- Email: [email protected]
Evaluation of developmental neurotoxicity of organotins via drinking water in rats: Monomethyl tin
NeuroToxicology 27 (2006) 409–420
Evaluation of developmental neurotoxicity of organotins via drinking water in rats: Monomethyl tin§ Virginia C. Moser a,*, Stanley Barone Jr.a,1, Pamela M. Phillips a, Katherine L. McDaniel a, Kimberly D. Ehman b,2 a
Neurotoxicology Division, National Health and Environmental Effects Research Laboratory, Office of Research and Development, US Environmental Protection Agency, Research Triangle Park, NC 27711, United States b University of North Carolina Chapel Hill, Curriculum in Toxicology, US Environmental Protection Agency, Research Triangle Park, NC, United States Received 30 September 2005; accepted 6 December 2005 Available online 25 January 2006
Abstract Organotins such as monomethyltin (MMT) are widely used as heat stabilizers in PVC and CPVC piping, which results in their presence in drinking water supplies. Concern for neurotoxicity produced by organotin exposure during development has been raised by published findings of a deficit on a runway learning task in rat pups perinatally exposed to MMT (Noland EA, Taylor DH, Bull RJ. Monomethyl and trimethyltin compounds induce learning deficiencies in young rats. Neurobehav Toxicol Teratol 1982;4:539–44). The objective of these studies was to replicate the earlier publication and further define the dose-response characteristics of MMT following perinatal exposure. In Experiment 1, female Sprague–Dawley rats were exposed via drinking water to MMT (0, 10, 50, 245 ppm) before mating and throughout gestation and lactation (until weaning at postnatal day [PND] 21). Behavioral assessments of the offspring included: a runway test (PND 11) in which the rat pups learned to negotiate a runway for dry suckling reward; motor activity habituation (PNDs 13, 17, and 21); learning in the Morris water maze (as adults). Other endpoints in the offspring included measures of apoptosis (DNA fragmentation) at PND 22 and as adults, as well as brain weights and neuropathological evaluation at PND 2, 12, 22, and as adults. There were no effects on any measure of growth, development, cognitive function, or apoptosis following MMT exposure. There was a trend towards decreased brain weight in the high dose group. In addition, there was vacuolation of the neuropil in a focal area of the cerebral cortex of the adult offspring in all MMT dose groups (1–3 rats per treatment group). In Experiment 2, pregnant rats were exposed from gestational day 6 until weaning to 500 ppm MMT in drinking water. The offspring behavioral assessments again included the runway task (PND 11), motor activity habituation (PND 17), and Morris water maze (as adults). In this second study, MMT-exposed females consumed significantly less water than the controls throughout both gestation and lactation, although neither dam nor pup weights were affected. As in Experiment 1, MMT-exposure did not alter pup runway performance, motor activity, or cognitive function. These results indicate that perinatal exposure to MMT, even at concentrations which decrease fluid intake, does not result in significant neurobehavioral or cognitive deficits. While mild neuropathological lesions were observed in the adult offspring, the biological significance of this restricted finding is unclear. Published by Elsevier Inc. Keywords: Developmental neurotoxicity; Organotin; Monomethyl tin; Behavior; Neuropathology
§ The information in this document has been reviewed by the National Health and Environmental Effects Research Laboratory and approved for publication. Approval does not signify that the contents necessarily reflect the views of the Agency nor does mention of trade names or commercial products constitute endorsement or recommendation for use. * Corresponding author at: Neurotoxicology Division, National Health and Environmental Effects Research Laboratory (MD B105-04), US Environmental Protection Agency, Research Triangle Park, NC 27711, United States. Tel.: +1 919 541 5075; fax: +1 919 541 4849. E-mail address: [email protected] (V.C. Moser). 1 Current address: National Center for Environmental Assessment, US Environmental Protection Agency, Washington, DC, United States. 2 Current address: RTI International, Research Triangle Park, NC, United States.
0161-813X/$ – see front matter. Published by Elsevier Inc. doi:10.1016/j.neuro.2005.12.003
410
V.C. Moser et al. / NeuroToxicology 27 (2006) 409–420
1. Introduction
2. Methods
Mono- and di-substituted organotins are commercially used as heat and light stabilizers for polyvinyl chloride (PVC) and chlorinated PVC (CPVC) products. PVC pipes are widely used in domestic water systems, and this has contributed to the presence of organotins in drinking water supplies and in the environment (Donard et al., 1993; Forsyth et al., 1993; Mersiowsky et al., 2001; Sadiki and Williams, 1999; Sadiki et al., 1996; Wilkinson, 1984). Both mono- and dimethyl tin have been detected in PVC-supplied drinking water at levels up to 49.1 ng Sn/l (dimethyl tin, DMT) or as high as 291 ng Sn/l (monomethyl tin, MMT; Sadiki and Williams, 1999). Because human exposure to organotins is widespread and the health consequences unknown, organotins were listed on the US EPA Candidate Contaminant List (CCL). This List includes nonregulated substances known or anticipated to occur in public water systems, with identified needs (data gaps) in terms of health effects (US EPA, 1998a). Considerable uncertainty exists in the evaluation of adverse health outcomes following developmental exposure to organotin compounds in drinking water. The neurotoxicity of trimethyl tin (TMT) has been extensively studied, but DMT and MMT are considered to be relatively non-toxic (Bouldin et al., 1981; Krigman and Silverman, 1984). Concern for neurotoxicity following developmental exposure to MMT was raised by published findings of a deficit on a cognitive task in rat pups (Noland et al., 1982). Noland and coworkers reported that exposure of rats to MMT (12– 120 mg Sn/l in drinking water) or TMT (0.15–1 mg Sn/l in drinking water) during pregnancy and postnatal development altered acquisition of a runway learning task at higher doses, and all dose levels resulted in a deficit of extinction. As such, Noland et al. (1982) could not report a no-observable effect level (NOEL) for MMT. The similar effects of both MMT and TMT on extinction were interpreted as a learning impairment which is consistent with limbic system damage, a known consequence of exposure to TMT; however, confirmatory pathological evaluations were not conducted (Noland et al., 1982). More recent studies have reported neuropathological lesions (loss of neuronal perikarya in hippocampus) following subchronic exposure of adult rats to mixtures of MMT and DMT (US EPA, 2004). Taken together, the available evidence suggests that both adult and developmental exposure to MMT may produce neurotoxicity. The current study was initiated to replicate the original findings of Noland et al. (1982) and to expand the neurological and neuropathological evaluation following developmental exposure to include other standard behavioral tests in the offspring during lactation and as adults. The first study was designed to mimic the Noland study as closely as possible, while the second, follow-up study was designed to be more consistent with the dosing paradigm of standard developmental neurotoxicity studies, such as those described in the US EPA test guidelines (US EPA, 1998b).
2.1. Chemicals 2.1.1. Experiment 1 Monomethyl tin trichloride (MMT; purity 97%; Aldrich, Milwaukee, WI, USA) was dissolved in distilled/deionized water at concentrations of 0, 10, 50 or 245 ppm MMT (0, 5, 25, or 120 mg Sn/l). These doses were chosen to match and extend those used in the Noland et al. (1982) study. The MMT was administered through polypropylene/polyethylene water bottles containing double ball-bearing sipper tubes. A concentrated stock solution was stored at 20 8C for no more than 10 days. Dilutions were made fresh twice a week and bottles refilled. Water bottles were changed and weighed twice weekly, and rats were weighed at the same time. 2.1.2. Experiment 2 Rats were exposed to either 0 or 500 ppm MMT (245 mg Sn/ l) in the drinking water as in Experiment 1. 2.2. Chemical analyses/speciation Analyses were conducted under contract (contract #4D6174-NTNX to Research Triangle Institute) to verify the concentration and speciation of the methyltins in the MMT solutions over days in the water bottles. It was of interest to determine if, and to what extent, interconversions in methyltin speciation occurred under conditions of our study. The stock solution and solutions of the low and high test concentrations were sampled daily for 5 days in water bottles under conditions of the animal exposure. Analysis of MMT, DMT, and trimethyl tin (TMT) were determined using ion chromatography (IC) with inductively coupled plasma mass spectrometer (ICP-MS). In addition, total tin levels were measured using inductively coupled plasma optical emission spectrometer (ICP-OES). Quality control checks were performed after each five to six runs; recoveries of 80–120% were considered acceptable. For MMT, the limit of detection (LOD) was 10 ng Sn/ml, and for DMT and TMT, LOD was 1 ng Sn/ml. The results of these analyses indicated that, as prepared, the high concentration was about 5% higher than the nominal concentration of 120 mg Sn/l, and the low concentration was indeed 5 mg Sn/l. The low concentration showed 4% decrease from days 4 to 5, but the high concentration did not change appreciably. Since bottles were changed every 3–4 days, we concluded there was no significant loss of Sn under our conditions of use. Furthermore, there was no detectable DMT or TMT in the MMT water samples at any time, indicating that if any changes in speciation occurred, it would represent only a negligible fraction of the nominal concentration. 2.3. Animals 2.3.1. Experiment 1 Nulliparous Sprague–Dawley (CD) female rats (Charles River, Raleigh, NC, USA), 53–54-days old, were housed in
V.C. Moser et al. / NeuroToxicology 27 (2006) 409–420
AAALAC-International accredited, temperature- and humidity-controlled rooms (19–21 8C and 40–60%, respectively) on Beta-Chip bedding. As in the Noland study (1982), rats were maintained on a reverse 12 h light:dark cycle (lights on at 1200 h). Upon arrival, 120 females were housed in pairs and provided food (Purina Rodent Chow 5001) and deionized water ad libitum. Following acclimation, and in the two weeks prior to MMT exposure, estrus cycles were monitored through daily vaginal smears. Females (30/dose) were then rearranged and paired according to synchronized estrus cycles (over 4 days) at which time MMT exposure began. Two weeks following initial MMT exposure, the 90-day-old females were bred by placing two receptive females (i.e., latestage proestrus) with a breeder male late in the afternoon and removed the next day at lights-on. Following breeding, females were individually housed and maintained on a high-calorie, high-protein diet formulation (Purina Rodent Chow 5008) throughout gestation and lactation. All females continued to be exposed to MMT through gestation and lactation. Nonpregnant females were euthanized and examined for fetal resorption. When possible, the litters were culled to eight males on postnatal day (PND) 1 (day of birth = PND 0). In a few cases, female pups were used to maintain equal litter sizes. On PND 21, the offspring were weaned and the littermates separated and housed individually on Beta-Chip bedding and provided food (Purina Rodent Chow 5001) and deionized water ad libitum. Only male offspring (one from each litter) were tested in the neurobehavioral tasks. Offspring were weighed at least weekly throughout the study. 2.3.2. Experiment 2 Thirty-five (n = 17 control, n = 18 MMT) timed-pregnant Sprague–Dawley (CD) female rats (Charles River, Raleigh, NC, USA), were received at gestational day (GD) 2 (spermpositive considered gestational day 0). Upon arrival, all females were housed individually as in Experiment 1, except for the light cycle (lights on at 0600 h). MMT exposure began at GD 6 and continued through gestation and lactation, as recommended in the US EPA Developmental Neurotoxicity Test Guidelines (US EPA, 1998b). When possible, litters (n = 17 control, n = 14 MMT) were culled to four males and four females on PND 4. On PND 21, the offspring were weaned and the littermates separated and housed individually on Beta-Chip bedding and provided food (Purina Rodent Chow 5001) and deionized water ad libitum. Both male and female offspring (one from each litter) were tested in neurobehavioral tasks with the exception of the runway task, in which only males were tested. 2.4. Neurobehavioral assessments 2.4.1. Runway learning test The runway was a Plexiglas testing apparatus consisting of an alley (38 cm long, 7.5 cm wide, 10 cm high) and a goal box (17 cm 26 cm 10 cm). Both the testing apparatus and the inter-trial interval holding cage were maintained at 37 8C
411
through the use of water-circulating heating pads. The task was an appetitive learning paradigm in which a food-deprived PND 11 rat pup was trained to negotiate a runway for a dry suckling reward from its anesthetized dam in the goal box (Amsel et al., 1976). Training consisted of alternating reinforced (R) and non-reinforced (N) trials. The pup was placed at the end of the runway facing away from the dam and the time to reach the goal box was recorded. On (R) trials, the pup was given 15 s of dry suckling before being transferred to the holding cage. On (N) trials, the pup was promptly placed in the holding cage upon reaching the end of the runway. If the pup failed to find the dam within the allotted time, the experimenter guided it down the runway for either reinforcement or to be immediately placed in the holding cage. Extinction immediately followed acquisition. The specific parameters for each experiment are outlined below. 2.4.2. Experiment 1 Dams were anesthetized using a drug cocktail equivalent to Chloropent1 (Amsel et al., 1976). Because Chloropent1 was not available at the time of this study, the composition (42.5 mg chloral hydrate, 8.86 mg pentobarbital and 21.2 mg magnesium sulfate, in each ml of sterile aqueous solution containing water, 33.8% propylene glycol and 14.25% ethyl alcohol) was obtained from the Online Database System of FDA approved animal drug products. The dams were dosed with 2 ml/kg body weight i.p. of the Chloropent equivalent approximately 15 min before testing began. When fully anesthetized, the dam was placed in the goal box at the end of the runway to serve as the reinforcer for the pup. Pups (n = 10/treatment group, except n = 11 at 245 ppm) were food-deprived for 10 h prior to testing and tested on PND 11 during their dark cycle. Acquisition consisted of 25 trials with (R) and (N) trials alternating on every other trial, and the maximum time was set at 120 s. Between each trial, the pup was placed in the holding cage for 15 s. Extinction began on the 26th trial. During this time, the pup was removed to the holding cage without gaining access to the dam on every trial. The maximum time was set at 100 s. Criterion to extinction was two consecutive trials with latencies of 100 s. After this criterion was reached, the pup was no longer tested; however, all pups were tested until they met criterion, regardless of the number of trials required. 2.4.3. Experiment 2 Dams were anesthetized using Nembutal1 sodium solution (50 mg/ml) (Abbott Laboratories, IL, USA). The dams were dosed with 2 to 2.2 ml/kg body weight i.p. of Nembutal1. As before, testing began when the dam was fully anesthetized. In this paradigm, pups (n = 17 control, n = 14 MMT) were food-deprived for 8 h prior to testing and tested on PND 11 during their light cycle. (Pilot studies had indicated that 8 h deprivation was not different from 10 h in terms of motivation and task performance.) Testing began with a preliminary training session of five initial training (R) trials, followed by a 2 min retention interval in the holding cage. There were then 25 acquisition trials in which (R) and (N) trials alternated in blocks
412
V.C. Moser et al. / NeuroToxicology 27 (2006) 409–420
of five trials, beginning and ending with 5(R) trials. The intertrial interval was 8 s. During acquisition, the pup was permitted 100 s before the trial ended and, if unsuccessful, it was guided to the dam. Extinction began on the 26th trial. Criterion to extinction was one trial with a latency of 100 s, and a maximum of 10 extinction trials were run.
