Article pubs.acs.org/est
Tracing Bioavailability of ZnO Nanoparticles Using Stable Isotope Labeling Fiona Larner,*,† Yuktee Dogra,*,‡ Agnieszka Dybowska,§ Julia Fabrega,‡ Björn Stolpe,∥ Luke J. Bridgestock,† Rhys Goodhead,⊥ Dominik J. Weiss,†,§ Julian Moger,⊥ Jamie R. Lead,∥ Eugenia Valsami-Jones,§ Charles R. Tyler,‡ Tamara S. Galloway,‡ and Mark Rehkam ̈ per†,§ †
Department of Earth Science & Engineering, Imperial College London, SW7 2AZ, U.K. Biosciences, College of Life and Environmental Sciences, The Geoffrey Pope Building, University of Exeter, EX4 4QL, U.K. § Department of Mineralogy, Natural History Museum, London SW7 5BD, U.K. ∥ School of Geography, Earth & Environmental Sciences, University of Birmingham, Edgbaston, Birmingham B15 2TT, U.K. ⊥ School of Physics, University of Exeter, EX4 4QL, U.K. ‡
S Supporting Information *
ABSTRACT: Zinc oxide nanoparticles (ZnO NPs) are widely used in commercial products and knowledge of their environmental fate is a priority for ecological protection. Here we synthesized model ZnO NPs that were made from and thus labeled with the stable isotope 68Zn and this enables highly sensitive and selective detection of labeled components against high natural Zn background levels. We combine high precision stable isotope measurements and novel bioimaging techniques to characterize parallel water-borne exposures of the common mudshrimp Corophium volutator to 68ZnO NPs, bulk 68ZnO, and soluble 68ZnCl2 in the presence of sediment. C. volutator is an important component of coastal ecosystems where river-borne NPs will accumulate and is used on a routine basis for toxicity assessments. Our results demonstrate that ionic Zn from ZnO NPs is bioavailable to C. volutator and that Zn uptake is active. Bioavailability appears to be governed primarily by the dissolved Zn content of the water, whereby Zn uptake occurs via the aqueous phase and/or the ingestion of sediment particles with adsorbed Zn from dissolution of ZnO particles. The high sorption capacity of sediments for Zn thus enhances the potential for trophic transfer of Zn derived from readily soluble ZnO NPs. The uncertainties of our isotopic data are too large, however, to conclusively rule out any additional direct uptake route of ZnO NPs by C. volutator.
■
INTRODUCTION Nanotechnology now finds applications in many industrial processes, and an increasing range of nanomaterials are used as constituents in commercial products. There are a vast number of ecological processes that nanomaterials can become involved in, and hence varied and complicated interactions with the biosphere require consideration.1−3 The physical state and chemical form of the nanomaterials are of particular concern, as they will determine the pathways of nanoparticles (NPs) within ecological systems and their toxicity to a particular organism.4 Especially important are studies of aquatic systems as they are the ultimate repositories for most anthropogenic contaminants released into the environment, including nanomaterials.5,6 Such investigations are complex, as they must consider the interactions of nanomaterials with sediments, water, and suspended matter because such interactions can have a key impact on NP bioavailability.4 Of particular interest are studies that aim to establish the behavior and bioavailability of NPs at relatively low exposure concentrations, which are representative of real environmental conditions. Due to this complexity and © 2012 American Chemical Society
