- Email: [email protected]
Sensitive periods for lead-induced behavioral impairment (nonspatial discrimination reversal) in monkeys
TOXICOLOGY
AND
APPLIED
PHARMACOLOGY
102,10!-109
(1990)
Sensitive Periods for Lead-Induced Behavioral Impairment (Nonspatial Discrimination Reversal) in Monkeys DEBORAH C. RICE’ AND STEVEN G. GILBERT Tosicology
Research
Division,
M/plfare Canada,
Bureau Banting
ofChemical Safety. Food Directorate, Building,
Received
April
Tunnqv’s
Pasture.
21, 1989; accepted
Ottawa,
.4trgust
Health Ontario,
Protection Branch, Canada KIA OL2
Health
and
22, 1989
Sensitive Periods for Lead-Induced Behavioral Impairment (Nonspatial Discrimination Reversal) in Monkeys. RICE, D. C., AND GILBERT, S. G. (1990). Toxicol. Appl. Pharmacol. 102, 101-109. A total of 52 nursery-reared monkeys (Macacafascicularis) were dosed orally with 1.5 mg/kg/day of lead on one of four dosing regimens (13 monkeys/group): Group 1, vehicle only: Group 2, dosed with lead continuously from birth; Group 3, dosed with lead from birth to 400 days of age and vehicle thereafter: and Group 4, dosed with vehicle from birth to 300 days of age and lead thereafter. This dosing regimen allowed evaluation ofdifferential infant vulnerability as well as reversibility of the behavioral toxicity of lead. Blood lead concentrations averaged 3-6 pg/dl when monkeys were not being exposed to lead, 32-36 fig/d1 when being dosed with lead and having access to infant formula, and 19-26 pg/dl during lead exposure after weaning from infant formula. When monkeys were 5-6 years old, they were tested on a series of nonspatial discrimination reversal tasks: form, form with irrelevant color cues, color with irrelevant form cues, and alternating form and color. Group 2 exhibited the greatest degree of impairment compared to controls. Group 4 also exhibited impaired performance, although less marked than that of Group 2. Group 3 was not impaired on this series of tasks. These results confirm findings observed in other monkeys exposed continuously to lead and suggest that while exposure beginning after infancy produces impairment, exposure during infancy as well exacerbates the effeCt.
6 1990 Academic
Press,
Inc.
INTRODUCTION
Past research from this laboratory has documented reproducible behavioral impairment on a variety of tasks as a result of lead exposure continuously from birth in the monkey (Rice, 1988, 1985, 1984; Rice and Karpinski, 1988; Gilbert and Rice, 1987; Rice and Gilbert, 1985; Rice et al., 1979; Rice and Willes, 1979). In those studies, blood lead levels of treated monkeys ranged between 11 and 50 pg/dl for different groups of monkeys. Moreover, research from another laboratory reveals that exposure to higher doses of lead during the first year of life (blood lead levels ’ To whom correspondence should be addressed.
of 45-300 pg/dl) results in permanent behavioral impairment (Bushnell and Bowman, 1979; Levin and Bowman, 1986, 1983). Studies in rodents suggest that exposure beginning before maturity, but after weaning, results in behavioral impairment on intermittent schedules of reinforcement (CorySlechta et al., 1985, 1983). However, the issue of a potential sensitive period for lead-induced behavioral impairment has not previously been explored in monkeys. In a series of experiments, monkeys were exposed to lead at levels that have proved sufficient to produce clear behavioral impairment as a result of continuous exposure from birth (blood lead levels of 20-35 pg/dl), while still being in the relevant range for environ101
004 1-008X/90 $3.00 Copyright 0 1990 by Academic Press, Inc. All rights of reproduction m any form reserved.
102
RICE
AND
mentally exposed children. One group of monkeys was exposed continuously from birth, to replicate past experiments and serve as a positive control. Another group was exposed only early in life, while a third group was exposed beginning after infancy. These monkeys have been and are being tested at present using behavioral tasks previously determined to be sensitive to lead-induced behavioral impairment. The results of a series of nonspatial discrimination reversal tasks, with and without irrelevant cues, are reported here. METHODS Subjects and dosing. Fifty-two monkeys (Macaca fasciczdaris) used in the present study were born at the Health Protection Branch breeding colony over a period of 2 years, all from different mothers. Infants were assigned to one of four treatment groups (13 monkeys/ group) in such a way that groups were balanced for age, sex, and paternity. A total of 20 males fathered infants in this study with three occurrences of a male contributing two infants to the same dose group. Infants were separated from their mothers within 12 hr of birth and reared in the Neurotoxicology Primate Colony. They were dosed with a lead acetate dissolved is a 0.05 MNa#ZO, or NaZCOJ solution (vehicle) according to the following regimens: Group I. vehicle from birth onward (8 females, 5 males): Group 2. lead from birth onward (9 females, 4 males): Group 3. lead from birth to 400 days of age and vehicle thereafter (9 females, 4 males); and Group 4. vehicle to 300 days of age and lead thereafter (8 females, 5 males). Infants were dosed by mixing the lead solution with milk substitute formula and delivering it slowly into the back of the mouth using a syringe: by several months of age monkeys would reliably eat gelatin capsules, at which point dosing was accomplished by offering the monkey a capsule containing the dose. Monkeys were dosed 5 days per week at the equivalent of 1.5 mg/kg/day, or 2.1 mg/kg for each of 5 days. Monkeys were housed individually and were socialized with age and group-matched peers beginning within 3 weeks of birth. Monkeys were exercised and socialized for longer periods as they developed until they reached adulthood. Beginning at 3 to 4 years of age. males and females were no longer exercised together. Previous research in our laboratory indicates that dosing beginning at birth results in blood lead values that increase from birth to a peak at about 100 days of age. and decrease after withdrawal of infant formula to a new