In Experiment 1, only males were tested in the water maze (n = 9 control, n = 10 at 10 ppm, n = 8 at 50 ppm, and n = 7 at 245 ppm) whereas both males and females were tested in Experiment 2 (n = 16 control, n = 14 MMT 500 ppm).
2.5. Motor activity
For neuropathological evaluations, male rats (n = 6–7/dose at PND 1, n = 7–9/dose at PND 12, n = 7–11/dose at PND 22, and n = 7–10/dose at adult age, about 85–90-days old) were deeply anesthetized with pentobarbital and perfused via the left ventricle with buffered 4% formaldehyde:0.1% gluteraldehyde. Sagittal blocks of tissue were embedded in paraffin and sectioned to include all major structural landmarks of the brain in each section (e.g., olfactory bulb, striatum, cerebral cortex, hippocampus, thalamus, hypothalamus, brainstem, cerebellum). Twenty-four homologous sections of the brain from each rat, from each age, at each dose, were stained with hematoxylin and eosin. Brains from all control and high-dose rats, at all ages, were evaluated under contract by a qualified pathologist who was blind to the treatment (contract #68-D-02-102 to Experimental Pathology Laboratories). Step-down assessments, i.e., evaluation of the lower dose groups, were only conducted in the adult rats due to findings in the high-dose group; no abnormalities were observed at the younger ages. Any changes observed were subjectively scored in both number as well as nature of the change as follows: (1) minimal; (2) slight/mild; (3) moderate; (4) moderately severe; (5) severe.
Motor activity data were collected using an automated figure-eight chamber (Reiter, 1983). Activity was measured as interruptions (counts) of any of the eight photocell beams distributed around the maze. Counts were recorded over 5 min intervals for a 30 min testing session. In Experiment 1, motor activity was assessed in males at PNDs 13, 17 and 21 (n = 10 control, n = 9 at 50 ppm, n = 11 at 10 ppm, and n = 10 at 245 ppm). In Experiment 2, only PND 17 male and female offspring (one male and one female from each litter; n = 17 control, n = 14 MMT) were tested. 2.6. Morris water maze Spatial learning and memory was evaluated in adults (starting at about 70 days of age) using a Morris water maze (Morris, 1981) which consisted of a round galvanized steel tank 140 cm in diameter and filled with tap water (24–27 8C) to a depth of approximately 44 cm. A round, Plexiglas escape platform (9 cm in diameter) was positioned 2 cm beneath the surface of the water. The water was rendered opaque using black Tempra1 paint and was changed daily. The trials were videotaped and the image digitized for computer analysis using maze-tracking software (HVS Image, Ormond Cresent, Hampton, UK). Dependent variables included swim speed, latency and path length to find the platform, and time spent in the outer edge of the tank. The tank was digitally divided into three concentric zones (annuli) of equal width, and time spent in each zone was also analyzed. For the spatial training, the platform was placed in a fixed quadrant. Acquisition occurred during two trials each day over a total period of 9 days (5 days training, 2 days off, 4 days training). The rat was placed in the water at one of four starting points which varied such that the rat never started in the same place within 2 days. The trials continued until the rat mounted the platform, or for a maximum of 60 s. If the rat did not find the platform within the 60 s, it was guided there by the observer. After 15 s on the platform, the rat was placed in a holding cage for 5 min before the second trial began. On the 10th day, a probe trial was conducted in which the platform was removed and the rat’s tendency to search in the correct quadrant (where the platform had been) was measured over 60 s. The Gallagher proximity score, a measure of cumulative distance from the platform (Gallagher et al., 1993), and percent total path length within each quadrant, were used to assess quadrant bias. Following the memory probe trial, a visible target task (cued trial) was conducted using a raised platform of a contrasting color to confirm that the tested animals were not visually impaired.
2.7. Neuropathology (Experiment 1 only)
2.8. Neurochemical assessment and brain weight (Experiment 1 only) Male rats were decapitated under CO2-induced anesthesia on PND 1 (at culling), PND 22, and adult (80–90-days old). At culling, all pups within a litter were combined, and at older ages all rats came from different litters. Brains were removed from the calvaria, free of meninges, and weighed. The PND 22 and adult brains were dissected free-hand into the following regions: brainstem, neocortex, hippocampus and cerebellum (see White and Barone, 2001). Each region was immediately frozen in 2-methyl butane (99.5%, HPLC grade; Milwaukee, WI) on dry ice for 30 s, weighed and stored at 80 8C (White and Barone, 2001). Brain tissues were collected from Experiment 1 only (n = 4–9/dose/region). DNA fragmentation was used as an indicator of apoptotic cell death, and was quantified using a Cell Death ELISA (Roche Applied Science, Indianapolis, IN) procedure. Modifications to the procedure in the kit have been described previously (White and Barone, 2001). Specifically, the kit has been adapted for use with intact tissue and validated with both fresh and frozen brain tissue, and the results have been corroborated qualitatively by agarose gel and TUNEL data (White and Barone, 2001). In short, the enzyme-linked immunosorbent assay (ELISA) uses antibodies to bind fragmented DNA characteristic of apoptotic cell death. The bound fragments (i.e., nucleosomes) are then quantified photometrically.
V.C. Moser et al. / NeuroToxicology 27 (2006) 409–420
413
2.9. Statistical analyses Continuous data (e.g., body weight, water intake, activity counts, etc.) were analyzed using a general linear model ANOVA (SAS, Cary, NC). Extinction in the runway test was analyzed with a survival model for censored data (SAS), and count data (e.g., number of pups learning, pregnancy rate) were compared using Fisher’s exact test (SAS). When the same rat was used in repeated tests (e.g., repeated motor activity testing, body weights over time), the analyses included time as a repeated factor. For apoptosis data, all brain region values were compared to their respective controls, and arcsine transformed prior to statistical analyses. Following a significant overall analysis, Dunnett’s t-test was used to compare dose groups with the control. In all cases, resulting probability values <0.05 were considered significant. Fig. 1. Maternal body weights in Experiment 1 throughout exposure to 0, 10, 50, or 245 ppm MMT.
3. Results 3.1. Maternal fluid intake and weights
day-by-intake interaction was followed with step-down analyses, which indicated that only the intake measured 3 days post-parturition was not different. During gestation, MMT consumption was about 80–88% of control levels, and during lactation, 82–88% of control. Despite the lowered intake, body weight was not different in the treated group (Table 1). Table 1 also includes the calculated MMT intake during gestation and lactation. As in Experiment 1, the intake during lactation nearly doubled.
3.1.1. Experiment 1 Fluid intake increased somewhat during gestation (about 30–40%), and was much higher (more than double) during lactation. There were, however, no differences in fluid intake across treatment groups at any time during the study. Likewise, there were no treatment-related differences in body weight throughout exposure (Fig. 1). Table 1 lists the fluid intake, weight, and calculated MMT intake (mg/kg/day) during the three phases of the study: pre-breeding, gestation, and lactation. Predictably, MMT intake was much greater during lactation.
3.2. Reproductive/developmental parameters 3.2.1. Experiment 1 Despite the timed breeding, pregnancy rate was very low with only 44 of 120 (37%) rats delivering litters (10, 11, 11, and
3.1.2. Experiment 2 There was a significant depression of fluid intake across almost all days of treatment with MMT 500 ppm. A significant
Table 1 Fluid intake (average, and range of means, ml/day), body weight (average, and range of means, g), and calculated MMT intake (mg/kg/day) for each concentration of MMT (ppm) throughout exposure Concentration
Intake
Weight
MMT
0 10 50 245
28.9 26.5 28.2 25.7
(27.0–30.4) (24.5–27.7) (27.1–29.1) (23.6–26.8)
266.8 270.6 268.0 270.0
(253.0–271.1) (254.9–274.6) (255.5–271.6) (256.0–274.3)
– 1.0 5.3 23.3
Gestation
0 10 50 245
39.9 36.1 43.4 36.5
(36.1–48.1) (30.7–43.5) (35.1–50.6) (30.8–44.7)
330.8 330.6 336.3 336.7
(274.5–418.1) (280.8–414.5) (276.4–420.3) (278.4–422.3)
– 1.1 6.5 25.7
Lactation
0 10 50 245
68.4 65.2 75.0 62.8
(48.2–92.9) (44.7–86.8) (51.4–95.9) (47.8–82.1)
339.4 336.8 343.9 347.5
(336.2–346.3) (322.6–348.3) (332.3–356.6) (335.8–352.3)
– 1.8 10.6 41.6
0 500
38.2 (33.1–44.0) 32.4 (28.5–37.4)
284.7 (231–346.9) 290.5 (233.9–356)
– 55.8
0 500
71.7 (40.8–96.3) 60.8 (36.9–85.1)
314.7 (304.0–328.8) 322.5 (310.9–333.8)
– 94.3
Experiment 1 Pre-breeding
Experiment 2 Gestation Lactation
414
V.C. Moser et al. / NeuroToxicology 27 (2006) 409–420
12 litters for control, 10, 50, and 245 ppm, respectively). We typically find that most of our Sprague–Dawley rats deliver within a 24 h period which we specify as PND 0. In the present study, however, 11 of the 44 (25%) rats delivered later than that time window. Necropsy of all non-pregnant rats revealed resorptions in only two control rats and one rat from the lowdose group. Incidence of pregnancy, late delivery, and resorptions were not statistically different across treatment groups. 3.2.2. Experiment 2 All of the timed-pregnant females in the control group delivered, but two in the MMT group did not. These rats were not evaluated for implantation sites. All of the deliveries occurred when expected. In the MMT group, one litter consisted of all females and was not used, and another litter was killed by the dam shortly after birth. 3.3. Offspring number and growth 3.3.1. Experiment 1 There were no significant differences in the number of pups per litter across treatment groups, as follows (mean number per litter S.E.): control, 12.5 1.6; 10 ppm, 15.2 0.6, 50 ppm, 13.1 1.3, and 245 ppm, 13.4 1.5. Likewise, there were no treatment-related differences in the number of males and females per litter. Litter birth weights and body weights across time were similar across treatment groups throughout the entire study (data not shown). In addition, there were no differences in weights of the pups selected for each behavioral test. 3.3.2. Experiment 2 There were no differences in the number of pups per litter (control, 11.9 0.4; MMT, 12.2 0.8), or the sex ratio within the litters. Body weight changes during the lactation period showed a significant dose-by-sex interaction, but step-down analyses showed that on 1 day only (PND 11), male and female pups in the control group were different by about 4 g; this was considered to be within biological variability. There were no treatment effects on body weight after weaning. 3.4. Runway testing 3.4.1. Experiment 1 Almost all of the pups learned to traverse the runway to the goal box. Using a criterion of at least one latency (time to reach the goal box) less than the maximum time of 120 s, only one pup (high dose) did not learn the task. Mean latencies (excluding the data from the one non-learner) across acquisition trials are shown in Fig. 2. While highly variable, the latency data did not reveal a significant treatment-related difference. The time to extinction was also quite variable, with one pup (low-dose) requiring 46 trials to meet the criterion of two consecutive latencies of 100 s. Fig. 2 presents the cumulative percent of each treatment group reaching criterion at each trial.
Fig. 2. Performance in the runway task on PND 11 in Experiment 1. (Top) Latency to traverse the runway during each reinforced (R) or nonreinforced (N) trial which strictly alternated. (Bottom) Percent of each group reaching extinction (two consecutive latencies of 100 s) during consecutive nonreinforced trials.
