Environ. Sci. Technol. 2002, 36, 2884-2888
Influence of Chloride and Metals on Silver Bioavailability to Atlantic Salmon (Salmo salar) and Rainbow Trout (Oncorhynchus mykiss) Yolk-Sac Fry NICOLAS R. BURY* AND CHRISTER HOGSTRAND King’s College London, Division of Life Sciences, Franklin-Wilkins Building, 150 Stamford Street, London, SE1 9NN, U.K.
The effects of differing water chloride concentrations (0-10 mM) or competing metals [Cu(II), Cd(II), Zn(II), Pb(II), Co(II) (1-10000 nM)] on Ag(I) uptake in yolk-sac fry of two salmonid species, the Atlantic salmon (Salmo salar) and rainbow trout (Oncorhynchus mykiss), were studied. None of the metals tested were strong competitors of Atlantic salmon yolk-sac fry whole body Ag(I) influx. Inhibition of Ag(I) influx was only seen with a 100-fold excess of Cu(II) or Cd(II) or a 1000-fold excess of Pb(II) or Co(II). At these concentrations, the degree of competition appears to be directly proportional to the conditional stability constant of the competing metal to the gill (metal-gill log K). The range of [Cl-] allowed an assessment of Ag+, AgCl(aq), and AgCl2- bioavailability. The pattern of Ag(I) uptake was similar for each fish species. At AgCl(aq) >>> AgCl2-.
Introduction The gill is the first site for the interaction between waterborne pollutants and freshwater fish. Understanding how pollutants interact with this organ and how water chemistry influences this interaction will enable regulatory authorities to reliably forecast if polluted waterbodies are potentially deleterious to freshwater fish. A conceptual framework for a model that may predict metal toxicity was proposed by Pagenkopf (1) and assumes that metal toxicity is a function of gill-metal burden. The gill contains a number of different anionic moieties (e.g., sulfhydryl and carboxyl groups) to which metal ions may bind. Prevalent cations in the water (such as Ca2+, * Corresponding author telephone: +44 207 8484091; fax: +44 207 8484500; e-mail:
[email protected]. 2884
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 13, 2002
Mg2+, Na+, or K+) may compete with the metal for these binding sites, thus reducing gill-metal accumulation. Another possibility that has seldom been addressed is how metal mixtures influence, through competition, binding of individual metals. Anionic ligands, such as Cl-, HCO3-, or dissolved organic matter, may form complexes with the metal rendering that metal unavailable. These interactions can be described in terms of an equilibrium stability constant (log K) (2). This approach has enabled the gill to be considered as a “biotic ligand” that can be inserted into geochemical speciation modeling programs (3) to make a software-based tool for predicting gill-metal burdens (2, 4, 5) and metal toxicity (6) in chemically complex aquatic environments. In general, metals exert their acute toxic actions by interacting with biologically sensitive molecules within the cell. For example, Ag(I) and Cu(I) inhibit the basolateral membrane Na+/K+-ATPase activity of the freshwater fish gill (7, 8). Consequently, the metal must first enter the cell before a toxic effect is observed. Metals enter cells either via specific transporters (9, 10), via mimicry of other cation uptake processes (11-13), as anion metal complexes (14) via anion transporters, or via passive diffusion by neutral metal complexes (15, 16). Determining the mechanisms by which differing metal species enter gill cells and how they interact with biologically sensitive molecules may help explain differences in the relative acute toxicity of certain metal complexes. A conundrum regarding the