Aromatic Naphthenic Acids in Oil Sands Process-Affected Water

Ltd. West In pit settling basin in Fort McMurray, Alberta, Canada in spring 2009. ..... Bermuda Institute for Ocean Sciences, 17 Ferry Reach, St. ...
0 downloads 0 Views 1MB Size
Download PDF
Article pubs.acs.org/est

Aromatic Naphthenic Acids in Oil Sands Process-Affected Water, Resolved by GCxGC-MS, Only Weakly Induce the Gene for Vitellogenin Production in Zebrafish (Danio rerio) Larvae Helena C. Reinardy,†,‡,# Alan G. Scarlett,† Theodore B. Henry,‡,§,∥ Charles E. West,† L. Mark Hewitt,⊥ Richard A. Frank,⊥ and Steven J. Rowland*,† †

Petroleum and Environmental Geochemistry Group, Biogeochemistry Research Centre, University of Plymouth, Drake Circus, Plymouth PL4 8AA, U.K. ‡ School of Biomedical and Biological Sciences, Plymouth University, Plymouth, U.K. § Center for Environmental Biotechnology, University of Tennessee, Knoxville, Tennessee 37996, United States ∥ Department of Forestry, Wildlife and Fisheries, University of Tennessee, Knoxville, Tennessee 37996, United States ⊥ Aquatic Contaminants Research Division/Water Science & Technology Directorate, Environment Canada, 867 Lakeshore Road, Burlington, ON, Canada L7R 4A6 S Supporting Information *

ABSTRACT: Process waters from oil sands industries (OSPW) have been reported to exhibit estrogenic effects. Although the compounds responsible are unknown, some aromatic naphthenic acids (NA) have been implicated. The present study was designed to investigate whether aromatic NA might cause such effects. Here we demonstrate induction of vitellogenin genes (vtg) in fish, which is a common bioassay used to indicate effects consistent with exposure to exogenous estrogens. Solutions in water of 20−2000 μg L−1 of an extract of a total OSPW NA concentrate did not induce expression of vtg in larval zebrafish, consistent with earlier studies which showed that much higher NA concentrations of undiluted OSPW were needed. Although 20−2000 μg L−1 of an esterifiable NA subfraction of the OSPW NA concentrate did induce expression, this was of much lower magnitude to that induced by much lower concentrations of 17α-ethynyl estradiol, indicating that the effect of the total NAs was only weak. However, given the high NA concentrations and large volumes of OSPW extant in Canada, it is important to ascertain which of these esterifiable NA in the OSPW produce the effect. Up to 1000 μg L−1 of an OSPW subfraction containing only alicyclic NA, and considered by most authors to be NA sensu stricto, did not produce induction; but, as predicted, 10−1000 μg L−1 of an aromatic NA fraction did. Such effects by the aromatic acids are again consistent with those of only a weak estrogenic substance. These findings may help to focus studies of the most environmentally significant OSPW-related pollutants, if reproduced in a greater range of OSPW.

■

INTRODUCTION Numerous studies have been made of the toxicity of undifferentiated process-affected waters (OSPW) resulting from the oil sands industries of Alberta, Canada.1−3 Often the toxic effects have not been assigned to specific compounds in OSPW, but a group of compounds commonly referred to as naphthenic acids (NA) and which occur in OSPW in concentrations as high as 100 mg L−1 has been the subject of much debate.1−3 NA are considered, sensu stricto, to be complex mixtures of alicyclic carboxylic acids. For instance, several NA in OSPW have been shown to be tricyclic adamantane acids and isomers of pentacyclic diamantane acids.4,5 In developing practice, particularly in Canada, such NA are commonly referred to as ‘classical’ NA, by reference to the general formula, CnH2n+zO2. In the latter, n refers to the number of carbon atoms and z is © XXXX American Chemical Society

zero or a negative even integer, referring to the hydrogen deficiency. It is widely understood that the hydrogen deficiency (z) in the NA formula is a function of the number of alicyclic rings present. However, this is an oversimplification: hydrogen deficiency may also be due to aromaticity (or double bonds). Thus, aromatic NA are also included in the formulaic definition. Most toxicity assays to date have been made on OSPW or on dilutions of OSPW,1−3 and since chemical analysis has shown that the organic compounds in OSPW comprise very complex mixtures,4,5 it will no doubt continue to take considerable effort Received: November 23, 2012 Revised: March 1, 2013 Accepted: March 5, 2013

A

dx.doi.org/10.1021/es304799m | Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Environmental Science & Technology