GILBERT steady state level over the next IOO- 150 days. In the present study, Groups 2 and 3 were weaned from infant formula at 200 days of age as in previous studies. Group 4 was not weaned until 500 days of age. 200 days after dosing commenced. This ensured that the pattern of blood lead values would be similar between Groups 2 and 4, but displaced in time with respect to developmental period. Half the monkeys in Group I were weaned at 200 days, and half at 500 days. Blood sampling and anal.vsis. Blood was drawn from femoral vein for blood lead analysis every 2 weeks until 500 days of age, every month until 3 years of age. and every 3 months thereafter. The method of blood lead analysis and quality control procedures have been described previously (Rice, 1985). Free erythrocyte protoporphyrin analysis was scheduled to correspond with peak blood lead values, with the control group (Group I) sampled as long as the longest treated group. Group 2 was analyzed from birth to 450 days ofage, Group 3 from birth to 650 days of age, and Group 4 from birth to 550 days of age. The analytical method was a modification of a method described previously (Rice and Willes, 1979). Hematological parameters were assessed on the same schedule as blood lead levels, while blood biochemistry (SMA- 18) was analyzed every 3 months. Blood lead and FEP data were averaged for each monkey individually and then for each group as a whole for days 100-300 after birth, which corresponded to highest levels for Groups 2 and 3, and before institution ofdosing for Group 4. Blood lead levels were then averaged for Group 2 after they stabilized postweaning, for Group 3 after they stabilized following withdrawal from lead, and Group 4 at peak blood lead levels (comparable to Days 100-300 for Groups 2 and 3). Data for Group 1 spanned the range for the other three groups. Blood lead values were also averaged for 3-4 and 4-5 years of age, to complete the blood lead history prior to initiation of the present study. FEP values were averaged for all groups for 100-300 days of age, and for each treated group over the last 100 days of sampling for FEP. which were chosen to include the highest levels for Groups 2 and 4 and to allow for decline after cessation of dosing in Group 3. Group 1 samples span the age range of all treated groups. Behavioral methods. When monkeys were 5-6 years old, they were tested on a series of nonspatial discrimination reversal tasks. Details of behavioral apparatus and testing procedures have been described previously (Rice. 1985). Briefly, the monkey sat facing a Plexiglas panel with two clear pushbuttons. placed at the same level vertically. which could be backlit with stimuli. A tube for delivery of juice was centered between them. Monkeys were tested on a series of four nonspatial descrimination reversal tasks (Table 1). For each task, when the monkey responded with 0 or I error in a nonoverlapping lo-trial block, the positive and negative stimuli were reversed for the next block. A total of 15 reversals plus the initial acquisition were run for each task. Task 1 was a form dis-
SENSITIVE
103
PERIODS FOR LEAD NEUROTOXICITY TABLE
I
SUMMARYOFDOSINGREGIMENSANDDISCRIMINATIONREVERSALTASKS Correct stimulis Type of task Task I
Form discrimination
Task 2
Form discrimination with irrelevant color cues
Task 3
Color discrimination with irrelevant form cues
Task 4
Alternate form and color
Acquisition, even number reversals
Stimuli Square and triangle with red surround Square and triangle with red or green surround, all permutations Same as Task 2 Square on gray or triangle on gray vs red or green
Odd number reversals
Square on red
Triangle on red
Square on red or square on green
Triangle on red or triangle on green Green with square or green with triangle Red or green
Red with square or red with triangle Square or triangle
Dosing regimens GTOUP 1 Group 2 Group 3 Group 4
Dosed Dosed Dosed Dosed
with with with with
vehicle from birth onward lead from birth onward lead from birth to 400 days, vehicle thereafter vehicle from birth to 300 days of age, lead thereafter
crimination task with a white square and triangle superimposed on red backgrounds. Task 2 was a form discrimination with irrelevant color cues; the forms were the same as in Task I, with a red background on one button and a green background on the other. All four permutations of form and color were presented. Task 3 was a color discrimination task with irrelevant form cues. The stimuli were the same as in Task 2, but the relevant stimulus dimension was color rather than form. In Task 4, the positive stimulus alternated between color and form on successive reversals. Stimulus pairs were square on gray or triangle on gray vs red or green. Positive stimuli were therefore red or green for one reversal, square or triangle on the next. Monkeys were tested 100 trials per day. 5 days per week. Monkeys had previously performed on a multiple fixed interval-fixed ratio (Rice and Gilbert, 1987) and DRL schedules of reinforcement. One female monkey from Group 3, who had a constant perineal swelling, learned to press the pushbuttons after a month of intensive training. However, this monkey refused to perform the multiple fixed internal-fixed ratio schedule, and so was eliminated from the study. This is the only monkey to be dropped from a study in this laboratory for this reason out of more than 150 individuals. Two male control monkeys died from acute illnesses before being tested on this series of discrimination reversal tasks. Dala analysis. Data analysis was by means of Good’s statistic (Good, 1979). This is a nonparametric randomization test that includes both differences in the mean
and the group variability in the analysis, in order to take into account that there may be both responders and nonresponders in the treated group(s). Since past experience in this laboratory revealed such an effect in lead-exposed monkeys, equal weight was given to differences in mean and variability in the present analysis (i.e., 19= 0.5). The main variable of interest was the total number of incorrect responses across all reversals of each task. Group 1 was compared to Group 2 first, since Group 2 represents a replication of previous positive results on nonspatial discrimination reversal performance, and would presumably be impaired to the greatest degree. Moreover, impairment in Groups 3 or 4 in the absence of impairment of Group 2 would be difficult to interpret. If that comparison was significant for a Task (p < 0. lo), Group 1 was compared to Group 3 and Group 4. If the total number of errors across all reversals of a task was significant @ < O.lO), reversals l-5. 6-10, and I l-15 were analyzed as separate blocks to determine whether treated monkeys were more impaired on early or late blocks. Analogous comparisons were also made for all tasks combined to increase the sensitivity of the analysis (Edgington, 1980). In addition, acquisition of performance was assessed separately, again comparing Group 1 to Group 2 before comparison of Group I to the other groups. For any comparison that was significant (p < 0. IO) for the total number of errors across all reversals, further analysis was performed to assessattention to irrelevant cues. Such an analysis has proved useful in the past in
104
RICE
AND
GILBERT
TABLE
BLOOD LEAD
FEP
AND
2
CONCENTRATIONS
Blood lead (mean f SE) (pg/lOO ml whole blood) Group Age 100-300 days
Selectedages
3-4years
FEP (mean + SE) (&ml RBC) 4-5
years
100-300
days
Selectedages
Age (days) I. Control 2.