The median number of trials to reach criterion for each dose group was as follows: control, 17; 10 ppm, 15.5, 50 ppm, 20.5; 245 ppm, 20.5. 3.4.2. Experiment 2 Fewer pups were able to learn the different paradigm used in the second study. One control and five treated pups did not meet the criterion of having at least one latency less than 100 s during acquisition; however, this difference did not attain statistical significance (Fisher’s exact p = 0.067). Furthermore, one pup in each group did not extinguish within the 10 trials. A different pattern of responding was elicited by the procedure of batched reinforced (R) and non-reinforced (N) trials. Fig. 3 shows that, in general, latencies became shorter during the R trials, and longer during the N trials. This pattern indicates repeated acquisition and extinction, and was most evident in the last two sets of trials. Trend analyses were conducted on each set of trials to determine if the changes within each block were significant (either increasing or decreasing). For the control group, the last two blocks showed significant slopes in the latencies (mean S.E. slopes for the last N-trials, 8.7 3.0; last R trials, 7.8 3.3). On the other
V.C. Moser et al. / NeuroToxicology 27 (2006) 409–420
415
and 17. By PND 21, habituation was evident. There were no group differences on any of these measures. 3.5.2. Experiment 2 Motor activity testing was only conducted on PND 17, at which time there were no differences between the control and MMT-treated groups (mean S.E. for each group: control, 99.6 7.0 total counts; 500, 108.3 8.5). Habituation was evident in both dose groups. 3.6. Morris water maze 3.6.1. Experiment 1 All rats learned the location of the hidden platform over the two weeks of training, as evidenced by decreasing latency to find the platform with repeated training (Fig. 4). Daily blocks and training trials showed significant changes over testing. Latency, path length, swim speed, and most spatial distribution variables were not different across treatment groups. The only significant finding was increased time in the middle zone in the 50 ppm group on day 6 only; this was not supported by other findings and is probably not biologically relevant.
Fig. 3. Performance in the runway task on PND 11 in Experiment 2. (Top) Latency to traverse the runway during each reinforced (R) or nonreinforced (N) trial which were grouped with five trials each. (Bottom) Percent of each group reaching extinction (one latency of 100 s) during consecutive nonreinforced trials.
hand, the MMT group showed significant slopes during the third and fourth blocks (middle R trials, 12.0 2.9; last N trials 11.0 4.0), but not the fifth (last R trials, 6.0 5.3, p = 0.074). Using ANOVA to evaluate group differences, only one trial showed a significant difference between the groups (fourth trial of the middle R-trials). Cumulative percentages of the groups reaching criterion (one trial with latency of 100 s) across each extinction trial are shown in Fig. 3; there were no significant differences. The median number of trials to reach criterion were: control, 4 trials; MMT, 5.5 trials. 3.5. Motor activity 3.5.1. Experiment 1 Total motor activity counts during 30-min sessions were lowest at PND 13 (means ranging from 20.3 to 31.2), and increased over PND 17 (means 46.5 to 72.7) and 21 (means 98.7 to 120.3). Analysis of the within-session activity (in 5 min intervals) showed no habituation during the session at PNDs 13
Fig. 4. Performance in the Morris water maze in adults in Experiment 1. (Top) Latency to find the platform in daily training trials, with two trials each day. (Bottom) Time in each quadrant (Q4 = quadrant where platform had been) during the probe trial (no platform present).
416
V.C. Moser et al. / NeuroToxicology 27 (2006) 409–420
During the memory probe, all rats spent the most time searching in the quadrant where the platform had been located (Fig. 4). There were no differences in quadrant bias, Gallagher index, or spatial distribution variables. For the visual function test, there were no differences in the latency to reach a raised, visible platform, as follows (mean S.E.): control, 27.4 4.7 s; 10 ppm, 25.3 2.6 s; 50 ppm, 31.4 7.0 s; 245 ppm, 24.0 4.4 s. 3.6.2. Experiment 2 As in Experiment 1, all rats learned the spatial task, indicated by significant changes in daily blocks and individual training trials. There were, however, no interactions with treatment or sex on any endpoint. Latency to find the platform is shown in Fig. 5, and other dependent variables showed similar patterns. Likewise, there were no differences on any measures during the memory probe, or during the visual probe (visible platform). 3.7. Brain weight (Experiment 1 only) There was a marginal effect on brain weight, although the overall dose effect at p < 0.07 did not reach the criterion for statistical significance. There was no interaction of treatment and age, and collapsed across ages, the high dose group showed a 5% decrease in weight. 3.8. Neuropathology (Experiment 1 only) No histological alterations were observed in the brains of MMT-treated offspring killed on PND 1, 12, or 22. There were, however, changes in the brains of the offspring evaluated as adults. The cerebral cortical lesion was characterized by 2– 4 mm diameter, round vacuoles of varying size, in the gray matter neuropil, localized to the region of the orbital cortex. Furthermore, the focal location of the lesion was very similar across animals, supporting the conclusion of a treatmentrelated effect. Myelin was not affected. Fig. 6 depicts representative lesions in a control and a low-dose rat.
Fig. 5. Performance in the Morris water maze in adults in Experiment 2. (Top) Latency to find the platform in daily training trials, with two trials each day. (Bottom) Time in each quadrant (Q4 = quadrant where platform had been) during the probe trial (no platform present).
Two of eight (25%) adult high-dose offspring had vacuolation scored as ‘‘slight/mild’’ (‘2’), whereas one of seven (14%) of the mid-dose group, and three of ten (30%) lowdose rats, had similar vacuolation scored as ‘‘minimal’’ (‘1’).
Fig. 6. High (40) magnification of the area of the cerebral cortex (H&E stain) from adults in Experiment 1. (A) Female control rat, no neuropil vacuolation present. (B) Female rat exposed to 10 ppm MMT, with minimal neuropil vaculation present. Bar = 25 mm.
V.C. Moser et al. / NeuroToxicology 27 (2006) 409–420
Fig. 7. DNA fragmentation from cell death ELISA on cortical tissues from PND 22 and adults in Experiment 1.
There were no such findings in control rats. Although there was no increase in incidence or more widespread localization of this lesion with increasing dose, the change was scored as slightly greater severity in the high-dose offspring (‘2’s) than that of the lesions in the lower dose groups (‘1’s). There were no histopathological findings in any other major brain region other than this specific area of the cerebral cortex. These adult rats had been tested previously in the Morris water maze. Review of the water maze data for the rats showing vacuolation revealed that their performance was not different from the unaffected rats. Thus, no within-individual correlations between vacuolation and behavioral performance could be made. 3.9. Apoptosis (Experiment 1 only) The data for PND 22 brainstem were not analyzed due to insufficient sample size. Data from PND 22 and adult cortex are presented in Fig. 7, showing no treatment-related difference in DNA fragmentation. Similarly, the PND 22, and adult, neocortex and hippocampus remained unaffected by perinatal MMT exposure, as did the adult brainstem (data not shown). 4. Discussion There were very few effects of MMT in offspring exposed during gestation and lactation. The highest concentration (500 ppm, providing intake of 56–94 mg/kg/day) lowered fluid intake, and we felt that using even higher concentrations would be unreasonably high as well as produce maternal toxicity. There were no neurobehavioral effects on any endpoint, at any time, nor were there developmental changes as measured by apoptosis. The presence of vacuolation in the neuropil was observed only in treated rats, but the incidence was minimal (1– 3 rats per treatment group). The overall effect on brain weight was also minimal, and neither of these findings could be related to other biological effects. Unfortunately, neuropathology and brain weights were not evaluated in the second study, which could have confirmed or disputed the findings.
417
One impetus for this study was to replicate and expand the findings of Noland et al. (1982) of increased trials to extinction in all treated groups. To that end, we attempted to duplicate as many details as possible in Experiment 1. First, our choice of concentrations (5, 25, 120 ppm Sn) extended the dose-range used by Noland (12, 40, 120 ppm Sn). Although the source of chemical was different, the purity was verified (but not stated) in the Noland study, and ours was obtained at 97% purity. The runway was built to the specifications described in Taylor et al. (1982), which was referenced in the Noland paper. Other variables that were the same included: rat strain, age, light cycle, mating and culling procedures, frequency of water change, and runway paradigm. The only variable we know to be different was bedding (sawdust versus our hardwood chips). It should be noted that even though Sprague–Dawley rats were obtained from Charles River in both studies, it is unknown which facility was used by Noland et al. (1982), and in the early 1980s the Raleigh facility (used in the present experiment) was not open. As such, it is possible that the rats were actually quite different, as genetic drift in outbred rats as well as differences in suppliers are well-documented (Festing, 1979, 1987; Glick et al., 1986; Kacew et al., 1995). Furthermore, there could be genetic changes among Sprague–Dawley rats that resulted from a comprehensive restructuring (i.e., International Genetic Standard) of breeding programs at Charles River Laboratories in 1992 (White and Lee, 1998). The current breeding program was instituted to minimize the degree of variation associated with genetic drift. Thus, it is likely that disparity among Sprague–Dawley animals produced today, in comparison to 20–25 years ago, may be partially responsible for the observed behavioral differences. There could also be other factors that were not reported in the original paper that could have contributed to our different results. Of course, it could simply be that the findings were not sufficiently robust to be replicated. The runway was originally developed by Amsel and colleagues in the 1970s to demonstrate the ontogeny of learning and extinction in preweanling rats, using dry suckling (anesthetized dam) as the reward (Amsel et al., 1976). Since the original publication, many experimental and parametric manipulations of the runway task have been described (e.g., Amsel and Chen, 1976; Amsel et al., 1977; Chen and Amsel, 1980; Highfield et al., 1996; Lilliquist et al., 1999; Lobaugh et al., 1985, 1989; Nair and Gonzalez-Lima, 1999; Nair et al., 2001; Stanton, 1982; Wigal et al., 1988a,b). PND 11 pups can learn the strict alternation (reinforced/non-reinforced), and show a go-no-go response pattern, but only after many more training trials than presented in the current study (e.g., 60–100 trials compared to 25; Lilliquist et al., 1999; Lobaugh et al., 1985; Stanton, 1982). In Experiment 2, the schedule of reinforcement was altered to demonstrate that we could control the behavior and be more consistent with Amsel’s early work (e.g., Amsel et al., 1976, 1977). After the first two to three blocks of five trials, it was evident that responding (latencies) got faster during the reinforced trials, but slower during the non-reinforced trials. There was substantial variability in the present studies, and while we used sample sizes typically used by Amsel’s group (n = 10–15), it is difficult to compare
418
V.C. Moser et al. / NeuroToxicology 27 (2006) 409–420
variability across laboratories since almost no other papers present estimates of variability in the figures or text. Despite the considerable literature on this test, there are relatively few uses of the test for evaluating chemical effects, and again, most of these are from Amsel’s laboratory. Gestational ethanol exposure produced changes in the runway extinction when using a partial reinforcement, but not single patterned alternation schedule of reinforcement (e.g., Wigal et al., 1988a,b). Direct dosing of MK-801 to preweaning rat pups impaired learning in this task, but only at longer inter-trial intervals (Highfield et al., 1996). The only non-Amsel papers we could find which used this specific apparatus came from Taylor’s laboratory; these included a perinatal study of lead acetate (Taylor et al., 1982) and the already-described study of MMT (Noland et al., 1982). A comparison of our control data to both of these studies showed that, while the acquisition curves appeared to be quite similar, the extinction curve for our control (as well as treated groups) was much like their MMT and TMT curves. Indeed, 50% of their treated groups had reached the extinction criterion by 15 trials (their cut-off), and half of our rats reached that level after 15–20 trials, thus showing a difference in control performance. In contrast to the lack of database for the runway, the other behavioral tests included in these studies have been widely used for developmental neurotoxicity evaluations. Motor activity is an apical test of neuronal function, and an evaluation of its preweanling ontogeny reflects maturation of the nervous system. This test is required in the current US EPA Developmental Neurotoxicity Test Guidelines (US EPA, 1998b). The Morris water maze was developed as a test of visual spatial navigation with both a learning and memory component. It can demonstrate the rodent’s ability to learn the position of a hidden platform using extra-maze spatial cues. There are many papers describing chemical effects on this behavior following acute, repeated, and developmental exposures (e.g., Brandeis et al., 1989; D’Hooge and DeDeyn, 2001; Jett et al., 1997; Kuhlmann et al., 1997; McNamara and Skelton, 1993; Moser et al., 2001). On these more standard and accepted behavioral assays, MMT had no effect. Apoptosis is widespread in both the fetal and postnatal CNS, and any disturbance in either the progression or total amount of apoptosis could be detrimental for a developing organism (e.g., learning and memory impairments, or progression of neurodegenerative diseases) (Bennett et al., 1998; Leist and Nicotera, 1998; White and Barone, 2001). Moreover, the process of apoptosis is closely linked, and potentially co-occurs, with other developmental processes such as proliferation and migration (Liesi, 1997; White and Barone, 2001). Using the same brain regions examined in the current study, the cell death ELISA procedure has been successfully used to detect changes in apoptosis following perinatal exposure to another organotin (dibutyltin, DBT) (Jenkins et al., 2004). The lack of effect observed on DNA fragmentation with MMT exposure agrees with the in vitro MMT data (Jenkins et al., 2004). Furthermore, the lack of effect in vivo, even in a similar brain region, does not obviate the pathological findings of vacuolation, since these are distinct processes.