■

Article

MATERIALS AND METHODS The OSPW was a 3 L subsample of a concentrate of thousands of liters of oil sands tailings pond water collected from Syncrude Canada Ltd. West In pit settling basin in Fort McMurray, Alberta, Canada in spring 2009. The subsequent treatment to isolate a concentrated NA (sodium salts) mixture was conducted in spring 2011 by Environment Canada by methods which have been described fully.7 Subsamples (3 × 1 L; NA concentration determined by ESI-MS, 2000 mg L−1; Headley, personal communication) of this concentrate as received (pH >10) were acidified with hydrochloric acid to pH < 2 and extracted with ethyl acetate, evaporated to dryness under a steady stream of nitrogen at 40 °C, heated with fresh boron trifluoride- methanol complex (BF3-MeOH 3h, 70 °C), re-extracted into hexane, and evaporated to dryness under a steady stream of nitrogen at 40 °C. This provided the methyl esters of NA esterifiable by the treatment, including ‘classical’ alicyclic and aromatic NA. Prior to toxicity testing, methyl esters were demethylated to free acids based on methods described by Frank et al.7 until demethylation was complete, as determined by Fourier Transform infrared spectroscopy. The samples were then rotary evaporated until a stable weight was achieved and then redissolved in HPLC-grade water to produce their sodium salts. Prior to zebrafish exposures, the test solutions were adjusted to pH 7.9 ± 0.1. Argentation solid phase extraction (Ag+ SPE) provided ten fractions (Fractions F1−F10) of the esterifiable OSPW. The method was based on a scaled up version of a method described previously.8,15 System blanks were created for each fraction from solvents or by eluting the sorbent with corresponding solvents in the absence of OSPW extract and were reduced in volume as described above. OSPW extracts in hexane (300 mg) were loaded onto the phase which was subsequently eluted using hexane (F1−F4, 39% by weight), 95% hexane:5% diethyl ether (F5−F7, 34% by weight), 90% hexane:10% diethyl ether (F8), 100% diethyl ether (F9, ∼12% by weight), and finally 100% methanol (F10,∼ 16% by weight).8,15 Fractions were collected, reduced to dryness under a steady stream of nitrogen at 40 °C, and weighed. Fractions 1−3 were shown by GCxGCMS of each fraction as previously15 to comprise the alicyclic ‘classical’ NA and typically differed only with the volume of hexane collected and so can essentially be considered equivalent. Fractions 5−7 were shown by GCxGC-MS to comprise aromatic NA and differed mainly in the volume of hexane:ether collected and so can be considered ‘aromatic’ NA fractions. The results of assays of F3 (as representative of alicyclic NA) and F6 (as representative of aromatic NA) are discussed herein. F3 and F6 were selected as they contained the largest amounts (∼17% by weight in each case) of materials in the alicyclic (F1−4) and aromatic (F5−7) NA fractions. Assays of F8−10 will be discussed elsewhere. GCxGC-MS analyses were conducted using a GCxGC column combination combining an apolar stationary phase in the primary column as described previously4 but also a combination in which the primary column was coated with an ionic liquid Supelco SLB-IL111 stationary phase (SigmaAldrich, Bellefonte, PA, USA) with dimensions 30 m × 0.25 mm × 0.20 μm. This was coupled via a 1 m × 0.25 mm deactivated fused silica column (SGE, Milton Keynes, UK) to a 3.2 m × 0.25 mm × 0.15 μm BPX50 secondary column (SGE). The initial temperature of the primary column was 40 °C and held for 1 min then ramped at 7 °C min−1 to 110 °C, then at 1

to discover which components might be responsible for the toxic effects.6 A study of distilled fractions of an esterified OSPW (after subsequent de-esterification) by the Microtox toxicological screening method showed that such acidic fractions dominated by unidentified NA, were toxic.7 Effective concentrations causing a 50% suppression of chemiluminescence in the bacterium Vibrio f ischeri were between 40 and 65 mg L−1. More recently, a fraction of OSPW containing alicyclic, ‘classical’ NA including the adamantane and diamantane acids4,5 was shown to be acutely toxic to larval zebrafish, Danio rerio.8 The concentration causing a lethal response in 50% of the experimental groups was ∼13 mg L−1. Interestingly however, an aromatic NA fraction from the OSPW, containing compounds such as dehydroabietic acid, was slightly more toxic8 (LC50 ∼8 mg L−1). Whether alicyclic and aromatic NA in OSPW are responsible for more specific toxicological actions, now remains to be examined. Certainly fish exposed to OSPW containing ∼11−41 mg L−1 NA showed altered concentrations of sex steroid hormones, impaired reproductive performance and less prominent secondary sexual characteristics.9 The estrogen receptor- (ER) and androgen receptor- (AR) mediated effects of an OSPW containing ∼20 mg L−1 NA measured in vitro were significantly greater than those of controls.10 Exposure to OSPW also produced significant antiandrogenicity and potentiation of androgen receptor-mediated effects.10 It was speculated that OSPW could cause estrogenic and antiandrogenic effects through receptor mediated pathways.10 In a further study, an in vivo investigation of the endocrinedisrupting effects of OSPW male and female fathead minnow (Pimephales promelas) was made.11 The results indicated that this OSPW had endocrine-disrupting effects at all levels of the Brain-Gonad-Liver axis, but the compounds responsible for the activities were not assigned or delimited. Gagné et al.12 exposed Oncorhynchus mykiss trout hepatocytes to OSPW extracts and used the quantitative reverse transcriptase polymerase chain reaction (qRT-PCR) technique to monitor the effects. This study identified a suite of gene targets that responded specifically to OSPW extracts, which could serve as toxicogenomic fingerprints of OSPW contamination, but again the compounds responsible for these activities in OSPW were not assigned. If the compounds in OSPW causing these sublethal responses could be delimited or identified, a more targeted approach to detoxification by methods such as ozonation could perhaps be designed, or at least the effects of such attempts may possibly be predicted more accurately. Previously some tentatively identified aromatic NA in OSPW were suggested by computer modeling to produce estrogenic effects,13 but until reference compounds are available for structural confirmation and toxicity testing, as with the alicyclic acids4,5 these remain largely untested hypotheses. So, in order to delimit and help to identify possible OSPW components associated with these estrogenic effects, in particular the aromatic NA, in the present study we studied extracts and fractions of OSPW NA, rather than whole OSPW. The induction of the vtg gene transcripts (vtg) that encode for VTG proteins in larval zebrafish, D. rerio which is a common bioassay used to indicate effects consistent with exposure to exogenous estrogens,14 was used as a sensitive measure of estrogenicity. B

dx.doi.org/10.1021/es304799m | Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Environmental Science & Technology

