Effects of imidacloprid on detoxifying enzyme ...

Environ Sci Pollut Res DOI 10.1007/s11356-016-6686-1

RECENT ADVANCES IN CHEMISTRY AND THE ENVIRONMENT

Effects of imidacloprid on detoxifying enzyme glutathione S-transferase on Folsomia candida (Collembola) Panwad Sillapawattana 1 & Andreas Schäffer 1

Received: 23 November 2015 / Accepted: 12 April 2016 # Springer-Verlag Berlin Heidelberg 2016

Abstract Chemical analyses of the environment can document contamination by various xenobiotics, but it is also important to understand the effect of pollutants on living organisms. Thus, in the present work, we investigated the effect of the pesticide imidacloprid on the detoxifying enzyme glutathione S-transferase (GST) from Folsomia candida (Collembola), a standard test organism for estimating the effects of pesticides and environmental pollutants on non-target soil arthropods. Test animals were treated with different concentrations of imidacloprid for 48 h. Changes in steady-state levels of GST messenger RNA (mRNA) and GST enzyme activity were investigated. Extracted proteins were separated according to their sizes by sodium dodecyl sulfatepolyacrylamide gel electrophoresis, and the resolved protein bands were detected by silver staining. The size of the glutathione (GSH) pool in Collembola was also determined. A predicted protein sequence of putative GSTs was identified with animals from control group. A 3-fold up-regulation of GST steady-state mRNA levels was detected in the samples treated with 5.0 mg L−1 imidacloprid compared to the control, while a 2.5- and 2.0- fold up-regulation was found in organisms treated with 2.5 and 7.5 mg L−1 imidacloprid, respectively. GST activity increased with increasing imidacloprid amounts from an initial activity of 0.11 μmol min−1 mg−1 protein in the control group up to 0.25 μmol min−1 mg−1 protein in the sample treated with the 5.0 mg L−1 of pesticide. By contrast, the total amount of GSH decreased with increasing Responsible editor: Philippe Garrigues * Panwad Sillapawattana [email protected]

1

Institute for Environmental Research (Biology V), RWTH Aachen University, Worringerweg 1, 52074 Aachen, Germany

imidacloprid concentration. The results suggest that the alteration of GST activity, steady-state level of GST mRNA, and GSH level may be involved in the response of F. candida to the exposure of imidacloprid and can be used as biomarkers to monitor the toxic effects of imidacloprid and other environmental pollutants on Collembola. Keywords Detoxifying enzyme . Springtail . Pesticide . Toxicological testing . Biomarker

Introduction A proper environmental risk assessment of pollutants requires knowledge of the concentration to which organisms are exposed and of their potentially toxic effects. We here present our investigations on several biomarkers that may be used for sublethal effect testing of the insecticide imidacloprid on Collembola. Imidacloprid, an insecticide neurotoxic, belongs to the chemical class of neonicotinoids, which contains a chlorinated pyridyl group or another heterocyclic group that withdraws electrons from an adjacent amino moiety in the molecule providing a partially positive charge without protonization (Stenersen 2004). Imidacloprid penetrates easily into the nervous system of insects and binds selectively to the nicotinic acetylcholine receptors (nAChR), causing false signals in cholinergic synapses which leads to overstimulation with tremors and paralysis (Stenersen 2004). Due to steric conditions at the nAChR, imidacloprid has much lower toxicity to mammals (Tomlin 2000). The use of imidacloprid and other similar neonicotinoids was restricted for 2 years (starting in December 2013) in 25 EU countries including Germany, Italy, France, and Slovenia as a result of a study that showed a link between imidacloprid and bee death (EASAC 2015). Related study about an effect of the insecticide imidacloprid

