indospicine (2) in feral camel meat samples collected from central Australia and has confirmed the presence of indospicine in them, but at levels far lower than has been previously associated with canine fatalities.1,16
■
AUTHOR INFORMATION
Corresponding Author
*Tel: +61 7 32766089. Fax: +61 7 326 6565. E-mail: mary. fl[email protected]. Funding
This study was partly funded by the Academic Training Scheme for Institutions of Higher Education (SLAI) Scholarship sponsored by the Malaysian Government. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS We are grateful for the support of the Australian Feral Camel Management Project, administered by NintiOne, and Jordan Hampton and Corissa Miller (Ecotone Wildlife Veterinary Services), who collected camel tissue samples. Dr. Maria Jose Gomez Ramos (The University of Queensland) performed accurate MS and MS/MS analysis, and UPLC−MS/MS technical advice was provided by Cindy Giles, Dennis Webber, and Warwick Turner (Department of Agriculture, Fisheries and Forestry, Queensland, Australia).
■
REFERENCES
(1) Hegarty, M. P.; Kelly, W. R.; McEwan, D.; Williams, O. J.; Cameron, R. Hepatotoxicity to dogs of horse meat contaminated with indospicine. Aust. Vet. J. 1988, 65, 337−40. 1978
dx.doi.org/10.1021/jf4052495 | J. Agric. Food Chem. 2014, 62, 1974−1979
Journal of Agricultural and Food Chemistry
Article
CSIRO Australia, Division Tropical Crops and Pastures: Brisbane, Australia, 1987. (23) Australia Feral Camel Management Project. http://www. feralcamels.com.au/ (Accessed October 14, 2013). (24) The Council of Heads of Australasian Herbaria Australia’s Virtual Herbarium. http://avh.chah.org.au (Accessed September 3, 2013). (25) Guo, D.; Mant, C. T.; Hodges, R. S. Effects of ion-pairing reagents on the prediction of peptide retention in reversed-phase highresolution liquid chromatography. J. Chromatogr. 1987, 386, 205−222. (26) The Bureau of Meteorology Australia. http://www.bom.gov.au/ (Accessed October 11, 2013)
1979
dx.doi.org/10.1021/jf4052495 | J. Agric. Food Chem. 2014, 62, 1974−1979
Determination of Hepatotoxic Indospicine in Australian Camel Meat by Ultra-Performance Liquid Chromatography−Tandem Mass Spectrometry Eddie T. T. Tan,†,§ Mary T. Fletcher,*,† Ken W. L. Yong,† Bruce R. D’Arcy,‡ and Rafat Al Jassim‡ †
Queensland Alliance for Agriculture and Food Innovation (QAAFI), The University of Queensland, Health and Food Sciences Precinct, Coopers Plains, QLD 4108 Australia ‡ School of Agriculture and Food Sciences, Faculty of Science, The University of Queensland, Gatton, QLD 4343 Australia § Food Technology Department, Faculty of Applied Sciences, Universiti Teknologi MARA, Shah Alam, Malaysia ABSTRACT: Indospicine is a hepatotoxic amino acid found in Indigofera plant spp. and is unusual in that it is not incorporated into protein but accumulates as the free amino acid in the tissues (including muscle) of animals consuming these plants. Dogs are particularly sensitive to indospicine, and secondary poisoning of dogs has occurred from the ingestion of indospicinecontaminated horse meat and more recently camel meat. In central Australia, feral camels are known to consume native Indigofera species, but the prevalence of indospicine residues in their tissues has not previously been investigated. In this study, a method was developed and validated with the use of ultra-performance liquid chromatography−tandem mass spectrometry (UPLC−MS/MS) to determine the level of indospicine in camel meat samples using isotopically labeled indospicine as an internal standard. UPLC−MS/MS analysis showed that the method is reproducible, with high recovery efficiency and a quantitation limit of 0.1 mg/kg. Camel meat samples from the Simpson Desert were largely contaminated (≈50%) by indospicine with levels up to 3.73 mg/kg (fresh weight) determined. However, the majority of samples (95%) contained less than 1 mg/kg indospicine. KEYWORDS: indospicine, Indigofera, camel, Camelus dromedarius, ultra-performance liquid chromatography−tandem mass spectrometry
■