Article

°C min−1 to 270 °C, and held for 10 min. The secondary column was held at 40 °C above the primary oven and the hot jet at 60 °C above the primary oven. The modulation period was 8 s with a 0.35 s hot jet pulse. Injection volume was 1 μL, and the flow rate was 1.2 mL min−1. Inlet temperature was 300 °C, transfer line 290 °C, and ion source 280 °C. Data processing was conducted using GC Image v2.1 (Zoex) with peak detection (referred to as “blob detection” within GCImage software) settings of peak apex >5× standard deviation of noise and volume >100000. Third, a GCxGC column combination combining a polar wax stationary phase in the primary column, as described previously,15 was used (as shown in Table of Contents graphic). Zebrafish Larvae Assay. Zebrafish were reared in the Zebrafish Research Facility, Plymouth University, and maintained under routine approved animal welfare protocols. Photoperiod was 12 h, temperature was 28 ± 1 °C, and stock fish were fed three times daily with live brine shrimp nauplii, Artemia sp., or dry fish flake mix (equal proportions ZM Systems flake, brine shrimp, Spirulina, and TetraMin stable flake). Larvae were routinely bred from bulk spawning of stock fish. Developing larvae were kept in plastic dishes (90 mm diameter, 50 mL) with daily water changes to remove unfertilized or undeveloping eggs and debris. Hatched embryos, 72 h postfertilization (hpf), were used for all larval exposures. Double concentration stocks of OSPW fraction exposure treatments were made up in zebrafish water (Plymouth tapwater, buffered and aerated for min. 24 h) and pH determined. Larvae were exposed (96 h) in triplicate glass beakers of 200 mL (final 1:1 dilution with stock oil fractions), n = 40 larvae per beaker. After exposure, larvae were collected by sieve, transferred to 1.5 mL microcentrifuge tube, and placed on ice for rapid euthanasia. Excess water was removed, and samples were stored at −80 °C until further analysis. Exposures were carried out in 4 experiments with 4 different batches of larvae (exposure 1: F3 and F6 blank fractions, exposure 2: F3 and OSPW fractions, exposure 3: F6 and OSPW-DM fractions, exposure 4: dehydroabietic acid, DHAA). Each experiment contained triplicate zebrafish water controls (no added OSPW fraction), triplicate ethanol (EtOH) controls (0.001% EtOH), and 0.1 and 1 μg L−1 EE2 (17α-ethynyl estradiol) as withinbatch positive controls. For RNA extraction, total RNA was extracted from each sample of 40 larvae following the manufacturer’s protocol (RNeasy MiniKit for animal tissue, Qiagen) with initial sonication (3−5 s), additional tissue break-up (QiaShredder column,Qiangen), and a 15 min DNase treatment. RNA was eluted into 30 μL, and the concentration and quality of total RNA were determined by spectrophotometer (NanoDrop, ND1000 Spectrophotometer). Quality control of RNA purity was maintained by acceptance of 260/280 and 260/230 ratios within the range 1.8−2.3. Samples out with the range (1% of total number of samples) were rewashed to improve quality ratios. All samples were diluted to 100 ng μL−1 total RNA, and 800 ng was used to synthesize cDNA following the manufacturer’s protocol for ImProm-IITM Reverse Transcription System (Promega), with random hexamer primers and deoxynucleotide mix (Sigma-Aldrich). cDNA was synthesized under the following conditions: annealing at 25 °C, extending at 42 °C, and heat-inactivating transcriptase at 70 °C (GeneAmp PCR System, 9700, Applied Biosystems). cDNA was stored at −80 °C until q-RT-PCR gene expression analysis.