Environ Sci Pollut Res

on the arthropod community has shown that imidacloprid prevented population increase of springtail (Marquini et al. 2003). In addition, Alves et al. (2014) found that imidacloprid reduced significantly reproduction of Folsomia candida (EC20 = 0.02 mg kg−1). F. candida, colloquially called springtails, are arthropods found in soil throughout the world and have been used as standard test organisms for more than 40 years to estimate the effects of pesticides and environmental pollutants on non-target soil arthropods because the species is parthenogenetic and can be affordably maintained in the laboratory (Fountain and Hopkin 2001; Fountain and Hopkin 2005). Mainly, it is used to estimate toxic effects of pesticides and environmental pollutants on non-target soil arthropods as prominent members of the soil fauna. Glutathione (GSH) is a tripeptide of glutamic acid, cysteine, and glycine found in all aerobic cells. GSH plays a significant role in protecting cells against the destructive effects of free radicals, as well as in detoxifying endogenous metabolites and xenobiotics, such as environmental pollutants and drugs (Fournier et al. 1992). In a detoxification reaction, a suitable substituent bound to an electrophilic atom of the xenobiotic is replaced by GSH; this reaction is commonly designated as GSH conjugation. For example, in many resistant insect strains, pesticides such as lindane, dimethyl phosphorothionates, atrazine, alachlor, and dichlorodiphenyltrichloroethane (DDT) are metabolized and detoxified via this reaction. Though suitable xenobiotics and glutathione can react spontaneously, formation of glutathione conjugate in organisms is catalyzed by diverse enzyme family called glutathione S-transferases (abbreviated as GSTs) (Stenersen 2004). The conjugation of the substrate with glutathione, mostly after oxidative activation, is accomplished by nucleophilic attack of the thiol group (as thiolate anion) of the central cysteine moiety of glutathione to the substrate. The conjugates from this catalytic reaction, depending on the organism, are removed from the organism by transport and excretion or, in case of plants, are imported into vacuoles and subsequently degraded. Various isoforms of GST, with different substrate specificities covering a broad spectrum of xenobiotics, are found in organisms, giving GSTs a central role in the detoxification of organic substances. GSTs have been studied in many species such as mussels (Fitzpatrick and Sheehan 1993; Fitzpatrick et al. 1995), insects (Kostaropoulos and Papadopoulos 1998; Kostaropoulos et al. 2001; Shen and Chien 2003), earthworms (Aly and Schroeder 2008; Ribera et al. 2001), and especially mammals (Johnson et al. 1993; Antunes et al. 2009). A recent classification system separates insect GSTs into six classes including Delta, Epsilon, Omega, Theta, Sigma, and Zeta (Agianian et al. 2003; Enayati et al. 2005). The first two classes are found only in insects, whereas Sigma GSTs are found to be the most abundant and conserved among insect orders (Gewande et al. 2015). Furthermore, Sigma GSTs were

found to play an important role in detoxifying lipid peroxides (Gewande et al. 2015). GSTs are known to be implicated in the detoxification of xenobiotics in F. candida. One isozyme was sequenced and submitted to GenBank (AB509261.1) as completed complementary DNA (cDNA) sequence (Nakamori et al. 2010). Since GSTs are, in general, often inducible by xenobiotics and other substances, it has been suggested that levels of GSTs might be a promising indicator of exposure of an organism to chemical pollution (Fitzpatrick and Sheehan 1993). We here report an investigation, whether GSTs (with emphasis on the Sigma class) and GSH levels in Collembola exposed to the model chemical stressor imidacloprid can be used as biomarkers in ecotoxicity testing. Total GSH in the animals was determined in order to study the relation between GSTs and GSH regarding the response of collembolan to chemical stress.

Material and methods Collembola culture F. candida (Collembola) of European origin was propagated in synchronized laboratory culture. Collembolans were kept in plastic boxes filled with 50 g of a mixture of calcium sulfate hemihydrate (Sigma-Aldrich, Taufkirchen, Germany) and activated charcoal (Fluka, Buchs, Switzerland) in a ratio of 9:1 (w/w) and cultured at 20 ± 1 °C, a light/dark cycle of 12:12 h. The plastic box is of food grade and made of polypropylene. The study of Lithner et al. (2012) shows that leaching of polypropylene in deionized water for 3 days at 50 °C did not show any toxic effect on Daphnia magna. With respect to any effects on the test substance, imidacloprid has a low log Kow (0.57) (Hoffmann 2008) and therefore shows negligible adsorption to the plastic material used. Containers were kept moist at all times, ensuring that relative humidity of the air within the containers is close to the saturation, which can be seen by the presence of free water within the porous plaster, but avoiding generating a water film on the plaster surface. Any dead individuals were removed from the containers (OECD 2009). Collembola were fed with 0.01 g of granulated dried baker’s yeast (Dr. Oetker, Bielefeld, Germany) placed directly on the surface of the culture substrate in a small heap twice a week. For all experiments, test animals aged 40– 60 days were used (adapted from OECD guideline to obtain only adults, which are more tolerant to the sucking device during collection than juveniles). Exposure of Collembola to imidacloprid Fifty grams of fluid culture substrate was poured into the plastic container and left for hardening for 2 h. Consequently, 5 mL of imidacloprid Pestanal® (99.9 %

