Interpreting Nonracemic Ratios of Chiral Organochlorines Using

Oct 4, 2001 - Enantioselective elimination by our test fish (Fundulus sp.) ...... Kimmel, Lenka; Coelhan, Mehmet; Leupold, Guenther; Vetter, Walter; P...
6 downloads 0 Views 84KB Size
Environ. Sci. Technol. 2001, 35, 4444-4448

Interpreting Nonracemic Ratios of Chiral Organochlorines Using Naturally Contaminated Fish W A L T E R V E T T E R , * ,† KELLY L. SMALLING,‡ AND KEITH A. MARUYA‡ Department of Food Chemistry, Friedrich-Schiller-University Jena, Dornburger Strasse 25, D-07743 Jena, Germany, and Skidaway Institute of Oceanography, 10 Ocean Science Circle, Savannah, Georgia 31411

Although reports of nonracemic proportions of chiral organochlorine pollutants in the environment are widespread, the interpretation of such data is not well developed. Using GC/MS and a chiral stationary phase consisting of 25% tert-butyldimethylsilylated β-cyclodextrin in PS086 (βBSCD), we followed the change in the enantiomeric signature of 2-exo,3-endo,6-exo,8,9,10-hexachlorobornane (B6-923) in naturally contaminated fish maintained under toxaphene-free conditions. Whereas the enantiomeric ratio (ER) of B6-923 was near racemic at the start of the elimination experiment, it had increased severalfold by the end of 60 d. On the basis of first-order kinetics, one enantiomer of B6-923 was eliminated twice as fast as its mirror image, resulting in half-lives of 7 and 13 d, respectively. Enantioselective elimination by our test fish (Fundulus sp.) strongly suggests active biotransformation of B6-923; however, bioprocessing throughput estimates suggest a very low in situ rate of natural attenuation. These results confirm that the relatively constant ERs observed for chiral organochlorines in a given species are the result of competing processes, e.g., uptake vs elimination. Our experiments also further illustrate the utility of enantioselective analysis in characterizing the biotransformation of persistent organochlorine pollutants.

Introduction The enantioselective processing of chemicals is of interest to a broad scientific audience, including environmental chemists, biochemists, natural product chemists, and pharmaceutical scientists (1, 2). Prior to the introduction of chiral stationary phases (CSPs) based on modified cyclodextrins some 10 years ago (3, 4), cumbersome isolation and concentration procedures were necessary to achieve the sensitivity and compound purity required for detailed analysis. With recent improvements in enantioselective technology, however, the enantiospecific determination of chiral compounds in environmental samples is now possible at ultra trace levels (5, 6). Since many organohalogen compounds including those in technical products such as DDT (e.g., o,p′-DDT), hexachlorocyclohexane or HCH (R-HCH), toxaphene, and chlor* Corresponding author telephone: +3641-949-657; fax: +3641949-652; e-mail: [email protected]. † Friedrich-Schiller-University Jena. ‡ Skidaway Institute of Oceanography. 4444

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 22, 2001

dane are chiral, enantioselective analyses are well-suited for characterizing their environmental processing and fate. Whereas enantiomers of organochlorine pollutants have identical physical/chemical properties in an achiral environment (e.g., melting points, mass spectra, and chromatographic retention times on achiral stationary phases), their response to biological processes may differ. If biological processing of a chiral contaminant dominates, it follows that one enantiomer would be depleted relative to its mirror image. One application of this technology is the determination of enantiomeric ratios (ERs)sthe concentration of one enantiomer divided by the concentration of its mirror images for organochlorines in environmental samples. Enantiospecific data for wildlife show that ERs vary by compound and trophic level; however, the ranges of species- and/or genusspecific ERs are relatively narrow (6-8). For instance, the ERs of R-HCH in cetaceans ranged from 1 to 4, while the concentrations (sum of both enantiomers) in individual animals varied by 3 or 4 orders of magnitude (Table 1). Although it is often reported that organochlorine levels in higher organisms increase with age (particularly for male individuals), ERs tend to be independent of age within a population. On the other hand, nonracemic proportions (i.e., ERs deviating from unity) of chiral organochlorines clearly indicate that enantioselective bioprocessing occurs in nature. Furthermore, it can be concluded that metabolic capability differs among species. However, the implications of ERs deviating from the racemate have not been fully explored and thus understood. Over a long period of time, for example, one might expect enantioselective biotransformation to result in the exclusive presence of the more stable enantiomer. This is, however, rarely observed in wild specimens (Table 1). The goal of the present study was to determine if ERs of organochlorine compound are the result of equilibrium between uptake and elimination processes. To test this hypothesis, we constrained uptake and followed the biotransformation/elimination of a chiral organochlorine analogue using a teleost. The model compound selected is derived from toxaphene, a nonsystemic organochlorine pesticide with a large global inventory (9, 10) that was recently classified as a persistent, bioaccumulative, and toxic (PBT) chemical of primary concern by the U.S. Environmental Protection Agency (11).