effect of increasing [Cl-] on Ag(I) toxicity to rainbow trout exists. Although Cl- mitigates Ag(I) toxicity (17-19), it does not prevent branchial Ag(I) accumulation in freshwater rainbow trout (18, 20). Hence, there is disparity between Ag(I) bioavailability and bioreactivity in water of differing [Cl-]. A number of studies have suggested that both Ag+ and AgCl(aq) are readily accumulated by freshwater fish, although by different mechanisms (18, 20, 21). Ag+ uptake may occur via a Na+ uptake pathway (13), while the circumneutral AgCl(aq) may enter via passive diffusion or other nonspecific mechanisms. However, the relative bioavailability of these Ag(I) species to freshwater fish has not been ascertained. In the euryhaline microalga Thalassiosira weisflogii and the grass shrimp Palaemonetes pugio, the increase in Ag(I) uptake associated with a rise in aqueous [Cl-] can best be correlated to the [AgCl(aq)] (15, 16). Fortin and Campbell (22) also observed that Ag(I) uptake in the green alga Chlamydomonas reinhardtii was a function of aqueous [Cl-]; they found no evidence for passive diffusion of AgCl(aq) into the algal cells. They attributed the increase in Ag(I) uptake to enhanced diffusion across the unstirred boundary layer that surrounds the cell (22). The present investigation has two aims. First, to assess the pattern of Ag(I) uptake in two salmonid fish, Atlantic salmon (Salmo salar) and rainbow trout (Oncorhynchus mykiss), over a wide range of aqueous [Cl-] (0-10 mM Cl-). This range of [Cl-] provides water conditions with vastly varying [Ag+], [AgCl(aq)], and [AgCl2-], enabling a thorough investigation into the bioavailabilty of differing Ag(I) species to freshwater fish. Second, to determine the influence of other metals (Cu(II), Cd(II), Zn(II), Pb(II), Co(II)) on Ag+ uptake in Atlantic salmon. Ag+ has been shown to mimic Cu+ in CuATPases (23-25), and other metal transporters have also been shown to be promiscuous in allowing the passage of various metals (26, 27). Both of these aims serve to fill gaps of information in current biotic ligand models. 10.1021/es010302g CCC: $22.00
2002 American Chemical Society Published on Web 05/22/2002
Materials and Methods Fish Husbandry. Atlantic salmon (S. salar) were obtained as eyed eggs from Landcatch Ltd, Scotland, U.K., and were reared in synthetic ion-poor water generated by combining reverse osmosis water with the addition of a small quantity of Exeter dechlorinated tap water; the final cation concentrations being (in mM) 0.098 Na+, 0.014 K+, 0.180 Ca2+, and 0.049 Mg2+ with pH 7.19 and temperature 11 °C. Rainbow trout (O. mykiss) were obtained from Exmoor trout, Brompton Regis, Devon, and were reared in similar conditions. All experiments were performed at The University of Exeter on yolk-sac fry just prior to first feeding. Influence of Chloride on Silver and Sodium Uptake. All silver uptake studies were performed in ultrapure water (MilliQ, 18.2 MΩ) to which was added 150 µM NaHCO3, 500 µM Ca(NO3)2, and 100 µM MgSO4 at pH 7.3. For each experiment, eight fish were placed in 250 mL of water in blackened polyethylene containers. Containers used for Ag(I) uptake studies were pretreated for 24 h with a 50 nM AgNO3 solution to ensure partial saturations of Ag(I) binding sites on the container walls, and 110mAg levels, as determined by radioactivity in the water, dropped by 11 ( 1.8% (n ) 72) during the flux period. Water was continuously aerated throughout the experiment. Silver uptake was determined by the addition of 0.15 MBq 110mAgNO3 L-1 (specific activity 379 GBq/g Ag(I), Risoe isotoplaboratoriet, Roskilde, Denmark), with