Liquid chromatography tandem mass spectrometry ... - Springer Link

0 downloads 0 Views 788KB Size Report
Oct 3, 2014 - David L. Sedlak & Tyrone Hayes. Received: 11 June 2014 /Revised: 12 September 2014 /Accepted: 19 September 2014 /Published online: 3 ...
Anal Bioanal Chem (2014) 406:7677–7685 DOI 10.1007/s00216-014-8207-8

RESEARCH PAPER

Liquid chromatography tandem mass spectrometry method using solid-phase extraction and bead-beating-assisted matrix solid-phase dispersion to quantify the fungicide tebuconazole in controlled frog exposure study: analysis of water and animal tissue Martin Hansen & Rikke Poulsen & Xuan Luong & David L. Sedlak & Tyrone Hayes

Received: 11 June 2014 / Revised: 12 September 2014 / Accepted: 19 September 2014 / Published online: 3 October 2014 # Springer-Verlag Berlin Heidelberg 2014

Abstract This paper presents the development, optimization, and validation of a LC-MS/MS methodology to determine the concentration of the antifungal drug and fungicide tebuconazole in a controlled exposure study of African clawed frogs (Xenopus laevis). The method is validated on animal tank water and on tissue from exposed and non-exposed adult X. laevis. Using solid-phase extraction (SPE), the analytical method allows for quantification of tebuconazole at concentrations as low as 3.89 pg mL−1 in 10 mL water samples. Using bead-beating-assisted matrix solid-phase dispersion

Electronic supplementary material The online version of this article (doi:10.1007/s00216-014-8207-8) contains supplementary material, which is available to authorized users. M. Hansen (*) : R. Poulsen : X. Luong : T. Hayes Laboratory for Integrative Studies in Amphibian Biology, Department of Integrative Biology, University of California, Berkeley, CA 94720, USA e-mail: [email protected] M. Hansen : D. L. Sedlak Department of Civil and Environmental Engineering, University of California, Berkeley, CA 94720, USA M. Hansen Department of Civil and Environmental Engineering, Stanford University, Stanford, CA 94305, USA M. Hansen Department of Growth and Reproduction, Copenhagen University Hospital, Blegdamsvej 9, 2100 Copenhagen, Denmark M. Hansen : R. Poulsen Toxicology Laboratory, Section of Advanced Drug Analysis, Department of Pharmacy, Faculty of Health and Medical Sciences, University of Copenhagen, Universitetsparken 2, 2100 Copenhagen, Denmark

(MSPD), it was possible to quantify tebuconazole down to 0.63 pg mg−1 wet weight liver using 150 mg tissue. The deuterated analogue of tebuconazole was used as internal standard, and ensured method accuracy in the range 80.6– 99.7 % for water and 68.1–109 % for tissue samples. The developed method was successfully applied in a 4-week X. laevis repeated-exposure study, revealing high levels of tebuconazole residues in adipose and liver tissue, and with experimental bioconcentration factors up to 18,244 L kg−1. Keywords Sample preparation . Fungicide . Agrochemicals . Endocrine disruptor . Bioconcentration factor . CAS 107534-96-3

Introduction Pesticides are heavily used worldwide and carefully designed to incapacitate unwanted agricultural pests, such as fungi, insects, and plants. The environmental dispersion of these agrochemicals has caused concern, as some of these xenobiotics are suspected of impairing immune function and interfering with the endocrine system in wildlife. Studies have demonstrated the impact of pesticides (e.g., atrazine, endosulfan, and organophosphate insecticides) on amphibians [1–3]. It has been hypothesized that endocrine disruptors, such as dichlorodiphenyltrichloroethane (DDT), may act as tumor initiators or promoters leading to urogenital carcinogenesis in California sea lions [1–4]. Tebuconazole is a fungicide used in agriculture and is mainly involved in the production of cereals, fruits, and vegetables [5]. Additionally, the fungicide is used as a preservative biocide for wood, film, and masonry [6, 7]. In 2011, it was reported that 33,493 kg and 47,080 kg