Materials and Methods Model Compound. B6-923 (2-exo,3-endo,6-exo,8,9,10-hexachlorobornane), a moderately hydrophobic (log Kow ∼ 4-5) relatively minor component of technical toxaphene (CTT) (12), was selected as our model compound. Along with 2-endo,3-exo,5-endo,6-exo,8,9,10-heptachlorobornane (B71001), B6-923 is a metabolite of anaerobic dechlorination of higher chlorinated bornanes (13) and is a typical contaminant in toxaphene-contaminated reducing soils and sediments (14-16). Assignment of structures for B6-923 and B7-1001 was carried out in previous work (15). Additionally, B6-923 does not biomagnify in aquatic food webs (17). Model Fish Species. The mummichog (Fundulus sp.) is a small, minnow-like fish (2-10 cm; 1-5 yr life span) that has a limited home range in shallow tidal creeks of coastal estuaries. Fundulus is widespread along the eastern United States and is a major component in terms of biomass in these systems. Because of its abundance, ecological impor10.1021/es015514s CCC: $20.00

 2001 American Chemical Society Published on Web 10/04/2001

TABLE 1. Enantiomeric Ratios (ERs) of r-HCH in the Blubber of Cetaceans species harbor porpoise

white-beaked dolphin Dall’s porpoise Baird’s beaked whale northern right whale dolphin Pacific white-sided dolphin common dolphin striped dolphin melon-headed whale Fraser’s dolphin spinner dolphin spinner dolphin humpbacked dolphin

sampling site

tissue

ER (n)

ref

North Sea North Sea Baltic Iceland North Sea Bering Sea northeastern North Pacific Japan Sea Japan northern North Pacific northern North Pacific northwestern North Pacific Japan Japan Japan Bay of Bengal eastern tropical Pacific Bay of Bengal

blubber blubber blubber blubber blubber blubber blubber blubber blubber blubber blubber blubber blubber blubber blubber blubber blubber blubber

1.7-3.9 (4) 2.8-2.9 (2) 1.4-2.5 (7) 1.9-3.9 (4) 1.1-1.5 (4) 2.0-2.1 (3) 1.6-1.8 (5) 1.9 (1) 2.1-2.8 (3) 1.8-1.9 (2) 2.2-2.8 (3) 1.7-1.8 (2) 1.7-1.9 (4) 1.7-2.1 (3) 1.6-1.8 (4) 2.4-2.8 (2) 2.0-2.6 (2) 1.6-1.7 (2)

