Nano-sized zeolites as modulators of thiacloprid toxicity on Chironomus riparius

Zeolites

Zeolites are microporous, crystalline aluminosilicates, abundant in numerous natural and synthetic modifications with the generalized formula M_x∕n[(AlO₂) _x(SiO₂)_y] . z H₂O (n = charge of M) (Baerlocher, Olson & Meier, 2001). Their framework consists of AlO${}_4^{-}$—and SiO₄—tetrahedra with aluminum- and silica-atoms interconnected through oxygen-bridges (Masters & Maschmeyer, 2011). Zeolites show a very homogeneous pore size distribution with a high (up to 1,000 m²/g) internal pore surface area. Due to their highly microporous structure and structural anionic exchange sites, zeolites serve as cation exchangers and molecular sieves.

The structure of the zeolite Y30 (Thermo Fisher Scientific, USA) is of faujasite type. Faujasite type zeolites consist of sodalite cages and their pores are orthogonal to each-other. The surface area of our Y30 zeolites is 780 m²/g, the SiO₂/Al₂O₃ mole ratio is 30:1 and the pore volume is 0.56 cm³/g (based on N₂ physisorption). The average pore-diameter of Y-type zeolite is 0.74 nm and the size of a unit cell is ∼3 nm, depending on Si/Al ratio, concentration of counter-cations and degree of hydration (Scherzer, 1978). To obtain nano-sized particles we milled the purchased Y30 particles using a planetary mill (Pulverisette 5; Fritsch GmbH) for 30 min in zirconium oxide beakers using zirconium oxide milling balls of 0.1 mm diameter (Retsch GmbH). After dry sieving through a 0.063 mm grid to recover the milling balls, the milled zeolite was washed thrice in distilled water and methanol and centrifuged (Herolab HiCen 21) for 10 min at 5,000 rpm to eliminate larger particles by sedimentation. Sedimentation time and sedimentation velocity were calculated using Stoke’s law (Batchelor, 2000).

Beta(OH)-III zeolites were synthesized at the Center for Ordered Nanoporous Materials Synthesis, POSTECH (Korea). The reagents for the Beta(OH)-III zeolites synthesis included tetraethylammonium hydroxide (TEAOH, 35% aqueous solution, Aldrich), aluminum metal (99.5%, Alfa Aesar), and fumed silica (100%, Degussa). The aluminum metal source was first mixed with a solution of TEAOH in water. Then, fumed silica was added to this mixture. The chemical composition of the synthesis mixtures used here was 43.2TEAOH⋅0.8Al₂O₃⋅80SiO₂⋅1200H₂O (Camblor, Corma & Valencia, 1998). After stirring at room temperature for 24 h, the synthesis mixture was loaded into a Teflon-lined 45 mL autoclave and heated at 140 °C for 6 d with agitation (60 rpm) under autogenous pressure. The solid products were then recovered by centrifugation (15,000 rpm for 10 min), washed repeatedly with distilled water, and dried overnight at room temperature. The resulting particles with a size of 50 nm were calcinated in air at 500 °C for 8 h to remove residues of the organic template TEAOH and were refluxed twice in 1 mol/L ammonium nitrate aqueous solution at 80 °C for 6 h. After a secondary calcination in air at 550 °C for 2 h the particles (H-Beta(OH)-III) were obtained for further tests.

Test concentrations and preparation of test solutions

We chose three concentrations of Y30 zeolite (low, medium and high) to obtain 30, 60 or 97% sorption of the total thiacloprid, as calculated by the following equation: (1)${\displaystyle r_{SW}=\frac{\frac{1}{f_W}-1}{K_d}}$

Where

r_SW –concentration of adsorbent (mass of sorbent per volume of aqueous phase)

f_W –fraction of the adsorbate present in the water phase

K_d –distribution coefficient.

