- Email: [email protected]
Developmental Neurotoxicity Evaluation of the Avermectin Pesticide, Emamectin Benzoate, in Sprague–Dawley Rats
Neurotoxicology and Teratology, Vol. 19, No. 4, pp. 315–326, 1997 Copyright © 1997 Elsevier Science Inc. Printed in the USA. All rights reserved 0892-0362/97 $17.00 1 .00
PII S0892-0362(97)00002-0
Developmental Neurotoxicity Evaluation of the Avermectin Pesticide, Emamectin Benzoate, in Sprague–Dawley Rats L. DAVID WISE, HENRY L. ALLEN, CHAO-MIN L. HOE, DAVID R. VERBEKE AND RONALD J. GERSON1 Merck Research Laboratories, Department of Safety Assessment, West Point, PA 19486 Received 28 October 1996; Accepted 3 February 1997 WISE, L. D., H. L. ALLEN, C.-M. L. HOE, D. R. VERBEKE AND R. J. GERSON. Developmental neurotoxicity evaluation of the avermectin pesticide, emamectin benzoate, in Sprague–Dawley rats. NEUROTOXICOL TERATOL 19(4) 315–326, 1997.—The potential of emamectin benzoate (EB) to cause developmental neurotoxicity in Sprague–Dawley rats was assessed using a study design proposed by the US EPA. Dosages of 0 (deionized water), 0.1, 0.6, or 3.6 mg/kg/day were administered at 5 ml/kg by oral gavage from gestational day (GD) 6 to lactational day (LD) 20 to groups of 25 mated females each. Between GD 17 and 20 the high dose was reduced to 2.5 mg/kg/day because of pup tremors observed at this dose level in a concurrent two-generation study. Females were allowed to deliver and the young were evaluated for survival, growth, development, behavior, and histological changes to brain, spinal cord, peripheral nerve, and skeletal muscle. Behavioral assessment of the offspring consisted of open field motor activity, auditory startle habituation, and passive avoidance tests; each was conducted on weanling and adult animals (one animal/sex/litter). Histopathological examination of the CNS and PNS was conducted on one animal/sex/litter on postnatal days (PND) 11 and 60. There were significant increases in average F0 maternal body weight gains during gestation in the 0.6 and 3.6/2.5 mg/kg/day groups, but no other effects were observed in pregnant females of these or the low-dose groups during the study. Beginning on PND 6, tremors were observed in high-dose pups, and this was followed by hindlimb splay in all high-dose pups by PND 15–26. Both of these physical signs disappeared by PND 34 (i.e., 10–11 days after weaning). There were no compound-related deaths in F1 offspring. Beginning on PND 11, progressive decreases in preweaning average weights were observed in the high-dose group (to 42% below control in females on PND 21). Average weight gain during the postweaning period was significantly decreased in the 3.6/2.5 mg/kg/day group. There were EB-related effects in behavioral tests only in the high-dose group. A significant increase in PND 13 average horizontal motor activity was due to stereotypical movements. Average horizontal activity was decreased on PND 17 and in adult females, but there were no effects on PND 21. Average peak auditory startle response amplitude was decreased on PND 22 and in adults. There were no EB-related effects in the passive avoidance test, relative brain weights, or in the histological examination (including morphometry) of the nervous system. These results demonstrate that the high-dose EB exposure during gestation and lactation to rats produced evidence of neurotoxicity in the F1 offspring, and a clear No Observed Adverse Effect Level (NOAEL) for developmental neurotoxicity of EB was determined to be 0.6 mg/kg/day. © 1997 Elsevier Science Inc. Emamectin benzoate
Developmental neurotoxicity
Avermetcin pesticides
EMAMECTIN benzoate (EB) is the 49-deoxy-49-epi-methylamino derivative of avermectin B1, and is similar in structure to abamectin, a natural fermentation product of Streptomyces avermitilis. The avermectins as a class are well known for their insecticidal, acaricidal, and anthelmintic activities (1), and em-
Sprague–Dawley rats
amectin is being developed as a broad-spectrum pesticide for vegetables with a very low application rate (i.e., Z6 g/acre). As reviewed by Turner and Schaeffer (13), it is likely that all avermectins share a common mode of action, which involves binding to high-affinity receptors and a resultant increase in
Requests for reprints should be addressed to L. David Wise, Merck Research Laboratories, Safety Assessment, W45-1, West Point, PA 19486. Tel: (215) 652-6974; Fax: (215) 652-3423; E-mail: [email protected] 1Present address: DuPont Merck Pharmaceutical Co., Stine-Haskell Research Center, P.O. Box 30, Bldg. 320, Wilmington, DE 19714.
315
316 membrane chloride–ion permeability. The actions of avermectins on vertebrate and invertebrate systems are distinct due to a variety of differences; mammalian species are much less sensitive due to lower receptor affinities and relative inability to cross the blood–brain barrier (13). The genetic and mammalian toxicology of the related compounds, ivermectin and abamectin, has been reviewed (5). In terms of developmental toxicity, both of these compounds were evaluated in mice, rats, and rabbits by conventional oral developmental toxicity studies. In both rodent species, developmental toxicity was observed only at or above the dose levels that produced evidence of maternal toxicity. In rabbits, however, ivermectin (but not abamectin) produced a low but dose-related increase in cleft palate and clubbed forepaws at dosages slightly below those dosages that produced maternotoxicity. The effects of ivermectin on preweaning rat pup mortality, growth, and behavior were examined in a study that orally administered 1, 2, or 4 mg/kg/day from gestation days 6 to 20 and lactation days 2 to 20 (8). Increased pup mortality, delayed growth, and effects on various reflex and behavioral tests were observed in the 2 and 4 mg/kg/day groups, whereas in the 1 mg/kg/day group only cliff avoidance and locomotion were affected. There are no published studies examining the developmental toxicity potential of EB; however, unpublished work from our laboratory has shown the NOAEL for developmental toxicity to be 4 mg/kg/day in rats and 6 mg/kg/day in rabbits. In these studies maternal effects were evident at dosages >2 mg/kg/day in rats and >3 mg/kg/day in rabbits. Rats given 2–5 mg/kg/day had transient increases in maternal body weight gain during the first week of gestational treatment, and decreases in body weight gains are seen at dosages of 4 mg/kg/ day and greater later in gestation. Similar increases in average body weight gain have been noted in other rat toxicity studies with this compound (our unpublished data); however, such increases are not considered to represent a toxic response because they did not appear to affect the health and well-being of the animals or their offspring. In adult animal models, EB produces dose-dependent behavioral and neuropathological changes in all species studied (rat, dog, rabbit, and CD-1 mouse, unpublished work from this laboratory). The behavioral effects are characterized by
WISE ET AL. tremors and changes in motor activity (either lethargy, hyperactivity, and/or ataxia). Neuronal lesions are observed in the CNS and/or PNS and are characterized by degeneration and vacuolation of the neuronal cytoplasm. The cytoplasmic vacuoles are associated with cellular debris, macrophages, and pyknotic nuclei. In the rat these lesions are observed at dosages down to 2.5 mg/kg/day; however, the behavioral signs have not been observed below 5 mg/kg/day. The lowest toxic dose level in dogs for both histological and behavioral effects is 0.5 mg/kg/day during a 1-year study. Because EB, like other avermectins, has neurotoxic potential in laboratory species it was deemed necessary to assess its potential to produce developmental neurotoxicity. The study design follows the guidelines put forward by the US EPA (14) and is intended to evaluate the developmental, behavioral, and neuropathological effects in offspring from female rats treated during gestation and lactation. METHOD
Chemicals Emamectin benzoate (Fig. 1), also known as MK-0244, was supplied by Merck Research Laboratories (Rahway, NJ). Purity was approximately 98% based on HPLC. Suspensions of EB were prepared daily in deionized water during the dosing period. Concentrations were adjusted to deliver the desired dose levels of the base compound in a volume of 5 ml/kg. Animals Nulliparous Sprague–Dawley rats [Crl:CD®(SD) BR], obtained from Charles River Laboratories, Inc. (females from Raleigh, NC; males from Kingston, NY) were housed individually (except during mating) in hanging metal cages. Mated females were moved to polycarbonate cages with hardwood chip bedding (Alpha-dri, Shepherd Specialty Papers, Inc., Kalamazoo, MI). Animal rooms were maintained at 19–22°C, 40–60% relative humidity, and a 12-h light/dark cycle. Food (Purina Certified Rodent Chow #5002) and drinking water were provided ad lib. F0 female animals were approximately 11 weeks of age at the beginning of gestation. Experimental Design This study was conducted under and met Good Laboratory Practices requirements. Females were cohabited with males in a 1:1 ratio. The day of finding seminal plugs below the wire-
TABLE 1 TESTS AND PROCEDURES APPLIED TO F1 ANIMALS Test/Procedure
Pupillary function Brain weight Brain histology Vaginal canalization Preputial separation Open field activity FIG. 1. The structure of EB is shown. The compound, supplied in the benzoate hydrate form, was approximately 98% pure, consisting of approximately 94% B1a and 4% B1b.