in the tissues of animals fed Indigofera plant material.9 Such residues have been shown to be highly residual and persist for several months after the cessation of feeding in horses,10 goats,11 cattle,12 and rabbits.13 Indospicine (2) has been shown experimentally to cause liver damage in a range of animals, including mice,2 rats,14 and dogs.15 Dogs are particularly sensitive to indospicine hepatotoxicity, and of concern, from a food safety viewpoint, is that fatalities have been reported when naturally contaminated meat was used as pet food. Secondary poisoning of dogs has occurred from the ingestion of indospicine-contaminated horse meat10 and, more recently, camel meat16 from animals that had grazed in regions where Indigofera species are prevalent. The population of feral camels in Australia has been estimated at 1 million camels, predominantly present in arid inland regions, with a predicted population growth rate of 7− 8% per annum (an increase of approximately 80 000 feral camels per year).17 Investigations of dietary preferences of camels in arid central Australia have provided evidence of the consumption range of Indigofera spp.18 Of 243 plant species observed to be consumed by camels in this region, six Indigofera species were noted, including I. colutea, I. georgei, I. helmsii, and I. hirsuta, which were assessed to be the “main food plant at
INTRODUCTION
In Australia there are more than 40 species of Indigofera, both native and introduced, with a number of these known to contain hepatotoxin indospicine (2) (Figure 1), including Indigofera circinella, Indigofera colutea, Indigofera hendecaphylla, Indigofera hirsuta, Indigofera linnaei, Indigofera suffruticosa, Indigofera spicata, and Indigofera trita.1−8 The amino acid indospicine (2) is an arginine (1) (Figure 1) analogue and is unusual in that it is not incorporated into proteins but is present as the free amino acid and accumulates in the free form
Received: Revised: Accepted: Published:
Figure 1. Chemical structure of arginine (1); indospicine (2), an analogue of arginine; and D3-L-indospicine (3), an internal standard. © 2014 American Chemical Society
1974
November 21, 2013 January 15, 2014 January 16, 2014 January 16, 2014 dx.doi.org/10.1021/jf4052495 | J. Agric. Food Chem. 2014, 62, 1974−1979
Journal of Agricultural and Food Chemistry
Article
Figure 2. Tandem mass spectrometry fragmentation spectra of (A) indospicine and (B) D3-L-indospicine. Indospicine Extraction. Prior to analysis, Amicon Ultra 0.5 mL, 3K centrifugal filter units (Merck Millipore, Kilsyth, Australia) were prerinsed and centrifuged (10 000 rpm, 20 min) with reverse osmosis water (2 × 300 μL) to remove glycerine and then inverted and spun for 1 min at 1000 rpm. Camel muscle samples were thawed, finely chopped, weighed (0.5 g), and mixed with 0.1% heptafluorobutyric acid (HFBA, 25 mL), follow by homogenization with a Polytron T25 Basic homogenizer (Labtek, Brendale, Australia) for 15 s. The homogenized samples were chilled (4 °C) for 20 min and then centrifuged at 4500 rpm for 20 min at 18 °C. Aliquots of supernatants (1.0 mL) spiked with the internal standard D3-L-indospicine (3) (1 mg/L in 0.1% HFBA, 100 μL) were vortexed for 10 s. A portion (450 μL) was transferred into prerinsed Amicon Ultra 0.5 mL, 3K centrifugal filters, which were centrifuged (10 000 rpm, 20 min), with the filtrates transferred to a limited volume insert for UPLC− MS/MS analysis. UPLC−MS/MS Equipment and Conditions. Sample data were collected by MassLynx 4.1 and processed by TargetLynx application managers. Separation of indospicine was done by a Waters ACQUITY UPLC system liquid chromatography (Waters, Rydalmere, Australia). LC separations were performed using a 100 mm × 2.1 mm i.d., 1.7 μm, BEH C18 column (Waters, Rydalmere, Australia) at 30 °C with flow rate of 0.2 mL/min. The mobile phase was a mixture of (A) H2O with 0.1% HFBA (v/v) (pH 2.15) and (B) acetonitrile with 0.1% HFBA, which were prepared freshly for each analysis. The following gradient was used: 0 min, 99% A; 4 min, 70% A; 7 min, 70% A; 8 min, 99% A; 10 min, 99% A. MS/MS detection was made using a Waters Micromass Quattro Premier triple quadrupole mass spectrometer with an electrospray ionization (ESI) source (Waters, Rydalmere, Australia), operated in the positive mode. Eluted indospicine (2) was quantitated by utilizing selected reaction monitoring (SRM) transitions of m/z 174.2 → 111.0 (verified by transition of m/z 174.2 → 157.1) for indospicine (2) and m/z 177.1 → 114.0 (verified by transition of m/z 177.1 → 113.0) for D3-L-indospicine (3) as internal standard. The capillary voltage was 2.79 kV, cone gas flow was 50 L/h, and desolvation gas flow was 600 L/h. The source and desolvation temperatures were set at 150 and 350 °C, respectively. The argon gas collision energies for indospicine (2) (15 and 12 eV) and D3-L-indospicine (3) (15 and 15 eV) were set with cone voltage at 25 V. Synthesized indospicine (2) (external standard) and D3-Lindospicine (3) (deuterium-labeled internal standard, >99% pure) were used in UPLC−MS/MS analysis. Internal (1 mg/L) and external standard (0.002−2 mg/L) solutions for indospicine (2) were prepared in H2O with 0.1% HFBA. Solutions were frozen for no longer than 1 month.