7678

tebuconazole was used in California and Denmark, respectively [8–10]. This heavy usage underlines the need to investigate the fate of the compound in the environment and its possible effects on the ecosystems. Tebuconazole belongs to the group of fungicides called conazoles, which inhibit fungal growth by specifically blocking the cytochrome P450-mediated C14 α-demethylation step in the biosynthesis of ergosterol, a biomolecule unique to fungi. Little is known regarding tebuconazole’s biological fate, accumulation, and physiological effects. The molecular structure and physicochemical properties for tebuconazole can be found in Fig. 1. The compound has a log Kow value above three, suggesting possible bioaccumulation; however, in 2013, the European Food Safety Authority concluded that bioaccumulation is unlikely to occur [11]. This statement has recently been challenged by results from Smalling et al. [12], who found tebuconazole to be one of the most frequently detected pesticides in whole-body extracts of wild Pacific chorus frogs in the Sierra Nevada Mountains. These contradictory results indicate need for further studies and the development of reliable analytical methods for quantification of tebuconazole in environmental compartments and tissue. Newly developed analytical methodologies used in research, e.g., to answer in-depth biological hypotheses, are typically sketchily described in papers focused on investigating other research questions. In such cases, the analytical chemistry receives little attention and is hardly discussed in detail. Consequently, the aim of this study was to develop, validate, and describe a LC-MS/MS methodology to determine tebuconazole in water using solid-phase extraction (SPE), and in exposed frog tissue using bead-beatingassisted matrix solid-phase dispersion (MSPD). Recently, a few analytical methods have been developed to quantify tebuconazole in fish and crab tissue, vegetables, fruit juice,

M. Hansen et al.

wine, soil, sludge, and wood [13–20]. More importantly, knowledge from these efforts has been used in the present work; such as Carpinteiro et al., who well-demonstrated obtained clean extracts from red wine by trapping matrix interfering components on ion-exchange solid-phase columns [20]. Others have developed vortex-assisted MSPD methods to analyze tebuconazole and other fungicides in fish liver [13], however, with a relatively high quantification limit (limit of quantification (LOQ), 125 pg mg−1). Farajzadeh and coworkers focused on analyzing surface water and juice using liquid-liquid micro-extraction in combination with GC-MS, however, also the authors obtained fairly high LOQs (2.3–13 ng mL−1, [14]).

Experimental Chemicals, reagents, and materials (R, S)-tebuconazole and the internal standard d6-tebuconazole (IS) were obtained as racemic mixtures from Crescent Chemicals (Islandia, NY, USA), with a purity level of 98.5 % and isotopic purity 98.0 %, respectively. All utilized solvents (methanol, acetonitrile, ethanol) and buffers (formic acid and ammonia hydroxide) were of analytical grade and all obtained from Fisher Scientific (Fair Lawn, NJ, USA). Ammonia formate was obtained from Sigma-Aldrich (St. Louis, MO, USA). Fine analytical balance (AT261, Mettler Toledo, Columbus, OH, USA). Standard solutions Stock solutions of approximately 1000 ng μL−1 in methanol were prepared for tebuconazole. Working dilutions were prepared in MilliQ water (Millipore, Jaffrey, NH, USA) in the concentration range 0.2 ng mL−1 to 2 μg mL−1. The internal standard d6-tebuconazole was obtained in a reference solution 100 μg mL −1 in acetone, and a working solution of 20 ng mL−1 was also prepared in MilliQ water. Stock solutions were stored in darkness at −18 °C while aqueous working dilutions were stored in darkness at 4 °C. Sample preparation of water

Fig. 1 Chemical structure, CAS number, synonyms and physicochemical properties (molar mass (Mw), acidity constant (pKa), water solubility (Sw), partition coefficient between octanol and water phase (log P) of the neutral species), and vapor pressure (Vp) for tebuconazole (data from ChemIdplus, US NLM). Chemical structure for d6-tebuconazole (internal standard, IS) is also shown

An overview of the sample preparation procedure is shown in Fig. 2 and is further described in detail in the sections below. To avoid in-lab contamination, all utilized glassware were rinsed with soap, water and ethanol followed by heating to 500 °C for at least 6 h prior to use. Water sample sizes (0.1– 10 mL) were adjusted to pH 6.3 by diluting the sample 1:1 with 100 mM ammonia formate solution (NH4OOCH) in the SPE barrel (3 mL) and 1.00 ng of IS was added (50 μL of 20 ng mL−1 solution). The samples were loaded at a rate of

Tebuconazole in controlled frog exposure study

7679

performed by repeating the process. The two supernatant aliquots were combined, and a 500-μL aliquot was mixed with 500 μL mobile phase A (see the next section). Liquid chromatography-mass spectrometry

Fig. 2 The developed sample preparation method. A 500-μL aliquot of combined supernatant was added mobile phase A (500 μL) and transferred to HPLC vial for LC-MS/MS analysis