Auditory startle habituation Passive avoidance
Postnatal Day(s)
No. Animals/Sex/Litter
Once/week after weaning 11, 60 11, 60 31, 34, 37 39, 43, 47 13, 17, 21, 59 6 1 22, 59 6 1
1 (arbitrarily selected)
24, 59 6 1
1
>10/sex/group >6/sex/group 3–4 3–4 1 1
DEVELOPMENTAL NEUROTOXICITY OF EMAMECTIN BENZOATE bottom cage floor and/or in the vagina was defined as gestation day (GD) 0. Mated females were assigned to four groups of 25 animals each based on a weight-balanced random allocation scheme. Dose levels of 0 (deionized water, the vehicle control group), 0.1, 0.6, or 3.5 mg/kg/day were administered once daily by oral gavage with a metal catheter from GD 6 through lactation day (LD) 20. Between GD 17 and 20 the high-dose level was reduced to 2.5 mg/kg/day after the appearance of pup tremors in the 3.6 mg/kg/day group of a concurrent two-generation study in rats with EB. The reduction of the dose level was intended to lessen the severity and incidence of the pup tremors, and to ensure adequate survival of pups. Animals were observed at least once daily prior to dosing for mortality and adverse physical signs; an additional observation was made 1–5 h after each dose. Body weights were recorded periodically during gestation and lactation. From GD 21 until completion of delivery, each presumed-pregnant female was observed frequently for completion and/or difficulties in parturition. Thereafter, females were observed for signs of poor maternal care at least twice per day. Females with pups were euthanized by CO2 asphyxiation between LD 24 and 29 and the uterus of each female was examined to count metrial glands. Mated females that did not deliver pups were euthanized on presumed GD 24. All pups were counted, sexed, examined for external anomalies, and individually weighed on postnatal day (PND) 0 (day of birth), and 10 pups per litter (five per sex when possible) were randomly selected and identified with a foot tattoo. All pups were observed daily for mortality and physical signs of toxicity. Body weights were recorded on PND 4, 11, 17, 21, and weekly thereafter. On PND 4 each litter was reduced to four previously identified animals per sex. Nonselected pups were euthanized and discarded without further
317
examination. On the days of weighing and behavioral tests all animals were observed for physical signs of toxicity without knowledge of treatment group. One animal/sex/litter was arbitrarily selected and examined for pupillary function once per week from weaning through sacrifice using a small flashlight in subdued lighting. An outline of the various tests and procedures applied during the preweaning and postweaning periods is presented in Table 1. Pupillary function was assessed using a small flashlight in subdued lighting. The brains of animals on PND 11 were immersion-fixed in 10% neutral-buffered formalin, and those from the control and high-dose groups were embedded in paraffin, sectioned at 5 mm, stained with hematoxylin and eosin, and examined for histomorphological changes. Three to four animals per sex per litter were removed from the dams on PND 23 or 24 and housed a maximum of two per sex per cage until postnatal week 5–6 when they were then singly housed. These pups were tail tattooed for identification. The PND 60 necropsy was limited to an examination of the brain, and body and brain weights were recorded. Six animals/sex/group were perfused using 4% formaldehyde/1% glutaraldehyde with the necropsy limited to removal of brains, spinal cords, optic and sciatic nerves, and skeletal muscle. These tissues from the control and high-dose groups were processed by routine methods for hematoxylin and eosin histology. Further, samples of Gasserian ganglion, dorsal (with ganglion) and ventral roots, and sural, tibial, and peroneal branches of the sciatic nerve were embedded in plastic and sections from control and high-dose groups were prepared by routine methods. All brains were sectioned to include (insofar as mechanically possible) the following structures: cerebral cortex (fronto-parietal, piriform), basal ganglia (nucleus caudatus putamen), hip-
FIG. 2. Average 6 SEM body weights of F0 females during gestation and lactation. F(1, 93) 5 36.392, p , 0.0001 for rate of weight gain over GD 6–20, using trend analysis through the 3.6/2.5 mg/kg/day dose group, adjusted by analysis of covariance for GD 6 weight. Trend analysis through the 0.6 mg/kg/ day dose group yielded F(1, 93) 6 5.902, p 5 0.017, and trend analysis through the 0.1 mg/kg/day dose group yielded F(1, 93) 5 0.008, p 5 0.930. F(1, 92) 5 0.163, p 5 0.69 for rate of weight gain over LD 0–20, using trend analysis through the 3.6/2.5 mg/kg/day dose group.
318
WISE ET AL.
pocampus, thalamic and hypothalamic regions, midbrain (tectum, tegmentum, and cerebral peduncle), cerebellum and metencephalon, and myelencephalon (brain stem). Anatomical landmarks were confirmed according to König and Klippel (3). Bilateral measurements of cerebral cortical depth (parietal region), hippocampus major, and single cerebellar folia (typically in the ansiform, lunate, or parafloccular lobes) were recorded using a micrometer eyepiece at 2503 (total magnification) for PND 11 brains and at 253 for adult brains. Remaining F1 animals were euthanized by CO2 asphyxiation and discarded without further examination after completion of the designated behavioral test. Behavioral Assessments
interruptions in each of six consecutive 10-min periods (5-min periods for PND 13 pups) was recorded. The testing period for PND 13 pups was shortened because a previous study showed no significant changes in activity after 30 min (16). The number of beams broken is termed horizontal activity. Naive animals were assessed for auditory startle habituation on PND 22 whereas the PND 59 6 1 animals had been previously tested for passive avoidance behavior. Two computer-controlled auditory startle chambers were used (San Diego Instruments, Inc.). Each chamber consists of four compartments, and each chamber was used exclusively for one sex. All compartments were calibrated to a constant vibrating source prior to testing. Different dosage groups were tested sequentially in each compartment. The floor of each compart-
Open field motor activity was assessed for the same pups on PND 13, 17, and 21, whereas the PND 59 6 1 animals had been previously tested for auditory startle habituation on PND 22. Computer-controlled activity monitors [Digiscan Animal Activity Monitor, RXYCM (model 8), Omnitech Electronics, Inc., Columbus, OH] were used. Each clear square acrylic activity monitor has a floor area of approximately 1640 cm2 and each monitor was used exclusively by one sex. Different dosage groups were tested sequentially in each monitor. The floor of each activity cage was cleaned with water and dried between animals. The test was performed in a room with red light illumination and background white noise (Z70 dB). Rats were adapted to the red light illumination at least 10 min prior to testing. The sides and top of each activity cage were covered during testing. The total number of beam
TABLE 2 REPRODUCTIVE PERFORMANCE OF F0 FEMALE RATS Dose Level (mg/kg/day)
Mated females Females with live pups Length of gestation (days) Implants per female Live pups/litter on PND 0 % live pups on PND 0, LM Percent postimplantation survival to PND 0 Percent pup deaths (LM) PND 1–4 PND 5–21 a Dose
0
0.1
0.6
3.6/2.5a
25 23 22.4 16.6 14.8 99.4
25 25 22.3 16.6 15.8 99.0
25 25 22.4 16.4 15.0 95.4
25 25 22.5ns, b 16.8ns, c 15.5ns, d 99.5
88.9
95.0
92.2
92.2ns, e
1.3 0.5
1.7 0
1.5 0
1.9ns, f 1.0ns, g
level reduced between GD 17 and 20 analysis through the 3.6/2.5 mg/kg/day dose group yielded F(1,94) 5 0.9954, p 5 0.32. c Trend analysis through the 3.6/2.5 mg/kg/day dose group yielded F(1,93) 5 0.3226, p 5 0.57. d Trend analysis through the 3.6/2.5 mg/kg/day dose group yielded F(1,94) 5 0.5216, p 5 0.47. e Trend analysis through the 3.6/2.5 mg/kg/day dose group yielded F(1,94) 5 0.0424, p 5 0.58. f Trend analysis through the 3.6/2.5 mg/kg/day dose group yielded F(1,93) 5 0.2969, p 5 0.29. g Trend analysis through the 3.6/2.5 mg/kg/day dose group yielded F(1,93) 5 2.3444, p 5 0.06. ns Trend not statistically significant ( p . 0.05) through indicated dose level. PND 5 postnatal day. LM 5 litter mean. b Trend
FIG. 3. Average 6 SD body weights of F1 female (top) and male (bottom) pups from birth to weaning. Asterisks indicate statistically significant decreases in average pup weights. Adjustment of preweaning body weight by ANCOVA was done whenever including a covariate reduced the error mean square significantly (p < 0.05). Length of gestation and number of live pups per litter on PND 0 were used as a covariate for both sexes on PND 0, 4, and 11. No adjustments were done for either sex on PND 17 or 21. On PND 0, trend analyses yielded F(1, 93) < 3.686, p > 0.058 through the 3.6/2.5 mg/kg/day group. On PND 4, trend analyses yielded F(1, 93) < 1.979, p > 0.16 through the 3.6/2.5 mg/kg/day dose group. On PND 11, 17, and 21, trend analyses of average body weight of both sexes was statistically significant through the 3.6/2.5 mg/kg/day dose group, F(1, 93) > 19.971, p , 0.0001, but not through the 0.6 mg/kg/day group for either sex, F(1, 93) < 1.438, p > 0.23.
DEVELOPMENTAL NEUROTOXICITY OF EMAMECTIN BENZOATE ment was cleaned with water and dried between animals. Animals were placed in compartments and allowed to acclimate for 3 min in the presence of 72 dB white noise. Immediately following the acclimation period, each animal was exposed to a series of 50, 120-dB, 50-ms bursts of white noise, with a 5-s intertrial interval. The peak amplitude of each startle movement and the interval between acoustic stimulus and peak amplitude were recorded. Short-term learning using a passive avoidance paradigm was assessed in naive animals on PND 24 whereas PND 59 61 animals had been previously tested for open field motor activity, and long-term retention was then assessed in each of these sets of animals 1 week later. The tests were performed in a room with normal illumination and background white noise. Computer-controlled two-compartment shuttleboxes (San Diego Instruments, Inc.) were used. Each shuttlebox was used exclusively for one sex and all dosage groups were sequentially tested in each shuttlebox. Animals were placed in the compartment, which had a 25-W light attached to the outside (light turned off). The testing sequence consisted of a 10-s adaptation period, after which the light came on and the gate separating the lighted compartment from the dark compartment opened. Rats that entered the dark compartment broke an infrared beam, which automatically closed the gate. After the gate closed a scrambled shock [0.2 mA (young animals) or 1 mA (adult animals) for 1 s] was delivered to the cage floor. The maximum duration of each trial was 60 s. After each trial the animal was removed to a holding cage and held for a short interval (30–90 s) prior to retesting (if necessary). Animals were tested in consecutive trials until they reached criterion,
319
defined as remaining in the lighted compartment for two (pups) or three (adult) consecutive 60-s trials. The entire test using the same conditions was repeated 7 days later with the same animals. Statistical Methods Statistical analyses were done by an analysis of variance (ANOVA) or covariance (ANCOVA) [continuous variables (11)] or by an extended Mantel-Haenszel test [discrete variables (7)]. Nonparametric analyses were done in a rankit scale to normalize the data (2). Repeated measurements over time, specifically F0 and F1 body weights, were analyzed by orthogonal polynomials (9). For all other analyzed parameters, including average beams broken over the 30- or 60-min interval and average peak amplitude across the 10 trial blocks, a trend analysis was used to determine if there was a significant trend with increasing dosage across all treatment groups (12). Results were considered to be statistically significant if p > 0.05. When there was a significant trend, the data from the high dosage group were excluded and the trend test was repeated. This process was repeated until nonsignificance (p > 0.05) was achieved. RESULTS
F0 Generation There were no deaths and no adverse physical signs observed in F0 females during the study. There were statistically
FIG. 4. Average body weights and weight change of F1 males and females during the postweaning period (6 SD). For males, trend analysis of change in body weight from postnatal weeks 4 to 10 was statistically significant through the 3.6/2.5 mg/kg/day dose group, F(1, 93) 5 112.195, p , 0.0001, but not through the 0.6 mg/kg/day dose group, F(1, 93) 5 0.289, p 5 0.59. For females, trend analysis of change in body weight from postnatal weeks 4 to 10 was statistically significant through the 3.6/2.5 mg/kg/day, F(1, 93) 5 120.021, p , 0.0001, and 0.6 mg/kg/day, F(1, 93) 5 4.912, p 5 0.029, dose groups, but not statistically significant through the 0.1 mg/kg/day dose group, F(1, 93) 5 3.877, p 5 0.052. Because of the small magnitude in the effect on overall weight gain in the 0.6 mg/kg/day group (6% below control), this was not considered to be compound related.