Neuropathological lesions and brain weight changes are considered indicators of neurotoxicity (US EPA, 1998c); however, the MMT effects described here were relatively small. The localization of the vacuoles was consistent across rats, and there was a total lack of such findings in any control rats or at any earlier ages. While the incidence did not increase with dose, the severity of change did show a slight dose–response (only high-dose rats were scored as ‘slight/mild’ changes, while lesions in the lower dose groups were ‘minimal’). All this information taken together suggests that this finding is not an artifact. Whether the vacuoles were located in neuronal or glial tissue, or the nature of the vacuoles, cannot be fully determined with light microscopy. Clearly, additional studies of this outcome are warranted and would require greater resolution using plastic embedding and transmission electron microscopic examination. The extent of parallels between specific regions of prefrontal cortex in rat and human, such as the orbitofrontal cortex, is not clear (Dalley et al., 2004), and it is further unclear whether the area of these vacuoles corresponds to the orbitofrontal cortex area. This region of cortex is sensitive to chemical disruptions, and has been implicated in motivation, decision making, and working memory (Dalley et al., 2004; Schoenbaum and Setlow, 2001). The behavioral tests used herein do not tap into these functions, which could help explain the lack of behavioral effects observed. In humans, prefrontal cortical damage spares basic cognitive function while interfering with emotional signals (Bechara, 2004). On the other hand, the minimal degree of vacuolation may have no disruptive influence on these functions. Additional studies using other cognitive tasks could pursue and resolve these possibilities. It is clear, however, that the nature of these findings differ from lesions produced by adult or postnatal exposure to TMT or triethyltin (TET) (O’Shaughnessy and Losos, 1986). Acute TMT exposure produces neuronal cell death characterized by homogeneous eosinophilic cytoplasm and nuclear pyknosis, followed by gliosis including microglial activation and astrogliosis. The limbic system is most severely affected, but other affected areas include the cerebral cortex and brainstem. While TET produces neuronal vacuolization, another prominent feature is spongy myelin edema (O’Shaughnessy and Losos, 1986), which was not observed here. Despite the developmental exposure paradigm, no morphological changes were observed in younger rats, which have been shown to be more sensitive and have earlier onset of pathological indices to TMT and TET (Barone, 1993; Freeman et al., 1994). While we showed that MMT did not convert to other methyltin species in the rats’ water bottles, it is possible that such interconversions could take place in the body. We could find no data addressing this, but we feel it is unlikely that MMT converted to TMT since the pathology we observed was quite unlike that produced by TMT (see above). It is interesting to note, however, that similar findings were obtained with dimethyl tin (unpublished results). In summary, two separate developmental neurotoxicity studies of MMT indicated a lack of effects on the neuro-
V.C. Moser et al. / NeuroToxicology 27 (2006) 409–420
behavioral endpoints that were tested, no changes in the apoptosis assay, but a small yet reliable appearance of brain changes. This could indicate that the behaviors included were not sensitive to the pathological changes, or that the pathology was not sufficient to disrupt the functioning of the test subject. We could not replicate the effect of MMT on cognitive function. This information adds to the scant database of developmental effects of organotins. Acknowledgements The authors wish to acknowledge the invaluable assistance of Megan Newcomb (NC State University School of Veterinary Medicine), Joan Hedge, Rene´e Marshall, and Brian Robinette (all of Neurotoxicology Division, US EPA) for their help in the technical conduct of these labor-intensive tests. We also thank Drs. Lori White and Scott Jenkins for their help with the perfusions and apoptosis assay, and Drs. Karl Jensen and John Lipscomb who reviewed an earlier version of the manuscript. Dr. K. Ehman was supported by the EPA/UNC Toxicology Research Program, Training Agreement #T829472. References Amsel A, Chen JS. Ontogeny of persistence: immediate and long-term persistence in rats varying in training age between 17 and 65 days. J Comp Physiol Psychol 1976;90:808–20. Amsel A, Burdette DR, Letz R. Appetitive learning, patterned alternation, and extinction in 10-d-old rats with non-lactating suckling as reward. Nature 1976;262:816–8. Amsel A, Letz R, Burdette DR. Appetitive learning and extinction in 11-day-old rat pups: effects of various reinforcement conditions. J Comp Physiol Psychol 1977;91:1156–67. Barone S Jr. Developmental differences in neural damage following trimethyltin as demonstrated with GFAP immunohistochemistry. In: Johannessen JN, editor. Markers of neuronal injury and degeneration. New York: New York Academy of Science; 1993. p. 306–17. Bechara A. The role of emotion in decision-making: evidence from neurological patients with orbitofrontal damage. Brain Cogn 2004;55: 30–40. Bennett SA, Tenniswood M, Chen JH, Davidson CM, Keyes MT, Fortin T, et al. Chronic cerebral hypoperfusion elicits neuronal apoptosis and behavioral impairment. NeuroReport 1998;9:161–6. Bouldin TW, Goines ND, Bagnell CR, Krigman MR. Pathogenesis of trimethyltin neuronal toxicity: ultrastructural and cytochemical observations. Am J Pathol 1981;104:237–49. Brandeis R, Brandys Y, Yehuda S. The use of the Morris water maze in the study of memory and learning. Int J Neurosci 1989;48:29–69. Chen JS, Amsel A. Learned persistence at 11–12 days but not at 10–11 days in infant rats. Dev Psychobiol 1980;13:481–91. Dalley JW, Cardinal RN, Robbins TW. Prefrontal executive and cognitive functions in rodents: neural and neurochemical substrates. Neurosci Biobehav Rev 2004;28:771–84. D’Hooge R, DeDeyn PP. Applications of the Morris water maze in the study of learning and memory. Brain Res Rev 2001;36:60–90. Donard OF, Quevauviller P, Bruchet A. Tin and organotin speciation during wastewater and sludge treatment processes. Water Res 1993;27: 1085–1089. Festing MFW. Properties of inbred strains and outbred stocks, with special reference to toxicity testing. J Toxicol Environ Health 1979;5: 53–68. Festing MFW. Genetic factors in toxicology: implications for toxicological screening. CRC Crit Rev Toxicol 1987;18:2–36.
419
Forsyth DS, Dabeka R, Sun WF, Dalglish K. Speciation of organotins in poly(vinyl chloride) products. Food Addit Contam 1993;10:531–40. Freeman JH, Barone S Jr, Stanton ME. Triethyltin produces neural damage and cognitive deficits in developing rats that depend on age of exposure. Brain Res 1994;634:85–95. Gallagher M, Burwell R, Burchinal M. Severity of spatial learning impairment in aging: development of a learning index for performance in the Morris water maze. Behav Neurosci 1993;107:618–26. Glick SD, Shapiro RM, Drew KL, Hinds PA, Carlson JN. Differences in spontaneous and amphetamine-induced rotational behavior, and in sensitization to amphetamine, among Sprague–Dawley derived rats from different sources. Physiol Behav 1986;38:67–70. Highfield DA, Nixon K, Amsel A. The NMDA antagonist MK-801 affects nonspatial learning in preweanling rats. Behav Neurosci 1996;110: 300–4. Jenkins SM, Ehman K, Barone S Jr. Structure-activity comparison of organotin species: dibutyltin is a developmental neurotoxicant in vitro and in vivo. Dev Brain Res 2004;151:1–12. Jett DA, Kuhlmann AC, Farmer SJ, Guilarte TR. Age-dependent effects of developmental lead exposure on performance in the Morris water maze. Pharmacol Biochem Behav 1997;57:271–9. Kacew S, Ruben Z, McConnell RF. Strain as a determinant factor in the differential responsiveness of rats to chemicals. Toxicol Pathol 1995;23: 701–15. Krigman MR, Silverman AP. General toxicology of tin and its organic compounds. Neurotoxicology 1984;5:129–40. Kuhlmann AC, McGlothan JL, Guilarte TR. Developmental lead exposure causes spatial learning deficits in adult rats. Neurosci Lett 1997;233: 101–104. Liesi P. Ethanol-exposed central neurons fail to migrate and undergo apoptosis. J Neurosci Res 1997;48:439–48. Leist M, Nicotera P. Apoptosis, excitotoxicity, and neuropathology. Exp Cell Res 1998;239:183–201. Lilliquist MW, Nair HP, Gonzalez-Lima F, Amsel A. Extinction after regular and irregular reward schedules in the infant rat: influence of age and training duration. Dev Psychobiol 1999;34:57–70. Lobaugh NJ, Bootin M, Amsel A. Sparing of patterned alternation but not partial reinforcement effect after infant and adult hippocampal lesions in the rat. Behav Neurosci 1985;99:46–59. Lobaugh NJ, Greene PL, Grant M, Nick T, Amsel A. Patterned (single) alternation in infant rats after combined or separate lesions of hippocampus and amygdale. Behav Neurosci 1989;103:1159–67. McNamara RK, Skelton RW. The neuropharmacological and neurochemical basis of place learning in the Morris water maze. Brain Res Rev 1993;18:33–49. Mersiowsky I, Brandsch R, Ejlertsson J. Screening for organotin compounds in European landfill leachates. J Environ Qual 2001;30:1604–11. Morris RGM. Spatial localization does not require the presence of local cues. Learn Motiv 1981;12:239–60. Moser VC, Shafer TJ, Ward TR, Meacham CA, Harris MW, Chapin RE. Neurotoxicological outcomes of perinatal heptachlor exposure in the rat. Toxicol Sci 2001;60:315–26. Nair HP, Gonzalez-Lima F. Extinction of behavior in infant rats: development of functional coupling between septal, hippocampal, and ventral tegmental regions. J Neurosci 1999;19:8646–9655. Nair HP, Berndt JD, Barrett D, Gonzalez-Lima F. Maturation of extinction behavior in infant rats: large-scale regional interactions with medial prefrontal cortex, orbitofrontal cortex, and anterior cingulated cortex. J Neurosci 2001;21:4400–7. Noland EA, Taylor DH, Bull RJ. Monomethyl and trimethyltin compounds induce learning deficiencies in young rats. Neurobehav Toxicol Teratol 1982;4:539–44. O’Shaughnessy DJ, Losos GJ. Peripheral and central nervous system lesions caused by triethyl- and trimethyltin salts in rats. Toxicol Pathol 1986;14: 141–8. Reiter LW. Chemical exposures and animal toxicity: utility of the figure-eight maze. In: Hayes AW, Schell RC, Miya TS, editors. The science and practice of toxicology. New York: Elsevier; 1983. p. 73–84.
420
V.C. Moser et al. / NeuroToxicology 27 (2006) 409–420
Sadiki A-I, Williams DT. A study on organotin levels in Canadian drinking water distributed through PVC pipes. Chemosphere 1999;38:1541–8. Sadiki A-I, Williams DT, Carrier R, Thomas B. Pilot study on the contamination of drinking water by organotin compounds from PVC materials. Chemosphere 1996;32:2389–98. Schoenbaum G, Setlow B. Integrating orbitofrontal cortex into prefrontal theory: common processing themes across species and subdivisions. Learn Memory 2001;8:134–47. Stanton ME. Performance of 11- and 14-day-old rats on a working memory problem. Behav Neural Biol 1982;36:304–10. Taylor DH, Noland EA, Brubaker CM, Crofton KM, Bull RJ. Low level lead (Pb) exposure produces learning deficits in young rat pups. Neurobehav Toxicol Teratol 1982;4:311–4. U.S. Environmental Protection Agency (US EPA). US environmental protection agency announcement of the drinking water contaminant candidate list. Fed. Register 63; March 2, 1998a, p. 10274–87. U.S. Environmental Protection Agency (US EPA). Developmental neurotoxicity study, health effects test guidelines, OPPTS 870.6300, EPA 712-C-98239; 1998b. http://www.epa.gov/opptsfrs/publications/OPPTS_Harmonized/ 870_Health_Effects_Test_Guidelines/Series.
U.S. Environmental Protection Agency (US EPA). Guidelines for neurotoxicity risk assessment, EPA/630/R-95/001F; 1998c. http://cfpub.epa.gov/ncea/raf/ recordisplay.cfm?deid=12479. U.S. Environmental Protection Agency (US EPA). Memo re: TSCA 8(e) Submission; 2004. http://www.epa.gov/oppt/tsca8e/doc/8ehq/2004/february04/ 8ehq-0104-15511a.pdf. White LD, Barone S Jr. Qualitative and quantitative estimates of apoptosis from birth to senescence in the rat brain. Cell Death Differ 2001;8:345–56. White WJ, Lee CS. The development and maintenance of the Crl:CD1(SD)IGS BR rat breeding system, CD(SD) IGS; 1998, p. 8–14 (online). Wigal T, Greene PL, Amsel A. Effects on the partial reinforcement extinction effect and on physical and reflex development of short-term in utero exposure to ethanol at different periods of gestation. Behav Neurosci 1988a;102:51–3. Wigal T, Lobaugh NJ, Wigal SB, Greene PL, Amsel A. Sparing of patterned alternation but not partial reinforcement extinction effect after prenatal chronic exposure to ethanol in infant rats. Behav Neurosci 1988b;102: 43–50. Wilkinson RR. Technoeconomic and environmental assessment of industrial organotin compounds. Neurotoxicology 1984;5:141–58.