Thus, the applied zeolite concentrations were 5.2 mg/L (low), 18.2 mg/L (medium) and 391.7 mg/L (high), respectively, and we used these concentrations (±5% handling variability) also for the experiments with H-Beta(OH)-III particles. For each of the two zeolite sorbents, we tested the three particle concentrations solely, as well as in mixtures with 1.0 µg/L thiacloprid. Moreover, the effects of 1.0 µg/L thiacloprid solely were assessed and a control treatment with filtered and dechlorinated tap water (filtered by iron and active carbon filter) was run simultaneously. Consequently, eight different treatments were tested within one experiment, with n = 15. Test concentrations were prepared with filtered and dechlorinated tap water directly before the experiment started. A stock solution of 5.0 mg/L thiacloprid (analytical standard; Sigma Aldrich, Darmstadt, Germany) in demineralized water was prepared prior to testing and was stirred overnight in the dark at 7 °C. Y30 zeolites were available in an aqueous stock suspension of 2.32 g/L zeolites dispersed in aqua dest.. This stock suspension was stirred via ultrasonication (15 min) before it was diluted to test concentrations. H-Beta(OH)-III zeolites suspensions were prepared directly from dry zeolite powder and were dispered in the test medium by ultrasonication (15 min). After setting up the test concentrations for the experiments, all dispersions were left to rest for 1 h to allow adsorption of thiacloprid to zeolites in the mixture treatments.

Maintenance of Chironomus riparius

Chironomus riparius (Diptera, Chironomidae) belongs to one of the most abundant insects in freshwater systems and is of high ecological relevance (Armitage, Cranston & Pinder, 1995; Langer-Jaesrich, Köhler & Gerhardt, 2010). Larvae of this species are sediment-dwelling and feed on detritus (Armitage, Cranston & Pinder, 1995). Moreover, this midge is a well-established model organism in ecotoxicology (e.g., OECD, 2004a; OECD, 2004b; OECD, 2010; OECD, 2011) and toxic effects of thiacloprid on this organism have been described already (Langer-Jaesrich, Köhler & Gerhardt, 2010; Lorenz et al., 2017).

The animals of our study originated from a stock culture established at our laboratory at the University of Tübingen (Germany), originating from a C. riparius larvae stock obtained from Goethe University Frankfurt/Main (Germany) in 2013. The animals were kept in a climate chamber at a temperature of 21.0 ± 0.5 °C and a light–dark cycle of 16:8 h. Larvae were reared in basins (30 × 55 × 12 cm), containing a 3 cm thick layer of fine quartz sand (particle size: 0.1–0.3 mm) and filtered and dechlorinated tap water with a continuous, gentle aeration. They were fed every second day with ground TetraMin^® fish flakes (Tetra, Melle, Germany), and 50% of the water volume was changed once per week. Moreover, all basins were covered by breeding cages (55 × 65 × 120 cm; mesh material with a mesh size of 0.5 mm²) to allow swarming and mating of adult midges.

Acute toxicity test

Test vessels (glass, diameter: 7.0 cm, height: 6.5 cm) were saturated overnight with the corresponding test solution, even though only low, if any, adsorption of thiacloprid on glass was expected due to the chemical properties of this compound. Earlier studies under the same experimental conditions did not reveal any adsorption of thiacloprid to glass (Lorenz et al., 2017). As glass (SiO₂) is slightly negatively charged at the pH applied (7.8 ± 0.2) we also exclude the zeolites (zeta potential Y30: −42 mV; Beta(OH)-III. −40 mV) to bind the surface of the vessels. Subsequently, the test vessels the test vessels were emptied before 30 g (dry wt.) fine quartz sand (SiO₂ particle size: 0.1–0.3 mm; adsorption of zeolites and thiacloprid was negligible for the above-mentioned reasons) and 100 mL of the respective test solution were filled into them. Prior to filling the vessels, all dispersions and solutions were stirred by ultrasonication for 15 min before they were partitioned. So, zeolites were supposed to be equally distributed within the dispersions and, consequently, also in test vessels. To adjust the temperature of the test dispersions/solutions to those of the breeding basins, we incubated the test vessels for 2 h in the climate chamber before the temperature of the medium was measured and larvae were added. Five individuals of fourth instar larvae of C. riparius were added to each test vessel using a blunt glass pipette. The test vessels were covered with perforated Parafilm^® (Carl Roth GmbH, Karlsruhe, Germany) to minimize evaporation and placed in a random arrangement on a desk within the climate chamber. Larvae were exposed for 96 h and no food was added to avoid potential interactions between food and test substances. Every 24 h we checked the test vessels and collected data on mortality (i.e., the absence of movement up to 30 s after gentle stimulation, or untraceable larvae) and on behavioral disruption, i.e., convulsions correlated with the disability to bury themselves into the sediment. Moreover, we removed dead animals to avoid contamination of the test vessels by carcass decomposition products. Our criterion for test validity was a mortality rate of ≤10% in the control.