1–2 mL min−1 on mixed-mode anion-exchanger material SPE cartridges (60 mg PAX, Agilent Technologies, Palo Alto, CA, USA) mounted on vacuum manifold (Supelco, Bellefonte, PA, USA). Prior to enrichment, the cartridges were preconditioned with 2 mL methanol and 2 mL MilliQ water. Immediately after enrichment, the SPE cartridges were flushed with 1 mL 5 % ammonia hydroxide solution (NH4OH) followed by air drying by vacuum for 15 min. Finally analytes were eluted from the SPE with 1.00 mL acetonitrile into a test tube. To the extract, 1.00 mL mobile phase A (see “Liquid chromatography-mass spectrometry”) was added and an aliquot was transferred to a HPLC vial. Sample preparation of frog tissue Euthanized, previously non-exposed, Xenopus laevis adult male and female frogs were used for the method development. University of California animal care and use program, protocol, and procedures were followed and appropriate permits obtained. Pre-frozen (stored at −20 °C) tissue weighing 150 mg (wet weight, ww) was transferred to a 2-mL Eppendorf tube, and 200 mg ENVI-C18 material (Supelco, Bellefonte, PA, USA), 1.00 mL acetonitrile and 2.00 ng IS (100 μL from a 20 ng mL−1 solution) were added. Finally, two tungsten-carbide beads (3 mm, Qiagen, Valencia, CA, USA) were added to the tube and placed in the bead-beater (TissueLyser LT, Qiagen, Valencia, CA, USA) operated at 50 Hz for 10 min. The Eppendorf tube was then centrifuged at 16,900×g for 5 min (Eppendorf 5418, Hauppauge, NY, USA) at room temperature and the supernatant was transferred to a clean test tube. Another 1.00 mL acetonitrile was added to the bead-beating Eppendorf tube and re-extraction was

The HPLC separation was achieved using a reversed phase system consisting of PEEK frit guard (0.5 μm), a guard column (C18, 2.1×20 mm, 5 μm) and an analytical column (TARGA C18, 2.1×40 mm, 5 μm, all items from Higgins Analytical Inc, Mountain View, CA, USA). Mobile phase A consisted of 5 % methanol and 95 % MilliQ water, and phase B of 95 % methanol and 5 % MilliQ water. Both mobile phases contained 0.1 % formic acid by volume. The analytical system consisted of an Agilent 1260 series HPLC system (Agilent Technologies, Palo Alto, CA, USA), equipped with a degasser, autosampler using 50 μL injections. The column flow rate was 0.200 mL min−1 using the isocratic conditions (25:75, A:B), which ensured elution of tebuconazole and d6tebuconazole both at 3.30 min. Mass spectrometry detection was achieved using an Agilent 6460 triple-stage quadrupole equipped with a Jet Stream electrospray ionization (ESI) source (Agilent Technologies, Palo Alto, CA, USA). The instrument was operated in positive selected reaction monitoring mode (SRM), applying nitrogen as sheath, nebulizer, and collision gas. Using the isotopic pattern for chlorine as an advantage, the ion transitions (m/z values) were 308>70 and 310>70 for tebuconazole, and 314>72 and 316>72 for d6tebuconazole. Quadrupole 1 and 3 were operated at unit resolution. Other MS parameters were optimized to: dwell times, 200 ms; fragmentor, 120 V; collision energies, 20 V; collision cell acceleration voltage, 3 V. Collection and treatment of data were performed using Agilent MassHunter Workstation version B.05.00. Validation The LC-ESI-MS/MS response of tebuconazole was evaluated for linearity using calibration curve of standard solutions in a solvent 1:1 mixture between mobile phase A and acetonitrile. Tebuconazole concentrations were 0, 0.005, 0.010, 0.050, 0.100, 0.500, 1.00, 5.00, 10.0, 50.0, and 100.0 ng mL−1. In all vials, the IS concentration were 1.00 ng mL−1. Instrument limit of detection (LODinstr) and LOQinstr were obtained using Agilent MS MassHunter Workstation post-processing software. LODinstr was defined as lowest concentration level from calibration curve in three replicates having a softwareestimated ratio of signal-to-noise (S/N) above 3 for SRM-ion trace with lowest signal (310>70). LOQinstr was extrapolated from LODinstr multiplying with 10/3.3 [21, 22]. LODwater and LODtissue were defined as S/N=3 and both estimated from obtained S/N ratio in lowest spiked sample (i.e., 10 pg mL−1