Evaluation of developmental neurotoxicity of organotins via drinking water in rats: Monomethyl tin§ Virginia C. Moser a,*, Stanley Barone Jr.a,1, Pamela M. Phillips a, Katherine L. McDaniel a, Kimberly D. Ehman b,2 a
Neurotoxicology Division, National Health and Environmental Effects Research Laboratory, Office of Research and Development, US Environmental Protection Agency, Research Triangle Park, NC 27711, United States b University of North Carolina Chapel Hill, Curriculum in Toxicology, US Environmental Protection Agency, Research Triangle Park, NC, United States Received 30 September 2005; accepted 6 December 2005 Available online 25 January 2006
Abstract Organotins such as monomethyltin (MMT) are widely used as heat stabilizers in PVC and CPVC piping, which results in their presence in drinking water supplies. Concern for neurotoxicity produced by organotin exposure during development has been raised by published findings of a deficit on a runway learning task in rat pups perinatally exposed to MMT (Noland EA, Taylor DH, Bull RJ. Monomethyl and trimethyltin compounds induce learning deficiencies in young rats. Neurobehav Toxicol Teratol 1982;4:539–44). The objective of these studies was to replicate the earlier publication and further define the dose-response characteristics of MMT following perinatal exposure. In Experiment 1, female Sprague–Dawley rats were exposed via drinking water to MMT (0, 10, 50, 245 ppm) before mating and throughout gestation and lactation (until weaning at postnatal day [PND] 21). Behavioral assessments of the offspring included: a runway test (PND 11) in which the rat pups learned to negotiate a runway for dry suckling reward; motor activity habituation (PNDs 13, 17, and 21); learning in the Morris water maze (as adults). Other endpoints in the offspring included measures of apoptosis (DNA fragmentation) at PND 22 and as adults, as well as brain weights and neuropathological evaluation at PND 2, 12, 22, and as adults. There were no effects on any measure of growth, development, cognitive function, or apoptosis following MMT exposure. There was a trend towards decreased brain weight in the high dose group. In addition, there was vacuolation of the neuropil in a focal area of the cerebral cortex of the adult offspring in all MMT dose groups (1–3 rats per treatment group). In Experiment 2, pregnant rats were exposed from gestational day 6 until weaning to 500 ppm MMT in drinking water. The offspring behavioral assessments again included the runway task (PND 11), motor activity habituation (PND 17), and Morris water maze (as adults). In this second study, MMT-exposed females consumed significantly less water than the controls throughout both gestation and lactation, although neither dam nor pup weights were affected. As in Experiment 1, MMT-exposure did not alter pup runway performance, motor activity, or cognitive function. These results indicate that perinatal exposure to MMT, even at concentrations which decrease fluid intake, does not result in significant neurobehavioral or cognitive deficits. While mild neuropathological lesions were observed in the adult offspring, the biological significance of this restricted finding is unclear. Published by Elsevier Inc. Keywords: Developmental neurotoxicity; Organotin; Monomethyl tin; Behavior; Neuropathology
§ The information in this document has been reviewed by the National Health and Environmental Effects Research Laboratory and approved for publication. Approval does not signify that the contents necessarily reflect the views of the Agency nor does mention of trade names or commercial products constitute endorsement or recommendation for use. * Corresponding author at: Neurotoxicology Division, National Health and Environmental Effects Research Laboratory (MD B105-04), US Environmental Protection Agency, Research Triangle Park, NC 27711, United States. Tel.: +1 919 541 5075; fax: +1 919 541 4849. E-mail address: [email protected] (V.C. Moser). 1 Current address: National Center for Environmental Assessment, US Environmental Protection Agency, Washington, DC, United States. 2 Current address: RTI International, Research Triangle Park, NC, United States.
0161-813X/$ – see front matter. Published by Elsevier Inc. doi:10.1016/j.neuro.2005.12.003
410
V.C. Moser et al. / NeuroToxicology 27 (2006) 409–420
1. Introduction
2. Methods
Mono- and di-substituted organotins are commercially used as heat and light stabilizers for polyvinyl chloride (PVC) and chlorinated PVC (CPVC) products. PVC pipes are widely used in domestic water systems, and this has contributed to the presence of organotins in drinking water supplies and in the environment (Donard et al., 1993; Forsyth et al., 1993; Mersiowsky et al., 2001; Sadiki and Williams, 1999; Sadiki et al., 1996; Wilkinson, 1984). Both mono- and dimethyl tin have been detected in PVC-supplied drinking water at levels up to 49.1 ng Sn/l (dimethyl tin, DMT) or as high as 291 ng Sn/l (monomethyl tin, MMT; Sadiki and Williams, 1999). Because human exposure to organotins is widespread and the health consequences unknown, organotins were listed on the US EPA Candidate Contaminant List (CCL). This List includes nonregulated substances known or anticipated to occur in public water systems, with identified needs (data gaps) in terms of health effects (US EPA, 1998a). Considerable uncertainty exists in the evaluation of adverse health outcomes following developmental exposure to organotin compounds in drinking water. The neurotoxicity of trimethyl tin (TMT) has been extensively studied, but DMT and MMT are considered to be relatively non-toxic (Bouldin et al., 1981; Krigman and Silverman, 1984). Concern for neurotoxicity following developmental exposure to MMT was raised by published findings of a deficit on a cognitive task in rat pups (Noland et al., 1982). Noland and coworkers reported that exposure of rats to MMT (12– 120 mg Sn/l in drinking water) or TMT (0.15–1 mg Sn/l in drinking water) during pregnancy and postnatal development altered acquisition of a runway learning task at higher doses, and all dose levels resulted in a deficit of extinction. As such, Noland et al. (1982) could not report a no-observable effect level (NOEL) for MMT. The similar effects of both MMT and TMT on extinction were interpreted as a learning impairment which is consistent with limbic system damage, a known consequence of exposure to TMT; however, confirmatory pathological evaluations were not conducted (Noland et al., 1982). More recent studies have reported neuropathological lesions (loss of neuronal perikarya in hippocampus) following subchronic exposure of adult rats to mixtures of MMT and DMT (US EPA, 2004). Taken together, the available evidence suggests that both adult and developmental exposure to MMT may produce neurotoxicity. The current study was initiated to replicate the original findings of Noland et al. (1982) and to expand the neurological and neuropathological evaluation following developmental exposure to include other standard behavioral tests in the offspring during lactation and as adults. The first study was designed to mimic the Noland study as closely as possible, while the second, follow-up study was designed to be more consistent with the dosing paradigm of standard developmental neurotoxicity studies, such as those described in the US EPA test guidelines (US EPA, 1998b).
2.1. Chemicals 2.1.1. Experiment 1 Monomethyl tin trichloride (MMT; purity 97%; Aldrich, Milwaukee, WI, USA) was dissolved in distilled/deionized water at concentrations of 0, 10, 50 or 245 ppm MMT (0, 5, 25, or 120 mg Sn/l). These doses were chosen to match and extend those used in the Noland et al. (1982) study. The MMT was administered through polypropylene/polyethylene water bottles containing double ball-bearing sipper tubes. A concentrated stock solution was stored at 20 8C for no more than 10 days. Dilutions were made fresh twice a week and bottles refilled. Water bottles were changed and weighed twice weekly, and rats were weighed at the same time. 2.1.2. Experiment 2 Rats were exposed to either 0 or 500 ppm MMT (245 mg Sn/ l) in the drinking water as in Experiment 1. 2.2. Chemical analyses/speciation Analyses were conducted under contract (contract #4D6174-NTNX to Research Triangle Institute) to verify the concentration and speciation of the methyltins in the MMT solutions over days in the water bottles. It was of interest to determine if, and to what extent, interconversions in methyltin speciation occurred under conditions of our study. The stock solution and solutions of the low and high test concentrations were sampled daily for 5 days in water bottles under conditions of the animal exposure. Analysis of MMT, DMT, and trimethyl tin (TMT) were determined using ion chromatography (IC) with inductively coupled plasma mass spectrometer (ICP-MS). In addition, total tin levels were measured using inductively coupled plasma optical emission spectrometer (ICP-OES). Quality control checks were performed after each five to six runs; recoveries of 80–120% were considered acceptable. For MMT, the limit of detection (LOD) was 10 ng Sn/ml, and for DMT and TMT, LOD was 1 ng Sn/ml. The results of these analyses indicated that, as prepared, the high concentration was about 5% higher than the nominal concentration of 120 mg Sn/l, and the low concentration was indeed 5 mg Sn/l. The low concentration showed 4% decrease from days 4 to 5, but the high concentration did not change appreciably. Since bottles were changed every 3–4 days, we concluded there was no significant loss of Sn under our conditions of use. Furthermore, there was no detectable DMT or TMT in the MMT water samples at any time, indicating that if any changes in speciation occurred, it would represent only a negligible fraction of the nominal concentration. 2.3. Animals 2.3.1. Experiment 1 Nulliparous Sprague–Dawley (CD) female rats (Charles River, Raleigh, NC, USA), 53–54-days old, were housed in
V.C. Moser et al. / NeuroToxicology 27 (2006) 409–420
AAALAC-International accredited, temperature- and humidity-controlled rooms (19–21 8C and 40–60%, respectively) on Beta-Chip bedding. As in the Noland study (1982), rats were maintained on a reverse 12 h light:dark cycle (lights on at 1200 h). Upon arrival, 120 females were housed in pairs and provided food (Purina Rodent Chow 5001) and deionized water ad libitum. Following acclimation, and in the two weeks prior to MMT exposure, estrus cycles were monitored through daily vaginal smears. Females (30/dose) were then rearranged and paired according to synchronized estrus cycles (over 4 days) at which time MMT exposure began. Two weeks following initial MMT exposure, the 90-day-old females were bred by placing two receptive females (i.e., latestage proestrus) with a breeder male late in the afternoon and removed the next day at lights-on. Following breeding, females were individually housed and maintained on a high-calorie, high-protein diet formulation (Purina Rodent Chow 5008) throughout gestation and lactation. All females continued to be exposed to MMT through gestation and lactation. Nonpregnant females were euthanized and examined for fetal resorption. When possible, the litters were culled to eight males on postnatal day (PND) 1 (day of birth = PND 0). In a few cases, female pups were used to maintain equal litter sizes. On PND 21, the offspring were weaned and the littermates separated and housed individually on Beta-Chip bedding and provided food (Purina Rodent Chow 5001) and deionized water ad libitum. Only male offspring (one from each litter) were tested in the neurobehavioral tasks. Offspring were weighed at least weekly throughout the study. 2.3.2. Experiment 2 Thirty-five (n = 17 control, n = 18 MMT) timed-pregnant Sprague–Dawley (CD) female rats (Charles River, Raleigh, NC, USA), were received at gestational day (GD) 2 (spermpositive considered gestational day 0). Upon arrival, all females were housed individually as in Experiment 1, except for the light cycle (lights on at 0600 h). MMT exposure began at GD 6 and continued through gestation and lactation, as recommended in the US EPA Developmental Neurotoxicity Test Guidelines (US EPA, 1998b). When possible, litters (n = 17 control, n = 14 MMT) were culled to four males and four females on PND 4. On PND 21, the offspring were weaned and the littermates separated and housed individually on Beta-Chip bedding and provided food (Purina Rodent Chow 5001) and deionized water ad libitum. Both male and female offspring (one from each litter) were tested in neurobehavioral tasks with the exception of the runway task, in which only males were tested. 2.4. Neurobehavioral assessments 2.4.1. Runway learning test The runway was a Plexiglas testing apparatus consisting of an alley (38 cm long, 7.5 cm wide, 10 cm high) and a goal box (17 cm 26 cm 10 cm). Both the testing apparatus and the inter-trial interval holding cage were maintained at 37 8C
411
through the use of water-circulating heating pads. The task was an appetitive learning paradigm in which a food-deprived PND 11 rat pup was trained to negotiate a runway for a dry suckling reward from its anesthetized dam in the goal box (Amsel et al., 1976). Training consisted of alternating reinforced (R) and non-reinforced (N) trials. The pup was placed at the end of the runway facing away from the dam and the time to reach the goal box was recorded. On (R) trials, the pup was given 15 s of dry suckling before being transferred to the holding cage. On (N) trials, the pup was promptly placed in the holding cage upon reaching the end of the runway. If the pup failed to find the dam within the allotted time, the experimenter guided it down the runway for either reinforcement or to be immediately placed in the holding cage. Extinction immediately followed acquisition. The specific parameters for each experiment are outlined below. 2.4.2. Experiment 1 Dams were anesthetized using a drug cocktail equivalent to Chloropent1 (Amsel et al., 1976). Because Chloropent1 was not available at the time of this study, the composition (42.5 mg chloral hydrate, 8.86 mg pentobarbital and 21.2 mg magnesium sulfate, in each ml of sterile aqueous solution containing water, 33.8% propylene glycol and 14.25% ethyl alcohol) was obtained from the Online Database System of FDA approved animal drug products. The dams were dosed with 2 ml/kg body weight i.p. of the Chloropent equivalent approximately 15 min before testing began. When fully anesthetized, the dam was placed in the goal box at the end of the runway to serve as the reinforcer for the pup. Pups (n = 10/treatment group, except n = 11 at 245 ppm) were food-deprived for 10 h prior to testing and tested on PND 11 during their dark cycle. Acquisition consisted of 25 trials with (R) and (N) trials alternating on every other trial, and the maximum time was set at 120 s. Between each trial, the pup was placed in the holding cage for 15 s. Extinction began on the 26th trial. During this time, the pup was removed to the holding cage without gaining access to the dam on every trial. The maximum time was set at 100 s. Criterion to extinction was two consecutive trials with latencies of 100 s. After this criterion was reached, the pup was no longer tested; however, all pups were tested until they met criterion, regardless of the number of trials required. 2.4.3. Experiment 2 Dams were anesthetized using Nembutal1 sodium solution (50 mg/ml) (Abbott Laboratories, IL, USA). The dams were dosed with 2 to 2.2 ml/kg body weight i.p. of Nembutal1. As before, testing began when the dam was fully anesthetized. In this paradigm, pups (n = 17 control, n = 14 MMT) were food-deprived for 8 h prior to testing and tested on PND 11 during their light cycle. (Pilot studies had indicated that 8 h deprivation was not different from 10 h in terms of motivation and task performance.) Testing began with a preliminary training session of five initial training (R) trials, followed by a 2 min retention interval in the holding cage. There were then 25 acquisition trials in which (R) and (N) trials alternated in blocks
412
V.C. Moser et al. / NeuroToxicology 27 (2006) 409–420
of five trials, beginning and ending with 5(R) trials. The intertrial interval was 8 s. During acquisition, the pup was permitted 100 s before the trial ended and, if unsuccessful, it was guided to the dam. Extinction began on the 26th trial. Criterion to extinction was one trial with a latency of 100 s, and a maximum of 10 extinction trials were run.
