Article pubs.acs.org/est
Uptake and Speciation of Vanadium in the Benthic Invertebrate Hyalella azteca Madeleine Jensen-Fontaine,† Warren P. Norwood,‡,§ Mitra Brown,‡ D. George Dixon,§ and X. Chris Le*,†,∥ †
Department of Chemistry, University of Alberta, Edmonton, Alberta, Canada Aquatic Contaminants Research Division, Environment Canada, Burlington, Ontario, Canada § Department of Biology, University of Waterloo, Waterloo, Ontario, Canada ∥ Division of Analytical and Environmental Toxicology, Department of Laboratory Medicine and Pathology, University of Alberta, Edmonton, Alberta, Canada ‡
S Supporting Information *
ABSTRACT: Vanadium has the potential to leach into the environment from petroleum coke, an oil sands byproduct. To determine uptake of vanadium species in the biota, we exposed the benthic invertebrate Hyalella azteca with increasing concentrations of two different vanadium species, V(IV) and V(V), for seven days. The concentrations of vanadium in the H. azteca tissue increased with the concentration of vanadium in the exposure water. Speciation analysis revealed that V(IV) in the exposure water was oxidized to V(V) between renewal periods, and therefore the animals were mostly exposed to V(V). Speciation analysis of the H. azteca tissue showed the presence of V(V), V(IV), and an unidentified vanadium species. These results indicate the uptake and metabolism of vanadium by H. azteca. Because H. azteca are widely distributed in freshwater systems and are an important food supply for many fish, determining the uptake and metabolism of vanadium allows for a better understanding of the potential environmental effects on invertebrates.
■
is 2.5 m3 (16 barrels), though 80% is recycled.3 In 2003, 23 kg of coke was produced for every barrel of SCO produced, resulting in 2 million tons of coke produced yearly.9 The concentration of vanadium in the Athabasca Oil Sands is one of the highest in oil producing areas, the highest being in Venezuelan oils.10 The total concentration of vanadium in OSPW is 0.018 mg/L.11 The vanadium concentration in coke averages around 1680 ppm.3 It has been demonstrated that vanadium can leach from coke into water12 and therefore there is potential for vanadium to leach from the coke into the environment. Vanadium has six oxidation states; the most commonly found in nature are V(III), V(IV), and V(V).10,13 As the oxidation state of vanadium increases, so does its toxicity.14−17 It is wellknown that speciation studies for vanadium are imperative in detailed risk assessments. V(V) is a known carcinogen16,17 and it has been shown to be toxic to a range of test animals and cell lines,14−23 although certain vanadium species may be useful in the prevention and treatment of cancer.24 Vanadium toxicity
INTRODUCTION Oil sands are unconventional oil deposits. In Northern Alberta, Canada, the oil sands consist of three major deposits covering an area of 140 000 km2: the Athabasca, Cold Lake, and Peace River oil sands.1 Together, they contain 27 billion m3 of crude bitumen reserves.2 In 2010, the daily production of crude bitumen2 was 256 300 m3. That year, 46.1 million m3 or 290 million barrels of synthetic crude oil (SCO) were produced.2 To produce SCO, the bitumen undergoes three major processes: mining, extraction, and upgrading. During the mining and extraction processes, hot water is used for hydrotransport and to separate the bitumen from the silt, clay, and sand that compose the oil sands.3 The result is oil sands process-affected water (OSPW).4 It is stored on-site in tailings ponds, also known as settling basins, due to a “zero discharge” policy.5−7 The separated bitumen is upgraded to SCO by cracking, catalytic conversion, distillation, and hydrotreating. Petroleum coke, a byproduct of the cracking process,8 is used as fuel for the power plants in Fort McMurray,2 stockpiled for future use in land reclamation projects,1 or slurried to the settling basins with OSPW. To produce 1 m3 (6.3 barrels) of SCO, 11 tons of oil sands are required if the bitumen recovery rate is 90%.3 The amount of water used for hydrotransport and bitumen recovery © 2013 American Chemical Society
Received: Revised: Accepted: Published: 731
July 23, 2013 November 21, 2013 December 2, 2013 December 2, 2013 dx.doi.org/10.1021/es403252k | Environ. Sci. Technol. 2014, 48, 731−738
Environmental Science & Technology
Article