times”, and I. linnaei and I. basedowii, which are“preferred food plants”.18 Short-lived mass developments of forbs were reported to be seasonally utilized by camels, including Indigofera species in summer. Despite the reported dog poisoning incident,16 the extent of indospicine (2) contamination of camel meat in Australia remains largely unknown. Consequently, this study was undertaken to develop and validate a high-throughput ultra-performance liquid chromatography−tandem mass spectrometry (UPLC−MS/MS) method to quantitate indospicine (2) in camel meat tissue. The validated method was then used to determine the concentrations of indospicine (2) in a number of collected camel meat samples obtained from a high camel density area of Australia, the Simpson Desert in the Northern Territory of Australia. Indospicine (2) has previously been measured in plant material by amino acid analysis,1,19,20 HPLC,4,9 and, more recently, LC−MS/MS with phenylisothiocyanate derivitization.21 Analysis of indospicine (2) in meat is more challenging due to the lower levels of toxin and the complex matrix. Indospicine (2) in animal tissues have previously been analyzed by lengthy (nearly 2 h) amino acid analysis1,22 and by phenylisothiocyanate derivatization and UV detection, with resolution from other amino acids being problematic.9 The incorporation of D3-L-indospicine (3) (Figure 1) in this present study overcomes matrix effects noted in our previous LC−MS/ MS analysis of camel tissues.16
■
MATERIALS AND METHODS
Indospicine Standards. Synthesized indospicine (2) and D3-Lindospicine (3) were provided by Dr. James De Voss and Dr. Robert Lang, The University of Queensland. Meat Materials. A total of 99 camels were sampled as part of a cull exercise under the Australian Feral Camel Management Project23 in the Simpson Desert in central Australia. Rump muscle samples were excised from culled camels, placed in specimen containers, frozen, and transported to Health and Food Science Precinct, Brisbane, Australia. All of the samples were stored frozen (−30 °C) prior to extraction and analysis. Camel muscle samples free from indospicine (2) contamination (used for negative control and spike samples) were obtained from a pet food supplier, and these samples had been sourced from 200 km northwest of Rawlinna in Nullabor Plain, West Australia (an area relatively free from Indigofera species),24 and stored frozen. 1975
dx.doi.org/10.1021/jf4052495 | J. Agric. Food Chem. 2014, 62, 1974−1979
Journal of Agricultural and Food Chemistry
Article
Accurate Mass Analysis. Accurate mass analysis of both parent and daughter ions was performed on an AB SCIEX TripleTOF 5600+ system (AB SCIEX, Mt. Waverley, Australia) via direct injection and fragmentation of standard solutions of both indospicine and D3-Lindospicine.
camel meat extracts (spiked samples and Simpson Desert samples) were determined by interpolating the relative intensity with a calibration curve. Calibration curves were constructed for each batch and involved plotting the ratio of the intensity of the positive-ion ESI SRM transition m/z 174 → 111 relative to the internal standard (D3-L-indospicine, 3) SRM transition m/z 177 → 114, versus the indospicine (2) concentrations for standard solutions of 0.002−2.00 mg/L. Seven-point calibration curves with the overlaid chromatograms of internal and external standards are shown in Figure 3.