In Experiment 1, only males were tested in the water maze (n = 9 control, n = 10 at 10 ppm, n = 8 at 50 ppm, and n = 7 at 245 ppm) whereas both males and females were tested in Experiment 2 (n = 16 control, n = 14 MMT 500 ppm).
2.5. Motor activity
For neuropathological evaluations, male rats (n = 6–7/dose at PND 1, n = 7–9/dose at PND 12, n = 7–11/dose at PND 22, and n = 7–10/dose at adult age, about 85–90-days old) were deeply anesthetized with pentobarbital and perfused via the left ventricle with buffered 4% formaldehyde:0.1% gluteraldehyde. Sagittal blocks of tissue were embedded in paraffin and sectioned to include all major structural landmarks of the brain in each section (e.g., olfactory bulb, striatum, cerebral cortex, hippocampus, thalamus, hypothalamus, brainstem, cerebellum). Twenty-four homologous sections of the brain from each rat, from each age, at each dose, were stained with hematoxylin and eosin. Brains from all control and high-dose rats, at all ages, were evaluated under contract by a qualified pathologist who was blind to the treatment (contract #68-D-02-102 to Experimental Pathology Laboratories). Step-down assessments, i.e., evaluation of the lower dose groups, were only conducted in the adult rats due to findings in the high-dose group; no abnormalities were observed at the younger ages. Any changes observed were subjectively scored in both number as well as nature of the change as follows: (1) minimal; (2) slight/mild; (3) moderate; (4) moderately severe; (5) severe.
Motor activity data were collected using an automated figure-eight chamber (Reiter, 1983). Activity was measured as interruptions (counts) of any of the eight photocell beams distributed around the maze. Counts were recorded over 5 min intervals for a 30 min testing session. In Experiment 1, motor activity was assessed in males at PNDs 13, 17 and 21 (n = 10 control, n = 9 at 50 ppm, n = 11 at 10 ppm, and n = 10 at 245 ppm). In Experiment 2, only PND 17 male and female offspring (one male and one female from each litter; n = 17 control, n = 14 MMT) were tested. 2.6. Morris water maze Spatial learning and memory was evaluated in adults (starting at about 70 days of age) using a Morris water maze (Morris, 1981) which consisted of a round galvanized steel tank 140 cm in diameter and filled with tap water (24–27 8C) to a depth of approximately 44 cm. A round, Plexiglas escape platform (9 cm in diameter) was positioned 2 cm beneath the surface of the water. The water was rendered opaque using black Tempra1 paint and was changed daily. The trials were videotaped and the image digitized for computer analysis using maze-tracking software (HVS Image, Ormond Cresent, Hampton, UK). Dependent variables included swim speed, latency and path length to find the platform, and time spent in the outer edge of the tank. The tank was digitally divided into three concentric zones (annuli) of equal width, and time spent in each zone was also analyzed. For the spatial training, the platform was placed in a fixed quadrant. Acquisition occurred during two trials each day over a total period of 9 days (5 days training, 2 days off, 4 days training). The rat was placed in the water at one of four starting points which varied such that the rat never started in the same place within 2 days. The trials continued until the rat mounted the platform, or for a maximum of 60 s. If the rat did not find the platform within the 60 s, it was guided there by the observer. After 15 s on the platform, the rat was placed in a holding cage for 5 min before the second trial began. On the 10th day, a probe trial was conducted in which the platform was removed and the rat’s tendency to search in the correct quadrant (where the platform had been) was measured over 60 s. The Gallagher proximity score, a measure of cumulative distance from the platform (Gallagher et al., 1993), and percent total path length within each quadrant, were used to assess quadrant bias. Following the memory probe trial, a visible target task (cued trial) was conducted using a raised platform of a contrasting color to confirm that the tested animals were not visually impaired.
2.7. Neuropathology (Experiment 1 only)
2.8. Neurochemical assessment and brain weight (Experiment 1 only) Male rats were decapitated under CO2-induced anesthesia on PND 1 (at culling), PND 22, and adult (80–90-days old). At culling, all pups within a litter were combined, and at older ages all rats came from different litters. Brains were removed from the calvaria, free of meninges, and weighed. The PND 22 and adult brains were dissected free-hand into the following regions: brainstem, neocortex, hippocampus and cerebellum (see White and Barone, 2001). Each region was immediately frozen in 2-methyl butane (99.5%, HPLC grade; Milwaukee, WI) on dry ice for 30 s, weighed and stored at 80 8C (White and Barone, 2001). Brain tissues were collected from Experiment 1 only (n = 4–9/dose/region). DNA fragmentation was used as an indicator of apoptotic cell death, and was quantified using a Cell Death ELISA (Roche Applied Science, Indianapolis, IN) procedure. Modifications to the procedure in the kit have been described previously (White and Barone, 2001). Specifically, the kit has been adapted for use with intact tissue and validated with both fresh and frozen brain tissue, and the results have been corroborated qualitatively by agarose gel and TUNEL data (White and Barone, 2001). In short, the enzyme-linked immunosorbent assay (ELISA) uses antibodies to bind fragmented DNA characteristic of apoptotic cell death. The bound fragments (i.e., nucleosomes) are then quantified photometrically.
V.C. Moser et al. / NeuroToxicology 27 (2006) 409–420
413
2.9. Statistical analyses Continuous data (e.g., body weight, water intake, activity counts, etc.) were analyzed using a general linear model ANOVA (SAS, Cary, NC). Extinction in the runway test was analyzed with a survival model for censored data (SAS), and count data (e.g., number of pups learning, pregnancy rate) were compared using Fisher’s exact test (SAS). When the same rat was used in repeated tests (e.g., repeated motor activity testing, body weights over time), the analyses included time as a repeated factor. For apoptosis data, all brain region values were compared to their respective controls, and arcsine transformed prior to statistical analyses. Following a significant overall analysis, Dunnett’s t-test was used to compare dose groups with the control. In all cases, resulting probability values <0.05 were considered significant. Fig. 1. Maternal body weights in Experiment 1 throughout exposure to 0, 10, 50, or 245 ppm MMT.
3. Results 3.1. Maternal fluid intake and weights
day-by-intake interaction was followed with step-down analyses, which indicated that only the intake measured 3 days post-parturition was not different. During gestation, MMT consumption was about 80–88% of control levels, and during lactation, 82–88% of control. Despite the lowered intake, body weight was not different in the treated group (Table 1). Table 1 also includes the calculated MMT intake during gestation and lactation. As in Experiment 1, the intake during lactation nearly doubled.
3.1.1. Experiment 1 Fluid intake increased somewhat during gestation (about 30–40%), and was much higher (more than double) during lactation. There were, however, no differences in fluid intake across treatment groups at any time during the study. Likewise, there were no treatment-related differences in body weight throughout exposure (Fig. 1). Table 1 lists the fluid intake, weight, and calculated MMT intake (mg/kg/day) during the three phases of the study: pre-breeding, gestation, and lactation. Predictably, MMT intake was much greater during lactation.
3.2. Reproductive/developmental parameters 3.2.1. Experiment 1 Despite the timed breeding, pregnancy rate was very low with only 44 of 120 (37%) rats delivering litters (10, 11, 11, and
3.1.2. Experiment 2 There was a significant depression of fluid intake across almost all days of treatment with MMT 500 ppm. A significant
Table 1 Fluid intake (average, and range of means, ml/day), body weight (average, and range of means, g), and calculated MMT intake (mg/kg/day) for each concentration of MMT (ppm) throughout exposure Concentration
Intake
Weight
MMT
0 10 50 245
28.9 26.5 28.2 25.7
(27.0–30.4) (24.5–27.7) (27.1–29.1) (23.6–26.8)
266.8 270.6 268.0 270.0
(253.0–271.1) (254.9–274.6) (255.5–271.6) (256.0–274.3)
– 1.0 5.3 23.3
Gestation
0 10 50 245
39.9 36.1 43.4 36.5
(36.1–48.1) (30.7–43.5) (35.1–50.6) (30.8–44.7)
330.8 330.6 336.3 336.7
(274.5–418.1) (280.8–414.5) (276.4–420.3) (278.4–422.3)
– 1.1 6.5 25.7
Lactation
0 10 50 245
68.4 65.2 75.0 62.8
(48.2–92.9) (44.7–86.8) (51.4–95.9) (47.8–82.1)
339.4 336.8 343.9 347.5
(336.2–346.3) (322.6–348.3) (332.3–356.6) (335.8–352.3)
– 1.8 10.6 41.6
0 500
38.2 (33.1–44.0) 32.4 (28.5–37.4)
284.7 (231–346.9) 290.5 (233.9–356)
– 55.8
0 500
71.7 (40.8–96.3) 60.8 (36.9–85.1)
314.7 (304.0–328.8) 322.5 (310.9–333.8)
– 94.3
Experiment 1 Pre-breeding
Experiment 2 Gestation Lactation
414
V.C. Moser et al. / NeuroToxicology 27 (2006) 409–420
12 litters for control, 10, 50, and 245 ppm, respectively). We typically find that most of our Sprague–Dawley rats deliver within a 24 h period which we specify as PND 0. In the present study, however, 11 of the 44 (25%) rats delivered later than that time window. Necropsy of all non-pregnant rats revealed resorptions in only two control rats and one rat from the lowdose group. Incidence of pregnancy, late delivery, and resorptions were not statistically different across treatment groups. 3.2.2. Experiment 2 All of the timed-pregnant females in the control group delivered, but two in the MMT group did not. These rats were not evaluated for implantation sites. All of the deliveries occurred when expected. In the MMT group, one litter consisted of all females and was not used, and another litter was killed by the dam shortly after birth. 3.3. Offspring number and growth 3.3.1. Experiment 1 There were no significant differences in the number of pups per litter across treatment groups, as follows (mean number per litter S.E.): control, 12.5 1.6; 10 ppm, 15.2 0.6, 50 ppm, 13.1 1.3, and 245 ppm, 13.4 1.5. Likewise, there were no treatment-related differences in the number of males and females per litter. Litter birth weights and body weights across time were similar across treatment groups throughout the entire study (data not shown). In addition, there were no differences in weights of the pups selected for each behavioral test. 3.3.2. Experiment 2 There were no differences in the number of pups per litter (control, 11.9 0.4; MMT, 12.2 0.8), or the sex ratio within the litters. Body weight changes during the lactation period showed a significant dose-by-sex interaction, but step-down analyses showed that on 1 day only (PND 11), male and female pups in the control group were different by about 4 g; this was considered to be within biological variability. There were no treatment effects on body weight after weaning. 3.4. Runway testing 3.4.1. Experiment 1 Almost all of the pups learned to traverse the runway to the goal box. Using a criterion of at least one latency (time to reach the goal box) less than the maximum time of 120 s, only one pup (high dose) did not learn the task. Mean latencies (excluding the data from the one non-learner) across acquisition trials are shown in Fig. 2. While highly variable, the latency data did not reveal a significant treatment-related difference. The time to extinction was also quite variable, with one pup (low-dose) requiring 46 trials to meet the criterion of two consecutive latencies of 100 s. Fig. 2 presents the cumulative percent of each treatment group reaching criterion at each trial.
Fig. 2. Performance in the runway task on PND 11 in Experiment 1. (Top) Latency to traverse the runway during each reinforced (R) or nonreinforced (N) trial which strictly alternated. (Bottom) Percent of each group reaching extinction (two consecutive latencies of 100 s) during consecutive nonreinforced trials.