Baensch, Melle, Germany). The test media consisted of dechlorinated (DeChlor) tap water from the City of Burlington, Ontario, Canada, originating from Lake Ontario. The common water parameters (mean ±95% confidence interval (CI) were: dissolved organic carbon 3.8 ± 0.6 mg/L, dissolved inorganic carbon 23.4 ± 0.4 mg/L, Alk 97.5 ± 0.9 mg/L, Cl 35.7 ± 0.3 mg/L, SO4 31.8 ± 0.5 mg/L, Ca 39.1 ± 0.1 mg/L, Mg 9.8 ± 0.1 mg/L, Na 19.3 ± 0.2 mg/L, K 2.1 ± 0.1 mg/L, hardness 138 ± 1 mg/L (as CaCO3) pH 8.3 ± 0.1 and conductivity 404 ± 8 μs/cm). The analyses of major ions, alkalinity, and dissolved organic and inorganic carbon, were conducted on samples collected at the end of the first renewal period from all treatments (N = 10), by the National Laboratory for Environmental Testing, Environment Canada, Burlington, Ontario. Conductivity, pH, dissolved oxygen, and total ammonia were monitored throughout the experiment at the beginning and end of each renewal period in every treatment (N = 54). The tap water was dechlorinated by activated charcoal filtration and then vigorous aeration for two weeks. The containers were placed in an incubator maintained at 25 °C and operated with a 16:8 light:dark photoperiod. The test media and food were renewed on days 2 and 5. At the end of each renewal period, the contents of the containers were transferred into a sorting bowl and the surviving H. azteca were transferred to a Petri dish of fresh DeChlor, counted, and transferred into the appropriate fresh renewal solutions. Sample Collection and Preservation. Water. At the beginning and end of each renewal period, 10 mL water samples were collected from each of the four replicates of every treatment for vanadium speciation analysis. These samples were kept in 12 mL HDPE screw top vials, preserved with 0.5 mL of 52.5 mM Na2EDTA (Sigma-Aldrich, Oakville, Ontario, Canada) and refrigerated at 4 °C until shipment on ice to the University of Alberta, Edmonton, Alberta, Canada. In addition, a 1 mL aliquot was collected for total vanadium analysis at the beginning and end of each renewal period from replicate 1 of every treatment. These samples were kept in 1.2 mL HDPE Nalgene cryovials and preserved with 10 μL concentrated ultrapure HNO3 (J.T. Baker, Canada). Tissue. The exposure was terminated at day 7. The survivors from each replicate were transferred to a sample cup containing 60 mL of a 50 μM EDTA solution made with DeChlor water, a piece of gauze, and 2.5 mg Tetra-Min food, for a 24 h gut clearance.31 After 24 h, the animals were counted, transferred to a clean solution for rinsing, dried on a Kim-wipe and weighed. Two animals were removed, placed in a 1.2 mL Nalgene cryovial loosely capped, and dried at 60 °C for 72 h in preparation for total vanadium analysis by atomic absorption spectrometry (AAS) at Environment Canada, Burlington, Ontario, Canada. The remaining H. azteca were reweighed and frozen at −80 °C until shipment on dry ice to the University of Alberta, where they remained frozen until analysis. Analysis by HPLC-ICPMS. The high performance liquid chromatography (HPLC) setup consisted of a PerkinElmer Series 200 HPLC system (PE Instruments, Shelton, CT) equipped with an autosampler and a column heater. The column was a PRP-X100 strong anion exchange column (50 mm in length, 4.1 mm in diameter, 5 μm packing material; Hamilton, Reno, NV). The mobile phase was 3% acetonitrile, 2 mM EDTA, and 80 mM ammonium bicarbonate (pH 6). The flow rate of the mobile phase was 1 mL/min. The sample injection volume was 50 μL. The HPLC column outlet was
was suspected as the cause of a die-off of Canada geese at a Delaware refinery fly ash pond.22 Acute toxicity of vanadium to cattle was implicated in the death of 23 out of 98 cattle in northern Sweden.23 The cattle were exposed to alkaline steel slag that contained 3% vanadium. There have been extensive eco-toxicological studies involving vanadium and oil sands byproducts.7,25−31 In aquatic toxicity testing, the range of V(V) LC50 for juvenile rainbow trout (Salmo gairdneri R.) was 1.9− 6.0 mg/L,23 for zebra fish (Brachydanio rerio) 2−3 mg/L,24 and for guppies (Poecilia reticulata) 3.3 mg/L.24 A recently developed monitoring program for the Athabasca Oil Sands region, the Integrated Monitoring Plan for the Oil Sands, aims to determine what contaminants should be monitored, where and when the sampling should occur, and