320
WISE ET AL.
significant increases in average maternal weight gains between GD 6 and 20 in the 0.6 and 3.6/2.5 mg/kg/day groups (11% and 15% above control, respectively). Average body weights are shown in Fig. 2. There were no adverse effects on maternal body weight during this period in the 0.1 mg/kg/day
group. During the lactation dosing period, there were no EBrelated or statistically significant effects on average weight gain. There were no EB-related or statistically significant effects on F0 reproductive performance as assessed by the number of females with live pups, average length of gestation,
FIG. 5. Average open field motor activity 6 SEM in females and males for all 6 on various days of testing. Asterisks indicate statistically significant increases or decreases in average beams broken during the testing interval. For males on PND 13 and 17, trend analysis of overall horizontal activity was statistically significant through the 3.6/2.5 mg/kg/day dose group, F(1, 93) > 9.181, p , 0.003, but not through the 0.6 mg/kg/day dose group, F(1, 93) < 2.624, p > 0.055. For females on PND 13, 17, and 59, there were statistically significant trends through the 3.6/2.5 mg/kg/day dose group, F(1, 92–93) > 5.617, p < 0.020, but not through the 0.6 mg/kg/day dose group, F(1, 92–93) < 2.690, p > 0.053, except on PND 17, F(1, 93) 5 3.725, p 5 0.028.
DEVELOPMENTAL NEUROTOXICITY OF EMAMECTIN BENZOATE average implants per pregnant female, and percent postimplantation survival on PND 0 (Table 2).
F1 Generation: Physical Signs and Body Weight There were no EB-related effects on pup survival (Table 2) or external morphology (data not shown). Beginning on PND 6, intermittent head tremors were observed in some pups of the 3.5/2.5 mg/kg/day group. This sign progressed to whole-body tremors in all pups of all litters in this group by PND 25. In addition, hindlimb extension was first observed on PND 10 in some pups. This latter sign was followed by hindlimb splay between PND 15 to 26 in all high-dose pups. No
321
adverse physical signs were observed in the 0.6 or 0.1 mg/kg/ day groups. There were statistically significant decreases in average pup weights in the high-dose group during the preweaning period (Fig. 3). The effects were first apparent on PND 11 (14% and 11% below control in females and males, respectively), and pup body weights became progressively lower (relative to controls) with time (to 42% and 40% below control, respectively, on PND 21). There were no adverse effects on average preweaning body weights in the 0.6 or 0.1 mg/kg/day groups. During the postweaning period there were no deaths, but male and female animals in the 3.5/2.5 mg/kg/day group continued to demonstrate statistically significant decreases in average body weight gains (18% and 17% below control in fe-
322
WISE ET AL. males and males, respectively, between postnatal weeks 4 and 10, Fig. 4), such that at termination average body weights were 25% and 21% below control, respectively. Although there was a statistically significant (p 5 0.029) effect detected in the time–response trend analysis of average F1 female (but not male) body weight through the 0.6 mg/kg/day group, this was not considered to be EB related because overall weight gain during this postweaning period was only 6% below control. There were no treatment-related effects on average body weights during this period in the 0.1 mg/kg/day group. The adverse physical signs observed during the preweaning period continued for only a short time into the postweaning period. Specifically, whole-body tremors were observed in 7 of 174 animals up to at most PND 27 and hindlimb splay was observed in 45% of the animals up to at most PND 34. Thereafter, only one high-dose animal was observed to have any hindlimb abnormality. No adverse physical signs were observed during this time in animals of the 0.1 and 0.6 mg/kg/day groups, and no effects on pupillary function were observed in any of the EB-exposed group. F1 Generation: Developmental Signs and Behavioral Tests
FIG. 6. Average 6 SEM stereotypic time during open field motor activity testing in females (top) and males (bottom) on PND 13. Asterisks indicate statistically significant (p < 0.05) increases in overall stereotypic time in the 3.6/2.5 mg/kg/day group. Trend analysis through the high-dose group was statistically significant for both sexes, F(1, 93) > 5.336, p > 0.023, but not through the 0.6 mg/kg/ day group, F(1, 93) > 0.303, p > 0.493.
Preweaning and early postweaning period. During the preweaning and early postweaning period compound-related effects were detected only in the 3.5/2.5 mg/kg/day group. In the PND 13 open field motor activity test, there was an increase in overall average horizontal activity in male and female pups (Fig. 5). In view of the physical signs and body weight decrements discussed above, additional analysis of horizontal activity was conducted to better understand the increase in this parameter. One component of horizontal activity is stereotypic behavior, which is represented as repeated breaks of the same infrared photo beam(s). There was a significant increase in
FIG. 7. Historical control data for open field motor activity in PND 17 females. The graph points out the variablitiy of these data on this particular day and that the overall average value of the 0.6 mg/kg/day group fell between the values of control females from two studies conducted in this laboratory near the time of the present study.
DEVELOPMENTAL NEUROTOXICITY OF EMAMECTIN BENZOATE stereotypic time in both sexes (Fig. 6). This finding provides an explanation for the increase in horizontal activity and is consistent with the observed tremors in this high-dose group. Average horizontal activity was significantly decreased in both sexes on PND 17 (41% and 30% in females and males, respectively), but there were no statistically significant effects detected on PND 21 (Fig. 5). There was a statistically significant decrease in average horizontal activity of females in the 0.6 mg/kg/day group on PND 17, but this was not considered to be compound related because the value was within the range of historical control PND 17 females in this laboratory (Fig. 7). In the auditory startle habituation test on PND 22 there was a significant decrease in the average peak response amplitude across the 50 trials in the 3.6/2.5 mg/kg/day group (approximately 74% below control in both sexes, Fig. 8). There was no apparent effect on average time to peak response (data not shown). There were no effects on these parameters in the 0.1 and 0.6 mg/kg/day groups. There were no effects in any EB-exposed groups on learning/short-term retention or long-term retention as assessed on PND 24 and 31 in the passive avoidance test (Fig. 9). Postweaning period. The appearance of developmental landmarks was delayed only in the high-dose group. The estimated mean day of occurrence of vaginal canalization was delayed 3.7 days and preputial separation was delayed 3.6 days compared to the control group, but there were no effects in
323
the other EB-exposed groups (vaginal canalization: 33.7, 33.4, 33.1, and 37.4 days; preputial separation: 44.8, 44.9, 44.8, and 48.4 days in the control, low-, mid-, and high-dose groups, respectively). Average horizontal activity was significantly decreased only in females on PND 59 6 1 (18% below control, Fig. 5). In the auditory startle habituation test there were significant decreases of average peak response amplitude in both females and males (46% and 31% below control, respectively, Fig. 8), but no apparent effect on average time to peak response (data not shown). As in the preweaning period, there were no effects on adult animals in the passive avoidance test (Fig. 9).
Organ Weight and Histological Examinations There were slight decreases in average absolute brain weights of animals in the 3.6/2.5 mg/kg/day group on PND 11 (4–8% below control) and postnatal week 8 (4–10% below control, Table 3); however, these effects were considered to be secondary to the larger decreases in absolute body weight of these animals, because there were slight to moderate increases in average brain weights expressed as a percent of body weight. Morphometry of brain structures showed no mensurational differences between control and EB-exposed brains on PND 11 or 59–60 (Table 4). There were no com-
FIG. 8. Average peak startle response amplitude 6 SEM over 50 trials in females and males on two separate days of testing. On PND 22 and 58–60 for both males and females, the trend through the 3.6/2.5 mg/kg/day dose group was statistically significant, F(1, 91–93) > 5.905, p < 0.017, but not through the 0.6 mg/kg/day group, F(1, 91–93) < 1.392, p > 0.448.
324
WISE ET AL.
FIG. 9. Results of passive avoidance testing in weanlings (top) and as adults (bottom). There were no meaningful differences between control and EB-exposed groups.
TABLE 3 ABSOLUTE AND RELATIVE BRAIN WEIGHTS OF F1 ANIMALS Postnatal Day 11 Females
Brain weight (g) Control 0.1 mg/kg/day 0.6 mg/kg/day 3.6/2.5 mg/kg/day Relative brain (% body wt.) Control 0.1 mg/kg/day 0.6 mg/kg/day 3.6/2.5 mg/kg/day
Postnatal Day 59–60 Males
Females
Males
Mean
S.D.
(N)
Mean
S.D.
(N)
Mean
S.D.
(N)
Mean
S.D.
(N)
1.09 1.07 1.09 1.05
0.04 0.06 0.05 0.06
(10) (12) (12) (13)
1.15 1.10 1.13 1.06
0.08 0.06 0.06 0.09
(12) (13) (12) (11)
1.92 1.85 1.92NS 1.73S
0.06 0.12 0.06 0.05
(12) (12) (11) (11)
1.98 2.05 2.02NS 1.90S
0.11 0.10 0.07 0.08
(10) (12) (12) (13)
3.99 4.14 4.16 4.24
0.35 0.49 0.36 0.52
4.06 4.07 3.92 4.49
0.38 0.46 0.34 0.64
0.76 0.73 0.76NS 0.90S
0.07 0.05 0.04 0.07
0.50 0.50 0.52NS 0.60S
0.04 0.04 0.03 0.07
NS Trend not statistically significant through indicated dose level [females: F(1, 42) < 0.419, p > 0.52; males: F(1, 43) < 1.062, p > 0.31]. S Trend statistically significant through indicated dose level [females: F(1, 42) > 34.435, p , 0.0001; males: F(1, 43) > 15.415, p < 0.0003].