The median number of trials to reach criterion for each dose group was as follows: control, 17; 10 ppm, 15.5, 50 ppm, 20.5; 245 ppm, 20.5. 3.4.2. Experiment 2 Fewer pups were able to learn the different paradigm used in the second study. One control and five treated pups did not meet the criterion of having at least one latency less than 100 s during acquisition; however, this difference did not attain statistical significance (Fisher’s exact p = 0.067). Furthermore, one pup in each group did not extinguish within the 10 trials. A different pattern of responding was elicited by the procedure of batched reinforced (R) and non-reinforced (N) trials. Fig. 3 shows that, in general, latencies became shorter during the R trials, and longer during the N trials. This pattern indicates repeated acquisition and extinction, and was most evident in the last two sets of trials. Trend analyses were conducted on each set of trials to determine if the changes within each block were significant (either increasing or decreasing). For the control group, the last two blocks showed significant slopes in the latencies (mean S.E. slopes for the last N-trials, 8.7 3.0; last R trials, 7.8 3.3). On the other
V.C. Moser et al. / NeuroToxicology 27 (2006) 409–420
415
and 17. By PND 21, habituation was evident. There were no group differences on any of these measures. 3.5.2. Experiment 2 Motor activity testing was only conducted on PND 17, at which time there were no differences between the control and MMT-treated groups (mean S.E. for each group: control, 99.6 7.0 total counts; 500, 108.3 8.5). Habituation was evident in both dose groups. 3.6. Morris water maze 3.6.1. Experiment 1 All rats learned the location of the hidden platform over the two weeks of training, as evidenced by decreasing latency to find the platform with repeated training (Fig. 4). Daily blocks and training trials showed significant changes over testing. Latency, path length, swim speed, and most spatial distribution variables were not different across treatment groups. The only significant finding was increased time in the middle zone in the 50 ppm group on day 6 only; this was not supported by other findings and is probably not biologically relevant.
Fig. 3. Performance in the runway task on PND 11 in Experiment 2. (Top) Latency to traverse the runway during each reinforced (R) or nonreinforced (N) trial which were grouped with five trials each. (Bottom) Percent of each group reaching extinction (one latency of 100 s) during consecutive nonreinforced trials.
hand, the MMT group showed significant slopes during the third and fourth blocks (middle R trials, 12.0 2.9; last N trials 11.0 4.0), but not the fifth (last R trials, 6.0 5.3, p = 0.074). Using ANOVA to evaluate group differences, only one trial showed a significant difference between the groups (fourth trial of the middle R-trials). Cumulative percentages of the groups reaching criterion (one trial with latency of 100 s) across each extinction trial are shown in Fig. 3; there were no significant differences. The median number of trials to reach criterion were: control, 4 trials; MMT, 5.5 trials. 3.5. Motor activity 3.5.1. Experiment 1 Total motor activity counts during 30-min sessions were lowest at PND 13 (means ranging from 20.3 to 31.2), and increased over PND 17 (means 46.5 to 72.7) and 21 (means 98.7 to 120.3). Analysis of the within-session activity (in 5 min intervals) showed no habituation during the session at PNDs 13
Fig. 4. Performance in the Morris water maze in adults in Experiment 1. (Top) Latency to find the platform in daily training trials, with two trials each day. (Bottom) Time in each quadrant (Q4 = quadrant where platform had been) during the probe trial (no platform present).
416
V.C. Moser et al. / NeuroToxicology 27 (2006) 409–420
During the memory probe, all rats spent the most time searching in the quadrant where the platform had been located (Fig. 4). There were no differences in quadrant bias, Gallagher index, or spatial distribution variables. For the visual function test, there were no differences in the latency to reach a raised, visible platform, as follows (mean S.E.): control, 27.4 4.7 s; 10 ppm, 25.3 2.6 s; 50 ppm, 31.4 7.0 s; 245 ppm, 24.0 4.4 s. 3.6.2. Experiment 2 As in Experiment 1, all rats learned the spatial task, indicated by significant changes in daily blocks and individual training trials. There were, however, no interactions with treatment or sex on any endpoint. Latency to find the platform is shown in Fig. 5, and other dependent variables showed similar patterns. Likewise, there were no differences on any measures during the memory probe, or during the visual probe (visible platform). 3.7. Brain weight (Experiment 1 only) There was a marginal effect on brain weight, although the overall dose effect at p < 0.07 did not reach the criterion for statistical significance. There was no interaction of treatment and age, and collapsed across ages, the high dose group showed a 5% decrease in weight. 3.8. Neuropathology (Experiment 1 only) No histological alterations were observed in the brains of MMT-treated offspring killed on PND 1, 12, or 22. There were, however, changes in the brains of the offspring evaluated as adults. The cerebral cortical lesion was characterized by 2– 4 mm diameter, round vacuoles of varying size, in the gray matter neuropil, localized to the region of the orbital cortex. Furthermore, the focal location of the lesion was very similar across animals, supporting the conclusion of a treatmentrelated effect. Myelin was not affected. Fig. 6 depicts representative lesions in a control and a low-dose rat.
Fig. 5. Performance in the Morris water maze in adults in Experiment 2. (Top) Latency to find the platform in daily training trials, with two trials each day. (Bottom) Time in each quadrant (Q4 = quadrant where platform had been) during the probe trial (no platform present).
Two of eight (25%) adult high-dose offspring had vacuolation scored as ‘‘slight/mild’’ (‘2’), whereas one of seven (14%) of the mid-dose group, and three of ten (30%) lowdose rats, had similar vacuolation scored as ‘‘minimal’’ (‘1’).
Fig. 6. High (40) magnification of the area of the cerebral cortex (H&E stain) from adults in Experiment 1. (A) Female control rat, no neuropil vacuolation present. (B) Female rat exposed to 10 ppm MMT, with minimal neuropil vaculation present. Bar = 25 mm.
V.C. Moser et al. / NeuroToxicology 27 (2006) 409–420
Fig. 7. DNA fragmentation from cell death ELISA on cortical tissues from PND 22 and adults in Experiment 1.
There were no such findings in control rats. Although there was no increase in incidence or more widespread localization of this lesion with increasing dose, the change was scored as slightly greater severity in the high-dose offspring (‘2’s) than that of the lesions in the lower dose groups (‘1’s). There were no histopathological findings in any other major brain region other than this specific area of the cerebral cortex. These adult rats had been tested previously in the Morris water maze. Review of the water maze data for the rats showing vacuolation revealed that their performance was not different from the unaffected rats. Thus, no within-individual correlations between vacuolation and behavioral performance could be made. 3.9. Apoptosis (Experiment 1 only) The data for PND 22 brainstem were not analyzed due to insufficient sample size. Data from PND 22 and adult cortex are presented in Fig. 7, showing no treatment-related difference in DNA fragmentation. Similarly, the PND 22, and adult, neocortex and hippocampus remained unaffected by perinatal MMT exposure, as did the adult brainstem (data not shown). 4. Discussion There were very few effects of MMT in offspring exposed during gestation and lactation. The highest concentration (500 ppm, providing intake of 56–94 mg/kg/day) lowered fluid intake, and we felt that using even higher concentrations would be unreasonably high as well as produce maternal toxicity. There were no neurobehavioral effects on any endpoint, at any time, nor were there developmental changes as measured by apoptosis. The presence of vacuolation in the neuropil was observed only in treated rats, but the incidence was minimal (1– 3 rats per treatment group). The overall effect on brain weight was also minimal, and neither of these findings could be related to other biological effects. Unfortunately, neuropathology and brain weights were not evaluated in the second study, which could have confirmed or disputed the findings.
417
One impetus for this study was to replicate and expand the findings of Noland et al. (1982) of increased trials to extinction in all treated groups. To that end, we attempted to duplicate as many details as possible in Experiment 1. First, our choice of concentrations (5, 25, 120 ppm Sn) extended the dose-range used by Noland (12, 40, 120 ppm Sn). Although the source of chemical was different, the purity was verified (but not stated) in the Noland study, and ours was obtained at 97% purity. The runway was built to the specifications described in Taylor et al. (1982), which was referenced in the Noland paper. Other variables that were the same included: rat strain, age, light cycle, mating and culling procedures, frequency of water change, and runway paradigm. The only variable we know to be different was bedding (sawdust versus our hardwood chips). It should be noted that even though Sprague–Dawley rats were obtained from Charles River in both studies, it is unknown which facility was used by Noland et al. (1982), and in the early 1980s the Raleigh facility (used in the present experiment) was not open. As such, it is possible that the rats were actually quite different, as genetic drift in outbred rats as well as differences in suppliers are well-documented (Festing, 1979, 1987; Glick et al., 1986; Kacew et al., 1995). Furthermore, there could be genetic changes among Sprague–Dawley rats that resulted from a comprehensive restructuring (i.e., International Genetic Standard) of breeding programs at Charles River Laboratories in 1992 (White and Lee, 1998). The current breeding program was instituted to minimize the degree of variation associated with genetic drift. Thus, it is likely that disparity among Sprague–Dawley animals produced today, in comparison to 20–25 years ago, may be partially responsible for the observed behavioral differences. There could also be other factors that were not reported in the original paper that could have contributed to our different results. Of course, it could simply be that the findings were not sufficiently robust to be replicated. The runway was originally developed by Amsel and colleagues in the 1970s to demonstrate the ontogeny of learning and extinction in preweanling rats, using dry suckling (anesthetized dam) as the reward (Amsel et al., 1976). Since the original publication, many experimental and parametric manipulations of the runway task have been described (e.g., Amsel and Chen, 1976; Amsel et al., 1977; Chen and Amsel, 1980; Highfield et al., 1996; Lilliquist et al., 1999; Lobaugh et al., 1985, 1989; Nair and Gonzalez-Lima, 1999; Nair et al., 2001; Stanton, 1982; Wigal et al., 1988a,b). PND 11 pups can learn the strict alternation (reinforced/non-reinforced), and show a go-no-go response pattern, but only after many more training trials than presented in the current study (e.g., 60–100 trials compared to 25; Lilliquist et al., 1999; Lobaugh et al., 1985; Stanton, 1982). In Experiment 2, the schedule of reinforcement was altered to demonstrate that we could control the behavior and be more consistent with Amsel’s early work (e.g., Amsel et al., 1976, 1977). After the first two to three blocks of five trials, it was evident that responding (latencies) got faster during the reinforced trials, but slower during the non-reinforced trials. There was substantial variability in the present studies, and while we used sample sizes typically used by Amsel’s group (n = 10–15), it is difficult to compare
418
V.C. Moser et al. / NeuroToxicology 27 (2006) 409–420
variability across laboratories since almost no other papers present estimates of variability in the figures or text. Despite the considerable literature on this test, there are relatively few uses of the test for evaluating chemical effects, and again, most of these are from Amsel’s laboratory. Gestational ethanol exposure produced changes in the runway extinction when using a partial reinforcement, but not single patterned alternation schedule of reinforcement (e.g., Wigal et al., 1988a,b). Direct dosing of MK-801 to preweaning rat pups impaired learning in this task, but only at longer inter-trial intervals (Highfield et al., 1996). The only non-Amsel papers we could find which used this specific apparatus came from Taylor’s laboratory; these included a perinatal study of lead acetate (Taylor et al., 1982) and the already-described study of MMT (Noland et al., 1982). A comparison of our control data to both of these studies showed that, while the acquisition curves appeared to be quite similar, the extinction curve for our control (as well as treated groups) was much like their MMT and TMT curves. Indeed, 50% of their treated groups had reached the extinction criterion by 15 trials (their cut-off), and half of our rats reached that level after 15–20 trials, thus showing a difference in control performance. In contrast to the lack of database for the runway, the other behavioral tests included in these studies have been widely used for developmental neurotoxicity evaluations. Motor activity is an apical test of neuronal function, and an evaluation of its preweanling ontogeny reflects maturation of the nervous system. This test is required in the current US EPA Developmental Neurotoxicity Test Guidelines (US EPA, 1998b). The Morris water maze was developed as a test of visual spatial navigation with both a learning and memory component. It can demonstrate the rodent’s ability to learn the position of a hidden platform using extra-maze spatial cues. There are many papers describing chemical effects on this behavior following acute, repeated, and developmental exposures (e.g., Brandeis et al., 1989; D’Hooge and DeDeyn, 2001; Jett et al., 1997; Kuhlmann et al., 1997; McNamara and Skelton, 1993; Moser et al., 2001). On these more standard and accepted behavioral assays, MMT had no effect. Apoptosis is widespread in both the fetal and postnatal CNS, and any disturbance in either the progression or total amount of apoptosis could be detrimental for a developing organism (e.g., learning and memory impairments, or progression of neurodegenerative diseases) (Bennett et al., 1998; Leist and Nicotera, 1998; White and Barone, 2001). Moreover, the process of apoptosis is closely linked, and potentially co-occurs, with other developmental processes such as proliferation and migration (Liesi, 1997; White and Barone, 2001). Using the same brain regions examined in the current study, the cell death ELISA procedure has been successfully used to detect changes in apoptosis following perinatal exposure to another organotin (dibutyltin, DBT) (Jenkins et al., 2004). The lack of effect observed on DNA fragmentation with MMT exposure agrees with the in vitro MMT data (Jenkins et al., 2004). Furthermore, the lack of effect in vivo, even in a similar brain region, does not obviate the pathological findings of vacuolation, since these are distinct processes.
