Environ. Sci. Technol. 2003, 37, 3298-3302
Uptake of Waterborne Tributyltin in the Brain of Fish: Axonal Transport as a Proposed Mechanism C L A U D E R O U L E A U , * ,† ZHENG-HU XIONG,‡ GRAZINA PACEPAVICIUS,† AND GUO-LAN HUANG‡ National Water Research Institute, Aquatic Ecosystem Protection Research Branch, 867 Lakeshore Road, P.O. Box 5050, Burlington, Ontario, Canada L7R 4A6, and Nankai University, College of Environmental Sciences and Engineering, Tianjin 300071, People’s Republic of China
In previous studies, it was shown that waterborne Hg(II), Cd(II), and Mn(II) enter nerves innervating water-exposed sensory organs of fish and are transported to the brain by axonal transport. However, it is not known if organometals, such as tributyltin (TBT), can reach the brain of fish via the same route. In this work, we exposed rainbow trout (Oncorhynchus mykiss) to waterborne [113Sn]-TBT (4.2 kBq/ L). Three fish were sampled after a 2-week exposure, and three others were sampled after a 2-week depuration period. Another group of four fish received an intravenous injection of [113Sn]-TBT and were sampled after 2 and 14 d. Distribution of the radiolabel was visualized and quantified by quantitative whole-body autoradiography. The brain accumulated a significant amount of 113Sn, with hot spots being found in parts receiving sensory nerves from waterexposed sensory organs, such as eminentia granulares (lateral lines organs). Labeling of the brain was also seen for i.v.injected fish, indicating that the blood-brain barrier is not impervious to TBT or its metabolites. Nevertheless, the distribution of radioactivity in the brain was much more uniform, with no evident hot spot. Though the transfer [water f gills f blood stream f blood-brain barrier f brain] may account for a certain proportion of the radiolabel accumulation in fish brain, exposure to [113Sn]-TBT via water resulted in higher accumulation in some areas of the brain, of which the specific location strongly suggests that it was taken up in different water-exposed sensory nerve terminals and transported directly to the brain by axonal transport, as the parent compound or as a metabolite. The resulting local enhancement of the accumulation of butyltins might jeopardize the integrity of nervous system. Further work is needed to assess the toxicological significance of this process.
Introduction Axonal transport in neurons is a normal physiological process for the transport of organelles and dissolved neuro* Corresponding author present address: Institut Maurice-Lamontagne, 850 Route de la Mer, C.P. 1000, Mont-Joli, Que´bec, Canada G5H 3Z4; telephone: (418)775-0734; fax: (418)775-0718; e-mail:
[email protected]. † National Water Research Institute. ‡ Nankai University. 3298
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 15, 2003
nal constituents along the axons to the nerve terminals (anterograde axonal transport) and back to the cell bodies (retrograde axonal transport). It also constitutes a route by which foreign materials, among them toxic trace metals (Pb, Cd, Hg, Tl, Ag), can circumvent the blood-brain barrier to reach the central nervous system (1). In previous studies on the axonal transport of metals in fish, we showed that 109 Cd(II) and 54Mn(II) applied in the nostrils of pike (Esox lucius) were taken up in receptor cells of the olfactory epithelium and transported along olfactory nerve neurons toward the brain (2, 3). We also observed that waterborne 203Hg(II), 109Cd(II), and 54Mn(II) were taken up in the olfactory epithelium and transported to the brain of brown trout (Salmo trutta) by axonal transport (4-6), even at concentrations at the lower end of the range found in freshwater environments (6). It is also noteworthy that 203Hg(II) can also enters the central nervous system via other water-exposed sensory nerve terminals, such as those innervating lateral line organs and gustatory papilla (4). Accumulation of metals from water in sensory systems may disturb fish behavior. For example, accumulation of metals in the olfactory system may be injurious for the olfactory sense of fish and may disturb processes relying on olfaction, such as predator avoidance, social interactions, and migration (7). For instance, it has been shown recently that exposure to waterborne Cd inhibits the normal predator avoidance behavior of rainbow trout in response to the presence of an alarm substance (8) and has an impact on the establishment of dominance hierarchies (9), which appeared to be due to the inhibition of olfaction resulting from the accumulation of the metal in the olfactory system. Organometals are extremely toxic chemicals characterized by the presence of one or more covalent metal-carbon bonds. It is not known if waterborne organometals, such as the antifouling biocide tributyltin (TBT), can reach the brain of fish via water-exposed sensory nerves. To verify this, we determine the fine-scale body distribution of 113Sn-radiolabeled TBT in rainbow trout upon exposure to waterborne TBT and after an intravenous injection, using whole-body autoradiography.