3.
4.
4.9
+ 0.7
2.9
t 0.2
(400-800)
f
1.3*
(500-700)
Lead from birth
32.3
2
3.1'
18.9
Leadto 400 days
36.3
k
2.3*
5.6
+ 0.7+
(600-800)
32.9
+ 2.1*
(400-600)
Lead from 300 days
6.0?
1.1
Age (days) 3.0 kO.4
23.8
3.8
22.2
k
l.8*
+ 0.6 t 1.0’
2.7
25.7
3.3
? 0.5
75?
9
69k
8
(350-650)
+ 1.5:
252+55+
?z0.6
431*71*
aa+
(550-650)
89t16
270?50*
(450-550,
22.0* 1.6*
132 +30
(350-450)
* Different from control. p < 0.000 1, Wilcoxin test. + Different from control, p < 0.005. determining the types of errors made by lead-exposed monkeys on discrimination reversal tasks (Rice, 1985; Gilbert and Rice, 1987). The ratio of incorrect responses on each button and, for Tasks 2 and 3, the ratio of incorrect responses on irrelevant stimuli was calculated by number incorrect on button (or stimulus) number incorrect on button (or stimulus) + number correct on button (or stimulus) The ratio of trials on which a monkey avoided a particular button or irrelevant stimulus was calculated by number incorrect on other button (or stimulus) number incorrect on other button (or stimulus) + number correct on button (or stimulus) The totals of the larger ratio for each pair (i.e.. right and left incorrect; triangle avoid and square avoid, etc.) over the course of each task were calculated in order to circumvent the fact that different individuals may show preference for different irrelevant stimuli. A combined measure of attention to irrelevant cues was calculated by adding the individual measures. This overall measure was compared using Good’s statistic; for significant overall comparisons each individual measure was compared.
RESULTS Blood lead values were approximately 3236 pg/dl when monkeys were dosed with lead and fed milk formula, and 19-26 pg/dl after the milk formula diet was discontinued (Table 2). When monkeys were not exposed to
lead, blood lead values were 3-6 pg/dl. FEP values were statistically significantly elevated when lead values were in the 32-36 pg/dl range. At blood lead values of 19 pg/dl, some but not all individuals of Group 2 (the only group for which FEP values were determined at this blood lead value) exhibited elevated FEP’s, resulting in an increased value that was not statistically significant for the group as a whole. Hematology and blood biochemistry were within normal limits for all monkeys; group comparisons have not been performed. Analysis of the total number of errors revealed that Group 2 was impaired relative to Group 1 over the entire experiment (Table 3), while Group 4 was impaired to a lesser extent than Group 2, and Group 3 was unimpaired. Group 2 was impaired over all reversals for Tasks 1, 2, and 3, while Group 4 was impaired at the beginning of the experiment (Task I), and on Task 3, when the relevant stimulus dimension was changed from form to color (Fig. 1). Task 4 proved extremely easy for all monkeys, with 15 reversals completed in 3-4 sessions. Group 2 was not impaired under these circumstances. Analyzing the performance by blocks of reversals revealed that on Task 1, Group 2 was impaired on the first two blocks, but not the
SENSITIVE
PERIODS
FOR TABLE
P-VALUES
’ (-)
Represents
3
l-15
l-5
6-10
11-15
0.03 NS 0.07
0.07 -
0.005
0.08
0.02
0.05 NS
0.01 NS 0.06
0.04 -
0.02 0.09
NS NS
0.03 NS NS
0.05 NS NS
0.04 -
0.08 -
NS -
0.006 NS 0.05
0.004 -
0.00 1 0.009
NS -*
NS
NS -
0.08
NS
NS
comparisons
105
NEUROTOXICITY
(GOOD’S RANDOMIZATION TEST) FOR PERFORMANCE ON A SERIES OF NONSPATIAL DISCRIMINATION REVERSAL TASKS
Acquisition Overall 1 vs2 1 vs3 1 vs4 Task 1 1 vs2 1 vs3 1 vs4 Task 2 1 vs2 1 vs3 1 vs4 Task 3 1 vs2 1 vs3 1 vs.4 Task 4 I vs2 I vs3 I vs4
LEAD
-
-
NS 0.05
-
not performed.
third, while Group 4 was marginally impaired with the same pattern (Fig. 2). On Task 2, when irrelevant cues were introduced, Group 2 was impaired on all three blocks of reversals (Fig. 3). On Task 3, Group 2 was impaired on blocks 1 and 2 as in Task 1, while Group 4 was impaired on blocks 2 and 3 (Fig. 4). When all tasks were combined, there is evidence that Group 2 was impaired over all three blocks of reversals, while Group 4 was impaired only on the first two blocks. Acquisition of the task was retarded only for Group 2 on Task 2, when irrelevant cues were introduced. Analysis of attention to irrelevant cues revealed that there was a tendency for Group 2 to have more total perseverative errors than Group 1 over the three tasks for which the total number of errors differed (Table 4). There was no strong evidence that this was true for Group 4, however (Task 1, p = 0.33, Task 3, p = 0.11). Further comparison of
Group 1 vs Group 2 revealed that on Task 1, treated monkeys had greater ratios of responding incorrectly on the left button but not for avoiding either button. For Task 2, treated monkeys responded on the left and avoided the right button, and attended to the irrelevant color stimuli more than controls. For Task 3, treated monkeys switched position strategies, responding more on the right and avoiding the left relative to controls; there was no strong indication that they attended to the irrelevant form stimuli more than controls. DISCUSSION The decrease in blood lead levels observed after withdrawal of infant formula replicates previous research from this laboratory (Rice et al., 1979; Rice, 1985). The fact that the same pattern of blood lead values was ob-
106
RICE AND GILBERT TASK
1
TASK
2
3500-
v) 2 2
2750-
:
2250-
8 g Y
1750. 1250-
8
7504
-
j
25;$:...