Primers were selected by Primer Blast (NCBI). The amplicons were designed to span 1 intron junction and were checked to avoid secondary structure, self-annealing, complementarity, and potential hairpins by DNA calculator (SigmaAldrich) and OligoCalc (Northwestern University, USA). Amplicon size was verified on a 2% agarose gel after PCR amplification. The forward primer 5′-ACACAGCCATGGATGAGGAAATCG and the reverse primer 3′-TCACTCCCTGATGTCTGGGTCGT were used to amplify β-actin cDNA (ref seq. NM_131031.1), which was used as a housekeeping gene, and vtg1 (ref seq. NM_0010044897.2) cDNA was amplified by use of the forward primer 5′-ATCAGTGATGCACCTGCCCAGATTG and the reverse primer 3′-ACGCAAGAGCTGGACAAGCTGAA. Lyophilized primers (Eurofins MWG Operon, Ebersberg, Germany) were reconstituted to 100 μmol with RNase-free water and mixed with SYBR Green JumpStart Taq ReadyMix to give a final reaction concentration of 375 nmol in 20 μL of total volume. Fluorescence was detected (StepOnePlus Real-Time PCR System, Applied Biosystems) over 40 cycles, cycling conditions of 94 °C for denaturing, primer-specific annealing (55 °C beta actin, 60 °C vtg), and extension at 72 °C. For analysis, the cycle threshold was set to 25,000 for all qRT-PCR runs. A standard curve of cDNA template was run on each plate to calculate efficiency of each reaction. For gene expression analysis, the efficiency of qRT-PCR was calculated (E = 10(−1/slope)) from the standard curve for each plate. Only efficiencies between 90 and 110% were accepted for further analysis, and across-plate normalization was carried out by resolving for slope and intercept of the standard curves. Efficiency-adjusted comparative quantification was used for calculating fold-changes in the gene of interest compared with the reference housekeeping gene, β-actin efficiency-adjusted 2−ΔΔCt.19 EtOH-exposed larvae (0.001% EtOH, n = 3 beakers per experiment) were used as controls for relative quantification. There were no differences in β-actin expression between fractions or concentrations (GLM, p > 0.05), and some exposures resulted in lower control (water and ethanol) samples by up to 2 CT values. Using the lowest fraction concentrations as controls for fold change calculation did not alter the pattern or magnitude of fold change results (data not shown); therefore, β-actin was considered a suitable housekeeping gene, with EtOH control values suitable for normalization. Statistical analyses of results were performed using Statgraphics Plus 5.1 (Herndon, VA, USA). Following checks for variance using Cochran’s C test, data were assessed by analysis of variance. Where there was a significant difference (p < 0.05) of means, the data were further analyzed by Student− Newman−Keuls test to determine significant differences (p < 0.05) between treatments.

■

RESULTS AND DISCUSSION Numerous studies have shown that different, non- or necessarily incompletely characterized, unfractionated whole OSPW or occasionally, OSPW extracts, produce nonlethal toxicological effects (e.g. refs 1−3 and 9−12). These include altered concentrations of sex steroid hormones and impaired reproductive performance, antiandrogenicity, and potentiation of androgen receptor-mediated effects and endocrine-disrupting effects in fish (e.g., refs 10 and 12). The egg-yolk protein C

dx.doi.org/10.1021/es304799m | Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Environmental Science & Technology

Article

Figure 1. Bar chart illustrating relative fold changes (efficiency-adjusted 2−ΔΔCt;28 bars indicate 1 standard deviation) in vtg expression in larval zebrafish after treatment with esterifiable acids in oil sands process water extract (OSPW-DM) and argentation solid phase extraction (Ag+ SPE) fractions eluting with 100% hexane (F3; ‘classical’ alicyclic naphthenic acids), 95% hexane:5% diethyl ether (F6; ‘aromatic’ naphthenic acids) and dehydroabietic acid (DHAA) plus system blank controls (demethylation solvent; Ag+ SPE column fraction blanks, F3, F6; EtOH blank for DHAA defined as fold change = 1, dotted line). Concentrations (increasing greyscale) were 20, 200, and 2000 μg L−1 for the OSPW-DM, 10, 100, and 1000 μg L−1 for SPE fractions F3 and F6, and 1, 10, and 100 μg L−1 for DHAA (since 1000 μg L−1 is approximately the 50% lethal concentration of DHAA to larval zebrafish).8 Insert shows average response (log scales) of 0.1 and 1 μg L−1 of 17α(H)-ethynyl estradiol (EE2) in all assays plus responses of 10 to 1000 μg L−1 F3 and F6 fractions on the same log scale, indicating a weak relative response of the latter.

vitellogenin (VTG) gene was expressed to a maximum of ∼30fold by unfractionated OSPW.12 None of these studies investigated whether esterifiable NA in the OSPW actually caused the effects- and if so, which NA were responsible. It is well-known that OSPW samples from different sources vary in elemental composition16 and in the composition of individual isomers of NA.17 Unfortunately there is, at present, no accepted reference OSPW sample with which comparisons of the results of different studies can be made. Commercial NA mixtures are certainly not suitable references for OSPW NA.16 Neither is there any accepted reference method of analysis for OSPW (though even if there were, a reference sample would still be needed for meaningful comparisons). Furthermore, the availability of samples of OSPW for study outside of Canada has been limited to date. Therefore, in the absence of an agreed reference OSPW, a NA concentrate isolated from a large volume of OSPW, from a source sampled numerous times (ref 7 and references therein) and NA of which have been wellstudied previously by a variety of analytical methods, including elemental, IR, UV, ESI-Orbitrap-MS, GCxGC-MS, GCxGCHRMS, and GCxGC-SCD,18 was chosen for study herein. In addition we conducted some further detailed GCxGC-MS analyses of those NA fractions which were demonstrated herein to exhibit estrogenic activity. It is accepted that not all components in OSPW will be esterifiable but many previous studies have implicated NA as the compounds of concern in OSPW (reviewed in ref 19), so these were the focus of the current study. Estrogenic Activity. The only treatments with OSPW and OSPW fractions that produced fold changes in expression of vitellogenin genes (vtg) in larval zebrafish statistically significantly different from those induced by EtOH controls in the vtg assay were those conducted with the demethylated OSPW ≥ 200 μg L−1 and with the aromatic NA fraction F6 ≥ 100 μg L−1. Thus, relative to the fold change in the controls and transfer solvents, 20−2000 μg L−1 of an ethyl acetate extract of the total OSPW NA concentrate did not induce expression of