Materials and Methods Rainbow trout (Oncorhynchus mykiss) from a local hatchery were acclimatized for 1 week to experimental conditions prior to contamination. Six fish, weighing ca. 10 g, were used for the water exposure experiment; three of them being sampled after 14 d of exposure, whereas the remaining three were allowed to depurate for 14 d before sampling. Four fish, weighing ca. 40 g, received an intravenous (i.v.) injection, and two fish were sampled 2 and 14 d after injection. Fish were fed three times a week with commercial chow about 6 h before water renewal (see below). Food was consumed within 1 min, and the remaining food pellets were immediately removed from water. We used a static exposure system, consisting of 15-L aquaria filled with dechlorinated tap water and maintained at 10 °C. Water (spiked or not) was replaced every Monday, Wednesday, and Friday in the afternoon. 113Sn-radiolabeled TBT chloride in pentane was synthesized from inorganic 113Sn(IV) [t1/2 ) 115 d, decay mode 100% electron capture, γ-ray emission at 392 keV (64%), X-ray emissions at 24-27 keV (≈120%), purchased from New England Nuclear] according to Rouleau (10). Chemical purity of [113Sn]-TBT was >98%, as assayed by thin-layer chromatography. For water exposure experiment, water was spiked with the [113Sn]-TBT pentane solution to a nominal con10.1021/es020984n CCC: $25.00
2003 American Chemical Society Published on Web 06/28/2003
centration of 4.2 kBq/L (0.07 µg of TBT/L), 30 min before the introduction of fish. Radioactivity levels decreased by 75 ( 8% between water changes, probably due to uptake by fish and adsorption on the glass walls of the aquaria. This resulted in an average concentration of 2.9 kBq/L for the 14 d of exposure. For the injection experiment, the pentane from the stock solution of [113Sn]-TBT was evaporated, and TBT was redissolved in 0.9% saline to a concentration of 37 kBq/ 100 µL (2.1 µg of TBT/100 µL). Each fish was injected in the caudal vein with 100 µL of this solution and then placed in clean water. Radioactivity in water was monitored every day during the depuration period of water-exposed fish and in the water of i.v.-injected fish. None of the water samples had radioactivity above the detection limit (0.25 Bq/3 mL of sample, 10 min counting time on a Wallac Wizard model 1480 γ-counter). Sampled fish were anaesthetized (MS-222, 100 mg/L), embedded in a 5% carboxymethylcellulose gel, and quickly frozen in a slurry of dry ice in hexane. From each fish, 15-20 pairs of 20-µm-thick sections were sampled at different locations with a specially designed cryomicrotome (Leica CM3600). Sections were then freeze-dried and placed on phosphor screens for 4-7 d. After exposure, the screens were scanned with a Cyclone Phosphor Imager (Packard Bioscience). 113Sn activity in the tissues and organs was quantified with the software Optiquant (Canberra-Packard). Data from 6-10 sections per animal were collected, corrected for exposure time and decay, and expressed as digital light unit per mm2 of section surface (DLU/mm2). Variability of the response of the phosphor screens was evaluated by exposing four different 14C-labeled autoradiography standard strips (0.002-35 µCi/g in 15 steps, American Radiolabeled Chemicals) for periods of time ranging from 1 to 203 h. The average coefficient of variation of the response of phosphor screens, corrected for exposure time, was 9%. The distribution of 113Sn in fish was expressed as a concentration index relative to whole body (IC) with (11)
IC ) (DLU/mm2)tissue/(DLU/mm2)whole body
(1)
False-color images presented in Figure 3 were produced with the software Dplot.