final Ag(I) concentrations being adjusted to the desired level by the addition of nonradioactive AgNO3. The effect of different water [Cl-] (50, 250, 500, 1000, 2000, 5000, and 10000 µM) was assessed by the addition of KCl. The K+ concentration was maintained constant at 10000 µM by the addition of KNO3. Ag(I) uptake was determined at two nominal Ag(I) concentrations, 10 or 87.5 nM AgNO3 in the case of the Atlantic salmon and at 50 nM AgNO3 in the case of the rainbow trout. Assessment of the concentrationdependent silver uptake at 0 and 5000 µM [Cl-] by Atlantic salmon was performed over a range of water [Ag(I)] (10, 50, 75, 100, 150, and 200 nM). A water sample was taken 15 min after the addition of the isotope as well as after a 3-h incubation period for radioactivity counting (Packard gamma counter). At this point, the fish underwent a wash protocol designed to remove loosely bound radiolabeled 110mAg(I): 1 min in 1 mg of AgNO3 L-1, followed by 1 further min in 500 mg of Na2S2O3 L-1, and 1 final min in Exeter dechlorinated tap water. Fish were weighed, and uptake was determined from the following equation:
whole body Ag(I) uptake ) cpm/(SA × wt × t) (1) where cpm is the counts per minute in the fish, SA is the specific activity of Ag(I) in the water (cpm/pmol), wt (g) is the weight of the fish, and t (h) is the duration of the flux. Rainbow trout yolk-sac fry Na+ influx rates were determined in a similar manner under the same water conditions. Instead of 110mAgNO3, 22Na+ (1 µCi L-1, Amersham Pharmacia, England, U.K.) was added to each container containing eight fish. A water sample was taken 15 min after the addition of the isotope as well as after a 3-h incubation period for radioactivity counting and analysis of Na+ content of the water samples determined by flame photometry. Fish were weighed, and Na+ uptake was determined from the following equation:
whole body Na+ uptake ) cpm/(SA × wt × t)
(2)
where cpm, wt, and t are the same as in eq 1; SA is the specific activity of Na+ in the water calculated from the following equation: +
+
SA ) [(cpmi/[Na ]I) + (cpmf/[Na ]f)]/2
(3)
FIGURE 1. (A) Ag(I) speciation at the differing water [Cl-] as determined by MINEQL+ (3) is given for Ag+ (s), AgCl(aq) (‚‚‚), and AgCl2- (- -). (B) Effect of altering water [Cl-] on whole body Ag(I) uptake in Atlantic salmon yolk-sac fry at 10 (b) and 87 nM (O) Ag(I), and rainbow trout yolk-sac fry at 50 nM Ag(I) (9). All values are mean ( SEM (n ) 8), and significant differences (one-way ANOVA followed by Tukey’s test, p < 0.005) in Ag(I) uptake rate at the different [Cl-] are denoted by different letters (a-c for the rainbow trout data; d-g and h-k for the Atlantic salmon data at 87.5 and 10 nM, respectively). (C) Whole body Na+ influx, in the absence of Ag(I) for Atlantic salmon at differing water [Cl-]. All values are mean ( SEM (n ) 10). There is no significant difference between Na+ influx at the different water [Cl-]. Influence of Metals on Silver Uptake. The effects of differing water metal concentrations (1-10000 nM), added as Pb(NO3)2, CuSO4, CdSO4, Zn(NO3)2, or CoSO4, on Ag(I) uptake in Atlantic salmon was assessed. Ag(I) was added to a concentration of 10 nM Ag(I) as 0.74 MBq/L110mAgNO3. Water chemistry and all other procedures followed exactly those described above. Statistical Analysis. Significant difference between uptake rates at each chloride concentration was determined by ANOVA followed by a Tukey’s post-hoc comparison (SPSS6 computer package), and calculations of silver speciation were performed using the computer-based geochemical speciation program MINEQL+ version 4.01 (3).