vitellogenin genes (vtg) in larval zebrafish (Figure 1S). This is perhaps unsurprising; an earlier study10 showed that much higher concentrations (∼20000 μg L−1) of NA in undiluted OSPW were needed to produce estrogenicity, as measured in a different assay. In contrast, 2000 μg L−1 of the esterifiable NA from the present OSPW acid concentrate (i.e., demethylated OSPW) caused an ∼8-fold change in induction expression relative to controls, and a dose-dependent relationship was observed between 20 and 2000 μg L−1 (Figure 1). Thus, some esterifiable NA in this OSPW acid concentrate are indeed responsible for at least some estrogenic effects in fish. The fact that the ethyl acetate extract of the total OSPW concentrate did not induce expression of vitellogenin genes (vtg) in larval zebrafish, whereas the same concentrations of esterifiable acids did is puzzling but may be due to the presence of estrogen antagonists in the total OSPW concentrate. Unfortunately this was not investigated during these experiments. A previous study indicated that estrogen antagonists were not present in a different OSPW,10 but OSPW samples are known to vary in composition.16 More relevant to the present aims, when fractions of the esterifiable NA were examined, a fraction eluting from Ag+ SPE with hexane (F3, 17% by weight) comprising up to 1000 μg L−1 ‘classical’ alicyclic (nonaromatic) acids produced no response or dose-dependent relationship relative to the controls (Figure 1). This fraction plus the similar adjacent fractions (i.e., F1−4 hexane eluates) represented about 39% of the recoverable acids by weight. Under the same assay conditions, 0.1 and 1 μg L−1 EE2, used as positive controls, produced >1000-fold vtg inductions (Figure 1, insert). This nonactivity of F3 is also consistent with the nonestrogenicity of some of the alicyclic adamantane acids in this fraction recorded in the CALUX estrogen alpha receptor assay and with the absence of computer-predicted estrogenicity13 predicted for the adamantane and diamantane acids identified in this fraction by GCxGC-MS (e.g., refs 4 and 5). Whether this result is typical of most, or all, OSPW samples remains to be seen; the acid distributions of OSPW from two different industrial operators D

dx.doi.org/10.1021/es304799m | Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Environmental Science & Technology

Article

complex mixture of acids (methyl esters), an ionic liquid stationary phase was used in the first dimension GC column. This allowed virtually complete chromatographic separation of the highly complex mixture of aromatic NA, apparently for the first time (Figure 2 and insert). The mixture was no longer a

and of different ages and from different locations were certainly broadly similar in composition, though the proportions of individual acids differed.17 In contrast, an aromatic NA fraction (F6, 17% by weight) eluting from SPE with 95% hexane:5% diethyl ether, which with similar fractions (i.e., F5−7 95% hexane:5% ether eluates) quantitatively represented >30% of the recoverable acids by weight,8 did exhibit a dose relationship for vtg expression between 10 and 1000 μg L−1, with the highest concentration causing an approximately 12-fold induction (Figure 1). Under the same assay conditions, 0.1 μg L−1 and 1 μg L−1 EE2, used as positive controls, produced on average >1000 -fold vtg inductions (Figure 1, insert). Where induction of vtg was detected, there was no corresponding induction of vtg in the SPE column blanks for F3 and F6 when made up to the same volumes (Figure 1). Thus the weak estrogenicity of this OSPW sample seems to be associated with the acids in the aromatic NA fraction (F6) and not, in this sample at least, with the ‘classical’ alicyclic NA (F3). This raises the question: what is it in the F6 aromatic acids which is weakly estrogenic? Numerous methods have been applied to the characterization of OSPW NA.19 The only method which has so far resulted in identification of any NA in OSPW is GCxGC-MS.4,5 Even then, relatively few acids have been identified firmly with reference acids, and several structures have only been assigned tentatively.20 However, the NA compositions of OSPW NA alicyclic fractions (e.g., F3) and aromatic fractions (e.g., F6) are very different. The alicyclic fraction F3 contains the numerous tricyclic adamantane,4 pentacyclic diamantane,5 and recently identified tetracylic acids,21 and the aromatic fractions (such as F6) contain DHAA15 and the tentatively identified aromatic acids20 containing dominant m/z 145 and m/z 310, 237 ions in the spectra, plus many others. Elemental analysis showed previously for a similar aromatic NA fraction of OSPW that no nitrogen or sulfur was present, and IR and UV spectroscopy confirmed the presence of esters and aromaticity.18 Therefore we concluded that characterization of the present weakly estrogenic fraction of aromatic acids in the present F6 by GCxGC-MS might be useful. The only aromatic acid identified in OSPW and eluting in the similar relevant SPE fractions to date is dehydroabietic acid, DHAA, identified by GCxGC-MS of the methyl ester and of an authentic sample.15 However, concentrations of 1 to 100 μg L−1 DHAA (a higher concentration of 1000 μg L−1 DHAA was toxic to the larvae8) examined in the present study caused no significant increase in relative fold change in vtg expression herein (Figure 1). This is also consistent with predictive models,13 so the observed estrogenic effects must be due to other constituents in the aromatic F6 fraction. The other dominant individual aromatic NA of a similar Ag+ SPE fraction to F6 were also tentatively assigned previously as monoaromatic acids.15,20 Electron ionization mass spectra of several contained a base peak ion m/z 145, possibly indicating a tetrahydroindan-type fragment,20 though this requires confirmation by synthesis and other studies. Other monoaromatic acids also tentatively identified in this fraction included other putative degradation products of aromatic steroid hydrocarbons:20 interestingly, some were modeled to be weakly estrogenic.20 GCxGC-MS studies of the methyl esters of F6 were therefore made herein to investigate whether such compounds were indeed present in this fraction. For improved separation of the