DEVELOPMENTAL NEUROTOXICITY OF EMAMECTIN BENZOATE
325
marks. In terms of the behavioral tests, open field motor activity was increased on PND 13 due to the detection of tremors by the testing apparatus. Decreased motor activity was then again observed in adult females only, but the decrease, although statistically significant, was fairly minor. In the auditory startle habituation task there were decreases in the average peak amplitude on PND 22 and as adults (PND 58–60). Although the etiology of this effect is not known, it should be noted that the decreases in average peak response amplitude in pups and adults of the high-dose group were observed in association with decreases in average body weight (approximately 40% and 25% below control, respectively). There were no morphological changes detected by light microscopy in the EB-exposed offspring on PND 11 or as adults. Other unpublished studies from this laboratory have shown histomorphological degenerative changes in the brains of adult animals orally administered EB in the dosage range used in the present study (see Introduction). We mention this only to make the point that this laboratory had the experience and aptitude to detect histomorphological brain changes, but none were observed. The lack of similar findings in the EBexposed offspring of the present study may be due to a lower total dosage received by the pups via the milk, and/or significantly different pharmacokinetics of the compound (and/or its metabolites). Alternatively, rat pups may be inherently less sensitive than adults to EB-induced brain degenerative changes. Addtional studies would be needed to answer these questions. In relation to the study of Poul (8), which assessed preweaning pup parameters in rats following exposure to ivermectin during gestation and lactation, the results of this study show that EB is less toxic to the F1 generation. In the study with ivermectin, dose levels of 2 and 4 mg/kg/day resulted in significant mortality before weaning (31% in the former group) whereas no such mortality occurred at EB dose levels up to 3.6 mg/kg/day administered during gestation and 2.5 mg/ kg/day during lactation. Poul (8) also demonstrated that preweaning behavioral development of pups, as assessed by the Cincinnati test battery (15), was impaired or delayed at all dose levels of ivermectin tested (1, 2, and 4 mg/kg/day). In the 1 mg/kg/day group there were delays in cliff avoidance, pivoting activity, and forward locomotion, but no other effects on growth or behavior were noted. The results of the present study showed that EB exposure of 3.6/2.5 mg/kg/day to the dam produced some preweaning and postweaning behavioral
pound-related changes in the histomorphology of brain, spinal cord, or peripheral nerve of PND 11 or adult animals in the 3.6/2.5 mg/kg/day group. DISCUSSION
The purpose of this study was to assess the potential of EB to produce developmental neurotoxicity in rats exposed during organogenesis and early postnatal life (i.e., GD 6 to LD 20). Maternal animals treated with EB showed significant increases in average weight gain during gestation in the midand high-dose groups, but there were no other signs of toxicity. The toxicological significance of these increased weight gains in the 3.6/2.5 mg/kg/day group is not known, but slightly increased body weight is a consistent feature of EB exposure via oral gavage to pregnant as well as nonpregnant rats. The NOEL (No Observed Effect Level) for maternal F0 animals in the present study was 0.1 mg/kg/day based on the increased weight gains, but the NOAEL is considered to be 3.6 mg/kg/ day (dose level was lowered to 2.5 mg/kg/day on GD 17–20, after the effect had occurred). This NOAEL for maternal toxicity is considered to be fairly accurate because in the oral gavage rat developmental toxicity study there were doserelated decreases in average weight gain during the latter portion of gestational exposure at dose levels of 4 and 8 mg/kg/ day (unpublished results). This is the first study from our laboratory to examine the behavioral effects of prenatal and early postnatal exposure of EB, or any avermectin, in the F1 generation of rats. Multigeneration studies in rats were conducted in this laboratory with ivermectin (6) and abamectin (5). Postnatal toxicity characterized by decreased weight gain and mortality was found in F1 offspring at doses < 0.4 mg/kg/day of both compounds. Further studies using radiolabeled ivermectin indicated that high drug concentrations in the milk of exposed dams led to high drug levels in the plasma and brain of the F1 offspring (relative to adult rats) and were probably responsible for the observed toxicity (6). The neurotoxicological effects in the F1 offspring of the present study were observed only in the 3.6/2.5 mg/kg/day group and consisted of tremors with hindlimb extensions/ splaying and effects in two of three automated behavioral tests. These effects were observed in conjuction with moderate to marked decreases in preweaning weights, postweaning weight gains, and delays in attainment of developmental land-
TABLE 4 MORPHOMETRY ON NEURAL STRUCTURES IN F1 RATS Control
PND 11 Cerebral cortex Hippocampus Cerebellum PND 59–60 Cerebral cortex Hippocampus Cerebellum a Mean
3.6/2.5 mg/kg/day
Females
Males
Females
Males
3.2a (2.8–3.7)b 2.8 (2.3–3.4) 1.1 (1.0–1.4)
3.1 (2.8–3.9) 3.0 (2.7–3.6) 1.2 (0.9–1.5)
3.6 (2.7–4.7) 2.9 (2.5–3.3) 1.2 (1.0–1.4)
3.2 (2.8–3.8) 2.9 (2.3–3.3) 1.1 (0.9–1.2)
4.0 (3.6–4.7) 4.6 (3.8–5.0) 1.9 (1.5–2.5)
4.2 (3.7–4.8) 5.1 (4.7–5.5) 1.9 (1.5–2.1)
4.4 (4.1–4.7) 4.7 (4.4–5.1) 1.9 (1.7–2.1)
4.1 (3.5–4.6) 4.9 (4.2–5.4) 2.0 (1.4–2.6)
value in micrometer units; scale is different between days. of values.
b Range
326
WISE ET AL.
effects in addition to clear effects on growth, but a NOAEL for developmental neurotoxicity was determined to be 0.6 mg/ kg/day. It is clear from these studies that the avermectins are able to disrupt postnatal development in rats when administered during gestation and lactation. The susceptibility of rat pups to the avermectins may be related to the levels of P-glycoprotein as suggested by work in mice. The sensitivity of mice to ivermectin neurotoxicity was shown to be due to the presence of a P-glycoprotein synthesized by the multidrug resistance gene mdr1a (10). Mice homozygous for a disruption of the mdr1a gene have no physiological, anatomical, or histological abnormalities but are 50to 100-fold more sensitive to orally administered ivermectin than genetically matched mdr1a (1/1) mice. This increased sensitivity was correlated with a near absence of immunohistochemically detectable P-glycoprotein in the brains of these mice (10). More recently, our laboratory has shown that P-gly-
coprotein in brain capillary endothelia and intestine of rats is nearly absent at birth and reaches mature levels at about 5 weeks of age (4). The behavioral test battery utilized in this study was designed to fulfill the US EPA guidelines for the assessment of developmental neurotoxicity (14). The battery revealed effects on motor activity and sensorimotor functions that were concomitantly associated with tremors and depressed body weight. These results, together with the results of a companion study with acrylamide (16), suggest that for some compounds the developmental neurotoxicity protocol outlined by the EPA does not necessarily identify behavioral or neurohistological effects at dose levels lower than those that produce more routine or obvious endpoints of toxicity. However, the utility of this study design, and potential modifications of it, can only be assessed after a larger number of different compounds have been similarly tested in such a rigorous study design.
REFERENCES 1. Campbell, W. C.: Ivermectin and abamectin. New York: SpringerVerlag Inc.; 1989. 2. Harter, H. L.: Expected values of normal order statistics. Biometrika 48:151–165; 1961. 3. Köig, J. F. R.; Klippel, R. A.: The rat brain. Huntington, NY: R.E. Krieger Publishing Co., Inc.; 1963. 4. Lankas, G. R.; Cartwright, M.; Minsker, D.; Cukierski, M.: Postnatal development of P-glycoprotein, a component of the blood– brain and intestinal barriers, in rats. Fundam. Appl. Toxicol. Suppl. 30:263; 1996. 5. Lankas, G. R.; Gordon, L. R.: Toxicology. In: Campbell, W. C., ed. Ivermectin and abamectin. Springer-Verlag New York Inc.; 1989:89–112. 6. Lankas, G. R.; Minsker, D. H.; Robertson, R. T.: Effect of ivermectin on reproduction and neonatal toxicity in rats. Food Chem. Toxicol. 27:523–529; 1989. 7. Mantel, N.: Chi-square tests with one degree of freedom. J. Am. Stat. Assoc. 58:690–700; 1963. 8. Poul, J.-M.: Eeffects of perinatal ivermectin exposure on behavioral development in rats. Neurotoxicol. Teratol. 10:267–272; 1988. 9. Robson, D. S.: A simplified method for constructing orthaogonal polynomials when independent variable is unequally spaced. Biometrics 15:187–191; 1959 10. Schinkel, A.; Smit, J.; van Tellingen, O.; Beijnen, J.; Wagenaar,
11. 12. 13. 14.
15. 16.
E.; van Deemter, L.; Mol, C.; van der Valk, M.; Robanus-Maandag, E.; te Riele, H.; Berns, A.; Borst, P. Disruption of the mouse mdr1a P-glycoprotein gene leads to a deficiency in the blood– brain barrier and to increased sensitivity to drugs. Cell 77:491– 502; 1994. Snedecor, G. W.; Cochran, W. G.: Statistical methods, 7th ed. Ames, IA: Iowa State University Press; 1980. Tukey, J. W.; Ciminera, J. L.; Heyse, J. F.: Testing the statistical certainty of a response to increasing doses of a drug. Biometrics 41:295–301; 1985. Turner, M. J.; Schaeffer, J. M.: Mode of action of Ivermectin. In: Campbell, W. C., ed. Ivermectin and abamectin. New York: Springer-Verlag Inc.; 1989:73–88. U.S. Environmental Protection Agency: Pesticide assessment guidelines—Subdivision F, Hazard evaluation: Human and domestic animals. Addendum 10—Neurotoxicity Series 81, 82, and 83. Publication No. PB-91-154617, National Technical Information Service, Department of Commerce, Spingfield, VA, 1991. Vorhees, C. V.: Behavioral teratogenicity testing as a method of screening for hazards to human health: A methodological proposal. Neurobehav. Toxicol. Teratol. 5:469–474; 1983. Wise, L. D.; Gordon, L. R.; Soper, K. A.; Duchai, D. M.; Morrissey, R. E.: Developmental neurotoxicity evaluation of acrylamide in Sprague–Dawley Rats. Neurotoxicol. Teratol. 17:189–198; 1995.