!I.
2
TASK
i
1100
E
900
500 300 100
. 1.
“m
Fe*
3
:
**a ir
!i:
1. . *
: : i:
birth, with peak blood levels of 15 or 25 pg/ dl, and steady state postweaning levels of 1 1 or 13 pg/dl, were impaired on a nonspatial form discrimination reversal task and a color discrimination reversal task that followed immediately. They were not impaired on the third task, a form discrimination reversal task with irrelevant color cues. In the present experiment, Group 4, dosed beginning at 300 days of age, was impaired on two of the tasks, to a lesser degree than Group 2. Group 3, dosed from birth to 400 days, was not impaired. In general, the increased total number of errors over all reversals was the result of more errors toward the beginning rather than at the end of the task. This has been observed by
:
2750
i.
2250
GROUP
FIG. 1. Number oferrors over all 15 reversals for Tasks 1, 2, and 3. Each dot represents an individual monkey. Group 1, dosed with vehicle from birth onward: Group 2, dosed with lead from birth onward; Group 3, dosed with lead from birth to 400 days of age and vehicle thereafter; Group 4, dosed with vehicle from birth to 300 days of age and lead thereafter.
served in Groups 2 and 4, associated with diet but not age, indicates that the decrease in blood lead levels observed routinely in our laboratory after weaning is an effect of diet rather than age. Similarly, the statistically significant increase in FEP levels at blood lead levels above 30 pg/dl, with some individuals exhibiting an increase at 20 pg/dl, replicates previous finding from this laboratory (Rice, 1985) and is consistent with data in children (CavaIleri et al., 1981; Bush et al., 1982). The deficits observed in Group 2 on the first three tasks replicates effects observed in monkeys with lower blood lead levels (Rice, 1985). In that study, monkeys dosed from
REVERSALS
l-5
REVERSALS
6-10
REVERSALS
11-15
1750 1250
900-
900 1 700
1
*
1
-
500 300
FIG. 2. Number oferrors for Reversals l-5,6-10, 1I - 15 on Task 1. Symbols as in Fig. 1.
and
SENSITIVE
107
PERIODS FOR LEAD NEUROTOXICITY
ACQUISITION
REVERSALS
1-5
i;[Z::. .$ :..... ;E”-‘;. 8: 1
2
3
4
GROUP
1
2
3
4
FIG. 3. Number oferrors for the Acquisition, Reversals l-5, 6-10, and 1 l-1 5 on Task 2. Symbols as in Fig. 1.
other investigators at higher blood lead levels (Bushnell and Bowman, 1979) as well as in the group of monkeys discussed previously (Rice, 1985). Group 2 was impaired on the acquisition of Task 2, when irrelevant cues were introduced. Such impairment in acquisition was not observed at lower doses on nonspatial discrimination reversal, but was observed in those same monkeys on a series of spatial discrimination reversal tasks when irrelevant cues were introduced (Gilbert and Rice, 1987). It is apparent from the figures that there are individual monkeys in treated Groups 2 and 4 that are severely impaired, while other individuals are not different from controls. This results in increased variability within these treated groups. In fact, the greater statistical significance in Group 2 than Group 4 appears to result from the increased number of individuals with aberrant performance in the former group, rather than a difference in the magnitude of the deficit within impaired individuals. This phenomenon has been observed repeatedly in our laboratory (Rice, 1985; Gilbert and Rice, 1987; Rice and Kar-
pinski, 1988; Rice et al., 1979) as well as in lead-exposed rats (Cory-Slechta et al., 1985, 1983). In the present study, the aberrant performances across tasks were typically due to different individuals, although this was less apt to be true within tasks. Impaired performance in the presence of distracting stimuli or changes in environmental contingencies (as in the requirement for reversal of the previously correct response) appears to be a hallmark of lead toxicity in the monkey (see Rice, in press, for review). Such procedures have proved more sensitive than the acquisition of a behavioral task, except perhaps where contingencies change, as in the deficit observed on Task 2 by Group 2. In the present experiment, Group 2 monkeys attended to position as an irrelevant stimulus over all three tasks for which they showed an increase in total number of errors. In addition, they attended to irrelevant colors when irrelevant stimuli were introduced in Task 2, but not to irrelevant forms in Task 3, when irrelevant cues were familiar. These results are consistent with previous results in monkeys with lower blood levels (Rice, 1985; Gil-
108
RICE AND GILBERT REVERSALS
Data from the present experiment suggest that lead exposure very early in life is not necessary for lead-induced behavioral impairment. Three hundred days of age in the monkey (when dosing of Group 4 commenced) corresponds roughly to 3.0-3.5 years of age in the child in terms of development. In the human, blood lead values peak at approximately 2 years of age (Mahaffey et al., 1982) so that exposure in the present study started after the equivalent of the usual peak in children. It is also evident, however, that exposure during this early period as well exacerbated the impairment observed in the present experiment, as is obvious from a comparison of Group 2 to Group 4. The data from the discrimination reversal task suggest that exposure early in life only, up to an equivalent brain age of 4.0-4.5 years in the human (Garey and de Courten, 1983; Booth et al., 1988; Teller et al., 1978) does not result in lasting impairment in the monkey. However, there is some indication from performance on intermittent schedules in these monkeys
l-5
350 .
it
0 g
w
REVERSALS
6-10 .