the analytical challenges that it poses, there are currently few published reports that address nanomaterial bioavailability within aquatic systems and the toxicological impact of NPs on relevant ecological species. Zinc oxide nanoparticles are widely used in many consumer products, and hence there is a particular interest in sensitive and selective tracing methods for such materials. The use of straightforward concentration changes, as have been applied for Au nanoparticles,7 to detect the presence of Zn-based nanomaterials at relevant exposure levels8−10 is prevented by the high natural background concentrations of Zn. The detection of ZnO nanomaterials is further complicated by the difficulty in differentiating between the NPs themselves and dissolved Zn species that are derived from particles.11,12 As a consequence of these challenges there is currently some Received: Revised: Accepted: Published: 12137
June 27, 2012 October 1, 2012 October 11, 2012 October 11, 2012 dx.doi.org/10.1021/es302602j | Environ. Sci. Technol. 2012, 46, 12137−12145
Environmental Science & Technology
Article
Table 1. Summary of Key Exposure Parameters and Experimental Results Obtained in This Studyg exposure system unit raw materials purity of 68Zn label in precursor material exposure concentration of 68Zn label max. 68ZnO solubility from dialysisb water 68 Zn/66Zn ratio total [Zn] [68Zn] from label sediment 68 Zn/66Zn ratio total [Zn] [68Zn] from label Corophium volutatorc 68 Zn/66Zn ratio total [Zn] [68Zn] from label hepatopancreasc 68 Zn/66Zn ratio [68Zn] from labele mass balance for 68Zn proportion of [68Zn] in water proportion of [68Zn] in sediment proportion of [68Zn] in C. volutator enrichment and bioconcentration factorsf hepatopancreas [68Zn]/C. volutator [68Zn] C. volutator [68Zn]/water [68Zn] hepatopancreas [68Zn]/water [68Zn] C. volutator [68Zn]/sediment [68Zn] hepatopancreas [68Zn]/sediment [68Zn] survival total Zn blank
n
% μg/kg μg/kg
control
bulk 68ZnO
soluble 68ZnCl2
nano 68ZnO
0
99.31 (2) ∼750a ∼29
98.91 (20) 254
99.11 (2) 1042 ∼650
0.6638 (6) 21 (8) 0.01 (0.01)
1.02 (11) 16 (6) 1.6 (0.8)
1.46 (4) 21 (5) 3.9 (1.2)
2.59 (37) 38 (3) 14 (1)
0.6631 (1) 74 (32) 0.00 (0)
0.889 (55) 91 (16) 5.8 (2.2)
0.757 (60) 87 (15) 2.2 (1.2)
1.220 (40) 29 (10) 4.0 (1.3)
0.6627 (3) 84 (19) 0.00 (0.01)
0.7162 (17) 69 (28) 1.0 (0.5)
0.7959 (51) 130 (109) 3.7 (0.9)
1.043 (47) 87 (3) 8.6 (0.8)
0.6652 (10)
0.764 (20) 114 (9)
0.865 (128) 330 (9)
1.152 (177) 897 (9)
0.2 (0.2) 99.7 (0.2) 0.04 (0.03)
1.3 (0.8) 98.4 (1.0) 0.3 (0.2)
2.1 (0.7) 97.5 (0.9) 0.4 (0.2)
3 μg/L μg/L 3 mg/kg mg/kg 6d mg/kg mg/kg 6d mg/kg 3 % % % 3
× 104
% ng
6d 9
84 (6)
129 (51) 88 (43) 796 (370) 830 (78) 9.4 (7.3) 7.4 (4.4) 0.2 (0.1) 1.9 (1.1) 20 (6) 203 (151) 86 (9) 72 (13) range = 0.4−10; average = 2.7
108 (53) 668 (37) 5.2 (2.2) 2.5 (0.8) 185 (74) 89 (7)
a
Approximate value, inferred from the measured mass balance of the exposure system (see the SI). bSolubilities expressed as maximum total dissolved Zn concentrations found after 148 h of equilibrium dialysis (Figure 1). cEach individual result was obtained for 4−5 pooled organisms. dn = 5 for controls. eThe 68Zn concentrations are estimated based on a nominal weight of 15 μg for each pooled hepatopancreas sample. The assumed weights (and hence the concentrations) have an estimated uncertainty of ±25%. fResults calculated from data for individual experiments where tissue, water, and sediment samples were analyzed. gn = number of results from individual experiments; the data given in the table are the averages and uncertainties for the least significant digits are given in parentheses. These uncertainties denote 1SD of the individual results for n ≥ 5 and the total range of data for n = 3.