Results Autoradiograms showed a similar general distribution picture of the radiolabel among the various groups (Figure 1). Radioactivity was found in all the tissues, with the liver generally showing the highest labeling. There was a labeling of the brain in all groups. However, some quantitative differences were observed (Figure 2). Fish that were exposed to [113Sn]-TBT in water lost 30% of the radioactivity accumulated during the depuration period. About 70% of the radiolabel injected intravenously was lost by fish. In the waterexposed group, IC values tended to be lower for the liver and higher for brain when compared to fish that received an intravenous injection. A time-dependent increase of the value of IC for liver occurred in both treatments as elimination proceeded, which was caused by the fact that the radiolabel concentration in that organ did not decrease or decreased at a slower rate as compared to the whole-body. It is noteworthy that the radiolabel concentration in the brain of fish exposed via water showed no variation, despite the 30% decrease of whole body radioactivity, whereas brain radioactivity decreased at a rate similar to that of the whole body in case of the i.v.-injected fish. As a result, the IC value of the brain tended to increase for water-exposed fish, while it remained approximately the same in the other group. A close examination of the brain revealed many interesting differences between water-exposed and i.v.-injected fish
FIGURE 1. Whole-body autoradiograms of trout exposed to [113Sn]-TBT. (A) 2-week water exposure, (B) 2-week water exposure followed by a 2-week depuration, and (C) 2 and (D) 14 d after an intravenous injection. Highest radiolabel concentrations corresponds to white areas. Images shown cover an area of 30 mm × 80 mm in panels A and B and of 50 mm × 132 mm in panels C and D. (Figure 3). In the latter, a uniform labeling of the brain was seen after 2 d (Figure 3A). After 14 d, distribution was still uniform, although the relative concentration of the radiolabel tended to be lower, as shown by IC values (Figure 3B). After 14 d of exposure via water (Figure 3C), radioactivity was also found in all the parts of the brain, but some areas, like the eminentia granulares, exhibited a highest labeling with IC values up to 4. Fish sampled after the depuration period (Figure 3D) exhibited a much more heterogeneous distribution of the radiolabel and higher values of IC, typically varying between 3 and 6. Hot spots are visible at different locations. This is best illustrated in Figure 3E, which is a montage of three sections from the same fish taken at 150-µm intervals. A high concentration of the radiolabel can be seen within VOL. 37, NO. 15, 2003 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
3299
FIGURE 2. Concentration of the radiolabel, expressed as digital light unit per mm2 (DLU/mm2), and concentration index (IC, eq 1) measured in whole-body sections of fish exposed to [113Sn]-TBT via water or after an intravenous injection. Key for the histogram on the right is the same as for the two others. the diencephalon and the eminentia granulares, with values of IC reaching 5 and 6, respectively.