■
RESULTS AND DISCUSSION UPLC−MS/MS Fragmentation of Indospicine and D3L-indospicine. Fragmentation of the protonated molecular ion of indospicine (2) ([M + H]+ m/z 174) in positive-ion ESI mode produces a number of daughter ions, notably m/z 157 (loss of NH3), m/z 129 (loss of NH3 and CO), m/z 111 (loss of NH3 and HCO2H), and m/z 84 (loss of NH3, CO2H, and HCN), as seen in Figure 2. Accurate mass MS and MS/MS data confirmed the composition of each of the daughter ions, particularly the perhaps unexpected composition of m/z 129 as C6H13N2O (observed 129.1029, calculated 129.1028). We postulate then that the required loss of NH3 and CO occurs via loss of NH3 from the amidine group followed by formation of a cyclic lactone intermediate, from which it is possible to lose CO while retaining the second oxygen atom. In the labeled standard D3-L-indospicine (3), deuterium atoms are incorporated at C-5 (two deuterium) and C-6 (one deuterium), and this is evident in the fragmentation of the protonated molecular ion [M + H]+ m/z 177, which produces corresponding daughter ions of m/z 160/159 (loss of NH3/ NH2D), m/z 132/131 (loss of NH3/NH2D and CO), m/z 114/113 (loss of NH3/NH2D and H2CO2H), and m/z 86/87 (loss of NH3/NH2D, H2CO2H, and HCN), as seen in Figure 2. The fragmentation to almost equal proportion of each of these ion pairs is consistent with the proposed initial loss of NH2 from the terminal amidine group with concomitant loss of either H or D from the monodeuterated C-6 position. Accurate mass MS and MS/MS data again confirmed that the composition of each of the daughter ions is consistent with the listed mass losses. The most dominant daughter ion observed at m/z 111 was chosen as the transition (m/z 174 → 111) for quantitation of indospicine (2) using the selected reaction monitoring (SRM) program on the UPLC−MS/MS, with the m/z 157 daughter ion selected as the transition (m/z 174 → 157) for verification. The efficiency of the ionization and the detection of the precursor and fragments ions were optimized by adjusting collision energy (15 and 12 eV) and cone voltage (25 V). Similarly, the transition m/z 177 → 114 was selected for SRM quantitation of D3-L-indospicine (3), with the transition m/z 177 → 113 for verification with optimized collision energy and cone voltage (15 eV and 25 V, respectively). Development and Validation of a UPLC−MS/MS Method for Indospicine in Camel Meat Samples. The UPLC−MS/MS method reported here is a high-throughput yet sensitive analysis method for indospicine (2) in camel meat tissue, and this is the first reported indospicine analysis in unprocessed camel meat tissue based on the developed and validated UPLC−MS/MS method. The method involves a simple tissue extraction into aqueous solution followed by physical deproteinisation using centrifugal ultrafiltration before UPLC−MS/MS detection. The described UPLC separation on BEH C18 column eluting with water and acetonitrile (0.1% HFBA) gradients provides good peak shape and column retention. Heptafluorobutyric acid (HFBA) is an effective ionpairing agent that is widely used in reverse-phase peptide/ amino acid separations.25 Concentrations of indospicine (2) in
Figure 3. Overlaid SRM chromatograms of indospicine standards (0.002−2.00 mg/L) and internal standard (A) monitored by UPLC− MS/MS with ESI.