Environ Sci Pollut Res

purity) (Sigma-Aldrich, Taufkirchen, Germany) solution was added onto the surface of culture substrate. At this stage, the surface of the substrate was covered with contaminated water. When water film formation disappeared and the culture substrate was very humid, test animals could then be introduced to the test vessels. Per assay, 100 synchronized individuals of F. candida were placed in a plastic box saturated with 5 mL of 2.5, 5.0, and 7.5 mg L−1 imidacloprid solution (0.25, 0.5, and 0.75 mg imidacloprid kg−1 culture substrate) and water controls for a period of 48 h (seven boxes per one concentration). The dose of imidacloprid can be calculated as follows: each test vessel contains 50 g of culture substrate. Five milliliters of imidacloprid solution (2.5, 5.0, and 7.5 mg L−1) was added on the surface of the substrate. Exemplarily for the 2.5 mg L−1 imidacloprid solution, 5 mL contains 0.0125 mg imidacloprid added to 50 g of culture substrate corresponding to 0.25 mg imidacloprid kg−1 culture substrate. The concentrations of imidacloprid used in our experiments are thus well below lethal concentrations (LC50 of imidacloprid = 20.96 mg kg−1 soil, EC20 = 0.02 mg kg−1 soil (data obtained from Alves et al. 2014)). According to OECD guideline (2009), artificial soil is used as media for reproduction test in F. candida. Our study, however, focused on the direct effect of imidacloprid on molecular and cellular levels of test animals within a short exposure time. Soil as test medium was tested but found not to be practical, since in this study, a large number of animals had to be collected alive and in good condition for the subsequent enzymatic and RNA analyses. This proved to be very difficult with the soil, as water needed to be added to separate the organisms from the soil matrix. Therefore, culture substrate, where the organism moved on the surface and could be easily collected, was used. Animals were fed on the first day with dried baker’s yeast (Dr. Oetker, Bielefeld, Germany) and kept in an incubator at 20 ± 1 °C, a light/dark cycle of 12:12 h. At the end of the test, animals were fixed in liquid nitrogen and used for RNA extraction, purification of GST, and determination of GSH pool. Visual inspection of F. candida Visual inspection was carried out under binocular microscope (Zeiss, Jena, Germany). Lethality of test animals could be assessed using needle. When test animals were touched with the needle and did not move or react, they should be defined as dead. However, if they could not move but could still slightly move their antenna, they should be considered as paralysis. Preparation of cytosolic protein fraction After exposure, frozen springtails (in liquid nitrogen) were pooled in a centrifuge tube and ground to fine powder in liquid nitrogen using a pestle. Four volumes (w/v) of potassium

phosphate buffer (100 mM, pH 7.8) containing 10 mM dithioerythritol, 5 mM EDTA, and 10 mg/mL polyvinylpyrrolidone K30 (all chemicals from Sigma-Aldrich, Taufkirchen, Germany) were added to the powdered sample. The slurry was incubated for 10 min on ice and centrifuged at 20,000×g for 20 min. The supernatant was collected and filtered through Miracloth (Calbiochem, Darmstadt, Germany). Solid ammonium sulfate (Merck Millipore, Darmstadt, Germany) was added to the filtrate to 40 % saturation. After shaking for 30 min at 4 °C on a rocking platform, the precipitated proteins were collected by centrifugation at 20,000×g for 20 min. The supernatant was brought to 80 % ammonium sulfate saturation and centrifuged (20,000×g for 20 min). The protein pellet was resuspended in 0.1 M Tris buffer, pH 8.0. Protein extracts were desalted using SpinTrap G-25 (GE Healthcare, Buckinghamshire, UK) pre-equilibrated with 0.1 M Tris buffer (pH 8.0) and stored at −20 °C. Affinity chromatography of the cytosolic protein fraction The desalted enzyme extract was passed through Pierce® Glutathione Spin Column pre-equilibrated with equilibration/wash buffer (50 mM Tris, 150 mM NaCl, pH 8.0). GST binds to the GSH immobilized on the beads, and the unbound substances were removed by passing two resinbed volumes of equilibration/wash buffer three times. Target proteins were eluted with a solution of 50 mM Tris and 150 mM NaCl (pH 8.0) containing 10 mM reduced glutathione. Glutathione and salt were eluted from the filtrate using a PD10 desalting column (GE Healthcare, Buckinghamshire, UK) pre-equilibrated with 0.1 M Tris buffer (pH 8.0). Enzyme assay and protein determination The activity of GSTs was determined according to Shen and Chien (2003). Briefly, assays were performed by recording the first 2 min of GS-CDNB conjugation at 25 °C in the reaction mixture consisting of 100 μL of 50 mM GSH (Sigma-Aldrich, Taufkirchen, Germany), 40 μL of 50 mM 1-chloro-2,4-dinitrobenzene (CDNB) (Sigma-Aldrich, Taufkirchen, Germany) dissolved in ethanol, 10 μL of enzyme extract, and 0.1 M potassium phosphate buffer, pH 6.5, in a total volume of 2 mL. The reaction was started by the addition of enzyme extract and monitored at 340 nm in Beckman Coulter DU® 640 spectrophotometer (Beckman Coulter, Krefeld, Germany). Activity was calculated using an extinction coefficient of 9.6 mM−1 cm−1 as described by Habig et al. (1974). Total protein was measured according to the method of Bradford (1976) using bovine serum albumin (Bio-Rad, Hercules, USA) as a standard. One milliliter of Bradford reagent (Bio-Rad, Hercules, USA) was added to 20 μL of protein standard and unknown samples. The mixtures were then incubated for at least 5 min at room temperature, and the