what methods are required for the sampling and testing.32 The scope of the program includes monitoring the air, water, sediment, and biodiversity, both terrestrial and aquatic, of the region. The amphipod Hyalella azteca is listed as a test organism for the toxicity assessment of sediment and water samples from the region as part of this monitoring program. Hyalella azteca are omnivorous, noncannibalistic epi-benthic freshwater shrimp that are sensitive to metals and other toxic substances and have been extensively used in toxicity studies.33−35 They are widely distributed in freshwater systems across North and Central America, from Guatemala to Inuvik, Northwest Territories, Canada34,36,37 and are an important food supply for many fish.38 There have been no studies on the exposure, uptake, or metabolism of different chemical species of vanadium to H. azteca. The objective of this study was to determine the total vanadium content as well as the vanadium species in the test water and tissues of H. azteca following 7-day exposure to water spiked with vanadium. Determining the uptake of vanadium allows for a better understanding of the effect of vanadium on an aquatic system. Speciation of the water and tissue samples was investigated to complement the standard practice of total vanadium analysis. The speciation analysis would determine which vanadium species the animals were exposed to and possible speciation changes that may have occurred after ingestion and metabolism. Vanadium speciation and bioaccumulation analyses may help in the assessment and future monitoring of other oil sands developments beyond the Athabasca region.
■
MATERIALS AND METHODS Reagents. Vanadyl sulfate hydrate (Sigma-Aldrich, St. Louis, MO) was used to prepare the test solutions for V(IV) and sodium metavanadate (Sigma-Aldrich, St. Louis, MO), for the V(V) solutions, for the exposure study. The vanadium standards for speciation analysis were prepared using the method described by Li et al.39 Vanadyl sulfate hydrate, ammonium metavanadate (Aldrich, St. Louis, MO), ethylenediaminetetraacetic acid (EDTA; Aldrich, St. Louis, MO), nitric acid (Fisher Scientific, Concord, Ontario, Canada) and ammonium hydroxide (Fisher Scientific, Concord, Ontario, Canada) were used. Hyalella azteca Bioaccumulation Test. The experimental setup was similar to that described for bioaccumulation studies in Norwood et al.40 A series of test concentrations (25, 141, 451, and 1410 μg V/L for V(IV), and 19, 106, 339, and 1060 μg V/L for V(V)) were used. Twenty 6−10 week old H. azteca were added to 400 mL of test medium containing a piece of 100% cotton gauze (5 × 5 cm) and 2.5 mg of TetraMin (Ulrich 732
dx.doi.org/10.1021/es403252k | Environ. Sci. Technol. 2014, 48, 731−738
Environmental Science & Technology
Article
For the determination of total vanadium in tissue samples, the two animals removed from each replicate for total analysis were dried at 60 °C for 72 h in a 1.2 mL Nalgene cryovial. Total dry weight was determined and then the tissues were digested following the method described by Norwood et al.41 Analysis was performed using GF-AAS as described above. Determination of V(IV) Stability in the Test Water. The stability of V(IV) in the test water was investigated to determine how fast V(IV) was oxidized in the exposure experiment. The initial conditions of the H. azteca exposure study were conducted in dechlorinated tap water from the City of Burlington, Ontario water. The dechlorinated tap water was spiked with fresh vanadyl sulfate solution simulating the addition of V(IV) to the exposure water. The vanadium speciation was monitored using HPLC-ICPMS for 85 min by successive injections from the vial directly into the HPLC system (50 μL injection volume). Each analysis took approximately 10 min. Test of V(V)−EDTA Complex. A Ca2+ solution was prepared using CaCl2·2H2O (BDH Inc., Toronto, Ontario, Canada). A 1 mg/L Ca2+ solution prepared in deionized water (DIW) was spiked with vanadium (20 μg/L V(IV), 20 μg/L V(V), or 20 μg/L of both V(IV) and V(V)) and allowed to sit overnight. A 0.5 g/L Ca2+ solution was prepared in either DIW or 2.5 mM Na2EDTA and spiked with 50 μg/L V(IV), 50 μg/L V(V), or 50 μg/L of both V(IV) and V(V), sonicated for 3 h or allowed to sit overnight. The solutions were analyzed for vanadium species using HPLC-ICPMS. No Ca−V(V)−EDTA complex was observed.