uncertainty concerning the toxicity of ZnO nanomaterials. While a few studies have shown that the degree of dissolution undergone by ZnO NPs generates a toxicity comparable to that of a soluble Zn source,11−13 other workers have reported an increased toxicity for ZnO nanomaterials.14,15 Zinc has five naturally occurring stable isotopes, which permit the application of stable isotope tracing methods for ZnO NPs. This methodology avoids the handling difficulties and time restrictions associated with radioactive tracing methods and enables highly selective and sensitive detection through the application of ZnO NPs that are purpose-made from and thus labeled with a single stable Zn isotope.9,10,16,17 Modeling studies6 have shown how to implement this stable isotope tracing technique for a range of isotope ratios and demonstrated that the most sensitive detection of stable isotope labeled ZnO nanomaterials is achieved by measurement of the corresponding isotope ratios using multiple-collector inductively coupled plasma mass spectrometry (MC-ICP-MS). This high precision technique allows the detection of Zn from ZnO NPs, at ecotoxicologically relevant exposure concentrations,
against high natural background levels. The Zn isotope compositions that are determined for organisms and exposure reservoirs8 provide constraints on the distribution of the NPderived isotope label within an experimental system, and they can be used to interrogate whether the uptake of Zn by organisms occurs as particulate (bulk/nano) ZnO and/or dissolved Zn that is released from NPs.8 Here, we describe the first study within an aquatic system that combines stable isotope tracing of ZnO NPs (using a 68Zn label) with high precision isotope analyses by MC-ICP-MS. These methods and novel bioimaging techniques are applied to investigate parallel water-borne exposures of the common mudshrimp Corophium volutator to 68ZnO NPs, bulk 68ZnO, and soluble 68ZnCl2 in the presence of sediment. C. volutator (Pallas, 1766) is a sediment dwelling amphipod that feeds on particulate matter from the sediment surface and from suspended particles.18 The organisms are an important component of coastal ecosystems and are widely used as a sediment toxicity test species in Europe.19,20 C. volutator are sensitive to soluble metals that can cross respiratory surfaces 12138
dx.doi.org/10.1021/es302602j | Environ. Sci. Technol. 2012, 46, 12137−12145
Environmental Science & Technology
Article
treatments ensure complete dissolution of any undissolved ZnO particles remaining in the samples. Preconcentration of Zn from the samples was then performed by anion exchange chromatography.25 The Zn fractions were dissolved in 0.1 M HNO3 for the subsequent determination of Zn concentrations and isotope compositions, using a Nu Instruments Nu Plasma HR MC-ICP-MS instrument.8 The mass balance between natural Zn and 68Zn derived from the labeled material in each sample was determined using calculations described previously.8 The biological uptake of Zn by C. volutator was explored using two complementary imaging techniques − coherent antiStokes Raman scattering (CARS) and scanning transmission electron microscopy (STEM) with energy dispersive X-ray spectroscopy (EDX). In particular, CARS was used to observe the uptake of ZnO into the gut of the exposed organisms.13,26,27 This method exploits the nonlinear optical susceptibility of ZnO to provide strong CARS signals27 and the vibrational mode of CH2 bonds abundant in phospholipids to delineate cellular structures. Hence, CARS was used to identify the distribution of Zn within tissues with submicrometer spatial resolution. The STEM-EDX system (JEOL 700F, Oxford Inca) was used to investigate the chemical composition of the granules formed in the hepatopancreas of the organisms.13 A full description of the imaging methods can be found in the SI.
and to the ingestion of metals accumulated in sediments. The organisms store accumulated trace metals in insoluble form in the cells of the ventral ceca (hepatopancreas), where they are bound to organic molecules in characteristic metal-rich sphaerites.13,21−23 The study of this well-characterized model species hence enables a comprehensive analysis of a coastal environment exposure and provides insights into the bioavailability, fate and uptake mechanisms of nanoparticulate 68 ZnO, compared to bulk material and dissolved Zn.