Environ Sci Pollut Res

absorbance was measured at 595 nm. The standard curves were generated by plotting the absorbance at 595 nm versus their concentration, and the unknown sample concentration was determined using the standard curves. Determining a predicted protein sequence from a cDNA sequence Collembola, which were not treated with imidacloprid, were pooled and ground in liquid nitrogen. Total RNA was extracted in a 1.5-mL tube using SV total RNA isolation system according to the supplier’s instructions (Promega, Madison, USA) and then treated with RNase-free DNase I (Promega, Madison, USA). RNA was reverse-transcribed to the firststrand cDNA using M-MLV reverse transcriptase (Promega, Madison, USA) according to the manufacturer’s protocol. PCR amplification of the cDNA sample was performed using primer pair targeting the 657-bp open reading frame of transcribed region. Primers (forward primer: ATGTCGACC TACAAGTTGAC, reverse primer: TTATTCGGAGT AATCGATGTG) were designed using reference nucleotide of F. candida messenger RNA (mRNA) for putative GST (GenBank accession number AB509261.1) (Nakamori et al. 2010). PCR was performed in a Mastercycler® gradient (Eppendorf, Hamburg, Germany) for 40 cycles exploiting HotMaster™ Taq DNA Polymerase (5 PRIME, Hamburg, Germany) using cycling parameters as suggested in the user manual. The PCR products were separated on 1 % agarose gel. An amplicon (657 bp) was excised and extracted from the gel via QIAquick® Gel Extraction Kit (QIAGEN, Hilden, Germany). The purified DNA was analyzed in agarose gel and sent for sequencing (Seqlab, Goettingen, Germany). The nucleotide sequence of the cDNA was translated in silico to obtain the protein sequence. qPCR and data analysis Two hundred nanograms of total RNA extracted from animals treated with different concentrations of imidacloprid was reverse transcribed to complementary DNA. Quantitative realtime PCR was performed on an ABI Prism®7000 Sequence Detection System (Applied Biosystems, MA, USA) with three technical replicates per sample. Relative quantification of GST expression was carried out using Platinum® SYBR® Green qPCR SuperMix-UDG with ROX (Invitrogen, Carlsbad, USA). Each reaction was done in a final volume of 10 μL containing 2 μL of 10-fold diluted cDNA and 8 μL of master mix. Cycling conditions were kept constant for all assays. Primers for target gene (forward primer: CATCAACATTG CCGACAAAC, reverse primer: ACGACAGCCTTGAA CTTTGC) were based on sequences published in GenBank AB509261.1 (Nakamori et al. 2010), and YWHAZ was used as reference gene (forward primer: TCGCCCTCAA

CTTTTCCGTT, reverse primer: TGCTATCGCTTCATC GAATGCT) (de Boer et al. 2009). The characterization of reaction products was done via melting curve analysis and gel electrophoresis to make sure that no formation of primerdimer exists. Threshold cycle (Ct) for the gene of interest in both the test samples and the control sample was adjusted in relation to a reference gene using the comparative quantification algorithms—ΔCt. Determination of glutathione in Collembola Pools of F. candida from each test condition were ground in liquid nitrogen, and five volumes (w/v) of sodium-phosphate buffer (143 mM, pH 7.5) were added to the samples. The samples were centrifuged at 20,000g for 10 min at 4 °C. The supernatants were collected and determined for total GSH using a standard enzymatic recycling assay in the presence of glutathione reductase (GR) (Carl Roth, Karlsruhe, Germany) as described by Griffith (1980) and Gruhlke et al. (2010) with minor modifications. Briefly, GSH was oxidized to GSSG by 5,5′-dithiobis-(2-nitrobenzoic acid) (DTNB) (Carl Roth, Karlsruhe, Germany) and catalyzed by GR in the presence of NADPH (Carl Roth, Karlsruhe, Germany). The formation rate of 2-nitro-5-thiobenzoic acid is proportional to the glutathione concentration and was monitored at 412 nm over the period of 2 min, and the size of the GSH pool was calculated from a prepared standard curve. Acrylamide gel electrophoresis Purity and molecular masses of GST were estimated by mean of SDS-PAGE. The affinity-purified extracts were loaded on the gel and separated according to their sizes on 4–20 % precast gel (Bio-Rad, Hercules, USA). Precision Plus Protein™ Standard (Bio-Rad, Hercules, USA) was used as a molecular weight marker. Proteins were diluted 1:4 with loading buffer (Invitrogen, Carlsbad, USA) and denatured at 98 °C for 7 min. The resolved protein bands were stained with silver.