The limit of quantitation (LOQ) determined for this work was at least 9:1 for the peak-to-peak signal-to noise ratio for the quantitation transition (m/z 174 → 111). Both quantitation and qualification ions were monitored for each analysis. The average ratio of the quantitation peak to that of the qualification peak for spiked samples (n = 21) was 4.49. Confirmation of indospicine (2) was also carried out by examining the quantitation/qualification ion ratio, with the requirement that the ratio be within ±20%. Hence, only samples that have the ratio of 3.59 to 5.39 are considered to contain actual indospicine (2). The calibration curves (n = 8) demonstrated good linearity with R2 values of 0.9960 ± 0.004 throughout the analysis. Recovery efficiency and reproducibility of this analytical method were demonstrated by replicate analyses (n = 7) of the matrix spiked at three concentration levels: 2.5, 5.0, and 50.0 mg/kg indospicine-free camel meat. The spiked matrices were then extracted by the method described in indospicine (2) extraction. These spiking studies had recoveries of 106, 109, and 102%, respectively. The mean concentrations (standard deviation (SD), relative standard deviation (RSD)) of seven analyses for each spiked sample were 2.65 mg/kg (SD 0.24, RSD 9.06%), 5.43 mg/kg (SD 0.42, RSD 7.82%), and 51.01 mg/kg (SD 5.02, RSD 9.85%), respectively. The relative standard deviation was in all cases below 10% for three spiked samples, and this indicated that the method is reproducible with good precision. A significant clinical challenge with UPLC−MS/MS is the potential for matrix effects that cause interferences or impact ionization efficiency. The matrices of 1976
dx.doi.org/10.1021/jf4052495 | J. Agric. Food Chem. 2014, 62, 1974−1979
Journal of Agricultural and Food Chemistry
Article
Figure 4. UPLC−MS/MS analysis of (A) individual and (B) overlaid SRM chromatograms for indospicine: (a) 0.05 mg/L standard, (b) 2.5 mg/L spiked sample, (c) Simpson Desert camel meat sample (SD_79) (1.41 mg/L), and (d) blank sample.
the extract can significantly (p ≤ 0.05) suppress the indospicine and isotope-labeled internal standard responses. However, the use of stable isotope-labeled internal standards in this study to compensate for matrix effects and to increase accuracy resulted in the quantitation of matrix and matrix-free extracts having no substantial differences. Reagent spike recovery efficiency was not done due to this internal standard compensation principle. A blank matrix sample, spiked matrix samples, and spiked check samples were analyzed for each batch of analyses as part of the quality control for the validity of the analyses. The validity of indospicine (2) present in camel meat samples was ensured by spiked check samples being reported within 70% ≤ spike ≤130%. The study showed that the developed and validated method is reproducible with good recovery. Figure 4 shows UPLC−MS/MS analysis of individual and overlaid SRM chromatograms corresponding to indospicine (2), standard, spiked, real, and blank samples. Trace amounts of glycerine present on the Amicon Ultra 0.5 mL, 3K centrifugal filter unit used in the extraction could interfere with the analysis by enhancing UPLC−MS/MS indospicine (2) responses. Hence, Amicon Ultra 0.5 mL, 3K centrifugal filter units are prerinsed twice to remove the interference. The response of indospicine (2) was then not significantly affected by the ultrafiltration step. To further minimize the effect of the Amicon Ultra 0.5 mL, 3K centrifugal filter, the internal standard was added prior to filtration in the sample preparation steps. While the use of an internal standard can correct for the effects of matrix on measured signals, as discussed earlier, an internal standard can also compensate for centrifugal filtration effects that occur when it cofilters with indospicine. In the method development process, recovery efficiency referrals to centrifugal filtered and noncentrifugal filtered external standards were carried out, and there was no significant difference between them. Therefore, the preparation of external standards for each batch of quantitation did not involve the ultrafiltration step. The indospicine (2) concentrations of the matrix spike samples were analyzed after 2 months of frozen storage and were still within standard deviation ranges. Thus, indospicine (2) is stable under cold