350
REVERSALS 3304 270
11-15
. i
i FIG. 4. Number 1 1- 15 on Task 3.
i
GROUP
’
3
i
of errors for Reversals l-5.6-10, Symbols as in Fig. 1.
TABLE 4 and
P-VALUES ATTENTION
(GOOD’S
RANDOMIZATION
TO IRRELEVANT
TEST)
STIMULI-GROUPS
Task
1
Task 2
bert and Rice, 1987). In those studies, treated Total perseveration 0.090 0.006 monkeys attended to irrelevant cues when Right incorrect 0.067 0.183 0.018 0.004 they were first introduced on a series of non- Left incorrect Right avoid 0.329 0.065 spatial discrimination reversal tasks: on a seLeft avoid 0.205 0.365 ries of spatial discrimination tasks, deficits Red incorrect N/A” 0.005 were observed in the presence but not the abGreen incorrect 0.014 N/A sence of irrelevant cues. In the previous study Red avoid 0.089 N/A of nonspatial discrimination reversal, unlike Green avoid 0.145 N/A the present study, the color discrimination Square incorrect N/A N/A with irrelevant form cues was tested before Triangle incorrect N/A N/A Square avoid the form discrimination with irrelevant color N/A N/A N/A N/A cues, while in the spatial discrimination re- Triangle avoid Position incorrect 0.017 0.011 versal study irrelevant form cues were also in0.195 0.011 troduced before irrelevant color cues. Thus it Position avoid Stimulus incorrect 0.005 N/A is unlikely that the results of the present study Stimulus avoid 0.032 N/A are the result of color cues being more distracting. ” N/A. analysis not applicable for the task.
FOR 1 vs 2
Task 3 0.090 0.044 0.175 0.105 0.034 N/A N/A N/A N/A 0.148 0.088 0.134 0.238 0.133 0.068 0.143 0.122
SENSITIVE
that early-exposed monkeys may behave differently from controls (Rice and Gilbert, 1987). It is clear that exposure to higher levels of lead than those in the present study during the first year of life only can result in later impairment on discrimination reversal performance (Bushnell and Bowman, 1979) as well as other tasks (Levin and Bowman, 1986, 1983) in the rhesus monkey. Whether the present group, exposed to lower doses of lead, has permanent behavioral sequelae as a result of early lead exposure awaits further research.
GILBERT, S. G., AND RICE, D. C. (1987). Low-level lifetime lead exposure produces behavioral toxicity (spatial discrimination reversal) in adult monkeys. Toxicol. Appl. Pharmacol.
ACKNOWLEDGMENTS thank Bruce Martin for testing the monanalysis, Stephen Hayward for statistical Tanner for lead and FEP analysis, and and Fred Bryce for review of the manu-
REFERENCES BOOTH,
R. G.,
KIORPES.
L., WILLIAMS,
R. A.. AND
TELLER, D. Y. ( 1988). Operant measurement of contrast sensitivity in infant macaque monkeys during normal development. Vision Res. 28,387-396. B~ISH, B., DORAN. D., AND JACKSON, K. ( 1982). Evaluation of erythrocyte protoporphyrin and zinc protoporphyrin as micro screening procedures for lead poisoning detection. .4nn. Clin. Biochem. 19,7 I-76. BUSHNELL. P. J., AND BOWMAN, R. E. (1979). Persistance of impaired reversal learning in young monkeys exposed to low levels of dietary lead. J. Toxicol. Environ. Health 5, 1015-1023. CAVALLERI, A., BARUFFINI, A., MINOIA. C., AND BIANCO, L. (198 1). Biologic response of children to low levels of inorganic lead. Environ. Res. 25,4 15-423. CORY-SLECHTA, D. A., WEISS, B., AND Cox, C. (1983). Delayed behavioral toxicity of lead with increasing exposure concentration. Toxicol. Appl. Pharmacol. 71, 342-352. CORY-SLECHTA. D. A., WEISS, B., AND Cox, C. (1985). Performance and exposure indices of rats exposed to low concentrations of lead. Toxicol. Appl. Pharmacol. 78.29 l-299. EDGINGTON, E. (1980). Randomization Tests, pp. 182184. Dekker, New York. GAREY, L. J., AND DE COURTEN, C. (1983). Structural development of the lateral geniculate nucleus and visual cortex in monkey and man. Behav. Brain Res. 10, 3-13.
19,484-490.
GOOD, P. (1979). Detection of a treatment effect when not all subjects will respond to treatment. Biometrics 35,483-489. LEVIN, E. D., AND BOWMAN. R. E. (1983). The effect of pre- or post-natal lead exposure on Hamilton search task in monkeys. Neurobehav. Toxicol. Teratol. 5, 391-394. LEVIN, E. D., AND BOWMAN, R. E. (1986). Long-term lead effects on the Hamilton search task and delayed spatial alternation in monkeys. Neurobehav. Toxicol. Teratol.
The authors keys and data analysis. Roy John Truelove script.
109
PERIODS FOR LEAD NEUROTOXICITY
8,2 19-224.
MAHAFFEY, K., ANNEST. J., ROBERTS, J., AND MURPHY, R. (1982). National estimates ofblood lead levels: United States, 1976- 1980. N. Engl. J. Med. 307,573579. RICE, D. C. ( 1984). Behavioral deficits (delayed matching to sample) in monkeys exposed from birth to low levels of lead. Toxicol. Appl. Pharmacol. 75,337-345. RICE, D. C. (1985). Chronic low-lead exposure from birth produces deficits in discrimination reversal in monkeys. Toxicol. Appl. Pharmacol. 77,20 l-2 10. RICE, D. C. (1988). Schedule-controlled behavior in infant and juvenile monkeys exposed to lead from birth. Neurotoxicology9,75-88.