7680

and 6.67 pg mg−1, respectively). LOQwater and LOQtissue were extrapolated from LOD multiplying with 10/3.3. Method accuracy was assessed using spike-recovery studies (pre-spike and post-spike approach [23–25]) in blank frog tank water (1–10 mL subsamples) and liver tissue from untreated adult frogs (150 mg homogenized subsamples). Method accuracies for water matrix were investigated at spike levels 0.010, 1.00, and 100 ng mL−1, and for liver at 0.67 and 6.67 pg mg−1 fw. For each fortification level, blank sample matrix (water or liver) was divided into three subgroups (A to C each with n=3). Subgroup A samples were spiked with tebuconazole prior to SPE enrichment or MSPD extraction. Subgroup B samples were spiked with tebuconazole after SPE (in SPE extract) or in MSPD extract. Moreover, in subgroups A and B, IS was spiked in SPE or MSPD extracts. Subgroup C samples were spiked with IS prior to SPE or MSPD and tebuconazole in the extracts. Consequently, absolute tebuconazole recoveries were obtained by dividing tebuconazole-to-IS ratios from subgroups A to B. Absolute IS recoveries were obtained by dividing IS-to-tebuconazole ratios from subgroups C to B. The relative recovery of tebuconazole to IS was obtained by dividing absolute recovery for tebuconazole with absolute recovery for IS [23–25]. The matrix effect was directly quantified by relating postspiked samples (subgroup B) to neat standards without matrix [26]. Extracting and analyzing fortified water samples (n=3) at two different days investigated reproducibility or day-today variation. Repeatability was assessed using coefficient of variation (CV%) from pre-spiked water samples (n=3).

Results and discussion LC-MS/MS development and optimization Tebuconazole was readily protonated and detected by the mass spectrometer when using the acidic mobile phase. The parent ion m/z 308 was selected as precursor and MS/MSproduct ion scan yielded m/z 70 and m/z 125 as product ions. These two MS/MS-product ions have been selected by others [16, 19, 20], however, we found that the product ion 125 was exceptionally low (Fig. 3). Consequently, we utilized the chlorine isotopic pattern monitoring ion transitions 308>70 and 310>70, the latter resulting in three to four times higher signal intensity and S/N when compared to 308>125 (see Electronic Supplementary Material (ESM) Fig. S1). The ratio between peak areas of 37Cl and 35Cl SRM-ion traces (310>70 and 308>70, respectively) were in the range 31–33 % in neat standards and samples (data not shown) and consistent with the theoretical isotopic pattern (31.97 %, [27]). Tebuconazole and IS were retained for 3.30 min using the guard and analytical column (2.1×20 and 40 mm, respectively) with C18-

M. Hansen et al.

material (Fig. 4a). Using the 0.5-μm PEEK frit guard did not change analyte retention (data not shown) and was changed for every 50–100 injections, ensuring longer column lifetimes. Different injection volumes (5–100 μL) were tested, and 50 μL was selected as appropriate injection volume (data not shown). Injection solvents (50 % methanol or 50 % acetonitrile) both combined with 50 % mobile phase A were tested, and no difference between solvents was observed. Acetonitrile was selected to ease sample preparation as acetonitrile is used as extraction and elution solvent. Sample preparation of water samples Using the polymeric anion exchange SPE (Plexa PAX) ensured a mixed-mode retention mechanism, as the backbone polymer is polystyrene/divinyl benzene combined with an anionic exchange mechanism (quaternary amine groups reaching out from SPE backbone). Washing the SPE immediately after extracting the water sample with 1 mL 5 % ammonia hydroxide ensured trapping of acidic and phenolic matrix components by the anion exchange mechanism. Eluting tebuconazole with acetonitrile ensured that acidic and phenolic matrix components were kept on the SPE and yielded clear extracts. Carpinteiro et al. have made similar observations for analyzing fungicides in wine. They obtained colorless extracts from red wine when using an Oasis MAX (also anionexchanger SPE, from Waters), but heavily matrix color extracts when using the neutral Oasis HLB SPE (polystyrene/ divinyl benzene polymer) [20]. In our study, water sample sizes between 1 and 10 mL were tested, and it was demonstrated that the developed method yielded good recoveries for tebuconazole and IS. The developed SPE procedure yielded quantitative absolute recoveries of 88.9±1.9, 91.8±5.3, and 84.7±3.3 % at a 10, 1000, and 100,000 pg mL−1 spike level, respectively (Table 1). A representative chromatogram from a water sample is shown in Fig. 4a. Low and slightly negative or no matrix effects were found (Table 1), this was probably caused by the aforementioned elimination of many matrix components from the SPE extract. As nicely described by others, using this approach, negative matrix effect values stand for signal suppression and positive values for signal enhancement [26, 28]. Matrix effects were further verified using post-HPLC column infusion of tebuconazole and injecting blank sample extracts containing IS [28]. Thereby, matrix effect profiles were obtained and verified merely a low ion-suppression or ion-enhancement for all matrices (Fig. 5). Importantly, dilutions of water samples (1:5, 1:50, and 1:250 in mobile phase A and acetonitrile) compared to SPE extracts verified that water matrix components were eliminated during the SPE procedure (Fig. 5). Notably, a water sample dilution greater than 1:50 could be used instead of the SPE procedure on the cost of increased HPLC guard column replacement frequency and if tebuconazole sample concentrations allows.