68
■
EXPERIMENTAL METHODS A detailed description of the experimental and analytical methods is provided in the Supporting Information (SI), but key aspects are summarized here. The acids for sample manipulation were prepared from quartz distilled AnalaR 6 M HCl and 15.4 M HNO3 by dilution with ultra pure water of >18 MΩ cm quality (Millipore), if necessary. All glassware, polyethylene, and Teflon vessels were precleaned prior to use, and Teflon was employed whenever practical to minimize Zn blank contributions. Bulk 68ZnO powder and 68Zn metal for the preparation of 68ZnO NP suspensions in DEG (diethylene glycol)10 and soluble 68ZnCl2 were purchased from Isoflex (USA). Individual stock solutions of soluble 68Zn as well as bulk and nanoparticulate 68ZnO were prepared immediately prior to use in exposures, by dilution of the precursor materials with dilute artificial seawater (dilute ASW, salinity ≈ 25 PSU; see the SI). Similar stock solutions were also employed for the equilibrium dialysis experiments that were carried out to study dissolution of bulk and nano 68ZnO particles over a period of almost 7 days in dilute ASW (see the SI). The methods that were employed for characterization of the pristine materials (crystal structure, zeta-potential, size) are discussed in the SI. Sediment and C. volutator organisms were collected from an intertidal area of the Otter estuary (Devon, UK), a site of low anthropogenic pollution. Total organic content, sand to silt ratio, and sediment Zn background concentrations were determined. An acute 10-day water exposure was performed on C. volutator, based on a modified static system20,24 at the University of Exeter. Parallel exposures of C. volutator to labeled 68 ZnO NPs (∼1000 ng/g Zn), bulk 68ZnO (∼750 ng/g Zn), and soluble 68ZnCl2 (250 ng/g Zn) were performed with the exposure stock solutions (150 mL) in the presence of approximately 26 g wet sediment (Table 1). The lower Zn content of the ZnCl2 exposures, relative to ZnO, was chosen to avoid the inhibitory growth affects of higher dissolved Zn concentrations on C. volutator for the time frame of the exposures.13 Sediment and water were added 24 h before the organisms to allow the test system to settle. On the day of exposure, adult organisms (4−7 mm) were harvested by processing the upper 3 cm of the collected sediment through a 300 μm nominal pore size sieve. Animals were not fed during the experiment. At the end of exposures, water and sediment sample aliquots were collected from 3 tanks per exposure type, and C. volutator were sifted from the sediment. Five specimens were dissected to isolate the hepatopancreas, and the remaining organisms were freeze-dried whole for analysis. All samples were prepared for isotope analysis in Class 10 laminar flow hoods in a Class 1000 clean room facility. Biological specimens and sediment were dissolved using appropriate mineral acids via microwave and hot plate digestion, respectively, while the water samples were simply acidified to 1 M HCl without any prior filtration. These
■
RESULTS AND DISCUSSION Particle Behavior in the Exposure Media. Equilibrium dialysis experiments were carried out for both bulk 68ZnO and the 68ZnO NPs immediately after suspension of the precursor materials in dilute artificial seawater to concentrations of about 1 mg/L. In both cases, dialysis revealed significantly lower initial 68Zn label concentrations (of less than 0.1 mg/kg) on the outside of the membrane (Figure 1), presumably due to the rapid agglomeration and subsequent sedimentation of ZnO particles at the high ionic strength of the exposure
Figure 1. The behavior of nano and bulk 68ZnO particles in the dilute artificial seawater (ASW) exposure medium was assessed by equilibrium dialysis over a 148 h period. Shown are the dissolved concentrations of the 68Zn label for experiments with bulk 68ZnO (purple circles) and 68ZnO NPs (green squares), as determined both within (open symbols) and outside (closed symbols) the dialysis cell. The results reveal rapid initial aggregation and sedimentation of ZnO particles, followed by slower dissolution of the precipitates. 12139
dx.doi.org/10.1021/es302602j | Environ. Sci. Technol. 2012, 46, 12137−12145
Environmental Science & Technology
Article
Figure 2. Morphology of the labeled nano and bulk 68ZnO particles that were applied in the exposures of this study. The images of 68ZnO NP in 18 MΩ cm H2O (a; scale: 100 nm, inset 20 nm) show ∼30 nm aggregates, composed of smaller sub-10 nm primary particles, while 68ZnO NP in dilute ASW exhibit more agglomeration and evidence of dissolution (b; scale: 2 μm). The 68ZnO bulk particles in 18 MΩ cm H2O (c; scale: 500 nm, inset 50 nm) and in dilute ASW (d; scale: 2 μm, inset 200 nm) show polydispersity in size and shape and significant agglomeration.