Chemical analyses

Thiacloprid analyses were conducted for water samples and larvae. The medium was analyzed to determine whether actual and nominal aqueous thiacloprid concentrations matched, and to confirm adsorption of thiacloprid to zeolites. For each tested concentration, one replicate was analyzed. Water samples were taken from the exposure experiments directly before larvae were introduced (0 h) and, additionally, at the end of the exposure (96 h). The samples were stored in 1.5 mL Eppendorf tubes (Eppendorf, Germany) at −18 °C. Prior to analysis, samples were centrifuged at 3,000 rpm for 3 min, and 1.0 mL supernatant was filtered through a PTFE syringe filter (pore size 0.45 µm, Chromafil; Macherey-Nagel, Düren, Germany). Thiacloprid concentrations were quantified by LC-MS. Internal thiacloprid concentrations of C. riparius larvae were determined for surviving individuals after termination of the exposure experiment. Animals originated from all treatments, excluding those which were exposed to zeolites solely. Moreover, larvae were analyzed both with gut contents (these larvae were frozen directly in liquid nitrogen after the termination of the exposure) as well as with depleted guts. For gut depletion, larvae were transferred to one Petri dish per treatment which was filled with 150 mL filtered and dechlorinated tap water after termination of the exposure experiment. Petri dishes were partially wrapped in aluminum foil to reduce light-induced stress (Hirabayashi & Wotton, 1998). Under these conditions, the larvae were allowed to empty their guts for 24 h, subsequently frozen in liquid nitrogen, and stored at −80 °C. To quantify thiacloprid within the larvae, a liquid-liquid extraction method was developed based on the QuEChERS extraction procedure that originally was used in food analysis but recently adjusted to analyze environmental samples and biota (Anastassiades et al., 2003; Berlioz-Barbier et al., 2014). For sample extraction and cleanup, 10 –30 mg of frozen larvae tissue was homogenized in liquid nitrogen with a Micro-homogenizer (Carl Roth GmbH, Karlsruhe, Germany) in a 2.5 mL tubes (Eppendorf, Hamburg, Germany). A volume of 20 µL deuterated internal standard thiacloprid-D4 in methanol (Sigma-Aldrich, St. Louis, MO, USA) was added, resulting in a final concentration of 8.0 µg/L in the analyzed sample. Subsequently, 0.5 mL water (chromasolv; Sigma-Aldrich, St. Louis, MO, USA) and 0.5 mL acetonitrile were added, and the sample was shaken for 20 s using a Vortex mixer. After addition of 25 mg NaCl (Sigma-Aldrich, St. Louis, MO, USA) and 75 mg anhydrous MgSO₄ (Sigma-Aldrich, St. Louis, MO, USA) the sample was immediately shaken for 20 s using a Vortex mixer and centrifuged for 2 min at 10,000 rpm. A volume of 0.5 mL of the acetonitrile phase was transferred to 12 mg primary-secondary amine (PSA; Agilent Technologies, Waldbronn, Germany) and 90 mg anhydrous MgSO₄, then shaken for 20 s and centrifuged for 2 min at 10,000 rpm. The supernatant was concentrated in a nitrogen stream at room temperature, and the residue was reconstituted in 250 µL methanol and filtered with a PTFE syringe filter (pore size 0.45 µm, Chromafil; Macherey-Nagel, Düren, Germany). Also these samples were analyzed by LC-MS.

For LC-MS analysis, a 1260 Infinity LC system coupled to a 6550 iFunnel QTOF LC/MS system (Agilent Technologies, Waldbronn, Germany) was used. Aliquots of 10 µL were injected onto a Zorbax Eclipse Plus C18 column (2.1 × 150 mm, 3.5-Micron narrow bore; Agilent Technologies, Santa Clara, CA, USA) with a narrow bore guard column (2.1 × 12.5 mm, 5-Micron; Agilent Technologies, Santa Clara, CA, USA). A jetstream electrospray ionization (ESI) source was operated with a nebulizer pressure of 35 psig, drying gas temperature of 160 °C, a flow rate of 16 L/min, and a fragmentor voltage of 360 V. In the positive mode, capillary voltage was set to −4,000 V, skimmer voltage to 65 V, and the nozzle voltage to −500 V. The mass range was 80–1,200 m/z with a data acquisition rate of 1 spectrum/s. The sheath gas temperature was set to 325 °C with a flow rate of 11 L/min. For internal calibration, purine and HP0921 (m/z  = 121.0508, 922.0097; Agilent Techologies, Santa Clara, CA, USA) were used, the mass range was set to 50 ppm. A gradient elution was performed at a flow rate of 0.3 mL/min using water and methanol both containing 0.1% formic acid. The initial content of 95% water was decreased after 1 min to 5% water over 7 min and, after another 7 min at 5%, increased to 95% water over 30 s. Data analysis was performed with MassHunter software (Agilent Technologies, Santa Clara, CA, USA). For quantification of water-samples, a calibration curve in the range of 0.2-10 µg/L thiacloprid was used (r² = 0.9991), the limit of detection was 0.2 µg/L. Thiacloprid concentrations in larvae were calculated based on peak area of the deuterated internal standard of 8.0 µg/L and the limit of detection was 1.0 µg/L.