Figure 2. GCxGC-MS total ion current (TIC) chromatogram of methyl esters of argentation solid phase extraction fraction F6 (‘aromatic’ naphthenic acids) using an ionic liquid stationary phase in the primary GC column, illustrating near complete separation of ∼2000 constituents, dominated by about eight acids characterized by a base peak ion m/z 145 in their electron ionization mass spectra21,22 (GCxGC-MS conditions are described in the text). Insert shows GCxGC-MS total ion current (TIC) chromatogram of the same sample using an apolar stationary phase in the primary GC column8 in which most acids were unresolved (GCxGC-MS conditions are described in the text).

chromatographic “hump” or unresolved complex mixture, as widely reported previously (e.g., Figure 2 insert), but comprised a series of resolved peaks. Over 2000 individual NA peaks were now separated, and mass spectra were obtained for each of them. Interpretation of these spectra will take much more effort, but should then allow the weakly estrogenic aromatic NAs to be identified. Most of the NA were only minor components (Figure 2): the dominant individual compounds were again those with the m/z 145 base peak ion.15,20 Integration of the total ion current response indicated that together eight aromatic NA comprised about 5% of the total aromatic NA (Figure 2) and thus ∼1% of the esterifiable acids (OSPW-DM). The dominance of these m/z 145 compounds in the GCxGC-MS chromatograms under all conditions (Figure 2) does not, of course, mean that they account for the observed estrogenic effects; this will require investigation of the effects of the individual acids (e.g., after isolation by preparative gas chromatography22). An additional observation about the OSPW NA mixtures is possible: the estrogenicity of a commercial NA mixture assayed previously was weak compared to an OSPW sample.10 Although commercial samples vary in composition, even within suppliers, it may be relevant that the gas chromatographic (GCMS) profiles of esterified commercial acids are often unimodal,23 in contrast to those of OSPW, which are usually bimodal.8,15,20 The later eluting node comprises the polycyclic aromatic acids such as DHAA.8,15 The OSPW NA, and E

dx.doi.org/10.1021/es304799m | Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Environmental Science & Technology

Article

acids (e.g., by GCxGC-high resolution-MS21) might be productive. Perhaps also, remediation of estrogenic effects will require study of the effects of ozonation and other proposed treatments, not only on the ‘classical’ alicyclic acids which were not estrogenic in the OSPW examined herein, but also on the structurally dissimilar aromatic acids which have thus far not been investigated by ozonation or other remediation methods, so far as we are aware. Clearly, much is still to be learned about these supercomplex mixtures.

especially the aromatic NA distributions and the composition of commercial NA mixtures, are thus sometimes (perhaps always) quite different,8,18,23 probably due to use of various feedstock oils and multiple isolation methods for commercial NA. Environmental Relevance. The vtg-inducing effects of esterifiable OSPW acids and particularly the aromatic OSPW acids demonstrated herein should be viewed in context. The anthropogenic steroid EE2 induced fold changes in zebrafish vtg expression herein at concentrations much lower than those of the aromatic OSPW acids (Figure 1, insert). It is known that the natural estrogens estradiol and estrone are also more active at their optimum concentrations.14,25 Previously an assay sensitive to up to 10 nM E2 required OSPW containing 20 mg L−1 naphthenic acids to elicit a statistically significant estrogenic response.10 Thus it seems that OSPW constituents are much less estrogenic than steroids such as E2 and EE2. However, the quantities of acids in OSPW are large, and there is considerable concern about the possible environmental effects of any unintended discharges. Once organic chemicals, such as the supercomplex mixtures found in OSPW, enter the environment, their fate is partly dependent on the physicochemical properties of the individual compounds rather than the whole. Some effects might be due to acids as a group and governed by, for example, their micellar or surfactant-like behavior.26 Other effects, such as those studied here, may be more structure-dependent. Interactive effects between the surfactant-like behavior and the estrogenic activity may also be possible. Harris et al.27 reported that the presence of the surfactant linear alkylbenzene sulfonate enhanced the estrogenic activity of a mixture of environmental estrogens. This effect was observed not only in a yeast estrogen screen assay but also with vtg induction in fathead minnow. The effect was found to be transient in both assays, but nevertheless this interactivity may be important since OSPW contains compounds with surfactant-like properties and estrogenic activity. It is possible that other fractions (e.g., F8−10) from the argentation SPE procedures will also contribute to the estrogenic and acutely toxic effects of OSPW, but these are still to be investigated. However, the estrogenic effects measured for the aromatic acid fraction seem sufficient to account for most of those of the esterifiable acids (Figure 1). There are many approaches to environmental monitoring; one approach to environmental monitoring might be to study those chemicals likely to cause the most deleterious effects. One approach of many, illustrated here, is to use the structural features (e.g., aromaticity) of the acids to direct relevant fractionation methods (e.g., Ag+ SPE)15 so that the effects of the individual chemicals and subclasses of mixtures containing them can be studied. Of course, studies of the whole OSPW samples are equally valid and useful, but in the environment, compounds may dissipate differentially. GCxGC-MS analysis of the aromatic acids using an ionic liquid stationary phase as the primary GC column revealed that relatively few compounds, including those with base peak m/z 145 in their mass spectra, dominated the chromatograms (Figure 2). With such improved chromatographic separations it is now possible to obtain GCxGC-MS accurate mass measurements of molecular and fragment ions of naphthenic acid esters21 and thereby to gain further structural information, which can be used to improve modeling of toxicological effects.13 From the results obtained thus far, it would seem that, if the weak estrogenic effects in fish associated with the esterifiable acids of OSPW demonstrated herein are deemed to be of interest, monitoring of the aromatic