medium.11,28,29 This interpretation is supported by TEM images, which show rapid agglomeration of both bulk and nanoparticulate 68ZnO in the dilute ASW (Figure 2). Following initial aggregation, the 68Zn label concentrations were observed to rise within the dialysis cells, as a consequence of particle dissolution (Figure 1). For the 68ZnO NPs experiment, this rise is most pronounced during the first 1 to 2 days, while no or only small changes in 68Zn content were detected within the cell after 3 days. At the end of the experiment (after 148 h) nearly identical 68Zn concentrations were furthermore determined within and outside the dialysis cells for the nano 68ZnO (Figure 1). Together, this suggests that a (near-) equilibrium state was established, whereby the solution is saturated in 68ZnO with a dissolved 68Zn content of about 0.60 mg/kg (equivalent to a total Zn concentration of ∼0.63 mg/kg). For the bulk 68ZnO particles, the concentration of the 68Zn label rises much less rapidly within the cell, and a nearly linear and continuous increase is observed for the duration of the experiment (Figure 1). After 148 h, essentially identical 68Zn label abundances of about 18 μg/kg (equivalent to a total Zn content of ∼29 μg/kg) were determined both within and outside of the dialysis cell. These results demonstrate that the bulk 68ZnO material features significantly slower dissolution kinetics in dilute ASW than the nano 68ZnO particles (as
expected), and saturation is therefore not established for the former during the duration of the experiment. Organism Survival and Imaging. After the 10-day exposure period, survival exceeded 80% for the control and the two 68ZnO exposures and 70% for the experiment with soluble 68ZnCl2. Hence, the results are all within the quality control limits for the assay.20 The CARS analyses showed Zn to be present as accumulates within both the alimentary tract and the hepatopancreas of C. volutator from all 68Zn exposures but not in control organisms (Figure 3; also see the SI). The resolution of CARS is insufficient to identify individual NPs, and therefore we used STEM-EDX to further analyze the gut, hepatopancreas, and surrounding tissue of C. volutator specimens. The examinations of the alimentary canal revealed the presence of Zn as accumulates within the gut, and Zn-enrichments associated with Si and Al-rich particles, presumably due to the adsorption of Zn to the surface of sediment particles (Figure 3). Analysis of the hepatopancreas, which is the main storage organ for accumulated metals in C. volutator, identified characteristic accumulations of Zn within sphearites30 in animals from all exposures (Figure 3). These sphaerites are typically composed of detoxified metals and organic materials. Taken together, the above data suggest that the ingestion and internal processing of Zn, either in the form of dissolved species or possibly as bulk/ 12140
dx.doi.org/10.1021/es302602j | Environ. Sci. Technol. 2012, 46, 12137−12145
Environmental Science & Technology
Article
Figure 3. CARS and STEM-EDX imaging and (at the bottom center) a schematic lateral view of the C. volutator22 alimentary canal. HP = hepatopancreas, G = gut. (a) CARS images of the hepatopancreas of an organism exposed to bulk 68ZnO. (b) STEM image of an organism exposed to bulk 68ZnO exhibiting sphaerites formed in hepatopancreatic cells. (c) EDX profile of sphaerites in the organism showing the presence of Zn. (d) CARS images of the gut of an organism exposed to bulk 68ZnO. (e) STEM image of an organism exposed to bulk 68ZnO, exhibiting Zn accumulates associated with Al- and Si-rich sediment particles within the lumen of the gut. (f) EDX profile of accumulates in the gut showing the presence of Zn. Similar CARS and STEM images were also recorded in organisms exposed to both 68ZnO NPs and 68ZnCl2 (see the SI). CARS scale bar = 50 μm, STEM scale bar = (b) 1 μm, (e) 200 nm, EDX scale = (c) 354 cps, (f) 25 cps/eV. The large Cu and Os peaks in EDX spectrum are due to use of copper TEM grids and osmium-rich fixative, respectively.