Tebuconazole in controlled frog exposure study

7681

Fig. 3 Mass spectra for tebuconazole (left) and the internal standard d6tebuconazole (right). The top panel is single MS full scan (range; m/z 100–500) from 200,000 pg mL−1 neat standards. The middle and lower panels display product ion scan (range; m/z 50–400) spectra when using

respectively 35Cl and 37Cl isotopes as precursor ions. Mass spectrometer parameters: positive ionization mode, fragmentor 120 V, collision cell acceleration 3 V, collision energy 20 V (only during MS/MS)

Sample preparation of frog tissue

ethyl acetate and acetonitrile as elution solvents when C18 was used as sorbent, and they found almost similar recovery from both solvents [13]. We therefore chose to use C18 in combination with acetonitrile and investigated number of required extraction cycles to obtain maximal recovery (Fig. 6). From this experiment, we selected two extraction cycles as sufficient for obtaining a recovery of approximately 90 % (Fig. 6). Finally, two fortified tebuconazole concentration levels (1.00 and 10.0 ng) in 150 mg ww tissue yielded 79.6±4.0 and 87.8±5.6 % in absolute recoveries using the optimized procedure (Table 1). Overall, the recoveries were

Bead-beating-assisted MSPD of frog tissue proved to be a rapid and accurate method. Others have combined vortex with MSPD to extract pesticides from fish liver [13], however, the authors obtained tebuconazole recoveries from 57 to 93 % and a fairly high quantification limit (125 pg mg−1). We believe that by combining MSPD with tissue bead-beating, higher absolute recoveries and a better quantification limit are ensured. In MSPD, the elution solvent and selection of sorbent are key factors. Souza Caldas et al. investigated effects of

7682

M. Hansen et al.

Fig. 4 Chromatograms of fortified A frog tank water (10 pg mL−1) and B frog liver (6.67 pg mL−1). Ion transitions: tebuconazole 308>70 and 310>70; d6-tebuconazole (IS) 314>72 and 316>72

found to be acceptable. Furthermore, the use of pre-frozen tissue will aid in lysis of tissue cells and increase release of harder matrix-bound tebuconazole, however, this was not well

investigated in the present study. As expected, we observed slightly higher matrix effects (5.8–9.9 %) when compared to water samples. This weak ion enhancement is evident from

Table 1 Method validation parameters. Frog tank water is validated using 1 mL as sample size for 1.00 and 100.0 ng mL−1 fortification levels, and a 10-mL sample size was used for 0.010 ng mL−1 level.

Frog tissue is validated using 150 mg liver fortified at two levels (6.67 and 66.7 pg mg−1)

Instrument Linear dynamic range (pg mL−1) Goodness of fit (R2) Tested concentration level (pg mL−1) Reproducibility (CV%) Repeatability (CV%)a Frog tank water Fortified concentration level (ng mL−1, n=3) Reproducibility (CV%)c Repeatability (CV%)d Absolute recovery (%) Relative recovery (%) Matrix effect (%) LOD (pg mL−1)e LOQ (pg mL−1)e Frog tissue Fortified concentration level (pg mg−1, n=3) Absolute recovery (%) Relative recovery (%) Matrix effect (%) LOD (pg mg−1)e LOQ (pg mg−1)e a

10 7.7 2.0

5–100,000 0.9992 1000 2.3 0.3

100,000 3.6 0.3

0.010b 6.9 5.4 88.9±1.9 80.6±3.5 −0.9

1.00 11.1 5.0 91.8±5.3 89.7±7.0 0.4

100 2.7 0.7 84.7±3.3 99.7±5.3 −12.0

1.28 3.89 6.67 79.6±4.0 68.1±13.0 9.9 0.21 0.63

66.7 87.8±5.6 109.1±8.3 5.8

Instrument repeatability was assessed using the coefficient of variation (%CV) from injecting neat standard six consecutive times

b

Using 10 mL as sample size

c

Method day-to-day variation, or reproducibility, was investigated in fortified samples at two different days

d

Method repeatability was assessed using CV % from the pre-spiked water samples (n=3) at three concentration levels

e

LODs are defined as the concentration where signal-to-noise ratio is 3 and are estimated from lowest fortification levels. LOQs are extrapolated as 10/3.3 times LOD. Standard deviation is given after ±