LA-ICP-MS imaging

To verify ingestion of zeolites by the larvae, LA-ICP-MS imaging techniques for the detection of aluminum within thin sections of larvae were used. Therefore, 10 surviving larvae, exposed to the highest zeolite concentrations or the control, were fixed in 2% glutardialdehyde (Sigma Aldrich, Darmstadt, Germany) buffered in 0.005 M cacodylate (sodium cacodylate trihydrate, pH 7.4; Sigma Aldrich, Darmstadt, Germany). Fixed larvae were stored at 4 °C for at least one week. Afterwards, samples were de-calcified in 5% TCA (trichloroacetic acid, ≥99% p.a.; Carl Roth GmbH, Karlsruhe, Germany) in formol (37% p.a., stabilized with methanol; Carl Roth GmbH, Karlsruhe, Germany), dehydrated with ethanol and routinely processed for Technovit^® embedding (Technovit^® 7100; Heraeus Kulzer GmbH, Germany). Embedded samples were cut by a rotation microtome (Leica RM2265; Leica Biosystems Nussloch GmbH, Wetzlar, Germany) into sections of 7 µm thickness which were used for LA-ICP-MS imaging experiments. These were performed with the LA system LSX 213 G2⁺ (CETAC Technologies, Omaha, NE, USA) incorporating a frequency quintupled Nd:YAG laser with a wavelength of 213 nm and equipped with Chromium 2.2 software. The laser system device was coupled to a triple quadrupole based inductively coupled plasma mass spectrometer (iCap TQ; Thermo Fisher Scientific, Darmstadt, Germany). The LA parameters relating to spot size, scanning speed, laser energy and carrier gas flow were optimized based on the best signal-to-noise ratio in combination with highest spatial resolution. The samples were ablated using a line by line scan with a laser energy density of 7,55 J/cm², 20 Hz laser shot frequency, 50 µm spot size and 250 µm/ s scan speed. The aerosol was transported to the ICP with a carrier gas of helium (0.45 L/min) passing the ablation chamber. For maximum sensitivity and to minimize possible interferences, the measurement was performed in kinetic energy discrimination mode (KED) with helium as cell gas. ²⁷Al was monitored with dwell times of 0.1 s. Data evaluation was performed using the software ImageJ (National Institutes of Health, Bethesda, MD, USA).

Data analyses

A generalized linear mixed model (GLMM) with binary data distribution was used, with the Petri dish identity as a random factor. The statistics were done with “R” (R Development Core Team, 2017) and an α- level set to 0.05. SigmaPlot 13 (Systat Software) was used to plot graphs.

Zeolite characteristics

The results of the particle size measurements by laser diffraction in the aqueous phase are shown in Table 1. As expected, the sizes of Y30 particles decreased with the processing of the particles and matched well with the SEM results (Fig. 1A). The measured particle size of Beta(OH)-III zeolites, however, was several micrometers. Although calcination reduced the particle size or aggregation/agglomeration, the particles were much larger than 50 nm, the reported particle size of Beta(OH)-III zeolite primary particles. The comparison of Y30 and H-Beta(OH)-III zeolites in SEM measurements revealed that Y30 zeolites, processed by milling and sedimentation, formed single primary particles (Fig. 1A). In contrast, H-Beta(OH)-III formed much larger aggregates/agglomerates of primary particles (Fig. 1B). The sorption experiments showed that sorption of thiacloprid on Y30 was strong and followed the Langmuir isotherm model. The K_d value at low concentrations within the linear range of the sorption isotherm was 82,500 L/kg.