Discussion Results obtained in this study show clearly that the fate of [113Sn]-TBT differs depending upon the administration route. An increase of the liver IC value as elimination proceeds is expected since fish metabolize TBT in the liver (12). However, the relative concentration of the radiolabel in the liver was higher in i.v.-injected fish. It has been shown recently that TBT in fish blood is bound to a glycoprotein with a molecular mass of 46.5 kDa (13). Such binding may affect the rate at which TBT is removed from the blood stream by the liver. When TBT is injected intravenously, various processes occurring when the chemical is taken up via natural routes (through the gills or the digestive system) may be circumvented. This may change the speciation of TBT in blood, resulting in an easier removal by the liver. Nevertheless, the labeling of the brain observed in autoradiograms from i.v.injected fish undoubtedly showed that [113Sn]-TBT or some metabolite was able to cross the blood-brain barrier under one chemical form or another. The other important point is that the distribution of the injected radiolabel in the brain remained uniform and that elimination proceeded at a rate similar to that of the whole body (Figure 2). It thus appears that radiolabel entering the brain of these fish was quite labile. The situation was different for water-exposed fish. The relative concentration of the radiolabel tended to be higher in some areas of the brain (Figure 3C), and the heterogeneity of the radiolabel distribution increased as depuration proceeded, as shown by the higher values of IC in the same areas (Figure 3, panels D and E). This increase of the relative concentration is due to the fact that radiolabel concentration remained constant in the brain while whole-body concentration decreased by 30% (Figure 2). This indicates that radiolabel accumulated in the brain of fish upon exposure to waterborne [113Sn]-TBT was somehow trapped or less mobile than in the brain of fish in i.v.-injected group, resulting in concentrations that are as high as in the liver. It is possible that some of the radiolabel entering fish body via the gills and transported within the blood stream have been able to cross the blood-brain barrier, although the extent of this process cannot be quantified with the present set of data. However, the hot spots observed are located in specific areas of the brain that receive nerve fibers innervating water-exposed sensory organs. Hence, eminentia granulares are the termination site of nerve fibers from the mechanoreceptors in lateral lines. Nerves innervating the gustatory papilla terminate in the rhombencephalon, in which lateral line nerve fibers also travel on their way to eminentia granulares (14). This strongly suggests that 3300
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 15, 2003
[113Sn]-TBT was taken up in those water-exposed sensory organs and transported toward the brain by axonal transport. The labeling of the diencephalon and optic tectum may also be the result of axonal transport in optic nerve fibers. The cell bodies of optic neurons are located in the innermost cell layer of the retina, the ganglion cell layer (15), and terminate in the diencephalon and the optic tectum. Though optic neurons are not directly exposed to water, it has been shown that many substances, such as amino acids, nucleotides, proteins, and metals, that are injected intraocularly can be taken up in optic neurons cell bodies and transported to the diencephalon and the optic tectum by axonal transport (16-20). A labeling of the retina of i.v.-injected fish could be seen after 2 d (IC ) 2-3), suggesting that the radiolabel diffused from blood. However, retina labeling became very faint 14 d after the injection (not shown). The labeling of the optic path (retina - optic nerve - diencephalon - optic tectum) of water-exposed fish remained high as depuration proceeded (Figure 3D,E), suggesting an accumulation process similar to the one leading to the accumulation of the radiolabel in eminentia granulares and rhombencephalon. A possibility is that waterborne [113Sn]-TBT may have diffused into the intraocular environment through the water-exposed cornea, followed by uptake in the optic neurons and axonal transport. A distribution pattern similar to that seen for TBT has also been observed in S. trutta after exposure to waterborne inorganic Hg(II), and it has been shown to result from the uptake of Hg in various water-exposed sensory organs and its subsequent transport in sensory neurons by axonal transport (4). Waterborne cadmium and manganese only accumulated in the olfactory system (5, 6). Mercury and cadmium have been shown to be unable to cross the synaptic junctions between primary and secondary sensory nerve fibers (2, 4, 21), whereas manganese can (3, 6). It is difficult here to determine if butyltins can cross or not synaptic junctions. Hence, mercury and cadmium taken up in the olfactory epithelium and transported along the olfactory nerve fibers accumulated in the proximal most rostral part of the olfactory bulb, yielding a typical crescent-like image in autoradiograms (2, 5, 6, 21). Though [113Sn]-TBT accumulated in the brain after water exposure appeared not to be very mobile and though it was also present in the olfactory system, the typical crescent shape was not seen in olfactory bulb region (Figure 3C). Further work will be needed to better understand the fate of TBT accumulated via axonal transport in terms of mobility and intracellular speciation. Though the transfer [water f gills f blood stream f blood-brain barrier f brain] likely accounted for a nonnegligible proportion of the radiolabel accumulation in fish brain, exposure to [113Sn]-TBT via water resulted in a specific distribution pattern of the radiolabel, strongly suggesting
FIGURE 3. False-color contour plots showing the fine-scale distribution of IC values in the brain of fish exposed to [113Sn]-TBT: (A) 2 and (B) 14 d after an intravenous injection, (C) 2-week water exposure, (D) and (E) 2-week water exposure followed by a 2-week depuration. Images shown cover an area of 16.9 mm × 25.4 mm in panels A and B and of 8.5 mm × 12.6 mm in panels C-E. Insert left of panel E is a schematic dorsal view of the brain with the location of the three sections shown. that waterborne TBT can be accumulated in water-exposed sensory organs and transported toward the brain by axonal transport, as the parent compound or as a metabolite. Furthermore, this process leads to the accumulation of high concentration of butyltins in brain tissues. Fish depend on
an intact nervous system for mediating relevant behavior such as searching for food, recognizing predators, communication, and orientation (22), and a significant enhancement of the accumulation of butyltins in the brain might jeopardized the integrity of the nervous system. Further work VOL. 37, NO. 15, 2003 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
3301
is needed to assess the toxicological significance of this process.