Results and Discussion The pattern of Ag(I) uptake at a fixed [Ag(I)], over a range of [Cl-], was similar in Atlantic salmon and rainbow trout yolksac fry (Figure 1). At [Cl-] below 1 mM, there was no VOL. 36, NO. 13, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
2885
FIGURE 2. Concentration-specific accumulation rate (CSAR), derived from the accumulation rate normalized for the Ag(I) species concentration, in rainbow trout at 50 nM total Ag(I). In each graph the CSAR for total Ag(I) is depicted by (b). The CSAR is given for (A) [Ag+] (3), (B) [AgCl(aq)] (3), (C) [AgCl2-] (3), (D) [Ag+ + AgCl(aq)] (3), (E) [Ag+ + AgCl2-] (3), and (F) [AgCl(aq) + AgCl2-] (3). Concentrations of all Ag(I) species were determined using MINEQL+ (3). statistically significant difference between Ag(I) uptake rates in either fish species, which supports previous findings that water [Cl-] up to 1 mM does not influence the accumulation of Ag(I) by the gills of rainbow trout (18, 20). However, neither of the latter studies assessed the effects of higher [Cl-] on gill Ag(I) burden. In the current study, where higher [Cl-] were examined, there was a rapid decline in Ag(I) uptake between 1 and 10 mM Cl- in both salmonid species. Increasing aqueous [Cl-] did not influence Na+ influx rates (Figure 1), excluding the possibility that the change in water chemistry indirectly affected Ag(I) uptake by altering branchial cation permeability. This is especially pertinent for Na+ influx, because Ag+ may enter rainbow trout via a Na+ uptake pathway (13). The decline in Ag(I) uptake occurs when the 2886
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 13, 2002
concentration of the circumneutral AgCl(aq) complex dominated the Ag(I) speciation and the concentration of negatively charged silver chloride species was increasing. Consequently, from our data in Figure 1, the order of bioavailability of Ag(I) species to salmonids appears to be Ag+ > AgCl(aq) >>> AgCl2-. Expressing whole body Ag(I) accumulation rate as a function of the concentration-specific accumulation rate (CSAR) shows that only the Ag+ + AgCl(aq) combination matches the pattern for total Ag(I) (Figure 2). The CSAR is indicative of bioavailability (28) and thus indicates that Ag+ and AgCl(aq) are the two species responsible for the Ag(I) uptake observed. It has generally been assumed that the AgCl(aq) complex may readily cross cell membranes (15, 16, 18,
comprises of a number of cell types that include those primarily involved in oxygen uptake (pavement cells), which make up approximately 90% of the cells, and also those involved in ion uptake (chloride or mitochondrial rich cells) (29). If, Ag+ entry is via a putative Na+ channel (13), then Ag(I) will preferentially accumulate in cells involved in the Na+ uptake pathway. This pathway involves Na+ extrusion from the cells via the Na+/K+-ATPase, an enzyme that is inhibited by Ag(I), and is one of the sites of Ag(I) toxic action in freshwater fish (7). In contrast, we propose that the AgCl(aq) complex will indiscriminately cross the membranes of all fish gill cells. This would lead to a lower effective dose of Ag(I) and thus a lower concentration of Ag(I) in the chloride cells. The reason for the dominant Ag+ bioavailability is probably a result of the passage of Ag+ across the gill being via active transport processes (13, 30). Ag+ uptake via the gills of rainbow trout is blocked by the drugs phenamil, a blocker of epithelial Na+ channels, and bafilomycin A, a proton pump inhibitor (13), as well as being reduced at low pH (N.R.B., personal observation). In addition, cation competition studies show that Na+, and not Ca2+ or K+, inhibits Ag(I) uptake (13). Taken together, these results indicate that the rapid branchial accumulation of Ag+ by freshwater