RICE, D. C. Behavioral impairment produced by developmental lead exposure: Evidence from primate research. In Human Lead Exposure (H. L. Needleman. Ed.). CRC Press, Boca Raton. In press. RICE, D. C., AND GILBERT. S. G. (1985). Chronic lowlevel lead exposure from birth produces behavioral toxicity (DRL) in monkeys. To.~icol. .4ppl. Pharmacol. 80,42 l-426. RICE, D. C.. AND GILBERT, S. G. (1987). Sensitive periods for lead behavioral toxicity in the monkey. The Toxicologist.
7, 249.
RICE, D. C.. GILBERT, S. G., AND WILLES, R. F. (1979). Neonatal low-level lead exposure in monkeys (Macaca fascicularis): Locomotor activity, schedule-controlled behavior, and the effects of amphetamine. Toxicol. Appl. Pharmacol.
51,503-5
13.
RICE, D. C., AND KARPINSKI, K. F. (1988). Lifetime lowlevel lead exposure produces deficits in delayed alternation in adult monkeys. Nezcrotoxicol. Teratol. 10, 207-2 14. RICE, D. C.. AND WILLES. R. F. (1979). Neonatal lowlevel lead exposure in monkeys (Macaca fascicularis): Effects on two-choice non-spatial form discrimination. J. Environ. Pathol. Toxicol. 2, 1195-1203. TELLER, D. Y.. REGAL, D. M., VIDEEN, T. 0.. AND PULOS, E. (1978). Development of visual acuity in infant monkeys (Macaca nemistrina) during the early postnatal weeks. Vision Res. l&561-566.
AND
APPLIED
PHARMACOLOGY
102,10!-109
(1990)
Sensitive Periods for Lead-Induced Behavioral Impairment (Nonspatial Discrimination Reversal) in Monkeys DEBORAH C. RICE’ AND STEVEN G. GILBERT Tosicology
Research
Division,
M/plfare Canada,
Bureau Banting
ofChemical Safety. Food Directorate, Building,
Received
April
Tunnqv’s
Pasture.
21, 1989; accepted
Ottawa,
.4trgust
Health Ontario,
Protection Branch, Canada KIA OL2
Health
and
22, 1989
Sensitive Periods for Lead-Induced Behavioral Impairment (Nonspatial Discrimination Reversal) in Monkeys. RICE, D. C., AND GILBERT, S. G. (1990). Toxicol. Appl. Pharmacol. 102, 101-109. A total of 52 nursery-reared monkeys (Macacafascicularis) were dosed orally with 1.5 mg/kg/day of lead on one of four dosing regimens (13 monkeys/group): Group 1, vehicle only: Group 2, dosed with lead continuously from birth; Group 3, dosed with lead from birth to 400 days of age and vehicle thereafter: and Group 4, dosed with vehicle from birth to 300 days of age and lead thereafter. This dosing regimen allowed evaluation ofdifferential infant vulnerability as well as reversibility of the behavioral toxicity of lead. Blood lead concentrations averaged 3-6 pg/dl when monkeys were not being exposed to lead, 32-36 fig/d1 when being dosed with lead and having access to infant formula, and 19-26 pg/dl during lead exposure after weaning from infant formula. When monkeys were 5-6 years old, they were tested on a series of nonspatial discrimination reversal tasks: form, form with irrelevant color cues, color with irrelevant form cues, and alternating form and color. Group 2 exhibited the greatest degree of impairment compared to controls. Group 4 also exhibited impaired performance, although less marked than that of Group 2. Group 3 was not impaired on this series of tasks. These results confirm findings observed in other monkeys exposed continuously to lead and suggest that while exposure beginning after infancy produces impairment, exposure during infancy as well exacerbates the effeCt.
6 1990 Academic
Press,
Inc.
INTRODUCTION
Past research from this laboratory has documented reproducible behavioral impairment on a variety of tasks as a result of lead exposure continuously from birth in the monkey (Rice, 1988, 1985, 1984; Rice and Karpinski, 1988; Gilbert and Rice, 1987; Rice and Gilbert, 1985; Rice et al., 1979; Rice and Willes, 1979). In those studies, blood lead levels of treated monkeys ranged between 11 and 50 pg/dl for different groups of monkeys. Moreover, research from another laboratory reveals that exposure to higher doses of lead during the first year of life (blood lead levels ’ To whom correspondence should be addressed.
of 45-300 pg/dl) results in permanent behavioral impairment (Bushnell and Bowman, 1979; Levin and Bowman, 1986, 1983). Studies in rodents suggest that exposure beginning before maturity, but after weaning, results in behavioral impairment on intermittent schedules of reinforcement (CorySlechta et al., 1985, 1983). However, the issue of a potential sensitive period for lead-induced behavioral impairment has not previously been explored in monkeys. In a series of experiments, monkeys were exposed to lead at levels that have proved sufficient to produce clear behavioral impairment as a result of continuous exposure from birth (blood lead levels of 20-35 pg/dl), while still being in the relevant range for environ101
004 1-008X/90 $3.00 Copyright 0 1990 by Academic Press, Inc. All rights of reproduction m any form reserved.