connected to an inductively coupled plasma mass spectrometer (ICPMS, PerkinElmer Elan 6100 DRCplus ICPMS; PE Sciex, Concord, Ontario, Canada) using a 38 cm piece of PEEK tubing (1/16 in. o.d., 0.007 in. i.d.; Supelco, Bellefonte, PA). The performance of the instrument was optimized daily using an atomic spectroscopy standard solution (Elan 6100 DRC Setup/Stab/Masscal Solution; PerkinElmer, Shelton, CT). Of the two naturally occurring isotopes of vanadium, V51 was chosen for monitoring because it is 99.75% abundant and has no elemental interferences. ClO+, the main isobaric interference in the samples, does not coelute with the vanadium standards.39 The ICPMS operating conditions are listed in Supporting Information (SI) Table S1. Vanadium Speciation in Water Samples. Water samples were mixed on a vortex mixer. A 1.00 mL aliquot was filtered using a 1 mL syringe (BD, Franklin Lakes, NJ) and 0.45 μm nylon filter membrane (PALL Life Sciences Acrodisc 13 mm Syringe Filter, Pall (Canada) Ltd., Mississauga, Ontario, Canada) and diluted in 2.5 mM Na2EDTA solution to a concentration that fell within the range of the calibration curve. Vanadium Speciation in Tissue Samples. Four replicate tissue samples for each treatment were combined and weighed in Cultube Sterile Culture Tubes (12 × 75 mm height; Simport, Beloeil, Quebec, Canada). A 1.00 mL aliquot of 2.5 mM Na2EDTA solution was added to the tube. The tissue was ground using a PowerGen 125 grinder (Fisher Scientific, Inc., Ottawa, Ontario, Canada). The grinder was rinsed with 1.00 mL of 2.5 mM Na2EDTA solution into the vial. The contents of the tube were transferred to a 15 mL conical vial (Corning Incorporated, Corning, NJ). The tube was rinsed with 1.00 mL 2.5 mM Na2EDTA that was added to the conical vial for a total of 3 mL of extract solution. The contents of the conical vial underwent sonication for 1 h (Sonicor Instrument Corporation, Copiague, NJ) then were centrifuged for 15 min at 3500 rpm (Sorvall Biofuge primo; Mandel Scientific Co. Ltd., Guelph, Ontario, Canada). The supernatant was removed. Another 1.00 mL portion of 2.5 mM Na2EDTA was added to the conical vial. The sample was vortexed, sonicated for an hour, centrifuged, and the supernatant removed and combined with the previous supernatant. This process was repeated one more time for a total of 3 h of sonication and 5 mL of supernatant. The supernatant was filtered through a 0.45 μm nylon filter (Pall (Canada) Ltd., Mississauga, Ontario, Canada) then analyzed in triplicate by HPLC-ICPMS. Analysis by Atomic Absorption. Total vanadium analyses were performed using graphite furnace atomic absorption spectrometry (GF-AAS). A Varian 400 SpectraAA with Zeeman background correction was used. Vanadium standard from O2SI Smart Solutions, Charleston, South Carolina (NH 4 VO 3 preserved in 2% HNO3) was used to generate standard calibration curves and to correct for any instrument drift in each run. Samples were modified with 20 g/L ascorbic acid (JT Baker Inc.) during injection. There was a 97.8 ± 1.9% (mean ±95% confidence) recovery of the certified reference water sample CRM-TMDW (High Purity Standards Inc., Charleston, SC) and a 97.1 ± 6.0% (mean ±95% confidence) recovery of the certified reference tissue sample TORT-2 (lobster hepatopancreas, National Research Council of Canada, Ottawa, Ontario, Canada). The detection limit was 1.20 μg/L (3 × std of blanks) for water analyses and 0.79−1.9 μg/g dry weight (dw) for tissue analyses based on the detection limit (3 × std of blank digests) divided by the minimum and maximum digest dry weights.