PII S0892-0362(97)00002-0
Developmental Neurotoxicity Evaluation of the Avermectin Pesticide, Emamectin Benzoate, in Sprague–Dawley Rats L. DAVID WISE, HENRY L. ALLEN, CHAO-MIN L. HOE, DAVID R. VERBEKE AND RONALD J. GERSON1 Merck Research Laboratories, Department of Safety Assessment, West Point, PA 19486 Received 28 October 1996; Accepted 3 February 1997 WISE, L. D., H. L. ALLEN, C.-M. L. HOE, D. R. VERBEKE AND R. J. GERSON. Developmental neurotoxicity evaluation of the avermectin pesticide, emamectin benzoate, in Sprague–Dawley rats. NEUROTOXICOL TERATOL 19(4) 315–326, 1997.—The potential of emamectin benzoate (EB) to cause developmental neurotoxicity in Sprague–Dawley rats was assessed using a study design proposed by the US EPA. Dosages of 0 (deionized water), 0.1, 0.6, or 3.6 mg/kg/day were administered at 5 ml/kg by oral gavage from gestational day (GD) 6 to lactational day (LD) 20 to groups of 25 mated females each. Between GD 17 and 20 the high dose was reduced to 2.5 mg/kg/day because of pup tremors observed at this dose level in a concurrent two-generation study. Females were allowed to deliver and the young were evaluated for survival, growth, development, behavior, and histological changes to brain, spinal cord, peripheral nerve, and skeletal muscle. Behavioral assessment of the offspring consisted of open field motor activity, auditory startle habituation, and passive avoidance tests; each was conducted on weanling and adult animals (one animal/sex/litter). Histopathological examination of the CNS and PNS was conducted on one animal/sex/litter on postnatal days (PND) 11 and 60. There were significant increases in average F0 maternal body weight gains during gestation in the 0.6 and 3.6/2.5 mg/kg/day groups, but no other effects were observed in pregnant females of these or the low-dose groups during the study. Beginning on PND 6, tremors were observed in high-dose pups, and this was followed by hindlimb splay in all high-dose pups by PND 15–26. Both of these physical signs disappeared by PND 34 (i.e., 10–11 days after weaning). There were no compound-related deaths in F1 offspring. Beginning on PND 11, progressive decreases in preweaning average weights were observed in the high-dose group (to 42% below control in females on PND 21). Average weight gain during the postweaning period was significantly decreased in the 3.6/2.5 mg/kg/day group. There were EB-related effects in behavioral tests only in the high-dose group. A significant increase in PND 13 average horizontal motor activity was due to stereotypical movements. Average horizontal activity was decreased on PND 17 and in adult females, but there were no effects on PND 21. Average peak auditory startle response amplitude was decreased on PND 22 and in adults. There were no EB-related effects in the passive avoidance test, relative brain weights, or in the histological examination (including morphometry) of the nervous system. These results demonstrate that the high-dose EB exposure during gestation and lactation to rats produced evidence of neurotoxicity in the F1 offspring, and a clear No Observed Adverse Effect Level (NOAEL) for developmental neurotoxicity of EB was determined to be 0.6 mg/kg/day. © 1997 Elsevier Science Inc. Emamectin benzoate
Developmental neurotoxicity
Avermetcin pesticides
EMAMECTIN benzoate (EB) is the 49-deoxy-49-epi-methylamino derivative of avermectin B1, and is similar in structure to abamectin, a natural fermentation product of Streptomyces avermitilis. The avermectins as a class are well known for their insecticidal, acaricidal, and anthelmintic activities (1), and em-
Sprague–Dawley rats
amectin is being developed as a broad-spectrum pesticide for vegetables with a very low application rate (i.e., Z6 g/acre). As reviewed by Turner and Schaeffer (13), it is likely that all avermectins share a common mode of action, which involves binding to high-affinity receptors and a resultant increase in
Requests for reprints should be addressed to L. David Wise, Merck Research Laboratories, Safety Assessment, W45-1, West Point, PA 19486. Tel: (215) 652-6974; Fax: (215) 652-3423; E-mail: [email protected] 1Present address: DuPont Merck Pharmaceutical Co., Stine-Haskell Research Center, P.O. Box 30, Bldg. 320, Wilmington, DE 19714.
315
316 membrane chloride–ion permeability. The actions of avermectins on vertebrate and invertebrate systems are distinct due to a variety of differences; mammalian species are much less sensitive due to lower receptor affinities and relative inability to cross the blood–brain barrier (13). The genetic and mammalian toxicology of the related compounds, ivermectin and abamectin, has been reviewed (5). In terms of developmental toxicity, both of these compounds were evaluated in mice, rats, and rabbits by conventional oral developmental toxicity studies. In both rodent species, developmental toxicity was observed only at or above the dose levels that produced evidence of maternal toxicity. In rabbits, however, ivermectin (but not abamectin) produced a low but dose-related increase in cleft palate and clubbed forepaws at dosages slightly below those dosages that produced maternotoxicity. The effects of ivermectin on preweaning rat pup mortality, growth, and behavior were examined in a study that orally administered 1, 2, or 4 mg/kg/day from gestation days 6 to 20 and lactation days 2 to 20 (8). Increased pup mortality, delayed growth, and effects on various reflex and behavioral tests were observed in the 2 and 4 mg/kg/day groups, whereas in the 1 mg/kg/day group only cliff avoidance and locomotion were affected. There are no published studies examining the developmental toxicity potential of EB; however, unpublished work from our laboratory has shown the NOAEL for developmental toxicity to be 4 mg/kg/day in rats and 6 mg/kg/day in rabbits. In these studies maternal effects were evident at dosages >2 mg/kg/day in rats and >3 mg/kg/day in rabbits. Rats given 2–5 mg/kg/day had transient increases in maternal body weight gain during the first week of gestational treatment, and decreases in body weight gains are seen at dosages of 4 mg/kg/ day and greater later in gestation. Similar increases in average body weight gain have been noted in other rat toxicity studies with this compound (our unpublished data); however, such increases are not considered to represent a toxic response because they did not appear to affect the health and well-being of the animals or their offspring. In adult animal models, EB produces dose-dependent behavioral and neuropathological changes in all species studied (rat, dog, rabbit, and CD-1 mouse, unpublished work from this laboratory). The behavioral effects are characterized by
WISE ET AL. tremors and changes in motor activity (either lethargy, hyperactivity, and/or ataxia). Neuronal lesions are observed in the CNS and/or PNS and are characterized by degeneration and vacuolation of the neuronal cytoplasm. The cytoplasmic vacuoles are associated with cellular debris, macrophages, and pyknotic nuclei. In the rat these lesions are observed at dosages down to 2.5 mg/kg/day; however, the behavioral signs have not been observed below 5 mg/kg/day. The lowest toxic dose level in dogs for both histological and behavioral effects is 0.5 mg/kg/day during a 1-year study. Because EB, like other avermectins, has neurotoxic potential in laboratory species it was deemed necessary to assess its potential to produce developmental neurotoxicity. The study design follows the guidelines put forward by the US EPA (14) and is intended to evaluate the developmental, behavioral, and neuropathological effects in offspring from female rats treated during gestation and lactation. METHOD
Chemicals Emamectin benzoate (Fig. 1), also known as MK-0244, was supplied by Merck Research Laboratories (Rahway, NJ). Purity was approximately 98% based on HPLC. Suspensions of EB were prepared daily in deionized water during the dosing period. Concentrations were adjusted to deliver the desired dose levels of the base compound in a volume of 5 ml/kg. Animals Nulliparous Sprague–Dawley rats [Crl:CD®(SD) BR], obtained from Charles River Laboratories, Inc. (females from Raleigh, NC; males from Kingston, NY) were housed individually (except during mating) in hanging metal cages. Mated females were moved to polycarbonate cages with hardwood chip bedding (Alpha-dri, Shepherd Specialty Papers, Inc., Kalamazoo, MI). Animal rooms were maintained at 19–22°C, 40–60% relative humidity, and a 12-h light/dark cycle. Food (Purina Certified Rodent Chow #5002) and drinking water were provided ad lib. F0 female animals were approximately 11 weeks of age at the beginning of gestation. Experimental Design This study was conducted under and met Good Laboratory Practices requirements. Females were cohabited with males in a 1:1 ratio. The day of finding seminal plugs below the wire-
TABLE 1 TESTS AND PROCEDURES APPLIED TO F1 ANIMALS Test/Procedure
Pupillary function Brain weight Brain histology Vaginal canalization Preputial separation Open field activity FIG. 1. The structure of EB is shown. The compound, supplied in the benzoate hydrate form, was approximately 98% pure, consisting of approximately 94% B1a and 4% B1b.