102
RICE
AND
mentally exposed children. One group of monkeys was exposed continuously from birth, to replicate past experiments and serve as a positive control. Another group was exposed only early in life, while a third group was exposed beginning after infancy. These monkeys have been and are being tested at present using behavioral tasks previously determined to be sensitive to lead-induced behavioral impairment. The results of a series of nonspatial discrimination reversal tasks, with and without irrelevant cues, are reported here. METHODS Subjects and dosing. Fifty-two monkeys (Macaca fasciczdaris) used in the present study were born at the Health Protection Branch breeding colony over a period of 2 years, all from different mothers. Infants were assigned to one of four treatment groups (13 monkeys/ group) in such a way that groups were balanced for age, sex, and paternity. A total of 20 males fathered infants in this study with three occurrences of a male contributing two infants to the same dose group. Infants were separated from their mothers within 12 hr of birth and reared in the Neurotoxicology Primate Colony. They were dosed with a lead acetate dissolved is a 0.05 MNa#ZO, or NaZCOJ solution (vehicle) according to the following regimens: Group I. vehicle from birth onward (8 females, 5 males): Group 2. lead from birth onward (9 females, 4 males): Group 3. lead from birth to 400 days of age and vehicle thereafter (9 females, 4 males); and Group 4. vehicle to 300 days of age and lead thereafter (8 females, 5 males). Infants were dosed by mixing the lead solution with milk substitute formula and delivering it slowly into the back of the mouth using a syringe: by several months of age monkeys would reliably eat gelatin capsules, at which point dosing was accomplished by offering the monkey a capsule containing the dose. Monkeys were dosed 5 days per week at the equivalent of 1.5 mg/kg/day, or 2.1 mg/kg for each of 5 days. Monkeys were housed individually and were socialized with age and group-matched peers beginning within 3 weeks of birth. Monkeys were exercised and socialized for longer periods as they developed until they reached adulthood. Beginning at 3 to 4 years of age. males and females were no longer exercised together. Previous research in our laboratory indicates that dosing beginning at birth results in blood lead values that increase from birth to a peak at about 100 days of age. and decrease after withdrawal of infant formula to a new
GILBERT steady state level over the next IOO- 150 days. In the present study, Groups 2 and 3 were weaned from infant formula at 200 days of age as in previous studies. Group 4 was not weaned until 500 days of age. 200 days after dosing commenced. This ensured that the pattern of blood lead values would be similar between Groups 2 and 4, but displaced in time with respect to developmental period. Half the monkeys in Group I were weaned at 200 days, and half at 500 days. Blood sampling and anal.vsis. Blood was drawn from femoral vein for blood lead analysis every 2 weeks until 500 days of age, every month until 3 years of age. and every 3 months thereafter. The method of blood lead analysis and quality control procedures have been described previously (Rice, 1985). Free erythrocyte protoporphyrin analysis was scheduled to correspond with peak blood lead values, with the control group (Group I) sampled as long as the longest treated group. Group 2 was analyzed from birth to 450 days ofage, Group 3 from birth to 650 days of age, and Group 4 from birth to 550 days of age. The analytical method was a modification of a method described previously (Rice and Willes, 1979). Hematological parameters were assessed on the same schedule as blood lead levels, while blood biochemistry (SMA- 18) was analyzed every 3 months. Blood lead and FEP data were averaged for each monkey individually and then for each group as a whole for days 100-300 after birth, which corresponded to highest levels for Groups 2 and 3, and before institution ofdosing for Group 4. Blood lead levels were then averaged for Group 2 after they stabilized postweaning, for Group 3 after they stabilized following withdrawal from lead, and Group 4 at peak blood lead levels (comparable to Days 100-300 for Groups 2 and 3). Data for Group 1 spanned the range for the other three groups. Blood lead values were also averaged for 3-4 and 4-5 years of age, to complete the blood lead history prior to initiation of the present study. FEP values were averaged for all groups for 100-300 days of age, and for each treated group over the last 100 days of sampling for FEP. which were chosen to include the highest levels for Groups 2 and 4 and to allow for decline after cessation of dosing in Group 3. Group 1 samples span the age range of all treated groups. Behavioral methods. When monkeys were 5-6 years old, they were tested on a series of nonspatial discrimination reversal tasks. Details of behavioral apparatus and testing procedures have been described previously (Rice. 1985). Briefly, the monkey sat facing a Plexiglas panel with two clear pushbuttons. placed at the same level vertically. which could be backlit with stimuli. A tube for delivery of juice was centered between them. Monkeys were tested on a series of four nonspatial descrimination reversal tasks (Table 1). For each task, when the monkey responded with 0 or I error in a nonoverlapping lo-trial block, the positive and negative stimuli were reversed for the next block. A total of 15 reversals plus the initial acquisition were run for each task. Task 1 was a form dis-
SENSITIVE
103
PERIODS FOR LEAD NEUROTOXICITY TABLE
I
SUMMARYOFDOSINGREGIMENSANDDISCRIMINATIONREVERSALTASKS Correct stimulis Type of task Task I
Form discrimination
Task 2
Form discrimination with irrelevant color cues
Task 3
Color discrimination with irrelevant form cues
Task 4
Alternate form and color
Acquisition, even number reversals
Stimuli Square and triangle with red surround Square and triangle with red or green surround, all permutations Same as Task 2 Square on gray or triangle on gray vs red or green
Odd number reversals
Square on red
Triangle on red
Square on red or square on green
Triangle on red or triangle on green Green with square or green with triangle Red or green
Red with square or red with triangle Square or triangle
Dosing regimens GTOUP 1 Group 2 Group 3 Group 4
Dosed Dosed Dosed Dosed
with with with with
vehicle from birth onward lead from birth onward lead from birth to 400 days, vehicle thereafter vehicle from birth to 300 days of age, lead thereafter