Tebuconazole in controlled frog exposure study

Fig. 5 Matrix profiles investigating ion-suppression and enhancement from blank samples. Tebuconazole (2090 ng mL−1 in MilliQ water) was post-column infused at 10 μL min−1. Blank extracts (SPE and liver), dilutions of water samples, and neat solvent was injected using optimized method (flow, 250 μL min−1). Internal standard is spiked in blank sample extracts, identifying peak retention time (gray peak) and area of interest. Vertical line at 2.80 min indicates when mass spectrometer diverted (valve) in final method. Selected reaction monitoring ion transition is 308>70 in all cases

the chromatographic profile of blank liver extract (Fig. 5), however it is not obvious from the chromatogram of a 6.67 pg mg−1 fortified frog liver extract (Fig. 4b). Validation and quality assurance

7683

or little matrix effects (Table 1 and Fig. 5). Accuracies, reported as relative recovery (tebuconazole relative to IS, e.g., see [25]), were very good for frog water samples and in the range 80.6–99.7 % over a wide concentration span (10–100,000 pg mL−1, Table 1). Previously published methods for water samples have reported LOQ in the range 2.3 to 13 ng mL−1 when using liquidliquid micro-extraction in combination with GC-MS [14]. We estimated LOD for water samples from three 10 pg mL−1 fortified water samples. In these samples, S/N was on average 23 (water sample depicted in Fig. 4a) and the derived LOD water is therefore 1.28 pg mL −1 and LOQ water is 3.89 pg mL−1 (Table 1). Hence, the presented method is orders of magnitude better than currently available methods. The tissue matrix was more problematic, however, still with acceptable relative recoveries between 68.1 and 109 % at a more narrow concentration range (6.67–66.7 pg mg−1, Table 1). Similarly, we estimated LOD for 6.67 pg mg−1 fortified frog liver tissue (Fig. 4b) by assessing the S/N ratio. Here, the S/N was on average 96 (n=3) and the derived LODtissue was 0.21 pg mg−1 and LOQtissue to 0.63 pg mg−1 (Table 1). This developed methodology offers improved LOQs when compared to existing methods (LOQ; 125 pg mg−1 in liver tissue [13]). It should be noted that the presented method quantifies the sum of the two tebuconazole enantiomers (R and S), and they have been successfully separated and described elsewhere [16].

Using d6-tebuconazole as internal standard, a linear calibration curve was obtained for tebuconazole at a vial concentration range between 5 and 100,000 pg mL−1. Higher concentrations were also tested and the curve was observed to flattening-off at the highest tested concentration probably caused by analyte self-induced ion-suppression (data not shown). Consequently, 100,000 pg mL−1 was selected as highest concentration level and a good fit was obtained (R2 =0.9992, Table 1). In addition, a matrix-matched calibration curve was considered redundant, as there were no

The developed methodology was applied on male frog (X. laevis) liver and adipose tissue from a 4-week tebuconazole repeated-exposure study. Three frogs where exposed to 10 μg L−1 tebuconazole in tank water, while a control group did not receive the fungicide. Water was exchanged every 3 days with fresh water (with or without

Fig. 6 Repeated extractions of fortified frog liver (150 mg ww, n=3, SD) spiked with 100 ng tebuconazole and post-spiked with 1.00 ng IS. Between each extraction cycle, acetonitrile was removed (after centrifugation) and analyzed, and clean acetonitrile was added and bead-beating repeated

Fig. 7 Tebuconazole levels in adult male X. laevis exposed to 10 μg L−1 for 4 weeks (n=3). Error bars display standard deviation

Application

7684

tebuconazole). We observed tebuconazole at 182.4±57.2 and 36.4±14.1 ng mg−1 fresh weight (fw, n=3) in adipose and liver tissue, respectively (Fig. 7). Used amount of tissue was 60–78 mg and 123–151 mg fw for adipose and liver, respectively. We did not find tebuconazole in controls (n=3), however for liver tissue one of three controls contained traces (0.40 ng mg−1) of tebuconazole (Fig. 7). This is likely due to contamination during sample preparation. Using our observed values, we find a bioconcentration factor (BCF) of 18,244 ± 5723 and 3644±1410 L kg−1 for adipose and liver tissue, respectively [29]. A recent study reported a median of 74 pg mg−1 tebuconazole in whole-body extracts from wild Pacific chorus frogs [12]. Our study shows a factor of around a thousand higher tebuconazole tissue levels, likely due to the relative high and repeated-exposure levels, and tissue-specific analysis. Importantly, four repeated extraction cycles of liver and adipose samples from one tebuconazole-treated frog demonstrated that we obtained quantitative recoveries 89.5 and 78.9 % of the fungicide in the first two extraction cycles from adipose and liver, respectively (see ESM Fig. S2). The latter value is lower when compared to the freshly spiked liver tissue study (90.7 %, Fig. 6), showing the importance of not only using freshly spiked matrix for assessing method performance.