Auditory startle habituation Passive avoidance
Postnatal Day(s)
No. Animals/Sex/Litter
Once/week after weaning 11, 60 11, 60 31, 34, 37 39, 43, 47 13, 17, 21, 59 6 1 22, 59 6 1
1 (arbitrarily selected)
24, 59 6 1
1
>10/sex/group >6/sex/group 3–4 3–4 1 1
DEVELOPMENTAL NEUROTOXICITY OF EMAMECTIN BENZOATE bottom cage floor and/or in the vagina was defined as gestation day (GD) 0. Mated females were assigned to four groups of 25 animals each based on a weight-balanced random allocation scheme. Dose levels of 0 (deionized water, the vehicle control group), 0.1, 0.6, or 3.5 mg/kg/day were administered once daily by oral gavage with a metal catheter from GD 6 through lactation day (LD) 20. Between GD 17 and 20 the high-dose level was reduced to 2.5 mg/kg/day after the appearance of pup tremors in the 3.6 mg/kg/day group of a concurrent two-generation study in rats with EB. The reduction of the dose level was intended to lessen the severity and incidence of the pup tremors, and to ensure adequate survival of pups. Animals were observed at least once daily prior to dosing for mortality and adverse physical signs; an additional observation was made 1–5 h after each dose. Body weights were recorded periodically during gestation and lactation. From GD 21 until completion of delivery, each presumed-pregnant female was observed frequently for completion and/or difficulties in parturition. Thereafter, females were observed for signs of poor maternal care at least twice per day. Females with pups were euthanized by CO2 asphyxiation between LD 24 and 29 and the uterus of each female was examined to count metrial glands. Mated females that did not deliver pups were euthanized on presumed GD 24. All pups were counted, sexed, examined for external anomalies, and individually weighed on postnatal day (PND) 0 (day of birth), and 10 pups per litter (five per sex when possible) were randomly selected and identified with a foot tattoo. All pups were observed daily for mortality and physical signs of toxicity. Body weights were recorded on PND 4, 11, 17, 21, and weekly thereafter. On PND 4 each litter was reduced to four previously identified animals per sex. Nonselected pups were euthanized and discarded without further
317
examination. On the days of weighing and behavioral tests all animals were observed for physical signs of toxicity without knowledge of treatment group. One animal/sex/litter was arbitrarily selected and examined for pupillary function once per week from weaning through sacrifice using a small flashlight in subdued lighting. An outline of the various tests and procedures applied during the preweaning and postweaning periods is presented in Table 1. Pupillary function was assessed using a small flashlight in subdued lighting. The brains of animals on PND 11 were immersion-fixed in 10% neutral-buffered formalin, and those from the control and high-dose groups were embedded in paraffin, sectioned at 5 mm, stained with hematoxylin and eosin, and examined for histomorphological changes. Three to four animals per sex per litter were removed from the dams on PND 23 or 24 and housed a maximum of two per sex per cage until postnatal week 5–6 when they were then singly housed. These pups were tail tattooed for identification. The PND 60 necropsy was limited to an examination of the brain, and body and brain weights were recorded. Six animals/sex/group were perfused using 4% formaldehyde/1% glutaraldehyde with the necropsy limited to removal of brains, spinal cords, optic and sciatic nerves, and skeletal muscle. These tissues from the control and high-dose groups were processed by routine methods for hematoxylin and eosin histology. Further, samples of Gasserian ganglion, dorsal (with ganglion) and ventral roots, and sural, tibial, and peroneal branches of the sciatic nerve were embedded in plastic and sections from control and high-dose groups were prepared by routine methods. All brains were sectioned to include (insofar as mechanically possible) the following structures: cerebral cortex (fronto-parietal, piriform), basal ganglia (nucleus caudatus putamen), hip-
FIG. 2. Average 6 SEM body weights of F0 females during gestation and lactation. F(1, 93) 5 36.392, p , 0.0001 for rate of weight gain over GD 6–20, using trend analysis through the 3.6/2.5 mg/kg/day dose group, adjusted by analysis of covariance for GD 6 weight. Trend analysis through the 0.6 mg/kg/ day dose group yielded F(1, 93) 6 5.902, p 5 0.017, and trend analysis through the 0.1 mg/kg/day dose group yielded F(1, 93) 5 0.008, p 5 0.930. F(1, 92) 5 0.163, p 5 0.69 for rate of weight gain over LD 0–20, using trend analysis through the 3.6/2.5 mg/kg/day dose group.
318
WISE ET AL.
pocampus, thalamic and hypothalamic regions, midbrain (tectum, tegmentum, and cerebral peduncle), cerebellum and metencephalon, and myelencephalon (brain stem). Anatomical landmarks were confirmed according to König and Klippel (3). Bilateral measurements of cerebral cortical depth (parietal region), hippocampus major, and single cerebellar folia (typically in the ansiform, lunate, or parafloccular lobes) were recorded using a micrometer eyepiece at 2503 (total magnification) for PND 11 brains and at 253 for adult brains. Remaining F1 animals were euthanized by CO2 asphyxiation and discarded without further examination after completion of the designated behavioral test. Behavioral Assessments
interruptions in each of six consecutive 10-min periods (5-min periods for PND 13 pups) was recorded. The testing period for PND 13 pups was shortened because a previous study showed no significant changes in activity after 30 min (16). The number of beams broken is termed horizontal activity. Naive animals were assessed for auditory startle habituation on PND 22 whereas the PND 59 6 1 animals had been previously tested for passive avoidance behavior. Two computer-controlled auditory startle chambers were used (San Diego Instruments, Inc.). Each chamber consists of four compartments, and each chamber was used exclusively for one sex. All compartments were calibrated to a constant vibrating source prior to testing. Different dosage groups were tested sequentially in each compartment. The floor of each compart-
Open field motor activity was assessed for the same pups on PND 13, 17, and 21, whereas the PND 59 6 1 animals had been previously tested for auditory startle habituation on PND 22. Computer-controlled activity monitors [Digiscan Animal Activity Monitor, RXYCM (model 8), Omnitech Electronics, Inc., Columbus, OH] were used. Each clear square acrylic activity monitor has a floor area of approximately 1640 cm2 and each monitor was used exclusively by one sex. Different dosage groups were tested sequentially in each monitor. The floor of each activity cage was cleaned with water and dried between animals. The test was performed in a room with red light illumination and background white noise (Z70 dB). Rats were adapted to the red light illumination at least 10 min prior to testing. The sides and top of each activity cage were covered during testing. The total number of beam
TABLE 2 REPRODUCTIVE PERFORMANCE OF F0 FEMALE RATS Dose Level (mg/kg/day)
Mated females Females with live pups Length of gestation (days) Implants per female Live pups/litter on PND 0 % live pups on PND 0, LM Percent postimplantation survival to PND 0 Percent pup deaths (LM) PND 1–4 PND 5–21 a Dose
0
0.1
0.6
3.6/2.5a
25 23 22.4 16.6 14.8 99.4
25 25 22.3 16.6 15.8 99.0
25 25 22.4 16.4 15.0 95.4
25 25 22.5ns, b 16.8ns, c 15.5ns, d 99.5
88.9
95.0
92.2
92.2ns, e
1.3 0.5
1.7 0
1.5 0
1.9ns, f 1.0ns, g
level reduced between GD 17 and 20 analysis through the 3.6/2.5 mg/kg/day dose group yielded F(1,94) 5 0.9954, p 5 0.32. c Trend analysis through the 3.6/2.5 mg/kg/day dose group yielded F(1,93) 5 0.3226, p 5 0.57. d Trend analysis through the 3.6/2.5 mg/kg/day dose group yielded F(1,94) 5 0.5216, p 5 0.47. e Trend analysis through the 3.6/2.5 mg/kg/day dose group yielded F(1,94) 5 0.0424, p 5 0.58. f Trend analysis through the 3.6/2.5 mg/kg/day dose group yielded F(1,93) 5 0.2969, p 5 0.29. g Trend analysis through the 3.6/2.5 mg/kg/day dose group yielded F(1,93) 5 2.3444, p 5 0.06. ns Trend not statistically significant ( p . 0.05) through indicated dose level. PND 5 postnatal day. LM 5 litter mean. b Trend
FIG. 3. Average 6 SD body weights of F1 female (top) and male (bottom) pups from birth to weaning. Asterisks indicate statistically significant decreases in average pup weights. Adjustment of preweaning body weight by ANCOVA was done whenever including a covariate reduced the error mean square significantly (p < 0.05). Length of gestation and number of live pups per litter on PND 0 were used as a covariate for both sexes on PND 0, 4, and 11. No adjustments were done for either sex on PND 17 or 21. On PND 0, trend analyses yielded F(1, 93) < 3.686, p > 0.058 through the 3.6/2.5 mg/kg/day group. On PND 4, trend analyses yielded F(1, 93) < 1.979, p > 0.16 through the 3.6/2.5 mg/kg/day dose group. On PND 11, 17, and 21, trend analyses of average body weight of both sexes was statistically significant through the 3.6/2.5 mg/kg/day dose group, F(1, 93) > 19.971, p , 0.0001, but not through the 0.6 mg/kg/day group for either sex, F(1, 93) < 1.438, p > 0.23.
DEVELOPMENTAL NEUROTOXICITY OF EMAMECTIN BENZOATE ment was cleaned with water and dried between animals. Animals were placed in compartments and allowed to acclimate for 3 min in the presence of 72 dB white noise. Immediately following the acclimation period, each animal was exposed to a series of 50, 120-dB, 50-ms bursts of white noise, with a 5-s intertrial interval. The peak amplitude of each startle movement and the interval between acoustic stimulus and peak amplitude were recorded. Short-term learning using a passive avoidance paradigm was assessed in naive animals on PND 24 whereas PND 59 61 animals had been previously tested for open field motor activity, and long-term retention was then assessed in each of these sets of animals 1 week later. The tests were performed in a room with normal illumination and background white noise. Computer-controlled two-compartment shuttleboxes (San Diego Instruments, Inc.) were used. Each shuttlebox was used exclusively for one sex and all dosage groups were sequentially tested in each shuttlebox. Animals were placed in the compartment, which had a 25-W light attached to the outside (light turned off). The testing sequence consisted of a 10-s adaptation period, after which the light came on and the gate separating the lighted compartment from the dark compartment opened. Rats that entered the dark compartment broke an infrared beam, which automatically closed the gate. After the gate closed a scrambled shock [0.2 mA (young animals) or 1 mA (adult animals) for 1 s] was delivered to the cage floor. The maximum duration of each trial was 60 s. After each trial the animal was removed to a holding cage and held for a short interval (30–90 s) prior to retesting (if necessary). Animals were tested in consecutive trials until they reached criterion,
319
defined as remaining in the lighted compartment for two (pups) or three (adult) consecutive 60-s trials. The entire test using the same conditions was repeated 7 days later with the same animals. Statistical Methods Statistical analyses were done by an analysis of variance (ANOVA) or covariance (ANCOVA) [continuous variables (11)] or by an extended Mantel-Haenszel test [discrete variables (7)]. Nonparametric analyses were done in a rankit scale to normalize the data (2). Repeated measurements over time, specifically F0 and F1 body weights, were analyzed by orthogonal polynomials (9). For all other analyzed parameters, including average beams broken over the 30- or 60-min interval and average peak amplitude across the 10 trial blocks, a trend analysis was used to determine if there was a significant trend with increasing dosage across all treatment groups (12). Results were considered to be statistically significant if p > 0.05. When there was a significant trend, the data from the high dosage group were excluded and the trend test was repeated. This process was repeated until nonsignificance (p > 0.05) was achieved. RESULTS
F0 Generation There were no deaths and no adverse physical signs observed in F0 females during the study. There were statistically
FIG. 4. Average body weights and weight change of F1 males and females during the postweaning period (6 SD). For males, trend analysis of change in body weight from postnatal weeks 4 to 10 was statistically significant through the 3.6/2.5 mg/kg/day dose group, F(1, 93) 5 112.195, p , 0.0001, but not through the 0.6 mg/kg/day dose group, F(1, 93) 5 0.289, p 5 0.59. For females, trend analysis of change in body weight from postnatal weeks 4 to 10 was statistically significant through the 3.6/2.5 mg/kg/day, F(1, 93) 5 120.021, p , 0.0001, and 0.6 mg/kg/day, F(1, 93) 5 4.912, p 5 0.029, dose groups, but not statistically significant through the 0.1 mg/kg/day dose group, F(1, 93) 5 3.877, p 5 0.052. Because of the small magnitude in the effect on overall weight gain in the 0.6 mg/kg/day group (6% below control), this was not considered to be compound related.