Results Predicted amino acid sequence from the PCR-amplified product The PCR-amplified GST product was sequenced and translated in silico. The sequence contains a 660-bp open reading frame encoding a GST protein of 220 amino acids (EC number 2.5.1.18). The protein sequence is shown in Fig. 1 and was submitted to GenBank (KF447154.1). The conserved N-terminal dom ain ( interval 4 –7 3) co nt a i ns t h e thioredoxin-fold domain, and a C-terminal domain (interval 83–184) consists of alpha helices. As a whole, the

Environ Sci Pollut Res Fig. 1 A complete nucleotide sequence of the tentative glutathione S-transferase cDNA. Amino acid sequence is exhibited below the nucleotide sequence. The region of a thioredoxin-like domain is underlined in black, and C-terminal domain is underlined in green

nucleotide sequence of GST cDNA mainly agrees with that published previously (Nakamori et al. 2010). In addition, some bases which had not been defined could be completely specified. GST activity and GSH pool in F. candida as affected by imidacloprid Imidacloprid exposure had a dose-dependent effect on the condition of Collembola. At the lowest concentration (2.5 mg L−1), no difference in behavior between animals in control and test groups was observed. However, at a concentration of 5.0 mg imidacloprid L−1, inactivity (immobility) of the test animals was observed, and at 7.5 mg L−1 imidacloprid, about 20 % of test animals were paralyzed and dead. In order to study F. candida GST and their reaction to chemical stress, after a 48-h pesticide exposure (5 mL/culture box of 2.5, 5.0,

and 7.5 mg L−1 imidacloprid), GST was extracted from test animals and activity was measured using CDNB as a model substrate. As shown in Fig. 2, compared to the control, increasing GST activity was detected with increasing imidacloprid concentration from 0 (control) to 5.0 mg L−1. The highest GST activity was detected in the 5.0 mg L−1 treated organisms at a rate of 0.25 μmol min−1 mg−1 protein. However, the lowest enzyme activity was detected in the assays with the highest imidacloprid concentration (7.5 mg L −1 ) at a rate of 0.12 μmol min−1 mg−1 protein, which is similar to the value obtained from the untreated control (0.11 μmol min−1 mg−1 protein). In the control group, GSH concentration in the extracts of Collembola was 20.59 μg mL−1. By contrast, 4.64 and 3.69 μg mL−1 GSH are found in extracts of Collembola treated with 2.5 and 5.0 mg L−1 imidacloprid, respectively. In test animals treated with 7.5 mg L−1 imidacloprid, GSH concentration was below the limit of quantification (see Fig. 3).

Environ Sci Pollut Res

Fig. 3 Size of GSH pool from F. candida treated with different concentrations of imidacloprid. The size of the GSH pool decreased with increasing concentration of imidacloprid. Tests were repeated three times. Error bars indicate standard deviations. Single asterisks indicate significance at P < 0.05

GST mRNA steady-state levels

Fig. 2 GST activity of protein extracts and GST steady-state mRNA levels of F. candida. a GST activity of protein extracts from test animals exposed to different concentrations of imidacloprid determined after ammonium sulfate precipitation by means of CDNB as a model substrate. The highest GST activity was found in 5.0 mg L−1 treated extract at a rate of 0.25 μmol min−1 mg−1 protein, whereas the lowest enzyme activity was detected in control extract at a rate of 0.11 μmol min−1 mg−1 protein. b mRNA expression of Collembola treated with imidacloprid normalized by reference gene YWHAZ (tyrosine 3-monooxygenase). The highest GST activity was found in 5.0 mg L−1 treated extract at a rate of 0.25 μmol min−1 mg−1 protein, whereas the lowest enzyme activity was detected in control extract at a rate of 0.11 μmol min−1 mg−1 protein. Measurements were conducted three times. Error bars indicate standard deviations. Single asterisk indicates significance at P < 0.05; double asterisks indicate significance at P < 0.01

Average threshold cycle (Ct) values obtained from qPCR varied between treatments with different concentrations of imidacloprid. The alteration of the steady-state level of GST mRNA is shown in Fig. 2. Collembola treated with 5.0 mg L−1 imidacloprid presented ca. a 3-fold up-regulation of a target gene compared to control, whereas a 2.5-fold up-regulation was observed in case of 2.5 mg L −1 imidacloprid. At 7.5 mg L−1 imidacloprid, the expression level was two times higher than that in the control.

Discussion In 48-h experiments, Collembola were directly exposed to imidacloprid-contaminated water on culture substrate (without soil). Imidacloprid concentrations used in the experiments are based on the field application rate (55–

Purification of GST GST in protein extracts (obtained from the animals as described in BMaterial and methods^ section) was immobilized onto agarose beads (6 % cross-linked agarose resin) and eluted using a buffer containing reduced glutathione. The protein fractions supposed to contain enzymes related to glutathione and glutathione metabolism were separated according to their molecular mass on a 4–20 % SDS-PAGE gel (see Fig. 4). The most intense band of GST subunits from each test condition appeared at the molecular mass of 25 kDa. New GST subunits were not found in protein extracts from Collembola treated with imidacloprid, and GST subunits were not able to separate on PAGE.