25 26 25 25 25 27 27 27 27 27 27 27 27 27 27 27 27 27

tance, and resistance to a wide variety of environmental stressors, Fundulus is studied extensively both in its natural setting and in laboratory studies. Experimental Section. Fundulus were collected in baited minnow traps from the cooling water discharge canal of a former toxaphene factory near Brunswick, GA (United States). Fish from this site were collected and analyzed previously for toxaphene (14). Toxaphene residues, assumed to be at steady state with source(s) of contamination in this system, were dominated by B6-923 (our model compound) and B71001 (15). Site fish were transported to the lab in coolers filled with aerated site water. Control fish (i.e., those with no detectable toxaphene residues) were collected from a reference brackish estuary. Contaminated and control fish were kept in separate, flow-through fiberglass tanks that received filtered seawater. Water temperature, pH, and salinity were monitored daily. Fish were maintained for 60 d on a diet of commercial fish food (TetraMin). Neither seawater nor fish food contained detectable levels of CTTs. Two to three individual fish were collected synoptically from the control and contaminated fish tanks 0, 3, 7, 14, 28, and 60 d into the elimination phase. Individual weight and length measurements were taken before they were frozen at -20 °C. Individual samples (whole body) were then extracted with organic solvents and analyzed as described in detail elsewhere (14). Fractionation of toxaphene compounds was carried out on 8 g of silica (1 cm i.d. glass column) with n-hexane as the solvent. After an initial volume of 48 mL was discarded, several fractions (25 mL each) containing toxaphene residues were collected. B6-923 was targeted in fraction 148-173 mL. Determination of ERs by GC/MS. ERs were determined in Fundulus tissue using GC/MS in the electron capture negative ion (ECNI) mode. GC/ECNI-MS/SIM analyses (reactant gas: methane) were performed with a HewlettPackard 5890 gas chromatograph interfaced to a 5989B mass spectrometer using previously published parameters (14). A 30 m × 0.25 mm i.d. column coated with 0.2 µm of 25% randomly tert-butyldimethylsilylated β-cyclodextrin (βBSCD) diluted in PS086 (BGB Analytik, Adliswil, Switzerland) was used for enantioseparation of CTTs (16). The GC oven was programmed as follows: isothermal at 80 °C (4-min hold), 20 °C/min to 180 °C (15-min hold), 20 °C/min to 200 °C (25-min hold), and 20 °C/min to 230 °C (15-min hold). After a solvent delay of 20 min, the following 10 ions were detected in parallel at 1.11 cycles/s: m/z 273/275 (pentachloro-CTTs), 307/309 (hexachloro-CTTs), 343/345 (heptachloro-CTTs), 377/379 (octachloro-CTTs), and 411/413 (nonachloro-CTTs).

GC/ECNI-MS full-scan analyses were performed using a Hewlett-Packard 6890 series II Plus gas chromatograph coupled to a 5973 mass selective detector, also using methane as the moderating gas. Conditions for full-scan analyses are described in Maruya et al. (18). GC/ECD Analysis. Concentrations of individual CTTs and total toxaphene were determined using a Varian 3400CX gas chromatograph with a 63Ni ECD. The injector and detector temperatures were 270 and 330 °C, respectively. The glass capillary column (DB-5, 30 m × 0.25 mm × 0.25 µm; J&W Scientific, Folsom, CA) was programmed as follows: 120 °C (1-min hold); increase to 260 °C at 2 °C/min (15-min hold). Individual CTT concentrations were computed from the average response of a 22-component mixture (TM2; Dr. Ehrenstorfer, Germany) (5 point calibration; r2 > 0.99). The response factors for B6-923 and B7-1001 were assumed to be equal to the mean of all 22 components in TM2 (19). Quality Assurance/Quality Control (QA/QC). Provisions for QA/QC, including the analysis of procedural blanks, spiked reference sediment and tissue, and replicate samples, are outlined elsewhere (18, 19). On the basis of blank values and instrument sensitivity, the nominal detection limit for individual congeners was ∼1 ng/g. For our chiral GC analyses, we detected an unsaturated hexachloro-CTT interference (Hx-ene) that coeluted with B6-923. The presence of Hx-ene was not surprising as unsaturated chlorinated monoterpenes are known constituents of technical toxaphene (20). Because GC oven programming modifications or the use of alternate CSPs did not rectify this problem, we separated B6-923 and Hx-ene prior to GC analyses by silica gel packed column chromatography (see Materials and Methods).