320
WISE ET AL.
significant increases in average maternal weight gains between GD 6 and 20 in the 0.6 and 3.6/2.5 mg/kg/day groups (11% and 15% above control, respectively). Average body weights are shown in Fig. 2. There were no adverse effects on maternal body weight during this period in the 0.1 mg/kg/day
group. During the lactation dosing period, there were no EBrelated or statistically significant effects on average weight gain. There were no EB-related or statistically significant effects on F0 reproductive performance as assessed by the number of females with live pups, average length of gestation,
FIG. 5. Average open field motor activity 6 SEM in females and males for all 6 on various days of testing. Asterisks indicate statistically significant increases or decreases in average beams broken during the testing interval. For males on PND 13 and 17, trend analysis of overall horizontal activity was statistically significant through the 3.6/2.5 mg/kg/day dose group, F(1, 93) > 9.181, p , 0.003, but not through the 0.6 mg/kg/day dose group, F(1, 93) < 2.624, p > 0.055. For females on PND 13, 17, and 59, there were statistically significant trends through the 3.6/2.5 mg/kg/day dose group, F(1, 92–93) > 5.617, p < 0.020, but not through the 0.6 mg/kg/day dose group, F(1, 92–93) < 2.690, p > 0.053, except on PND 17, F(1, 93) 5 3.725, p 5 0.028.
DEVELOPMENTAL NEUROTOXICITY OF EMAMECTIN BENZOATE average implants per pregnant female, and percent postimplantation survival on PND 0 (Table 2).
F1 Generation: Physical Signs and Body Weight There were no EB-related effects on pup survival (Table 2) or external morphology (data not shown). Beginning on PND 6, intermittent head tremors were observed in some pups of the 3.5/2.5 mg/kg/day group. This sign progressed to whole-body tremors in all pups of all litters in this group by PND 25. In addition, hindlimb extension was first observed on PND 10 in some pups. This latter sign was followed by hindlimb splay between PND 15 to 26 in all high-dose pups. No
321
adverse physical signs were observed in the 0.6 or 0.1 mg/kg/ day groups. There were statistically significant decreases in average pup weights in the high-dose group during the preweaning period (Fig. 3). The effects were first apparent on PND 11 (14% and 11% below control in females and males, respectively), and pup body weights became progressively lower (relative to controls) with time (to 42% and 40% below control, respectively, on PND 21). There were no adverse effects on average preweaning body weights in the 0.6 or 0.1 mg/kg/day groups. During the postweaning period there were no deaths, but male and female animals in the 3.5/2.5 mg/kg/day group continued to demonstrate statistically significant decreases in average body weight gains (18% and 17% below control in fe-
322
WISE ET AL. males and males, respectively, between postnatal weeks 4 and 10, Fig. 4), such that at termination average body weights were 25% and 21% below control, respectively. Although there was a statistically significant (p 5 0.029) effect detected in the time–response trend analysis of average F1 female (but not male) body weight through the 0.6 mg/kg/day group, this was not considered to be EB related because overall weight gain during this postweaning period was only 6% below control. There were no treatment-related effects on average body weights during this period in the 0.1 mg/kg/day group. The adverse physical signs observed during the preweaning period continued for only a short time into the postweaning period. Specifically, whole-body tremors were observed in 7 of 174 animals up to at most PND 27 and hindlimb splay was observed in 45% of the animals up to at most PND 34. Thereafter, only one high-dose animal was observed to have any hindlimb abnormality. No adverse physical signs were observed during this time in animals of the 0.1 and 0.6 mg/kg/day groups, and no effects on pupillary function were observed in any of the EB-exposed group. F1 Generation: Developmental Signs and Behavioral Tests
FIG. 6. Average 6 SEM stereotypic time during open field motor activity testing in females (top) and males (bottom) on PND 13. Asterisks indicate statistically significant (p < 0.05) increases in overall stereotypic time in the 3.6/2.5 mg/kg/day group. Trend analysis through the high-dose group was statistically significant for both sexes, F(1, 93) > 5.336, p > 0.023, but not through the 0.6 mg/kg/ day group, F(1, 93) > 0.303, p > 0.493.
Preweaning and early postweaning period. During the preweaning and early postweaning period compound-related effects were detected only in the 3.5/2.5 mg/kg/day group. In the PND 13 open field motor activity test, there was an increase in overall average horizontal activity in male and female pups (Fig. 5). In view of the physical signs and body weight decrements discussed above, additional analysis of horizontal activity was conducted to better understand the increase in this parameter. One component of horizontal activity is stereotypic behavior, which is represented as repeated breaks of the same infrared photo beam(s). There was a significant increase in
FIG. 7. Historical control data for open field motor activity in PND 17 females. The graph points out the variablitiy of these data on this particular day and that the overall average value of the 0.6 mg/kg/day group fell between the values of control females from two studies conducted in this laboratory near the time of the present study.
DEVELOPMENTAL NEUROTOXICITY OF EMAMECTIN BENZOATE stereotypic time in both sexes (Fig. 6). This finding provides an explanation for the increase in horizontal activity and is consistent with the observed tremors in this high-dose group. Average horizontal activity was significantly decreased in both sexes on PND 17 (41% and 30% in females and males, respectively), but there were no statistically significant effects detected on PND 21 (Fig. 5). There was a statistically significant decrease in average horizontal activity of females in the 0.6 mg/kg/day group on PND 17, but this was not considered to be compound related because the value was within the range of historical control PND 17 females in this laboratory (Fig. 7). In the auditory startle habituation test on PND 22 there was a significant decrease in the average peak response amplitude across the 50 trials in the 3.6/2.5 mg/kg/day group (approximately 74% below control in both sexes, Fig. 8). There was no apparent effect on average time to peak response (data not shown). There were no effects on these parameters in the 0.1 and 0.6 mg/kg/day groups. There were no effects in any EB-exposed groups on learning/short-term retention or long-term retention as assessed on PND 24 and 31 in the passive avoidance test (Fig. 9). Postweaning period. The appearance of developmental landmarks was delayed only in the high-dose group. The estimated mean day of occurrence of vaginal canalization was delayed 3.7 days and preputial separation was delayed 3.6 days compared to the control group, but there were no effects in
323
the other EB-exposed groups (vaginal canalization: 33.7, 33.4, 33.1, and 37.4 days; preputial separation: 44.8, 44.9, 44.8, and 48.4 days in the control, low-, mid-, and high-dose groups, respectively). Average horizontal activity was significantly decreased only in females on PND 59 6 1 (18% below control, Fig. 5). In the auditory startle habituation test there were significant decreases of average peak response amplitude in both females and males (46% and 31% below control, respectively, Fig. 8), but no apparent effect on average time to peak response (data not shown). As in the preweaning period, there were no effects on adult animals in the passive avoidance test (Fig. 9).
Organ Weight and Histological Examinations There were slight decreases in average absolute brain weights of animals in the 3.6/2.5 mg/kg/day group on PND 11 (4–8% below control) and postnatal week 8 (4–10% below control, Table 3); however, these effects were considered to be secondary to the larger decreases in absolute body weight of these animals, because there were slight to moderate increases in average brain weights expressed as a percent of body weight. Morphometry of brain structures showed no mensurational differences between control and EB-exposed brains on PND 11 or 59–60 (Table 4). There were no com-
FIG. 8. Average peak startle response amplitude 6 SEM over 50 trials in females and males on two separate days of testing. On PND 22 and 58–60 for both males and females, the trend through the 3.6/2.5 mg/kg/day dose group was statistically significant, F(1, 91–93) > 5.905, p < 0.017, but not through the 0.6 mg/kg/day group, F(1, 91–93) < 1.392, p > 0.448.
324
WISE ET AL.
FIG. 9. Results of passive avoidance testing in weanlings (top) and as adults (bottom). There were no meaningful differences between control and EB-exposed groups.
TABLE 3 ABSOLUTE AND RELATIVE BRAIN WEIGHTS OF F1 ANIMALS Postnatal Day 11 Females
Brain weight (g) Control 0.1 mg/kg/day 0.6 mg/kg/day 3.6/2.5 mg/kg/day Relative brain (% body wt.) Control 0.1 mg/kg/day 0.6 mg/kg/day 3.6/2.5 mg/kg/day
Postnatal Day 59–60 Males
Females
Males
Mean
S.D.
(N)
Mean
S.D.
(N)
Mean
S.D.
(N)
Mean
S.D.
(N)
1.09 1.07 1.09 1.05
0.04 0.06 0.05 0.06
(10) (12) (12) (13)
1.15 1.10 1.13 1.06
0.08 0.06 0.06 0.09
(12) (13) (12) (11)
1.92 1.85 1.92NS 1.73S
0.06 0.12 0.06 0.05
(12) (12) (11) (11)
1.98 2.05 2.02NS 1.90S
0.11 0.10 0.07 0.08
(10) (12) (12) (13)
3.99 4.14 4.16 4.24
0.35 0.49 0.36 0.52
4.06 4.07 3.92 4.49
0.38 0.46 0.34 0.64
0.76 0.73 0.76NS 0.90S
0.07 0.05 0.04 0.07
0.50 0.50 0.52NS 0.60S
0.04 0.04 0.03 0.07
NS Trend not statistically significant through indicated dose level [females: F(1, 42) < 0.419, p > 0.52; males: F(1, 43) < 1.062, p > 0.31]. S Trend statistically significant through indicated dose level [females: F(1, 42) > 34.435, p , 0.0001; males: F(1, 43) > 15.415, p < 0.0003].