(Table 1), and these are in accord with the expected magnitude of natural mass dependent isotope fractionations.8 Clearly, this natural variability is essentially negligible compared to the much larger isotopic changes that are observed for the three different exposure systems. In particular, all samples from exposures display readily resolvable positive deviations of the 68Zn/66Zn isotope ratios from natural values, and this demonstrates the presence of labeled 68Zn in all compartments (Table 1). For all exposures, the highest diagnostic isotope ratios are observed in the water samples, which hence feature the largest ‘excess’ of 68 Zn label relative to total Zn present. This reflects the relatively low (μg/kg level) background Zn content of the dilute ASW and the waters therefore contain only about 0.2 to 2% of the total 68Zn label present in the three exposure systems (Table 1). It is also of interest that the 68Zn/66Zn ratios of the aqueous compartment are much higher for the ZnO NP exposure than for the bulk ZnO experiment (Table 1). This result is expected from the dialysis experiments, as the ZnO NPs should incur more dissolution in the exposure medium than the bulk ZnO material (Figure 1). While the sediments feature lower 68Zn/66Zn ratios, and thus less excess 68Zn on a relative basis, than the water samples, they are nonetheless the dominant reservoir for the 68Zn label of all exposure systems. In detail, the sediments harbor more than
nano ZnO particles, is a consequence of the increased Zn concentrations of the exposure environments. Zinc Concentrations, 68Zn/66Zn Isotope Ratios, and the Mass Balance of Labeled 68Zn. In the following, we employ the isotope ratio 68Zn/66Zn for diagnostic purposes, because this is composed of the 68Zn label and a second Zn isotope (66Zn), which is present from natural sources only. The diagnostic ratio and the total Zn concentrations were determined by mass spectrometry for all compartments within each exposure system (Table 1). These data were used to calculate the 68Zn label abundances that are also shown in Table 1, whereby each result is the average determined from analyses of several individual samples (n = 3 to 6). Inspection of the total Zn concentrations (Table 1), which can be corrected for the addition of label-derived 68Zn for samples from exposures, shows both the considerable natural Zn background levels and the heterogeneous distribution of Zn in the water, sediment, and biological samples, in accord with published Zn concentrations for both geological and biological specimens.29 Clearly, this highlights the need for the application of stable isotope labeling to trace environmentally relevant concentrations of ZnO NPs. The control samples display only small (permil-level) variations in the Zn isotope compositions of the compartments 12141
dx.doi.org/10.1021/es302602j | Environ. Sci. Technol. 2012, 46, 12137−12145
Environmental Science & Technology
Article
97% of the total 68Zn label present in each exposure (Table 1). These observations are reconciled by the high natural Zn contents of the sediments (∼25 to 85 mg/kg as opposed to ∼20 μg/kg for the waters; Table 1). The C. volutator and separated hepatopancreas samples have 68Zn/66Zn ratios and Zn concentrations similar to or higher than those of sediments, but they feature only a small proportion of the total 68Zn (