Results and Discussion Figure 1 compares the gas chromatographic pattern of toxaphene residues in Fundulus at the start and at the end of the 60-d experiment using a CSP. The loss of earlier eluting, lower chlorinated toxaphene homologues by day 60 clearly indicates that congener-specific elimination had occurred. Among the congeners that were eliminated were B6-923 and B7-1001. Figure 2 shows that the elimination of B6-923 occurred enantioselectively. A stepwise increase in the ER of B6-923 was observed from day 0 to day 60. Enantiomer-2, the later eluting enantiomer, was completely eliminated after 30 d. Quantitative elimination of enantiomer-1 occurred by 60 d. To account for growth dilution and variable day 0 concentrations (97-987 µg/kg fish), we multiplied the measured concentration of B6-923 in individual fish at time VOL. 35, NO. 22, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

4445

FIGURE 1. GC/ECNI-MS total ion current chromatograms (β-BSCD column) of fish tissue (Fundulus sp.) extracts show the profound change in toxaphene residue profile from the (a) start to the (b) end of the 60-d elimination experiment. Arrows denote the range of octa- and nonachlorobornanes in these samples. Peaks corresponding to racemic B9-1679 are marked with asterisks (the second eluting enantiomer of B9-1679 contains an interfering peak). t (CB6-923,meas) by the ratio of the mean concentration of 2-endo,3-exo,5-endo,6-exo,8,8,9,10,10-nonachlorobornane (B91679) at time 0 (CB9-1679,t0) and the measured concentration of B9-1679 at time t (CB9-1679,t). (CB6-923,t ) CB6-923,t,meas[CB9-1679,t0/CB9-1679,t]). The latter congener is considered to be the most persistent nonachlorobornane in mammals (21, 22). Under strictly anaerobic conditions, B9-1679 can be dechlorinated to yield B7-1001, one of the two major deadend metabolites in sediments (Figure 3). However, B9-1679 exhibits much greater stability in biota (22), with half-lives in fish estimated at >1 yr (17). This was confirmed in our study as no shift in the ER of B9-1679 was observed throughout our experiment. Thus, we concluded that B91679 was not biotransformed by Fundulus during the elimination experiment, making it a suitable congener for the standardization of B6-923 concentrations (CB6-923,t, see above). The relative disappearance of B6-923 and B7-1001 was thus modeled on the basis of a constant B9-1679 level at each time point of the elimination study (Table 2). Assuming simple first-order kinetics, half-lives of 7 and 13 d were determined for B6-923 and B7-1001, respectively. Furthermore, enantiomer-2 of B6-923 was excreted twice as fast as enantiomer-1, with half-lives of enantiomer-1 and enantiomer-2 estimated as 9.3 and 4.7 d, respectively. This is consistent with our observation of complete elimination of both enantiomers within 60 d (Figure 1; Table 2). Note that our analysis was based on the analysis of whole fish, thus precluding tissue-specific variations in the ERs that have been reported previously (7, 8). Two observations are particularly noteworthy. The enantiomers of B6-923 were eliminated at different rates (roughly by a factor of 2). Also, Fundulus at the start of our elimination experiments exhibited constant ERs of ∼1.3 for B6-923. Since all sources of B6-923 were eliminated during our experiment, we conclude that maintenance of a relatively stable ER of 1.3 in wild fish requires a greater relative uptake rate for the less 4446

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 22, 2001

FIGURE 2. Severalfold increase in the enantiomeric ratio (ER) of B6-923, indicating faster elimination of the later eluting enantiomer (-2), was observed in Fundulus extracts using a chiral stationary phase (β-BSCD). GC/ECNI-MS/SIM traces are for m/z 307 (lower abundance) and 309 (higher abundance).

FIGURE 3. Structures of B6-923, B7-1001, and selected higher chlorinated components of technical toxaphene. Note the proposed dechlorination pathways that result in the formation of B6-923 and B7-1001 under reducing conditions.

TABLE 2. B6-923 Level and ER during the Elimination Studya day level (ng/g) ERb EFc enantiomer-1 (ng/g) enantiomer-2 (ng/g)

0 830 1.3 0.57 470 360

3 360 1.6 0.61 220 140

7 260 2.65 0.73 190 70

14 120 6 0.85 100 17

28 110 inf. 1.00 110 nd

60 nd nd nd

a B6-923 levels of individual fish are standardized to the level of B9-1679. b ER, enantiomer ratio is [E1]/[E2]. c EF, enantiomer fraction is defined as [E1]/[E1]+[E2] or ER/(1 + ER). nd ) Not detected.