Fig. 4 Silver-stained polyacrylamide gel electrophoresis of purified protein extract of F. candida after affinity chromatography. Bands: 1 protein standard; 3 control (without imidacloprid); 5 2.5 mg L−1 imidacloprid treatment; 7 5.0 mg L −1 imidacloprid treatment; 9 7.5 mg L−1 imidacloprid treatment

Environ Sci Pollut Res

140 g ha−1 (Cal 2006; EFSA 2008)). However, they could not directly be considered as environmental relevance, since the tests were performed on culture substrate instead of natural soil. Hence, the real concentrations can then be interpreted differently. Visible damage of Collembola was found in 5.0 and 7.5 mg L−1 imidacloprid-treated group. Nevertheless, only living animals were selected for subsequent analysis. With 2.5 mg L−1 of imidacloprid, visual inspection showed that movement as well as the vitality of test animals was the same as found in the control group, but cellular alterations (GST activity and GSH level) were already detected. The tendency shown for GST activity in F. candida was that it rose with an increasing concentration of pesticide and began to sink when test animals were treated with the highest dose probably because the acute toxicity threshold was surpassed (5–7 % mortality). An increase of GST activity is in agreement with the study of this species exposed to cadmium at 2, 4, and 10 days by Maria et al. (2014). Despite reports about non-inducible GSTs in other soil invertebrates such as earthworms during chemical exposure (Stokke and Stenersen 1993), GST activity in Collembola is inducible and changes of GST activity may be considered as a potential cellular endpoint used as a biomarker for environmental assessment of chemical stress by environmental pollutants. If GST activity was indeed being induced by imidacloprid as stressor, we reasoned that, at the gene level, it may be possible that GST mRNA is also induced. To assess this hypothesis, we analyzed the expression profile of GST mRNAs via qPCR. At the lowest concentration examined (2.5 mg L−1 imidacloprid), imidacloprid altered significantly the steady-state level of GST mRNA compared to the control. At this concentration, visible damage of test animals was not observed. Collembola treated with 5.0 mg L−1 imidacloprid presented a 3-fold increase in GST mRNA concentration compared to control Collembola, whereas 2.5- and 2-fold increase was shown in organisms treated with 2.5 and 7.5 mg L−1 imidacloprid, respectively. In case of the most intense dose, GST mRNA concentration was slightly higher than control. This result demonstrated that imidacloprid has an effect on GST enzyme activity and possible gene expression level because of the correlation between elevated GST activity and an up-regulation of GST steady-state mRNA levels. When test animals were exposed to concentrations between 2.5 and 5.0 mg L−1 imidacloprid, enzyme activity and expression level were increased, but oppositely at the highest imidacloprid concentration, both parameters declined possibly because of toxic effects of the insecticide. Likewise, Nakamori et al. (2010) found an expression of GST from F. candida exposed to cadmium-contaminated soil and stated that it may be used as a sensitive biomarker of environmental pollution. The relation between GST and GSH was also studied in this present work. The size of the GSH pool in F. candida treated with imidacloprid was less than in untreated

individuals. Moreover, the amount of GSH in treated test animals decreased as imidacloprid concentration rose. The change of total GSH amount is similar to the effect of allicin on the glutathione pool of yeast cells in which treated cells contain smaller amounts of total GSH than untreated cells (Gruhlke et al. 2010) as well as a decline in the total GSH concentration that has been reported in earthworms (Eisenia fetida andrei) exposed to benzo(a)pyrene spiked artificial soil (1 g kg−1 soil) (Saint-Denis et al. 1999) Our results indicated that three paradigms are possibly relevant to the effect of imidacloprid on cellular changes in Collembola. In the first scenario, part of the GSH pool would be used in a GST-catalyzed reaction to form imidacloprid conjugates. In the second scenario, imidacloprid metabolites produced after the oxidative cleavage by phase I enzymes undergo further metabolism via glutathione conjugation. This would be consistent with the proposed model of imidacloprid metabolism in rats (Klein 1987; NPIC 2015). Briefly, one of the major pathways is the oxidative cleavage of imidacloprid, yielding 6-chloronicotinic acid and imidazolidine. The former undergoes further metabolism via glutathione conjugation to form mercaptonicotinic acid and hippuric acid. Thirdly, GSTs might contribute to the removal of products formed by reaction of imidacloprid or its metabolites with reactive oxygen species (ROS) (Enayati et al. 2005). For this scenario, it is possible that imidacloprid acts as oxidative stressor and disturbs the normal redox state of cells and can cause toxic effects through the production of peroxides and free radicals. The metabolites from oxidative reactions are then glutathiolated by GSTs. In this case, Sigma GST might probably play a role due to the report that Drosophila glutathione S-transferase-2 (GST2) (Sigma GST) exhibits preference for catalyzing GSH conjugation to lipid peroxidation products, indicating an anti-oxidant role (Agianian et al. 2003). Moreover, in Arabidopsis, some GSTs are strongly induced by H2O2 (Wagner et al. 2002). It is accepted that GSTs share greater than 60 % identity within a class, while those less than 30 % identity are designated to separate classes (Kostaropoulos et al. 2001). The primary structure at the N-terminus is emphasized due to the conserved region including an important part of an active site. This region contains a thioredoxin-like domain (interval 4– 73), which is found in many proteins that interact with the thiol group of GSH (G-site) (Kostaropoulos et al. 2001) and thus lowers its pKa from around 9.0 to approx 6–7. This is considered to be a key component of catalysis in GSTs (Atkins et al. 1993). The bigger C-terminal domain (interval 83–184) contains a substrate-binding pocket (H-site) and a variable number of alpha helices. The cytosolic GSTs are hetero- or homo-dimeric proteins with a molecular weight of approx. 25 kDa (Stenersen 2004). The polypeptide chain of each monomer folds into two domains joined by a variable linker region. As illustrated in