storage conditions with low storage losses. This is in line with an indospicine stability study reported by Simpson and Hegarty.22 The current developed method is less time-consuming than the amino acid analyzer method,2,19,20 more accurate and specific than the HPLC method,9 and less tedious through the absence of derivatization in the sample preparation.21 The indospicine (2) method reported here and the method developed by Gardner and Riet-Correa21 are generally alike in their ease at achieving indospicine resolution from other components. Indospicine in Camel Meat. To demonstrate the applicability of the developed and validated method, an analysis of a large number of meat samples from feral camel sampled as part of a cull exercise in the Simpson Desert was carried out. UPLC−MS/MS analysis of the feral camel meat samples revealed the presence of indospicine (2) at concentrations of 0−3.73 mg/kg. Figure 5 shows the number of contaminated camel meat samples by range of indospicine (2) concentrations. The indospicine (2) concentrations have a skewed distribution toward a zero value. These results show that indospicine (2) does contaminate camel meat in central Australia and are consistent with the previous detection of indospicine (2) in minced camel meat.16 However, more than 50% of the samples analyzed in this study have concentrations that are below the LOQ (0.1 mg/kg), while less than 5% of the samples have concentrations ≥1 mg/kg. Even so, several samples contained concentrations as high as 3.73 mg/kg. These results suggest caution, since a clinical investigation found that dietary exposure to an indospicine-contaminated pet food diet, of sweet potato/camel meat (2:1) containing indospicine at 2.1 mg/kg, could cause severe hepatic disease in dogs when consumed over extended time periods.16 The camel meat used to make this diet would have contained 3 times this level (or 6.3 mg/kg), almost twice the maximal levels seen in the present study. The low prevalence of indospicine (2) residues in camels in this study may, however, have been affected by the below average and very much below average rainfall in central Australia for the past 12 months (April 2012−May 2013)26 1977
dx.doi.org/10.1021/jf4052495 | J. Agric. Food Chem. 2014, 62, 1974−1979
Journal of Agricultural and Food Chemistry
Article
(2) Hegarty, M. P.; Pound, A. W. Indospicine, a hepatotoxic amino acid from Indigofera spicata: Isolation, structure, and biological studies. Aust. J. Biol. Sci. 1970, 23, 831−42. (3) Charlwood, B. V.; Morris, G. S.; Grenham, M. J., A chemical database for the Leguminosae. In Database in Systematics, Alkin, R., Bisby, F. A., Eds.; Academic Press: London, 1984; pp 201−208. (4) Ossedryver, S. M.; Baldwin, G. I.; Stone, B. M.; McKenzie, R. A.; van Eps, A. W.; Murray, S.; Fletcher, M. T. Indigofera spicata (creeping indigo) poisoning of three ponies. Aust. Vet. J. 2013, 91, 143−149. (5) Morton, J. F. Creeping indigo (Indigofera spicata Forsk.) (Fabaceae): A hazard to herbivores in Florida. Econ. Bot. 1989, 43, 314−327. (6) Wilson, P. G.; Rowe, R. A revision of the Indigofereae (Fabaceae) in Australia. 1. Indigastrum and the simple or unifoliolate species of Indigofera. Telopea 2004, 10, 651−682. (7) Wilson, P. G.; Rowe, R. Three new species of Indigofera (Fabaceae: Faboideae) from Cape York Peninsula. Telopea 2008, 12, 285−292. (8) Wilson, P. G.; Rowe, R. A revision of the Indigofereae (Fabaceae) in Australia. 2. Indigofera species with trifoliolate and alternately pinnate leaves. Telopea 2008, 12, 293−307. (9) Pollitt, S.; Hegarty, M. P.; Pass, M. A. Analysis of the amino acid indospicine in biological samples by high performance liquid chromatography. Nat. Toxins 1999, 7, 233−240. (10) Hegarty, M. P., Non-metallic chemical residues in toxic plants with potential importance to animal and human health. In Vet Update ’92; Osborne, H. G., Ed.; University of Queensland. Continuing Professional Education: Brisbane, 1992; pp 323−332. (11) Young, M. P. Investigation of the toxicity of horsemeat due to contamination by indospicine. Ph.D. Thesis, University of Queensland, Brisbane, Australia, 1992. (12) AZRI, Arid Zone Research Institute. Indospicine in beef. In Northern Territory Department of Primary Industries and Fisheries Technical Annual Report 1987−1988; NTDPIF: Alice Springs, Australia, 1989; p 39. (13) Pollitt, S. Residue implications of indospicine, a toxic, nonprotein amino acid. Ph.D. Thesis, University of Queensland, Brisbane, Australia, 2001. (14) Christie, G. S.; Wilson, M.; Hegarty, M. P. Effects on the liver in the rat of ingestion of Indigofera spicata, a legume containing an inhibitor of arginine metabolism. J. Pathol. 