stable enantiomer-2. In other words, the uptake of B6-923 from all sources (see below) must be in an enantiomeric proportion that is closer to racemic (i.e., ER < 1.3). This verifies our hypothesis that constant ERs in an individual organism is the consequence of equilibrium between uptake and elimination processes. Major food items for Fundulus in their natural habitat are small crustaceans, including grass shrimp (Palaemonetes sp.). Together, Fundulus and Palaemonetes comprise a large fraction of total secondary productivity/biomass in coastal estuarine systems. In an earlier study, the ER of B6-923 in grass shrimp was determined to be ∼1.1 (14). Other potential sources of B6-923 for Fundulus are plankton/detritus, benthic invertebrates, contaminated sediment, and water. B6-923 was racemic in sediments from the Brunswick site (14), as it was in sediments of a Canadian Arctic lake treated with toxaphene (16). Assuming that the ER for B6-923 in tidal creek water will mirror that in sediments, we submit that the uptake of B6-923 from these sources by Fundulus will be largely racemic. Knowledge of ERs at the start of our experiment and our subsequent estimation of enantiospecific half-lives allow for an estimate of the processing “throughput” of B6-923, defined

as mass of B6-923 transformed/eliminated by a population of Fundulus in a contaminated marsh. Once again assuming first-order kinetics and a constant initial ER of 1.3, 91% of the initial pool of B6-923 was taken up as the racemate and subsequently eliminated within its half-life (i.e., 7 d). Assuming an initial steady-state B6-923 concentration of 990 ng/g or ∼1 ppm and an average weight of 5 g/fish under steady-state conditions, 2.5 µg of B6-923/fish is biotransformed and eliminated within 7 d (or 130 µg fish-1 yr-1). If one conservatively assumes a constant total fish biomass of 1000 kg in a 4-km2 brackish marsh (∼0.5 individuals/m2 of tidal creek area), a resident teleost species with similar metabolic capabilities as Fundulus could process ∼ 26 g of B6-923/yr. Although aquatic species such as fish may in fact reduce the inventory of an otherwise persistent chiral organochlorine compound like B6-923, our data and calculations suggest that this mechanism of natural attenuation is neither efficient nor rapid. The persistency of elevated levels of B6-923 at our study site (14), more than 20 yr after releases of toxaphene were regulated, is further proof that attenuation by fish populations at natural densities is not significant. Our results illustrate the utility of enantioselective analysis in understanding processes acting on chiral organic compounds in the environment. Modeling of elimination kinetics in our experiments demonstrated that ERs for individual organisms were not influenced by initial concentration. This helps explain the constancy of ERs among individuals of feral populations. In our study, we have shown that a change in environmental conditions (namely, the elimination of source contamination) had a strong impact on ERs, suggesting that atypical ERs in wildlife are a good indicator of disturbance in an ecosystem. A possible example of such a disturbance in the wild is disease. Different ERs for R-HCH in sick and healthy individuals were reported even though VOL. 35, NO. 22, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

4447

R-HCH levels in liver did not vary significantly (23, 24). In the diseased animals, there was strong evidence that (enantioselective) metabolic capacity was reduced (23). Starvation is another example of a change that may result in atypical ERs. It follows that animals that starve over long periods of time (e.g. polar bears and breeding penguins) may exhibit large seasonal variations with respect to ERs of chiral pollutants. Clearly, the health and nutritional status of individuals are important factors when interpreting enantioselective data. In conjunction with other tools such as stable isotope analysis, enantioselective analyses may prove useful in assessing other biological processes/phenomena, such as the health status and food quality associated with a captive or wild population.

Acknowledgments Support for K.L.S. and K.A.M. was provided in part by EPA Region 4 and NOAA Georgia Coastal Incentive Grant (9812). The views expressed herein are wholly those of the authors and do not reflect the views of supporting agencies, including the U.S. EPA. W.V. wishes to thank B. Luckas at the University of Jena for supporting this work.