fish gill can be explained by the fact that the apical entry step for Na+ appears unable to distinguish between Na+ and Ag+ (13). Once Ag(I) has entered the cell, branchial basolateral Ag(I) extrusion is via an active transport process that shows characteristics of a P-type ATPase (30). Whether this is a specific Ag(I) transporter or a transporter of an essential element that accidentally transports Ag(I) is unclear, but studies have shown that Ag+ can mimic Cu+ and be transported by Cu-ATPases (23-25). The present study suggests that Ag+ is more readily bioavailable than AgCl(aq), probably due to the presence of apical membrane transport proteins that are unable to discriminate between Ag+ and Na+ (13). The exact nature of these proteins is still not clear, and it is not known whether other metals compete with Ag+ for the same uptake site. None of the metals tested (Cu(II), Cd(II), Zn(II), Pb(II), Co(II)) were strong competitors of salmon yolk-sac fry whole body Ag(I) influx, and inhibition was only seen with a 100fold excess of Cu(II) or Cd(II) or a 1000-fold excess of Pb(II) or Co(II) (Table 1). Conditional equilibrium stability constants (log K) have been calculated for a number of metal-gill interactions (Table 1). By the use of a geochemical speciation modeling program (3) it is possible to predict gill-metal burden for single metal exposures (2, 4, 5, 31). If the metals are competing for the same binding site, then this model should be able to predict the metal gill burden in multiple metal conditions. Although the competition was weak, from the present study it may be possible to predict the degree of competition between these metals and Ag(I) accumulations based on the strength of their gill-metal bond [see Table 1 for gill-metal log K values, and the Ag-gill log K ) 10.0 (5)]. The only metal that did not fit this trend was Co(II), which despite having a low log K value was capable of reducing
FIGURE 3. Uptake rate of Ag(I) into Atlantic salmon yolk-sac fry at differing Ag(I) concentrations at 0 (b) and 5 mM Cl- (O). The pattern of uptake at 0 mM Cl- best fitted (r 2 ) 0.98) MichaelisMenten kinetics with the equation y ) 242 ( 56x/(216 ( 81 + x); at 5 mM Cl-, the uptake best fitted a linear regression (r 2 ) 0.97), equation y ) 3.1 + 0.17x. At 5 mM Cl-, the percent of Ag+ is 7, AgCl(aq) is 62, and AgCl2- is 31 as determined by MINEQL+ (3). 19, 21). However, recently, Fortin and Campbell (22) questioned this view. They found enhanced Ag(I) uptake in the presence of increasing aqueous [Cl-] in the green alga Chlamydomonas reinhardtii but found no evidence for the passive diffusion of AgCl(aq) across the cell membrane. The increase in Ag(I) uptake was attributed to diffusion across the unstirred surface boundary layer. At 5 mM Cl- (Figure 3), AgCl(aq) is the dominant Ag(I) species (AgCl(aq), 62%; AgCl2-, 31%; Ag+ 7%). The positive linear nature of Ag(I) accumulation over a range of [Ag(I)] at this [Cl-] suggests that the AgCl(aq) complex enters the fish gill cells by passive diffusion. In contrast, uptake of Ag+ (∼0 mM Cl-) followed Michaelis-Menten kinetics indicative of carrier-mediated transport (Figure 3). At each comparable [Ag(I)], Ag+ uptake exceeds AgCl(aq) accumulation (Figure 3). In salmonids, therefore, it would appear that, when Ag(I) is present as both Ag+ and AgCl(aq), the Ag(I) accumulation rate is dominated by Ag+ uptake (Figures 1 and 3). Interestingly, even though branchial and whole body Ag(I) accumulation are not reduced by water Cl- at 1 mM, Cl- does ameliorate Ag(I) toxicity (18, 20), and Ag(I) toxicity in freshwater rainbow trout correlates well to the [Ag+] (17-21). This phenomenon is very difficult to explain based on the assumption that AgCl(aq) enters the gills via the same cells as does Ag+. Because the AgCl(aq) bond is relatively weak [log K ) 3.3 (3)], once it has entered a cell it will probably behave in a manner similar to Ag+ and be bound to intracellular ligands with a strong affinity for Ag(I), such as glutathione and metallothionein (21). The reason for the protective effect of Cl- is thus proposed to be due to the functional morphology of the salmonid gill (21). The gill