Conclusions A novel method to extract and analyze tebuconazole in frog exposure studies was developed, validated, and applied. Compared to existing methodologies, the developed methodology yields a lower quantification limit, and uses a simpler and more exhaustive extraction approach to obtain better accuracy and precision. The combination of beadbeating-assisted matrix solid-phase dispersion extraction and highly sensitive LC-MS/MS apparatus yielded a tebuconazole quantification limit of 0.63 pg mg−1 wet weight frog tissue, which is orders of magnitude better than the previously published method (125 pg mg−1), with absolute recoveries between 79.6 and 87.8 %. Furthermore, water analysis using anion-exchanger solid-phase extraction to trap matrix components ensured tebuconazole a quantification limit of 3.89 pg mL−1, again an order of magnitude better than existing methods (2.3–13 ng mL−1), with absolute recoveries in the range 84.7–91.8 %. The methodology was successfully applied in a frog study investigating the environmental risk of tebuconazole. In our study, we find tebuconazole bioconcentrates to higher degree in adipose tissue compare to liver. Little is know about the fate, effects, and bioaccumulation of tebuconazole.

M. Hansen et al. Acknowledgments The authors acknowledge the assistance, input and discussions from research fellows in Hayes’ Lab. MH acknowledges financial support by the Danish Council for Independent Research | Natural Sciences, grant no. 12–131766. RP was gratefully supported by various travel and scholarship grants (Augustinus Fonden, Knud Højgaards Fond, Laura Bentzens Legat, Oticon Fonden, Niels Smed Søndergaards Thy Fond, Henry og Mary Skovs Fond, and finally Snedsted-Nørhå Sparekasses Jubilæumsfond). Furthermore, this study was supported in part by the US National Institute for Environmental Health Sciences (NIEHS) Superfund Research Program (Grant P42 ES004705) through the Superfund Research Center at University of California, Berkeley.

References 1. Mann RM, Hyne RV, Choung CB, Wilson SP (2009) Amphibians and agricultural chemicals: review of the risks in a complex environment. Environ Pollut 157:2903–2927. doi:10.1016/j.envpol.2009.05.015 2. Hayes TB, Khoury V, Narayan A et al (2010) Atrazine induces complete feminization and chemical castration in male African clawed frogs (Xenopus laevis). Proc Natl Acad Sci U S A 107: 4612–4617. doi:10.1073/pnas.0909519107 3. Hayes T, Haston K, Tsui M et al (2002) Herbicides: feminization of male frogs in the wild. Nature 419:895–896. doi:10.1038/419895a 4. Colegrove KM, Gulland FMD, Naydan DK, Lowenstine LJ (2009) Tumor morphology and immunohistochemical expression of estrogen receptor, progesterone receptor, p53, and Ki67 in urogenital carcinomas of California sea lions (Zalophus californianus). Vet Pathol 46:642–655. doi:10.1354/vp.08-VP-0214-C-FL 5. (2008) Food and Agriculture Organization, United Nations. Tebuconazole. Joint FAO/WHO Meeting on Pesticide Residues (JMPR). http://www.fao.org/fileadmin/templates/agphome/ documents/Pests_Pesticides/JMPR/Evaluation08/Tebuconazole.pdf 6. European Commission (2008) Council amending Directive 98/8/EC of 5 September 2008 to include tebuconazole as an active substance in Annex I thereto. Off J-Eur Union Legis L 239:9–11 7. European Commission (2013) Commission implementing regulation (EU) No 1038/2013 of 24 October 2013 concerning approving tebuconazole as an existing active substance for use in biocidal products for product- types 7 and 10 (98/8/EC). Off J Eur Union 208:40–41 8. Calza P, Baudino S, Aigotti R et al (2002) High-performance liquid chromatographic/tandem mass spectrometric identification of the phototransformation products of tebuconazole on titanium dioxide. Biol Mass Spectrom 37:566–576. doi:10.1002/jms.315 9. California Department of Pesticide Regulation (2013). http://calpip. cdpr.ca.gov 10. Danish Environmental Protection Agency, Pesticide Use 2011. http:// www2.mst.dk/Udgiv/publikationer/2012/10/978-87-92903-58-7.pdf 11. European Food Safety Authority (2014) Conclusion on the peer review of the pesticide risk assessment of the active substance tebuconazole. Eur Food Saf Authority 12:3485 12. Smalling KL, Fellers GM, Kleeman PM, Kuivila KM (2013) Accumulation of pesticides in Pacific chorus frogs (Pseudacris regilla) from California’s Sierra Nevada Mountains, USA. Environ Toxicol Chem 32:2026–2034. doi:10.1002/etc.2308 13. Souza Caldas S, Marian Bolzan C, Jaime de Menezes E et al (2013) A vortex-assisted MSPD method for the extraction of pesticide residues from fish liver and crab hepatopancreas with determination by GC-MS. Talanta 112:63–68. doi:10.1016/j.talanta.2013.03.054 14. Farajzadeh MA, Mogaddam MRA, Aghdam AA (2013) Comparison of air-agitated liquid-liquid microextraction technique and conventional dispersive liquid-liquid micro-extraction for determination of