Literature Cited (1) Novotny, M. V. Science 1989, 246, 51-55. (2) Lewis, D. L.; Garrison, A. W.; Wommack, K. E.; Whitlemore, A.; Staudler, P.; Mellio, J. Nature 1999, 401, 898-901. (3) Ko¨nig, W. A. Gas chromatographic enantiomer separation with modified cyclodextrines; Hu ¨thig Buch Verlag GmbH: Heidelberg, 1992. (4) Schurig, V. J. Chromatogr. A 1994, 666, 111-129. (5) Faller, J.; Hu ¨ hnerfuss, H.; Ko¨nig, W. A.; Ludwig, P. Mar. Pollut. Bull. 1991, 22, 82-88. (6) Vetter, W. Food Rev. Int. 2001, 71, 113-182. (7) Vetter, W.; Schurig, V. J. Chromatogr. A 1997, 774, 143-175. (8) Kallenborn, R.; Hu¨hnerfuss, H. Chiral Environmental Pollutants; Springer: Heidelberg, 2001. (9) Vetter, W.; Oehme, M. In The Handbook of Environmental Chemistry, Vol. 3, Part K; Paasivirta, J., Ed.; Springer: Berlin, 2000; pp 237-287.

4448

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 22, 2001

(10) Saleh, M. A. Rev. Environ. Contam. Toxicol. 1991, 118, 1-85. (11) U.S. EPA. PBT National Action Plan for the Level I Pesticides (Public Review Draft). Persistent, Bioaccumulative and Toxic Pollutants (PBT) Pesticides Working Group: 2000. (12) Fisk, A. T.; Norstrom, R. J.; Cymbalisty, C. D.; Muir, D. C. G. Environ. Toxicol Chem. 1998, 17, 951-961. (13) Fingerling, G.; Hertkorn, N.; Parlar, H. Environ. Sci. Technol. 1996, 30, 2984-2992. (14) Vetter, W.; Maruya, K. A. Environ. Sci. Technol. 2000, 34, 16271635. (15) Stern, G. A.; Loewen, M. D.; Miskimmin, B. M.; Muir, D. C. G.; Westmore, J. B. Environ. Sci. Technol. 1996, 30, 2251-2258. (16) Vetter, W.; Bartha, R.; Stern, G.; Tomy, G. Environ. Toxicol. Chem. 1999, 18, 2775-2781. (17) Delorme, P. D.; Muir, D. C. G.; Lockhart, W. L.; Mills, K. H.; Ward, F. J. Chemosphere 1993, 27, 1965-1973. (18) Maruya, K. A.; Wakeham, S. G.; Vetter, W.; Francendese, L. Environ. Toxicol. Chem. 2000, 19, 2198-2201. (19) Maruya, K. A.; Vetter, W.; Wakeham, S. G.; Lee, R. F.; Francendese, L. Am. Chem. Soc. Symp. Ser. 2000, No. 772, 164-175. (20) Kimmel, L.; Coelhan, M.; Leupold, G.; Vetter, W.; Parlar, H. Environ. Sci. Technol. 2000, 34, 3041-3045. (21) Stern, G. A.; Muir, D. C. G.; Ford, C. A.; Grift, N. P.; Dewailly, E.; Bidleman, T. F.; Walla, M. D. Environ. Sci. Technol. 1992, 26, 1838-1840. (22) 22. Vetter, W.; Scherer, G. Environ. Sci. Technol. 1999, 33, 34583461. (23) 23. Pfaffenberger, B.; Hu ¨ hnerfuss, H.; Kallenborn, R.; Ko¨hlerGu ¨ nter, A.; Ko¨nig, W. A.; Kru ¨ ner, G. Chemosphere 1992, 25, 719725. (24) 24. Mo¨ssner, S.; Spraker, T. R.; Becker, P. R.; Ballschmiter, K. Chemosphere 1992, 24, 1171-1180. (25) 25. Hummert, K.; Vetter, W.; Luckas, B. Chemosphere 1995, 31, 3489-3500. (26) 26. Mu ¨ ller, M. D.; Schlabach, M.; Oehme, M. Environ. Sci. Technol. 1992, 26, 566-569. (27) 27. Tanabe, S.; Kumaran, P.; Iwata, H.; Tatsukawa, R.; Miyazaki, N. Mar. Pollut. Bull. 1996, 32, 27-31.

Received for review May 4, 2001. Revised manuscript received August 29, 2001. Accepted August 30, 2001. ES015514S