Environ Sci Pollut Res

Fig. 4, affinity-purified extracts from test animals under different test conditions (control animals and those treated with 2.5, 5.0, and 7.5 mg L−1 imidacloprid) were identified by SDSPAGE. Different subunits could not be well separated resulting in broad protein bands on the gel. The molecular mass of the most intense band for every test condition is approx. 25 kDa and is presumed to be GST. Since each subunit has a similar molecular weight, the SDS-PAGE method alone was not sufficient to completely separate them. Thus, further separation by isoelectric focusing (IEF) was performed. By exploiting the fact that the net surface charge at a given pH is unique for a specific protein, affinity-purified proteins were run on a pH 5– 8 gradient gel. Two separated protein bands of each experimental condition appeared on the gel at the same positions, and no new GST subunit was induced (data not shown). Because of their distinct isoelectric points (pI), we assume that F. candida GSTs probably consist of two different subunits, which have a similar size around 25 kDa. The results suggested that imidacloprid could elevate GST activities but does not induce new GST subunits in F. candida, in accordance to the same findings in Drosophila melanogaster exposed to phenol that GSTs are inducible, whereas no new subunits were induced (Shen and Chien 2003).

Conclusions Imidacloprid has an effect not only on activity of stress protein (GSTs) but also possibly on the gene expression level of GST mRNA, which can be demonstrated by the correlation between an elevation of GST activity and an up-regulation of GST steady-state mRNA levels. The change of GST activity and GSH level in F. candida may serve as possible endpoints in ecotoxicological risk assessments. Moreover, the expression of GST might be a sensitive biomarker to study effects of chemicals on Collembola.

Acknowledgments This study was sponsored by the Ministry of Science and Technology, Thailand. We would like to thank Prof. Alan J. Slusarenko for advice and permission to use facilities in his laboratory and for critical reading of the manuscript. Thanks are also due to Dr. Martin C.H. Gruhlke and Dr. Marco Loehrer for their valuable discussions and technical help. We would like to express our gratefulness to Prof. Juliane Filser (University of Bremen, Germany) for kindly providing the F. candida culture.

References Agianian B, Tucker PA, Schouten A, Leonard K, Bullard B, Gros P (2003) Structure of a Drosophila Sigma class glutathione Stransferase reveals a novel active site topography suited for lipid peroxidase products. J Mol Biol 326:151–165

Alves PRL, Cardoso EJBN, Martines AM, Sousa JP, Pasini A (2014) Seed dressing pesticides on springtails in two ecotoxicological laboratory tests. Ecotoxicol Environ Saf 105:65–71 Aly MAS, Schroeder P (2008) Effect of herbicides on glutathione Stransferases in the earthworm, Eisenia fetida. Environ Sci Pollut Res 15:143–149 Antunes SC, Marques SM, Pereira R, Goncalves F, Nunes B (2009) Testing procedures for the determination of several biomarkers in different species, for environmental assessment of pollution. J Environ Monit 12:1625–1630 Atkins WM, Wang RW, Bird AW, Newton DJ, Lu AHW (1993) The catalytic mechanism of glutathione S-transferase (GST): spectroscopic determination of the pKa of Tyr-9 in rat alpha 1-1 GST. J Biol Chem 268:19188–19191 Bradford MM (1976) A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem 72:248–254 Cal EPA (2006) Imidacloprid: risk characterization document-dietary and drinking water exposure de Boer ME, de Boer TE, Marien J, Timmermans MJTN, Nota B, van Straalen NM, Ellers J, Roelofs D (2009) Reference genes for QRTPCR tested under various stress conditions in Folsomia candida and Orchesella cincta (Insecta, Collembola). BMC Mol Biol 10(54):1–11 EASAC (2015) Ecosystem services, agriculture and neonicotinoids. EASAC policy report 26 EFSA (2008) Conclusion regarding the peer review of the pesticide risk assessment of the active substance imidacloprid Enayati AA, Ranson H, Hemingway J (2005) Insect glutathione transferases and insecticide resistance. Insect Mol Biol 14:3–8 Fitzpatrick PJ, Sheehan D (1993) Separation of multiple forms of glutathione S-transferase from blue mussel, Mytilus edulis. Xenobiotica 23(8):851–861 Fitzpatrick PJ, Sheehan D, Livingstone DR (1995) Studies on isoenzymes of glutathione S-transferase in the digestive gland of Mytilus galloprovincialis with exposure to pollution. Mar Environ Res 39: 241–244 Fountain MT, Hopkin SP (2001) Continuous monitoring of Folsomia candida (Insecta: Collembola) in a metal exposure test. Ecotoxicol Environ Saf 48:275–286 Fountain MT, Hopkin SP (2005) Folsomia candida (Collembola): a standard soil arthropod. Annu Rev Entomol 50:201–222 Fournier D, Bride JM, Poire M, Berge JB, Plapp FW (1992) Insect glutathione S-transferases: biochemical characteristics of the major forms from houseflies susceptible and resistant to insecticides. J Biol Chem 267:1840–1845 Gewande ND, Subashini SS, Muragan M, Subbarayalu M (2015) Molecular screening of insecticides with sigma glutathione Stransferases (GST) in cotton aphid Aphis gossypii using docking. Bioinformation 10(11):679–683 Griffith OW (1980) Determination of glutathione and glutathione disulfide using glutathione reductase and 2-vinylpyridine. Anal Biochem 106:207–212 Gruhlke MCH, Portz D, Stitz M, Anwar A, Schneider T, Jacob C, Schlaich NL, Slusarenko AJ (2010) Allicin disrupts the cell’s electrochemical potential and induces apoptosis in yeast. Free Radical Biol Med 49:1916–1924 Habig WH, Pabst MJ, Jakoby WB (1974) Glutathione S-transferases: the first enzymatic step in mercapturic acid formation. J Biol Chem 249(22):7130–7139 Hoffmann EJ (2008) Identification and characterization of key insecticide performance mechanisms for the control of plum curculio (Conotrachelus nenuphar) in Michigan tart cherries. Dissertation, Michigan State University Johnson JA, Barbary AE, Kornguth SE, Brugge JF, Siegel FL (1993) Glutathione S-transferase isoenzymes in rat brain neurons and glia. J Neurosci 13(5):2013–2023