Literature Cited (1) Arvidson, B. Toxicology 1994, 88, 1. (2) Gottofrey, J.; Tja¨lve, H. Pharmacol. Toxicol. 1991, 69, 242. (3) Tja¨lve, H.; Meja`re, C.; Borg-Neczak, K. Pharmacol. Toxicol. 1995, 77, 23. (4) Rouleau C.; Borg-Neczak K.; Gottofrey J.; Tja¨lve H. Environ. Sci. Technol. 1999, 33, 3384. (5) Tja¨lve, H.; Gottofrey, J.; Bjo¨rklund, I. Toxicol. Environ. Chem. 1986, 12, 31. (6) Rouleau, C.; Tja¨lve, H.; Gottofrey, J.; Pelletier, EÄ . Environ. Toxicol. Chem. 1995, 14, 483. (7) Harden, Jones, F. R. Fish Migration; Edwards Arnold Publishers Ltd.: London, 1968. (8) Scott, G. R.; Sloman, K. A.; Rouleau, C.; Wood, C. M. J. Exp. Biol. 2003, 206, 1779. (9) Sloman, K. A.; Scott, G. R.; Diao, Z.; Rouleau, C.; Wood, C. M.; McDonald, D. G. Aquat. Toxicol. (in press). (10) Rouleau, C. Appl. Organomet. Chem. 1998, 12, 435. (11) Rouleau, C.; Gobeil, C.; Tja¨lve, H. Environ. Toxicol. Chem. 2000, 19, 631.
3302
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 15, 2003
(12) Fent, K. Crit. Rev. Toxicol. 1996, 26, 1. (13) Shimasaki, Y.; Oshima, Y.; Yokota, Y.; Kitano, T.; Nakao, M.; Kawabata, S. I.; Imada, N.; Honjo, T. Environ. Toxicol. Chem. 2002, 21, 1229. (14) Nieuwenhuys, R. Am. Zool. 1982, 22, 87. (15) Hoar, W. S.; Randall, D. J. Fish Physiology; Academic Press: New York, 1971; Vol. V. (16) Knull, H. R.; Wells, W. W. Brain Res. 1975, 100, 121. (17) Finger, T. E. J. Comp. Neurol. 1978, 181, 173. (18) McQuarrie, I. G.; Grafstein, B. Brain Res. 1982, 235, 213. (19) Edwards, D. L.; Grafstein, B. Brain Res. 1986, 364, 258. (20) Perry, G. W.; Burmeister, D. W.; Grafstein, B. J. Neurosci. 1990, 10, 3439. (21) Borg-Neczak, K.; Tja¨lve, H. Toxicol. Lett. 1996, 84, 107. (22) Baatrup, E. Comp. Biochem. Physiol. 1991, 100C, 253.
Received for review October 15, 2002. Revised manuscript received April 30, 2003. Accepted May 5, 2003. ES020984N