■

ASSOCIATED CONTENT

S Supporting Information *

Bar chart illustrating relative fold changes in vtg expression in larval zebrafish. This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Present Address #

Bermuda Institute for Ocean Sciences, 17 Ferry Reach, St. George’s GE01, Bermuda. E-mail:[email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS Funding of this research was provided by an Advanced Investigators Grant (no. 228149) awarded to S.J.R. for project OUTREACH, by the European Research Council, to whom we are extremely grateful. We thank Ms. G. Aguirre-Martinez for help with the DHAA assay.

■

REFERENCES

(1) Nero, V.; Farwell, A.; Lee, L. E. J.; Van Meer, T.; MacKinnon, M. D.; Dixon, D. G. The effects of salinity on naphthenic acid toxicity to yellow perch: Gill and liver histopathology. Ecotoxicol. Environ. Saf. 2006, 65 (2), 252−264. (2) Peters, L. E.; MacKinnon, M.; Van Meer, T.; van den Heuvel, M. R.; Dixon, D. G. Effects of oil sands process-affected waters and naphthenic acids on yellow perch (Perca flavescens) and Japanese medaka (Orizias latipes) embryonic development. Chemosphere 2007, 67 (11), 2177−2183. (3) Young, R. F.; Orr, E. A.; Goss, G. G.; Fedorak, P. M. Detection of naphthenic acids in fish exposed to commercial naphthenic acids and oil sands process-affected water. Chemosphere 2007, 68 (3), 518−527. (4) Rowland, S. J.; Scarlett, A.; West, C.; Jones, D.; Frank, R. Diamonds in the rough: identification of individual naphthenic acids in oil sands process water. Environ. Sci. Technol. 2011, 45 (7), 3154− 3159. (5) Rowland, S. J.; West, C. E.; Scarlett, A. G.; Jones, D.; Frank, R. A. Identification of individual tetra- and pentacyclic naphthenic acids in oil sands process water by comprehensive two-dimensional gas chromatography/mass spectrometry. Rapid Commun. Mass Spectrom. 2011, 25 (9), 1198−1204. (6) Clemente, J. S.; Fedorak, P. M. A review of the occurrence, analyses, toxicity, and biodegradation of naphthenic acids. Chemosphere 2005, 60 (5), 585−600. (7) Frank, R. A.; Kavanagh, R.; Burnison, B. K.; Arsenault, G.; Headley, J. V.; Peru, K. M.; Van Der Kraak, G.; Solomon, K. R. Toxicity assessment of collected fractions from an extracted naphthenic acid mixture. Chemosphere 2008, 72 (9), 1309−1314. (8) Scarlett, A. G.; Reinardy, H. C.; Henry, T. B.; West, C. E.; Frank, R. A.; Hewitt, L. M.; Rowland, S. J. Acute toxicity of aromatic and non-

F

dx.doi.org/10.1021/es304799m | Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Environmental Science & Technology