TABLE 1. Effect of Varying Metal Concentrations (nM) on Atlantic Salmon Whole Body Yolk-Sac Fry Ag(I) Influx at 10 nM 110mAgNO for 3 h Expressed as a Percentage of Control Influx 3 Cd(II) Cu(II) Pb(II) Zn(II) Co(II)
1
10
100
1000
10 000
log Ka
91.5 ( 9.4 87.5 ( 12.3 87.4 ( 3.7 87.9 ( 7.4 105.0 ( 7
93.6 ( 12 76.6 ( 9.8 93.2 ( 4.7 94.9 ( 6.2 118.0 ( 8.9
83.7 ( 4.3 99.2 ( 17.4 112.8 ( 4.1 70.7 ( 4.8 90.6 ( 10.4
20.6( 2.8* 49.3 ( 4.0* 98.9 ( 5.1 101.1( 6.3 99.2 ( 9.9
5.1 ( 1.6* 24.3 ( 4.0* 70.1 ( 3.3* 89.0 ( 3.4 61.9 ( 2.1*
8.6 (4) 7.4 (4) 6.0b 5.6 (29) 5.1 (30)
a Gill-metal equilibrium stability constants (log K) are included for comparison. Gill-Ag log K ) 10.0 (5). Values represent mean ( SEM (n ) 8-10). Asterisks represent significant difference between Ag(I) uptake rates of fish exposed to the metal and the control values. All statistical analysis were performed on un-transformed data (ANOVA followed by a Tukey’s test, p < 0.05). b R. Playle, personal communication.
VOL. 36, NO. 13, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
2887
Ag(I) accumulation. However, this was only observed at 1000fold excess of Co(II). It may be that competition reflects initial binding to the gill surface rather than a shared uptake protein. However, recently, Na+ has been shown to be a competitive inhibitor of Cu(II) uptake in fish (32), but whether Ag(I) and Cu(II) share entry via the same Na+-sensitive pathway awaits verification. The results demonstrate that the order of bioavailability of Ag(I) species to salmonids is Ag+ > AgCl(aq) >>> AgCl2-. This differs from the situation in a euryhaline alga and a crustacean, where AgCl(aq) is apparently taken up more readily (15, 16). The reason for this difference is probably because Ag+ is actively taken up from the water via a Na+ uptake pathway (13) and then passes from the gill cell into the plasma via an ATP-dependent enzyme (26). It is unclear whether this Ag(I) uptake pathway is shared with other metals, but the rapid uptake of Ag+ may in part explain the toxic nature of Ag+ as compared to AgCl(aq). The data from the present study also led to a different conclusion from previous reports on rainbow trout (21, 28), suggesting that AgCl(aq) is at least as bioavailable as Ag+ species. In the present study we used higher [Cl-] of the water, allowing a more precise evaluation of the effect of speciation (i.e., Ag+ versus AgCl(aq) versus AgCl2-) on Ag(I) uptake than previously published.
Acknowledgments N.R.B. was supported by a University of Exeter Fellowship and a grant from the Fisheries Society of the British Isles. We also thank the referees for their constructive comments and Drs. Sarah Bury and Chris Glover for re-reading the manuscript.
Literature Cited (1) Pagenkopf, G. K. Environ. Sci. Technol. 1983, 17, 342-347. (2) Playle, R. C. Sci. Total Environ. 1998, 219, 147-163. (3) Schecher, W. D.; McAvoy, D. C. Environ. Urban Syst. 1992, 16, 65-76. (4) Playle, R. C.; Dixon, D. G.; Burniston, K. Can. J. Fish. Aquat. Sci. 1993, 50, 2678-2687. (5) Janes, N.; Playle, R. C. Environ. Toxicol. Chem. 1995, 14, 18471858. (6) DiToro, D. M.; Allen, H. E.; Bergmann, H. L.; Meyer, J. S.; Paquin, P. R.; Santore, R. C. Environ. Toxicol. Chem. 2001, 20, 23832396. (7) Morgan, I. J.; Henry, R. P.; Wood, C. M. Aquat. Toxicol. 1996, 38, 145-163. (8) Laure´n, D. J.; McDonald, D. G. Can. J. Fish. Aquat. Sci. 1987, 44, 105-111.