■
RESULTS Vanadium in the Exposure Water. The sum of the vanadium species concentrations obtained from HPLC-ICPMS represented 92% of the total vanadium concentration determined by AAS (SI Figure S1). The agreement (R2 = 0.99) between the two sets of blind analyses, independently conducted in two laboratories, confirm the validity of the analytical methods. The speciation analysis of the exposure water indicated that the prevalent vanadium species in the V(IV) exposure was V(V) (Figure 1 and SI Figures S3, S4, S5). In contrast, V(IV)
Figure 1. Vanadium speciation in water from the 7 day bioaccumulation test with H. azteca. Samples for exposure to 25 μg/ L V(IV) from vanadyl sulfate. Day 0, 2.1, and 5.1 were fresh test solutions. Days 2, 5, and 7 were the end of each renewal period. Error bars represent one standard deviation. 733
dx.doi.org/10.1021/es403252k | Environ. Sci. Technol. 2014, 48, 731−738
Environmental Science & Technology
Article
was only observed at the beginning of each renewal period (Days 0, 2.1, and 5.1) and dropped to nondetectable levels by the end of each renewal period (Days 2, 5, and 7). These observations suggest that the V(IV) was oxidized completely between renewals. Some V(IV) was detected in samples collected at the end of each renewal period (Days 2, 5, and 7) in the higher test concentration experiments (SI Figures S4 and S5), but the concentrations of V(IV) were significantly less than those found in the initial samples collected for each treatment. The major vanadium species present in all the samples from the V(V) exposure test was V(V) (SI Figures S6−S9). To determine how fast V(IV) was oxidized in the exposure experiment, we tested the stability of V(IV) in the exposure water (dechlorinated tap water). Figure 2 shows chromato-
Figure 3. Total vanadium body concentrations in Hyalella azteca from individual exposures to V(IV) and V(V) in 7 day tests.
Figure 2. Change in vanadium speciation in DeChlor water spiked with 160 μg/L V(IV). Times on the chromatograms represent the time between the beginning of the first injection and the beginning of the subsequent injection.
grams from the HPLC-ICPMS analyses of exposure water repeatedly every 10 min. Although both V(IV) and V(V) were present initially, most of the V(IV) was oxidized to V(V) within 30 min (Figure 2). During the H. azteca experiment, the water samples collected for speciation were taken 3−4 h after the initial mixing of the fresh vanadium stock solution with the DeChlor water. During this time, oxidation of V(IV) could have occurred as the solution was exposed to air and no EDTA was present in the sample. This would explain the high percentage of V(V) and low percentage of V(IV) observed in the water samples from the H. azteca exposure experiment (Figure 1). Vanadium in the Tissue of H. azteca. The vanadium concentration in the tissue of the exposed H. azteca increased with increasing exposure concentration (Figure 3). Since V(IV) was oxidized to V(V) during the exposure period, it is understandable that the results from both sets of exposure experiments are similar. In both cases, the concentrations of vanadium in the H. azteca tissue increased with the concentrations of vanadium in the water (17−1002 μg/L) to which the H. azteca were exposed. HPLC-ICPMS analysis of the H. azteca tissue extracts shows the presence of V(IV), V(V), and a new vanadium species (Figure 4). All the eight tissue samples contained both V(IV) and V(V). In several extracts of H. azteca, there was also an unidentified peak corresponding to the retention time of 1.5 min (Figure 4). This peak has a greater retention time than Cl− (0.8 min), and therefore it is not an artifact of Cl− in the sample.39 Because vanadium is the only element with an m/z 51, the unknown peak was considered as a vanadiumcontaining species. This vanadium-containing species was
Figure 4. Chromatograms of vanadium speciation of tissue extracts of H. azteca that were exposed to either 1410 μg/L V(IV) or 1060 μg/L V(V).
present in six of the eight tissue samples. The possible identity of this vanadium species is considered in the Discussion section. The relative distribution of vanadium species in the H. azteca tissue are shown in Figure 5.