1975, 117, 195−205. (15) Kelly, W. R.; Young, M. P.; Hegarty, M. P.; Simpson, G. D. The hepatotoxicity of indospicine in dogs. In Poisonous Plants; James, L. F., Keeler, R. F., Bailey, E. M., Cheeke, P. R., Hegarty, M. P., Eds.; Iowa State University Press: Ames, IA, 1992; pp 126−130. (16) FitzGerald, L. M.; Fletcher, M. T.; Paul, A. E. H.; Mansfield, C. S.; O’Hara, A. J. Hepatotoxicosis in dogs consuming a diet of camel meat contaminated with indospicine. Aust. Vet. J. 2011, 89, 95−100. (17) Edwards, G.; Zeng, B.; Saalfeld, W.; Vaarzon-Morel, P.; McGregor, M. Managing the Impacts of Feral Camels in Australia: A New Way of Doing Business. DKCRC Report 47; Desert Knowledge Cooperative Research Centre: Alice Springs, Australia, 2008. (18) Dörges, B.; Heucke, J.; Dance, R. The palatability of Central Australian plant species to camels. Technote No. 116; Department of Primary Industry, Fisheries and Resources: Northern Territory Government, Alice Springs, Australia (www.nt.gov.au/d/Content/ File/p/Technote/TN116.pdf), July 11, 2013. (19) Aylward, J. H.; Court, R. D.; Haydock, K. P.; Strickland, R. W.; Hegarty, M. P. Indigofera species with agronomic potential in the tropics. Rat toxicity studies. Aust. J. Agric. Res. 1987, 38, 177−86. (20) Miller, R. W.; Smith, C. R., Jr. Seeds of Indigofera species. Their content of amino acids that may be deleterious. J. Agric. Food Chem. 1973, 21, 909−12. (21) Gardner, D. R.; Riet-Correa, F. Analysis of the toxic amino acid indospicine by liquid chromatography−tandem mass spectrometry. IJPPR 2011, 1, 20−27. (22) Simpson, G. D.; Hegarty, M. P. Determination of indospicine in biological material. Tropical Agronomy Technical Memorandum 57;
Figure 5. Number of camel meat samples contaminated by a range of indospicine concentrations (mg/kg).
before camel meat samples were being collected from Simpson Desert. Indigofera plants are known to be seasonally abundant, with the greatest density occurring after rain, which may lead to sporadic instances of higher Indigofera consumption by camels and potentially much higher indospicine (2) residues than found in the present study. In conclusion, this is the first nonderivatized UPLC−MS/MS method developed and validated for the quantitation of indospicine (2) in camel meat samples, with a limit of detection down to 0.1 mg/kg. It was used for the definitive identification of indospicine (2) in feral camel meat samples collected from central Australia and has confirmed the presence of indospicine in them, but at levels far lower than has been previously associated with canine fatalities.1,16
■
AUTHOR INFORMATION
Corresponding Author
*Tel: +61 7 32766089. Fax: +61 7 326 6565. E-mail: mary. fl[email protected]. Funding
This study was partly funded by the Academic Training Scheme for Institutions of Higher Education (SLAI) Scholarship sponsored by the Malaysian Government. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS We are grateful for the support of the Australian Feral Camel Management Project, administered by NintiOne, and Jordan Hampton and Corissa Miller (Ecotone Wildlife Veterinary Services), who collected camel tissue samples. Dr. Maria Jose Gomez Ramos (The University of Queensland) performed accurate MS and MS/MS analysis, and UPLC−MS/MS technical advice was provided by Cindy Giles, Dennis Webber, and Warwick Turner (Department of Agriculture, Fisheries and Forestry, Queensland, Australia).
■
REFERENCES
(1) Hegarty, M. P.; Kelly, W. R.; McEwan, D.; Williams, O. J.; Cameron, R. Hepatotoxicity to dogs of horse meat contaminated with indospicine. Aust. Vet. J. 1988, 65, 337−40. 1978
dx.doi.org/10.1021/jf4052495 | J. Agric. Food Chem. 2014, 62, 1974−1979
Journal of Agricultural and Food Chemistry
Article
CSIRO Australia, Division Tropical Crops and Pastures: Brisbane, Australia, 1987. (23) Australia Feral Camel Management Project. http://www. feralcamels.com.au/ (Accessed October 14, 2013). (24) The Council of Heads of Australasian Herbaria Australia’s Virtual Herbarium. http://avh.chah.org.au (Accessed September 3, 2013). (25) Guo, D.; Mant, C. T.; Hodges, R. S. Effects of ion-pairing reagents on the prediction of peptide retention in reversed-phase highresolution liquid chromatography. J. Chromatogr. 1987, 386, 205−222. (26) The Bureau of Meteorology Australia. http://www.bom.gov.au/ (Accessed October 11, 2013)
1979
dx.doi.org/10.1021/jf4052495 | J. Agric. Food Chem. 2014, 62, 1974−1979