Environ Sci Pollut Res Klein O (1987) [14C]-NTN 33893: biokinetic part of the general metabolism study in rat. Report No. PF2889, dated 9 November 1987, submitted to WHO by Bayer AG, Mannheim, Germany. INCHEM toxicological evaluations: imidacloprid; International Programme on Chemical Safety. World Health Organization, Geneva Kostaropoulos I, Papadopoulos AI (1998) Glutathione S-transferase isoenzymes expressed in the three developmental stages of insect Tenebrio molitor. Insect Biochem Mol 28:901–909 Kostaropoulos I, Papadopoulos AI, Metaxakis A, Boukouvala E, Papadopoulou-Mourkidou E (2001) Glutathione S-transferase in the defence against pyrethroids in insects. Insect Biochem Mol 31: 313–319 Lithner D, Nordensvan I, Dave G (2012) Comparative acute toxicity of leachates from plastic products made of polypropylene, polyethylene, PVC, acrylonitrite-butadiene-styrene, and epoxy to Daphnia magna. Environ Sci Pollut Res 19:1763–1772 Maria VL, Ribeiro MF, Amorim MJB (2014) Oxidative stress biomarkers and metallothionein in Folsomia candida—response to Cu and Cd. Environ Res 133:164–169 Marquini F, Picanco MC, Guedes RNC, Ferreira PSF (2003) Imidacloprid impact on arthropods associated with canopy of common bean. Neotrop Entomol 32(2):335–342 Nakamori T, Fujimori A, Kinoshita K, Ban-nai T, Kubota Y, Yoshida S (2010) mRNA expression of a cadmium-responsive gene is a sensitive biomarker of cadmium exposure in the soil collembolan Folsomia candida. Environ Pollut 158:1689–1695

NPIC (2015) Imidacloprid fact sheet. Oregon State University. http:// npic.orst.edu/reports/NPIC14AR.pdf. Accessed 3 Feb 2015 OECD (2009) Guidelines for testing chemicals no. 232: Collembolan reproduction test in soil Ribera D, Narbonne JF, Arnaud C, Saint-Denis M (2001) Biochemical responses of earthworm Eisenia fetida Andrei exposed to contaminated artificial soil, effects of carbaryl. Soil Biol Biochem 33:1123–1130 Saint-Denis M, Narbonne JF, Arnaud C, Thybaud E, Ribera D (1999) Biochemical responses of the earthworm Eisenia fetida Andrei exposed to contaminated artificial soil: effects of benzo(a)pyrene. Soil Biol Biochem 31:1837–1846 Shen S, Chien C (2003) Induction of glutathione S-transferases activities in Drosophila melanogaster exposed to phenol. Arch Insect Biochem 53:80–91 Stenersen J (2004) Chemical pesticides: mode of action and toxicity, 1st edn. CRC Press, Boca Raton Stokke K, Stenersen J (1993) Non-inducibility of the glutathione transferases of the earthworm Eisenia ferida andrei. Comp Biochem Physiol 106C:753–756 Tomlin CDS (2000) A pesticide manual: a world compendium, 12th edn. BCPC publication, Berkshire Wagner U, Edwards R, Dixon DP, Mauch F (2002) Probing the diversity of the Arabidopsis glutathione S-transferase gene family. Plant Mol Biol 49:515–532