Figure 5. Relative distribution of vanadium species in H. azteca tissue after 7 day exposure to vanadium in water. 734
dx.doi.org/10.1021/es403252k | Environ. Sci. Technol. 2014, 48, 731−738
Environmental Science & Technology
Article
Table 1. Concentration of Vanadium in H. azteca Tissue vanadium concentration in tissue (μg/g dry weight) vanadium species
measured total exposure concentration (μg/L)
V(IV)
V(V)
unknown peak
sum of three species
total by AAS
V(IV)
17 92 336 962
0.022 0.016 0.056 0.362
0.012 0.040 0.015 0.369
0.067 0.222 0.000 2.178
0.085 0.278 0.071 2.365
0.12 0.61 1.96 6.82
V(V)
20 95 336 1002
0.081 0.032 0.070 0.449
0.038 0.031 0.119 0.506
0.000 0.289 0.128 1.426
0.119 0.352 0.317 2.025
0.23 0.87 2.87 7.50
The concentrations of V(IV) and V(V) in the H. azteca tissue were on the same order of magnitude. The presence of V(V) in the animals exposed to both V(V) and V(IV) could be expected based on the analysis of the water samples. Most of the V(IV) was oxidized to V(V) before the animals were exposed to the test solution. There was no abiotic reduction of V(V) to V(IV) observed in the exposure water samples. The presence of V(IV) in the animals exposed only to V(V) suggests metabolism of vanadium species by the benthic invertebrate H. azteca. The sum of vanadium species in the extract of H. azteca is compared with the total concentration of vanadium as determined by AAS in the acid digest of H. azteca (Table 1). The tissue extraction efficiency of the sum of V(IV) species plus the V(V) species and the unknown vanadium species in the highest exposure concentrations of V(IV) was 34.7% (Table 1). Whereas the tissue extraction efficiency of the sum of V(V) species plus the V(V) species and the unknown species in the highest exposure concentrations of V(V) was 27.0% (Table 1).
investigation. It is also important to determine vanadium species in the natural environment to assess potential exposure by the biota in the environment. Speciation of vanadium in H. azteca that were exposed to predominantly V(V) revealed the presence of V(V), V(IV), and a new vanadium species (not yet identified) in the H. azteca tissue. Although this new vanadium species (the first peak in Figure 4) has not been identified, we have explored several possibilities. We first ruled out this as being a Ca−V(V)− EDTA complex.39,47 Analyses of mixtures containing 1 mg/L Ca2+ and vanadium species (20 μg/L V(IV), 20 μg/L V(V), or 20 μg/L of both V(IV) and V(V)) did not show the presence of a Ca−V(V)−EDTA complex. Further analyses of mixtures containing 0.5 g/L Ca2+, 2.5 mM Na2EDTA, and vanadium species (50 μg/L V(IV), 50 μg/L V(V), or 50 μg/L of both V(IV) and V(V)) again did not show any Ca−V(V)−EDTA complex. More likely possibilities of the new vanadium species are vanadium complexes with small cellular molecules and proteins. In cells, vanadium interacts with glutathione (GSH) and adenosine 5′-triphosphate (ATP).48 GSH and ATP are present in cells at mM concentrations, in high excess compared to vanadium concentrations. ATP is a stronger binder than GSH to vanadium species. As a structural analog of phosphate (PO 43‑ ), vanadate (VO43‑) inhibits the active sites of phosphatases and other related enzymes involved in the hydrolysis of phosphate esters. In vitro binding studies, using electron spin resonance (ESR) spectroscopy, have indicated formation of V(IV)-ATP, V(IV)-GSH, and V(IV)-ATP-GSH complexes, depending on the relative molar excess of V(IV), ATP, and GSH.48 Vanadium binds to proteins in blood, especially to transferrin and albumin.49−51 Transferrin maintains about 30% of its capacity to bind Fe3+ ions,52 and has the potential ability to bind and transport other metals ions, such as vanadium species.53 A number of studies51,53 have confirmed that V(IV) in blood exists predominantly as V(IV)−transferrin complexes. A smaller fraction (