Chemical analyses

The chemical analyses of the medium revealed the thiacloprid concentrations did not differ between the start and the end of the experiment. Moreover, it was shown that both zeolite types strongly adsorbed thiacloprid. Y30 particles in the toxicity assays adsorbed thiacloprid in the predicted manner, hence, the concentration of thiacloprid in mixtures was below the detection limit (<0.2 µg/L) when high Y30 concentrations were added (97% adsorption was calculated), at 0.5 µg/L when the medium Y30 concentration was added (60% adsorption was calculated) and at 0.65 µg/L for the lowest Y30 concentration (30% adsorption was calculated). H-Beta(OH)-III particles, however, adsorbed thiacloprid to a much higher extent, so that dissolved thiacloprid was not detectable in all mixture experiments containing H-Beta(OH)-III zeolites. When applied solely, the thiacloprid concentration was 1.0 ±0.09μg/L. The concentrations of thiacloprid associated with larval tissue were similar for all larvae exposed to thiacloprid, solely or in mixtures (58.41 ng/g ± 14.42 ng/g (wet weight) on the average), irrespective of the zeolite type added or the presence of gut contents in the sample.

Acute toxicity test

In general, control animals did not show any anomalies in behavior, and the average mortality rate was 4 ± 8% (Figs. 2 and 3). Consequently, the criterion of validity (mean mortality ≤ 10% in the control) was met. Moreover, both zeolites, applied as single substances, did not lead to behavioral disorders. In contrast, larvae exposed to thiacloprid as single compound or to mixtures with low or medium Y30 concentrations displayed behavioral impairments (Fig. 2). However, behavioral changes were found to be delayed in a concentration-dependent manner of Y30. Moreover, larvae exposed to the treatment with the highest Y30 concentration and, thus, highest sorption of thiacloprid did not show any behavioral disruption. In the experiment with H-Beta(OH)-III particles, no animals with behavioral disorders were observed. Experiments with thiacloprid solely showed an almost identical increase of convulsions over time as larvae of the identical treatment type in the experiment with Y30 did. Mortality predominantly occurred between 72 and 96 h of exposure. In the experiment with Y30, we did not find any statistical differences between animals exposed to Y30 solely and the control (GLMM, df = 3, F = 0.5351; p-values for comparisons between the control and low Y30 concentration: 0.676, medium concentration: 0.467, high concentration: 0.676). Nevertheless, larvae exposed to thiacloprid, solely or in the mixtures that comprised low and medium Y30 concentrations, suffered from significantly increased mortality rates compared to the control or the mixture with the highest Y30 concentration, but were not significantly different from one another (GLMM, df = 4, F = 10.849; p-values for the comparison between the control and thiacloprid solely, mix low and mix medium: <0.001, mix high: 0.4162; p-values for the comparisons between thiacloprid solely and mix low: 0.4314, mix medium: 0.0819, mix high: <0.001; p-values for the comparison between mix high and mix low, mix medium and thiacloprid solely: <0.001). However, the mean mortality rate decreased with increasing zeolite concentration (Fig. 3A). Larvae exposed to the mixture with the highest Y30 concentration showed a low mortality rate that did not differ to the control. Likewise, no statistical differences between animals solely exposed to H-Beta(OH)-III zeolites and the control were found (GLMM, df = 3, F = 0.3354; p values for comparisons between the control and low H-Beta-OH-III concentration: 1.0; medium concentration: 0.651, high concentration: 0.414). In further analyses, it was shown that the mortality rate of larvae exposed to the mixtures of thiacloprid and H-Beta(OH)-III did not differ from the control as well. In the contrary, animals exposed solely to 1.0 µg/L thiacloprid exhibited significantly higher mortality rates than larvae exposed to the mixtures or the control (GLMM, df = 4, F = 13.488; p-value for the comparison between the control and thiacloprid solely: <0.001, mix low: 0.433, mix medium: 0.660 and high: 0.660; p-values for all comparisons between thiacloprid solely and all three mixtures was <0.001; Fig. 3B).

LA-ICP-MS imaging

To evaluate whether zeolites were taken up by the larvae, LA-ICP-MS imaging technology with aluminum as marker for zeolites was used. The distribution of aluminum within the larvae confirmed that both zeolites were ingested by the larvae and were widely present in the gut lumen (Fig. 4). However, there was no indication that zeolites were taken up by the tissues of the organism, not even by the resorptive gut epithelium. As expected, larvae of the control did not show any aluminum-containing particles in the gut lumen or in any tissue.