Neuropathological lesions and brain weight changes are considered indicators of neurotoxicity (US EPA, 1998c); however, the MMT effects described here were relatively small. The localization of the vacuoles was consistent across rats, and there was a total lack of such findings in any control rats or at any earlier ages. While the incidence did not increase with dose, the severity of change did show a slight dose–response (only high-dose rats were scored as ‘slight/mild’ changes, while lesions in the lower dose groups were ‘minimal’). All this information taken together suggests that this finding is not an artifact. Whether the vacuoles were located in neuronal or glial tissue, or the nature of the vacuoles, cannot be fully determined with light microscopy. Clearly, additional studies of this outcome are warranted and would require greater resolution using plastic embedding and transmission electron microscopic examination. The extent of parallels between specific regions of prefrontal cortex in rat and human, such as the orbitofrontal cortex, is not clear (Dalley et al., 2004), and it is further unclear whether the area of these vacuoles corresponds to the orbitofrontal cortex area. This region of cortex is sensitive to chemical disruptions, and has been implicated in motivation, decision making, and working memory (Dalley et al., 2004; Schoenbaum and Setlow, 2001). The behavioral tests used herein do not tap into these functions, which could help explain the lack of behavioral effects observed. In humans, prefrontal cortical damage spares basic cognitive function while interfering with emotional signals (Bechara, 2004). On the other hand, the minimal degree of vacuolation may have no disruptive influence on these functions. Additional studies using other cognitive tasks could pursue and resolve these possibilities. It is clear, however, that the nature of these findings differ from lesions produced by adult or postnatal exposure to TMT or triethyltin (TET) (O’Shaughnessy and Losos, 1986). Acute TMT exposure produces neuronal cell death characterized by homogeneous eosinophilic cytoplasm and nuclear pyknosis, followed by gliosis including microglial activation and astrogliosis. The limbic system is most severely affected, but other affected areas include the cerebral cortex and brainstem. While TET produces neuronal vacuolization, another prominent feature is spongy myelin edema (O’Shaughnessy and Losos, 1986), which was not observed here. Despite the developmental exposure paradigm, no morphological changes were observed in younger rats, which have been shown to be more sensitive and have earlier onset of pathological indices to TMT and TET (Barone, 1993; Freeman et al., 1994). While we showed that MMT did not convert to other methyltin species in the rats’ water bottles, it is possible that such interconversions could take place in the body. We could find no data addressing this, but we feel it is unlikely that MMT converted to TMT since the pathology we observed was quite unlike that produced by TMT (see above). It is interesting to note, however, that similar findings were obtained with dimethyl tin (unpublished results). In summary, two separate developmental neurotoxicity studies of MMT indicated a lack of effects on the neuro-
V.C. Moser et al. / NeuroToxicology 27 (2006) 409–420
behavioral endpoints that were tested, no changes in the apoptosis assay, but a small yet reliable appearance of brain changes. This could indicate that the behaviors included were not sensitive to the pathological changes, or that the pathology was not sufficient to disrupt the functioning of the test subject. We could not replicate the effect of MMT on cognitive function. This information adds to the scant database of developmental effects of organotins. Acknowledgements The authors wish to acknowledge the invaluable assistance of Megan Newcomb (NC State University School of Veterinary Medicine), Joan Hedge, Rene´e Marshall, and Brian Robinette (all of Neurotoxicology Division, US EPA) for their help in the technical conduct of these labor-intensive tests. We also thank Drs. Lori White and Scott Jenkins for their help with the perfusions and apoptosis assay, and Drs. Karl Jensen and John Lipscomb who reviewed an earlier version of the manuscript. Dr. K. Ehman was supported by the EPA/UNC Toxicology Research Program, Training Agreement #T829472. References Amsel A, Chen JS. Ontogeny of persistence: immediate and long-term persistence in rats varying in training age between 17 and 65 days. J Comp Physiol Psychol 1976;90:808–20. Amsel A, Burdette DR, Letz R. Appetitive learning, patterned alternation, and extinction in 10-d-old rats with non-lactating suckling as reward. Nature 1976;262:816–8. Amsel A, Letz R, Burdette DR. Appetitive learning and extinction in 11-day-old rat pups: effects of various reinforcement conditions. J Comp Physiol Psychol 1977;91:1156–67. Barone S Jr. Developmental differences in neural damage following trimethyltin as demonstrated with GFAP immunohistochemistry. In: Johannessen JN, editor. Markers of neuronal injury and degeneration. New York: New York Academy of Science; 1993. p. 306–17. Bechara A. The role of emotion in decision-making: evidence from neurological patients with orbitofrontal damage. Brain Cogn 2004;55: 30–40. Bennett SA, Tenniswood M, Chen JH, Davidson CM, Keyes MT, Fortin T, et al. Chronic cerebral hypoperfusion elicits neuronal apoptosis and behavioral impairment. NeuroReport 1998;9:161–6. Bouldin TW, Goines ND, Bagnell CR, Krigman MR. Pathogenesis of trimethyltin neuronal toxicity: ultrastructural and cytochemical observations. Am J Pathol 1981;104:237–49. Brandeis R, Brandys Y, Yehuda S. The use of the Morris water maze in the study of memory and learning. Int J Neurosci 1989;48:29–69. Chen JS, Amsel A. Learned persistence at 11–12 days but not at 10–11 days in infant rats. Dev Psychobiol 1980;13:481–91. Dalley JW, Cardinal RN, Robbins TW. Prefrontal executive and cognitive functions in rodents: neural and neurochemical substrates. Neurosci Biobehav Rev 2004;28:771–84. D’Hooge R, DeDeyn PP. Applications of the Morris water maze in the study of learning and memory. Brain Res Rev 2001;36:60–90. Donard OF, Quevauviller P, Bruchet A. Tin and organotin speciation during wastewater and sludge treatment processes. Water Res 1993;27: 1085–1089. Festing MFW. Properties of inbred strains and outbred stocks, with special reference to toxicity testing. J Toxicol Environ Health 1979;5: 53–68. Festing MFW. Genetic factors in toxicology: implications for toxicological screening. CRC Crit Rev Toxicol 1987;18:2–36.
419
Forsyth DS, Dabeka R, Sun WF, Dalglish K. Speciation of organotins in poly(vinyl chloride) products. Food Addit Contam 1993;10:531–40. Freeman JH, Barone S Jr, Stanton ME. Triethyltin produces neural damage and cognitive deficits in developing rats that depend on age of exposure. Brain Res 1994;634:85–95. Gallagher M, Burwell R, Burchinal M. Severity of spatial learning impairment in aging: development of a learning index for performance in the Morris water maze. Behav Neurosci 1993;107:618–26. Glick SD, Shapiro RM, Drew KL, Hinds PA, Carlson JN. Differences in spontaneous and amphetamine-induced rotational behavior, and in sensitization to amphetamine, among Sprague–Dawley derived rats from different sources. Physiol Behav 1986;38:67–70. Highfield DA, Nixon K, Amsel A. The NMDA antagonist MK-801 affects nonspatial learning in preweanling rats. Behav Neurosci 1996;110: 300–4. Jenkins SM, Ehman K, Barone S Jr. Structure-activity comparison of organotin species: dibutyltin is a developmental neurotoxicant in vitro and in vivo. Dev Brain Res 2004;151:1–12. Jett DA, Kuhlmann AC, Farmer SJ, Guilarte TR. Age-dependent effects of developmental lead exposure on performance in the Morris water maze. Pharmacol Biochem Behav 1997;57:271–9. Kacew S, Ruben Z, McConnell RF. Strain as a determinant factor in the differential responsiveness of rats to chemicals. Toxicol Pathol 1995;23: 701–15. Krigman MR, Silverman AP. General toxicology of tin and its organic compounds. Neurotoxicology 1984;5:129–40. Kuhlmann AC, McGlothan JL, Guilarte TR. Developmental lead exposure causes spatial learning deficits in adult rats. Neurosci Lett 1997;233: 101–104. Liesi P. Ethanol-exposed central neurons fail to migrate and undergo apoptosis. J Neurosci Res 1997;48:439–48. Leist M, Nicotera P. Apoptosis, excitotoxicity, and neuropathology. Exp Cell Res 1998;239:183–201. Lilliquist MW, Nair HP, Gonzalez-Lima F, Amsel A. Extinction after regular and irregular reward schedules in the infant rat: influence of age and training duration. Dev Psychobiol 1999;34:57–70. Lobaugh NJ, Bootin M, Amsel A. Sparing of patterned alternation but not partial reinforcement effect after infant and adult hippocampal lesions in the rat. Behav Neurosci 1985;99:46–59. Lobaugh NJ, Greene PL, Grant M, Nick T, Amsel A. Patterned (single) alternation in infant rats after combined or separate lesions of hippocampus and amygdale. Behav Neurosci 1989;103:1159–67. McNamara RK, Skelton RW. The neuropharmacological and neurochemical basis of place learning in the Morris water maze. Brain Res Rev 1993;18:33–49. Mersiowsky I, Brandsch R, Ejlertsson J. Screening for organotin compounds in European landfill leachates. J Environ Qual 2001;30:1604–11. Morris RGM. Spatial localization does not require the presence of local cues. Learn Motiv 1981;12:239–60. Moser VC, Shafer TJ, Ward TR, Meacham CA, Harris MW, Chapin RE. Neurotoxicological outcomes of perinatal heptachlor exposure in the rat. Toxicol Sci 2001;60:315–26. Nair HP, Gonzalez-Lima F. Extinction of behavior in infant rats: development of functional coupling between septal, hippocampal, and ventral tegmental regions. J Neurosci 1999;19:8646–9655. Nair HP, Berndt JD, Barrett D, Gonzalez-Lima F. Maturation of extinction behavior in infant rats: large-scale regional interactions with medial prefrontal cortex, orbitofrontal cortex, and anterior cingulated cortex. J Neurosci 2001;21:4400–7. Noland EA, Taylor DH, Bull RJ. Monomethyl and trimethyltin compounds induce learning deficiencies in young rats. Neurobehav Toxicol Teratol 1982;4:539–44. O’Shaughnessy DJ, Losos GJ. Peripheral and central nervous system lesions caused by triethyl- and trimethyltin salts in rats. Toxicol Pathol 1986;14: 141–8. Reiter LW. Chemical exposures and animal toxicity: utility of the figure-eight maze. In: Hayes AW, Schell RC, Miya TS, editors. The science and practice of toxicology. New York: Elsevier; 1983. p. 73–84.
420
V.C. Moser et al. / NeuroToxicology 27 (2006) 409–420
Sadiki A-I, Williams DT. A study on organotin levels in Canadian drinking water distributed through PVC pipes. Chemosphere 1999;38:1541–8. Sadiki A-I, Williams DT, Carrier R, Thomas B. Pilot study on the contamination of drinking water by organotin compounds from PVC materials. Chemosphere 1996;32:2389–98. Schoenbaum G, Setlow B. Integrating orbitofrontal cortex into prefrontal theory: common processing themes across species and subdivisions. Learn Memory 2001;8:134–47. Stanton ME. Performance of 11- and 14-day-old rats on a working memory problem. Behav Neural Biol 1982;36:304–10. Taylor DH, Noland EA, Brubaker CM, Crofton KM, Bull RJ. Low level lead (Pb) exposure produces learning deficits in young rat pups. Neurobehav Toxicol Teratol 1982;4:311–4. U.S. Environmental Protection Agency (US EPA). US environmental protection agency announcement of the drinking water contaminant candidate list. Fed. Register 63; March 2, 1998a, p. 10274–87. U.S. Environmental Protection Agency (US EPA). Developmental neurotoxicity study, health effects test guidelines, OPPTS 870.6300, EPA 712-C-98239; 1998b. http://www.epa.gov/opptsfrs/publications/OPPTS_Harmonized/ 870_Health_Effects_Test_Guidelines/Series.
U.S. Environmental Protection Agency (US EPA). Guidelines for neurotoxicity risk assessment, EPA/630/R-95/001F; 1998c. http://cfpub.epa.gov/ncea/raf/ recordisplay.cfm?deid=12479. U.S. Environmental Protection Agency (US EPA). Memo re: TSCA 8(e) Submission; 2004. http://www.epa.gov/oppt/tsca8e/doc/8ehq/2004/february04/ 8ehq-0104-15511a.pdf. White LD, Barone S Jr. Qualitative and quantitative estimates of apoptosis from birth to senescence in the rat brain. Cell Death Differ 2001;8:345–56. White WJ, Lee CS. The development and maintenance of the Crl:CD1(SD)IGS BR rat breeding system, CD(SD) IGS; 1998, p. 8–14 (online). Wigal T, Greene PL, Amsel A. Effects on the partial reinforcement extinction effect and on physical and reflex development of short-term in utero exposure to ethanol at different periods of gestation. Behav Neurosci 1988a;102:51–3. Wigal T, Lobaugh NJ, Wigal SB, Greene PL, Amsel A. Sparing of patterned alternation but not partial reinforcement extinction effect after prenatal chronic exposure to ethanol in infant rats. Behav Neurosci 1988b;102: 43–50. Wilkinson RR. Technoeconomic and environmental assessment of industrial organotin compounds. Neurotoxicology 1984;5:141–58.