2888
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 13, 2002
(9) Pen ˜ a, M. M. O.; Lee, J.; Thiele, D. J. J. Nutr. 1999, 129, 12511260. (10) Andrews, N. C. Nat. Rev. Genet. 2000, 1, 208-217. (11) Verbost, P. M.; Van Rooij, J.; Flik, G.; Lock, R. A. C.; Wendelaar Bonga, S. E. J. Exp. Biol. 1989, 145, 185-197. (12) Hogstrand, C.; Verbost, P. M.; Wendelaar Bonga, S. E.; Wood, C. M. Am. J. Physiol. 1996, 270, R1141-R1147. (13) Bury, N. R.; Wood, C. M. Am. J. Physiol. 1999, 277, R1385R1391. (14) Fortin, C.; Campbell, P. G. C. Environ. Sci. Technol. 2001, 35, 2214-2218. (15) Reinfelder, J. R.; Chang, S. I. Environ. Sci. Technol. 1999, 33, 1860-1863. (16) Engel, D. W.; Sunda, W. G.; Fowler, B. A. In Biological Monitoring of Marine Pollutants; Vernberg, F. J., Caklabrese, A., Thurberg, F. P., Vernberg, W. D., Eds.; Academic Press: New York, 1981; p 127. (17) Galvez, F.; Wood, C. M. Environ. Toxicol. Chem. 1997, 16, 23632368. (18) McGeer, J. C.; Wood, C. M. Can. J. Fish. Aquat. Sci. 1998, 55, 2447-2454. (19) Bury, N. R.; Galvez, F.; Wood, C. M. Toxicol. Chem. 1999, 18, 56-62. (20) Bury, N. R.; McGeer, J. C.; Wood, C. M. Environ. Toxicol. Chem. 1999, 18, 49-55. (21) Wood, C. M.; Playle, R. C.; Hogstrand, C. Environ. Toxicol. Chem. 1999, 18, 71-8. (22) Fortin, C.; Campbell, P. G. C. Environ. Toxicol. Chem. 2000, 19, 2769-2778. (23) Solioz, M.; Odermatt, A. J. Biol. Chem. 1995, 270, 9317-9221. (24) Riggle, P. J.; Kumamoto, C. A. J. Bacteriol. 2000, 182, 48994905. (25) Weissman, Z.; Berdicevsky, I.; Cavari, B. Z.; Kormitzer, D. Proc. Natl. Acad. Sci. U.S.A. 2000, 97, 3520-3525. (26) Gunshin, H.; Mackenzie, B.; Berger, U. V.; Gunshin, Y.; Romero, M. F.; Boron, W. F.; Nussberger, S.; Gollan, J. L.; Hediger, M. A. Nature 1997, 388, 482-488. (27) Havelaar, A. C.; de Gast, I. L.; Snijders, S.; Beerens, C. E. M. T.; Mancini, G. M. S.; Veheijen, F. W. FEBS Lett. 1999, 436, 223227. (28) Hogstrand, C.; Wood, C. M. Environ. Toxicol. Chem. 1998, 17, 547-561. (29) Perry, S. F. Annu. Rev. Physiol. 1997, 59, 325-347. (30) Bury, N. R.; Grosell, M.; Grover, A. K.; Wood, C. M. Toxicol. Appl. Pharmacol. 1999, 159, 1-8. (31) Richards, J. G.; Playle, R. C. Comp. Biochem. Physiol. C 1998, 119, 185-197. (32) Grosell, M.; Wood, C. M. J. Exp. Biol. 2002, 205, 1179-1188.
Received for review November 21, 2001. Revised manuscript received April 18, 2002. Accepted April 19, 2002. ES010302G