Tebuconazole in controlled frog exposure study

15.

16.

17.

18.

19.

20.

21.

triazole pesticides in aqueous samples by gas chromatography with flame ionization detection. J Chromatogr A 1300:70–78. doi:10. 1016/j.chroma.2013.02.033 Wang J, Du Z, Yu W, Qu S (2012) Detection of seven pesticides in cucumbers using hollow fibre-based liquid-phase microextraction and ultra-high pressure liquid chromatography coupled to tandem mass spectrometry. J Chromatogr A 1247:10–17. doi:10.1016/j. chroma.2012.05.040 Wang X, Wang X, Zhang H et al (2011) Enantioselective degradation of tebuconazole in cabbage, cucumber, and soils. Chirality 24:104– 111. doi:10.1002/chir.21030 Caldas SS, Bolzan CM, Cerqueira MB et al (2011) Evaluation of a modified QuEChERS extraction of multiple classes of pesticides from a rice paddy soil by LC-APCI-MS/MS. J Agric Food Chem 59:11918–11926. doi:10.1021/jf202878s Šťávová J, Sedgeman CA, Smith ZT et al (2011) Method development for the determination of wood preservatives in commercially treated wood using gas chromatography-mass spectrometry. Anal Chim Acta 702:205–212. doi:10.1016/j.aca.2011.06.058 García-Valcárcel AI, Tadeo JL (2011) Determination of azoles in sewage sludge from Spanish wastewater treatment plants by liquid chromatography-tandem mass spectrometry. J Sep Sci 34:1228– 1235. doi:10.1002/jssc.201000814 Carpinteiro I, Ramil M, Rodríguez I, Cela R (2010) Determination of fungicides in wine by mixed-mode solid phase extraction and liquid chromatography coupled to tandem mass spectrometry. J Chromatogr A 1217:7484–7492. doi:10.1016/j.chroma.2010.09.080 Green JM (1996) Peer reviewed: a practical guide to analytical method validation. Anal Chem 68:305A–309A. doi:10.1021/ac961912f

7685 22. Vial J, Jardy A (1999) Experimental comparison of the different approaches to estimate LOD and LOQ of an HPLC method. Anal Chem 71:2672–2677. doi:10.1021/ac981179n 23. Hansen M, Krogh KA, Halling-Sørensen B, Björklund E (2011) Determination of ten steroid hormones in animal waste manure and agricultural soil using inverse and integrated clean-up pressurized liquid extraction and gas chromatographytandem mass spectrometry. Anal Methods 3:1087–1095. doi:10. 1039/c1ay00007a 24. Bester K (2008) Quantification with HPLC-MS/MS for environmental issues: quality assurance and quality assessment. Anal Bioanal Chem 391:15–20. doi:10.1007/s00216-008-1991-2 25. Hansen M, Björklund E, Krogh KA (2009) Analytical strategies for assessing ionophores in the environment. Trends in Analytical Chemistry 26. Stahnke H, Kittlaus S, Kempe G, Alder L (2012) Reduction of matrix effects in liquid chromatography-electrospray ionization-mass spectrometry by dilution of the sample extracts: how much dilution is needed? Anal Chem 84:1474–1482. doi: 10.1021/ac202661j 27. McLafferty FW (1993) Interpretation of mass spectra. Fred W. McLafferty. Google Books 28. Stahnke H, Reemtsma T, Alder L (2009) Compensation of matrix effects by postcolumn infusion of a monitor substance in multiresidue analysis with LC-MS/MS. Anal Chem 81(6): 2185–2192 29. Gobas F, Morrison HA (2000) Bioconcentration and biomagnification in the aquatic environment. Handbook of Property Estimation Methods for Chemicals: Environmental Health Sciences. CRC