Article

aromatic fractions of naphthenic acids extracted from oil sands process affected water to larval zebrafish. Submitted to Chemosphere. (9) Kavanagh, R. J.; Frank, R. A.; Oakes, K. D.; Servos, M. R.; Young, R. F.; Fedorak, P. M.; MacKinnon, M. D.; Solomon, K. R.; Dixon, D. G.; Van Der Kraak, G. Fathead minnow (Pimephales promelas) reproduction is impaired in aged oil sands process-affected waters. Aquat. Toxicol. 2011, 101 (1), 214−220. (10) He, Y. H.; Wiseman, S. B.; Hecker, M.; Zhang, X. W.; Wang, N.; Perez, L. A.; Jones, P. D.; El-Din, M. G.; Martin, J. W.; Giesy, J. P. Effect of ozonation on the estrogenicity and androgenicity of oil sands process-affected water. Environ. Sci. Technol. 2011, 45 (15), 6268− 6274. (11) He, Y.; Wiseman, S. B.; Wang, N.; Perez-Estrada, L. A.; El-Din, M. G.; Martin, J. W.; Giesy, J. P. Transcriptional responses of the brain−gonad−liver axis of fathead minnows exposed to untreated and ozone-treated oil sands process-affected water. Environ. Sci. Technol. 2012, 46 (17), 9701−9708. (12) Gagné, F.; Douville, M.; André, C.; Debenest, T.; Talbot, A.; Sherry, J.; Hewitt, L. M.; Frank, R. A.; McMaster, M. E.; Parrott, J.; Bickerton, G. Differential changes in gene expression in rainbow trout hepatocytes exposed to extracts of oil sands process-affected water and the Athabasca River. Comp. Biochem. Physiol., Part C: Toxicol. Pharmacol. 2012, 155 (4), 551−559. (13) Scarlett, A. G.; West, C. E.; Jones, D.; Galloway, T. S.; Rowland, S. J. Predicted toxicity of naphthenic acids present in oil sands processaffected waters to a range of environmental and human endpoints. Sci. Total Environ. 2012, 425, 119−127. (14) Henry, T. B.; McPherson, J. T.; Rogers, E. D.; Heah, T. P.; Hawkins, S. A.; Layton, A. C.; Sayler, G. S. Changes in the relative expression pattern of multiple vitellogenin genes in adult male and larval zebrafish exposed to exogenous estrogens. Comp. Biochem. Physiol., Part A: Mol. Integr. Physiol. 2009, 154 (1), 119−126. (15) Jones, D.; West, C. E.; Scarlett, A. G.; Frank, R. A.; Rowland, S. J. Isolation and estimation of the ‘aromatic’ naphthenic acid content of an oil sands process-affected water extract. J. Chromatogr., A 2012, 1247, 171−175. (16) Grewer, D. W.; Young, R. F.; Whittal, R. M.; Fedorak, P. M. Napthenic acids and other acid-extractables in water samples from Alberta: what is being measured? Sci. Total Environ. 2010, 408, 5997− 6010. (17) Rowland, S. J.; West, C. E.; Scarlett, A. G.; Ho, C.; Jones, D. Differentiation of two industrial oil sands process-affected waters by two-dimensional gas chromatography/mass spectrometry of diamondoid acid profiles. Rapid Commun. Mass Spectrom. 2012, 26 (5), 572− 576. (18) Jones, D.; Scarlett, A. G.; West, C. E.; Frank, R. A.; Gieleciak, R.; Hager, D.; Pureveen, J.; Tegelaar, E.; Rowland, S. J. Elemental and spectroscopic characterisation of fractions of an acidic extract of oil sands process water. Chemosphere, in press. (19) Hrudey, S. E.; Gosselin, P.; Naeth, M. A.; Plourde, A.; Therrien, R.; Van Der Kraak, G.; Xu, Z. Environmental and health aspects of Canada’s oil sands industry; Royal Society of Canada: 2010; p 22. (20) Rowland, S. J.; West, C. E.; Jones, D.; Scarlett, A. G.; Frank, R. A.; Hewitt, L. M. Steroidal aromatic ‘naphthenic acids’ in oil sands process-affected water: structural comparisons with environmental estrogens. Environ. Sci. Technol. 2011, 45 (22), 9806−9815. (21) West, C. E.; Scarlett, A. G.; Pureveen, J.; Tegelaar, E.; Rowland, S. J. Abundant naphthenic acids in oil sands process-affected water: Studies by synthesis, derivatisation and GCxGC-high resolution mass spectrometry. Rapid Commun. Mass Spectrom. 2013, 27, 357−365. (22) Sutton, P. A.; Lewis, C. A.; Rowland, S. J. Isolation of individual hydrocarbons from the unresolved complex hydrocarbon mixture of a biodegraded crude oil using preparative capillary gas chromatography. Org. Geochem. 2005, 36 (6), 963−970. (23) Rowland, S. J.; West, C. E.; Scarlett, A. G.; Jones, D.; Boberek, M.; Pan, L.; Ng, M.; Kwong, L.; Tonkin, A. Monocyclic and monoaromatic naphthenic acids: Synthesis & characterisation. Environ. Chem. Lett. 2011, 9, 525−533.

(24) Lange, A.; Katsu, Y.; Miyagawa, S.; Ogino, Y.; Urushitani, H.; Kobayashi, T.; Hirai, T.; Shears, J. A.; Nagae, M.; Yamamoto, J.; Ohnishi, Y.; Oka, T.; Tatarazako, N.; Ohta, Y.; Tyler, C. R.; Iguchi, T. Comparative responsiveness to natural and synthetic estrogens of fish species commonly used in the laboratory and field monitoring. Aquat. Toxicol. 2012, 109, 250−258. (25) Caldwell, D. J.; Mastrocco, F.; Anderson, P. D.; Lange, R.; Sumpter, J. P. Predicted-no-effect concentrations for the steroid estrogens estrone, 17 beta-estradiol, estriol, and 17 alpha-ethinylestradiol. Environ. Toxicol. Chem. 2012, 31 (6), 1396−1406. (26) Britton, L. N. Toxicity and Persistence of Surfactants used in the Petroleum Industry. In Surfactants: Fundamentals and Applications in the Petroleum Industry; Schramm, L. L., Ed.; Cambridge University Press: Cambridge, UK, 2000; pp 541−566. (27) Harris, C. A.; Brian, J. V.; Pojana, G.; Lamoree, M.; Booy, P.; Marcomini, A.; Sumpter, J. P. The influence of a surfactant, linear alkylbenzene sulfonate, on the estrogenic response to a mixture of (xeno)estrogens in vitro and in vivo. Aquat. Toxicol. 2009, 91 (1), 95− 98. (28) Pfaffl, M. W. A new mathematical model for relative quantification in real-time RT-PCR. Nucleic Acids Res. 2001, 29 (9), 2003−2007.

G

dx.doi.org/10.1021/es304799m | Environ. Sci. Technol. XXXX, XXX, XXX−XXX