DEVELOPMENTAL NEUROTOXICITY OF EMAMECTIN BENZOATE
325
marks. In terms of the behavioral tests, open field motor activity was increased on PND 13 due to the detection of tremors by the testing apparatus. Decreased motor activity was then again observed in adult females only, but the decrease, although statistically significant, was fairly minor. In the auditory startle habituation task there were decreases in the average peak amplitude on PND 22 and as adults (PND 58–60). Although the etiology of this effect is not known, it should be noted that the decreases in average peak response amplitude in pups and adults of the high-dose group were observed in association with decreases in average body weight (approximately 40% and 25% below control, respectively). There were no morphological changes detected by light microscopy in the EB-exposed offspring on PND 11 or as adults. Other unpublished studies from this laboratory have shown histomorphological degenerative changes in the brains of adult animals orally administered EB in the dosage range used in the present study (see Introduction). We mention this only to make the point that this laboratory had the experience and aptitude to detect histomorphological brain changes, but none were observed. The lack of similar findings in the EBexposed offspring of the present study may be due to a lower total dosage received by the pups via the milk, and/or significantly different pharmacokinetics of the compound (and/or its metabolites). Alternatively, rat pups may be inherently less sensitive than adults to EB-induced brain degenerative changes. Addtional studies would be needed to answer these questions. In relation to the study of Poul (8), which assessed preweaning pup parameters in rats following exposure to ivermectin during gestation and lactation, the results of this study show that EB is less toxic to the F1 generation. In the study with ivermectin, dose levels of 2 and 4 mg/kg/day resulted in significant mortality before weaning (31% in the former group) whereas no such mortality occurred at EB dose levels up to 3.6 mg/kg/day administered during gestation and 2.5 mg/ kg/day during lactation. Poul (8) also demonstrated that preweaning behavioral development of pups, as assessed by the Cincinnati test battery (15), was impaired or delayed at all dose levels of ivermectin tested (1, 2, and 4 mg/kg/day). In the 1 mg/kg/day group there were delays in cliff avoidance, pivoting activity, and forward locomotion, but no other effects on growth or behavior were noted. The results of the present study showed that EB exposure of 3.6/2.5 mg/kg/day to the dam produced some preweaning and postweaning behavioral
pound-related changes in the histomorphology of brain, spinal cord, or peripheral nerve of PND 11 or adult animals in the 3.6/2.5 mg/kg/day group. DISCUSSION
The purpose of this study was to assess the potential of EB to produce developmental neurotoxicity in rats exposed during organogenesis and early postnatal life (i.e., GD 6 to LD 20). Maternal animals treated with EB showed significant increases in average weight gain during gestation in the midand high-dose groups, but there were no other signs of toxicity. The toxicological significance of these increased weight gains in the 3.6/2.5 mg/kg/day group is not known, but slightly increased body weight is a consistent feature of EB exposure via oral gavage to pregnant as well as nonpregnant rats. The NOEL (No Observed Effect Level) for maternal F0 animals in the present study was 0.1 mg/kg/day based on the increased weight gains, but the NOAEL is considered to be 3.6 mg/kg/ day (dose level was lowered to 2.5 mg/kg/day on GD 17–20, after the effect had occurred). This NOAEL for maternal toxicity is considered to be fairly accurate because in the oral gavage rat developmental toxicity study there were doserelated decreases in average weight gain during the latter portion of gestational exposure at dose levels of 4 and 8 mg/kg/ day (unpublished results). This is the first study from our laboratory to examine the behavioral effects of prenatal and early postnatal exposure of EB, or any avermectin, in the F1 generation of rats. Multigeneration studies in rats were conducted in this laboratory with ivermectin (6) and abamectin (5). Postnatal toxicity characterized by decreased weight gain and mortality was found in F1 offspring at doses < 0.4 mg/kg/day of both compounds. Further studies using radiolabeled ivermectin indicated that high drug concentrations in the milk of exposed dams led to high drug levels in the plasma and brain of the F1 offspring (relative to adult rats) and were probably responsible for the observed toxicity (6). The neurotoxicological effects in the F1 offspring of the present study were observed only in the 3.6/2.5 mg/kg/day group and consisted of tremors with hindlimb extensions/ splaying and effects in two of three automated behavioral tests. These effects were observed in conjuction with moderate to marked decreases in preweaning weights, postweaning weight gains, and delays in attainment of developmental land-
TABLE 4 MORPHOMETRY ON NEURAL STRUCTURES IN F1 RATS Control
PND 11 Cerebral cortex Hippocampus Cerebellum PND 59–60 Cerebral cortex Hippocampus Cerebellum a Mean
3.6/2.5 mg/kg/day
Females
Males
Females
Males
3.2a (2.8–3.7)b 2.8 (2.3–3.4) 1.1 (1.0–1.4)
3.1 (2.8–3.9) 3.0 (2.7–3.6) 1.2 (0.9–1.5)
3.6 (2.7–4.7) 2.9 (2.5–3.3) 1.2 (1.0–1.4)
3.2 (2.8–3.8) 2.9 (2.3–3.3) 1.1 (0.9–1.2)
4.0 (3.6–4.7) 4.6 (3.8–5.0) 1.9 (1.5–2.5)
4.2 (3.7–4.8) 5.1 (4.7–5.5) 1.9 (1.5–2.1)
4.4 (4.1–4.7) 4.7 (4.4–5.1) 1.9 (1.7–2.1)
4.1 (3.5–4.6) 4.9 (4.2–5.4) 2.0 (1.4–2.6)
value in micrometer units; scale is different between days. of values.
b Range
326
WISE ET AL.
effects in addition to clear effects on growth, but a NOAEL for developmental neurotoxicity was determined to be 0.6 mg/ kg/day. It is clear from these studies that the avermectins are able to disrupt postnatal development in rats when administered during gestation and lactation. The susceptibility of rat pups to the avermectins may be related to the levels of P-glycoprotein as suggested by work in mice. The sensitivity of mice to ivermectin neurotoxicity was shown to be due to the presence of a P-glycoprotein synthesized by the multidrug resistance gene mdr1a (10). Mice homozygous for a disruption of the mdr1a gene have no physiological, anatomical, or histological abnormalities but are 50to 100-fold more sensitive to orally administered ivermectin than genetically matched mdr1a (1/1) mice. This increased sensitivity was correlated with a near absence of immunohistochemically detectable P-glycoprotein in the brains of these mice (10). More recently, our laboratory has shown that P-gly-
coprotein in brain capillary endothelia and intestine of rats is nearly absent at birth and reaches mature levels at about 5 weeks of age (4). The behavioral test battery utilized in this study was designed to fulfill the US EPA guidelines for the assessment of developmental neurotoxicity (14). The battery revealed effects on motor activity and sensorimotor functions that were concomitantly associated with tremors and depressed body weight. These results, together with the results of a companion study with acrylamide (16), suggest that for some compounds the developmental neurotoxicity protocol outlined by the EPA does not necessarily identify behavioral or neurohistological effects at dose levels lower than those that produce more routine or obvious endpoints of toxicity. However, the utility of this study design, and potential modifications of it, can only be assessed after a larger number of different compounds have been similarly tested in such a rigorous study design.
REFERENCES 1. Campbell, W. C.: Ivermectin and abamectin. New York: SpringerVerlag Inc.; 1989. 2. Harter, H. L.: Expected values of normal order statistics. Biometrika 48:151–165; 1961. 3. Köig, J. F. R.; Klippel, R. A.: The rat brain. Huntington, NY: R.E. Krieger Publishing Co., Inc.; 1963. 4. Lankas, G. R.; Cartwright, M.; Minsker, D.; Cukierski, M.: Postnatal development of P-glycoprotein, a component of the blood– brain and intestinal barriers, in rats. Fundam. Appl. Toxicol. Suppl. 30:263; 1996. 5. Lankas, G. R.; Gordon, L. R.: Toxicology. In: Campbell, W. C., ed. Ivermectin and abamectin. Springer-Verlag New York Inc.; 1989:89–112. 6. Lankas, G. R.; Minsker, D. H.; Robertson, R. T.: Effect of ivermectin on reproduction and neonatal toxicity in rats. Food Chem. Toxicol. 27:523–529; 1989. 7. Mantel, N.: Chi-square tests with one degree of freedom. J. Am. Stat. Assoc. 58:690–700; 1963. 8. Poul, J.-M.: Eeffects of perinatal ivermectin exposure on behavioral development in rats. Neurotoxicol. Teratol. 10:267–272; 1988. 9. Robson, D. S.: A simplified method for constructing orthaogonal polynomials when independent variable is unequally spaced. Biometrics 15:187–191; 1959 10. Schinkel, A.; Smit, J.; van Tellingen, O.; Beijnen, J.; Wagenaar,
11. 12. 13. 14.
15. 16.
E.; van Deemter, L.; Mol, C.; van der Valk, M.; Robanus-Maandag, E.; te Riele, H.; Berns, A.; Borst, P. Disruption of the mouse mdr1a P-glycoprotein gene leads to a deficiency in the blood– brain barrier and to increased sensitivity to drugs. Cell 77:491– 502; 1994. Snedecor, G. W.; Cochran, W. G.: Statistical methods, 7th ed. Ames, IA: Iowa State University Press; 1980. Tukey, J. W.; Ciminera, J. L.; Heyse, J. F.: Testing the statistical certainty of a response to increasing doses of a drug. Biometrics 41:295–301; 1985. Turner, M. J.; Schaeffer, J. M.: Mode of action of Ivermectin. In: Campbell, W. C., ed. Ivermectin and abamectin. New York: Springer-Verlag Inc.; 1989:73–88. U.S. Environmental Protection Agency: Pesticide assessment guidelines—Subdivision F, Hazard evaluation: Human and domestic animals. Addendum 10—Neurotoxicity Series 81, 82, and 83. Publication No. PB-91-154617, National Technical Information Service, Department of Commerce, Spingfield, VA, 1991. Vorhees, C. V.: Behavioral teratogenicity testing as a method of screening for hazards to human health: A methodological proposal. Neurobehav. Toxicol. Teratol. 5:469–474; 1983. Wise, L. D.; Gordon, L. R.; Soper, K. A.; Duchai, D. M.; Morrissey, R. E.: Developmental neurotoxicity evaluation of acrylamide in Sprague–Dawley Rats. Neurotoxicol. Teratol. 17:189–198; 1995.