crimination task with a white square and triangle superimposed on red backgrounds. Task 2 was a form discrimination with irrelevant color cues; the forms were the same as in Task I, with a red background on one button and a green background on the other. All four permutations of form and color were presented. Task 3 was a color discrimination task with irrelevant form cues. The stimuli were the same as in Task 2, but the relevant stimulus dimension was color rather than form. In Task 4, the positive stimulus alternated between color and form on successive reversals. Stimulus pairs were square on gray or triangle on gray vs red or green. Positive stimuli were therefore red or green for one reversal, square or triangle on the next. Monkeys were tested 100 trials per day. 5 days per week. Monkeys had previously performed on a multiple fixed interval-fixed ratio (Rice and Gilbert, 1987) and DRL schedules of reinforcement. One female monkey from Group 3, who had a constant perineal swelling, learned to press the pushbuttons after a month of intensive training. However, this monkey refused to perform the multiple fixed internal-fixed ratio schedule, and so was eliminated from the study. This is the only monkey to be dropped from a study in this laboratory for this reason out of more than 150 individuals. Two male control monkeys died from acute illnesses before being tested on this series of discrimination reversal tasks. Dala analysis. Data analysis was by means of Good’s statistic (Good, 1979). This is a nonparametric randomization test that includes both differences in the mean
and the group variability in the analysis, in order to take into account that there may be both responders and nonresponders in the treated group(s). Since past experience in this laboratory revealed such an effect in lead-exposed monkeys, equal weight was given to differences in mean and variability in the present analysis (i.e., 19= 0.5). The main variable of interest was the total number of incorrect responses across all reversals of each task. Group 1 was compared to Group 2 first, since Group 2 represents a replication of previous positive results on nonspatial discrimination reversal performance, and would presumably be impaired to the greatest degree. Moreover, impairment in Groups 3 or 4 in the absence of impairment of Group 2 would be difficult to interpret. If that comparison was significant for a Task (p < 0. lo), Group 1 was compared to Group 3 and Group 4. If the total number of errors across all reversals of a task was significant @ < O.lO), reversals l-5. 6-10, and I l-15 were analyzed as separate blocks to determine whether treated monkeys were more impaired on early or late blocks. Analogous comparisons were also made for all tasks combined to increase the sensitivity of the analysis (Edgington, 1980). In addition, acquisition of performance was assessed separately, again comparing Group 1 to Group 2 before comparison of Group I to the other groups. For any comparison that was significant (p < 0. IO) for the total number of errors across all reversals, further analysis was performed to assessattention to irrelevant cues. Such an analysis has proved useful in the past in
104
RICE
AND
GILBERT
TABLE
BLOOD LEAD
FEP
AND
2
CONCENTRATIONS
Blood lead (mean f SE) (pg/lOO ml whole blood) Group Age 100-300 days
Selectedages
3-4years
FEP (mean + SE) (&ml RBC) 4-5
years
100-300
days
Selectedages
Age (days) I. Control 2.
3.
4.
4.9
+ 0.7
2.9
t 0.2
(400-800)
f
1.3*
(500-700)
Lead from birth
32.3
2
3.1'
18.9
Leadto 400 days
36.3
k
2.3*
5.6
+ 0.7+
(600-800)
32.9
+ 2.1*
(400-600)
Lead from 300 days
6.0?
1.1
Age (days) 3.0 kO.4
23.8
3.8
22.2
k
l.8*
+ 0.6 t 1.0’
2.7
25.7
3.3
? 0.5
75?
9
69k
8
(350-650)
+ 1.5:
252+55+
?z0.6
431*71*
aa+
(550-650)
89t16
270?50*
(450-550,
22.0* 1.6*
132 +30
(350-450)
* Different from control. p < 0.000 1, Wilcoxin test. + Different from control, p < 0.005. determining the types of errors made by lead-exposed monkeys on discrimination reversal tasks (Rice, 1985; Gilbert and Rice, 1987). The ratio of incorrect responses on each button and, for Tasks 2 and 3, the ratio of incorrect responses on irrelevant stimuli was calculated by number incorrect on button (or stimulus) number incorrect on button (or stimulus) + number correct on button (or stimulus) The ratio of trials on which a monkey avoided a particular button or irrelevant stimulus was calculated by number incorrect on other button (or stimulus) number incorrect on other button (or stimulus) + number correct on button (or stimulus) The totals of the larger ratio for each pair (i.e.. right and left incorrect; triangle avoid and square avoid, etc.) over the course of each task were calculated in order to circumvent the fact that different individuals may show preference for different irrelevant stimuli. A combined measure of attention to irrelevant cues was calculated by adding the individual measures. This overall measure was compared using Good’s statistic; for significant overall comparisons each individual measure was compared.
RESULTS Blood lead values were approximately 3236 pg/dl when monkeys were dosed with lead and fed milk formula, and 19-26 pg/dl after the milk formula diet was discontinued (Table 2). When monkeys were not exposed to
lead, blood lead values were 3-6 pg/dl. FEP values were statistically significantly elevated when lead values were in the 32-36 pg/dl range. At blood lead values of 19 pg/dl, some but not all individuals of Group 2 (the only group for which FEP values were determined at this blood lead value) exhibited elevated FEP’s, resulting in an increased value that was not statistically significant for the group as a whole. Hematology and blood biochemistry were within normal limits for all monkeys; group comparisons have not been performed. Analysis of the total number of errors revealed that Group 2 was impaired relative to Group 1 over the entire experiment (Table 3), while Group 4 was impaired to a lesser extent than Group 2, and Group 3 was unimpaired. Group 2 was impaired over all reversals for Tasks 1, 2, and 3, while Group 4 was impaired at the beginning of the experiment (Task I), and on Task 3, when the relevant stimulus dimension was changed from form to color (Fig. 1). Task 4 proved extremely easy for all monkeys, with 15 reversals completed in 3-4 sessions. Group 2 was not impaired under these circumstances. Analyzing the performance by blocks of reversals revealed that on Task 1, Group 2 was impaired on the first two blocks, but not the
SENSITIVE
PERIODS
FOR TABLE
P-VALUES
’ (-)
Represents
3
l-15
l-5
6-10
11-15
0.03 NS 0.07
0.07 -
0.005
0.08
0.02
0.05 NS
0.01 NS 0.06
0.04 -
0.02 0.09
NS NS
0.03 NS NS
0.05 NS NS
0.04 -
0.08 -
NS -
0.006 NS 0.05
0.004 -
0.00 1 0.009
NS -*
NS
NS -
0.08
NS
NS
comparisons
105
NEUROTOXICITY
(GOOD’S RANDOMIZATION TEST) FOR PERFORMANCE ON A SERIES OF NONSPATIAL DISCRIMINATION REVERSAL TASKS
Acquisition Overall 1 vs2 1 vs3 1 vs4 Task 1 1 vs2 1 vs3 1 vs4 Task 2 1 vs2 1 vs3 1 vs4 Task 3 1 vs2 1 vs3 1 vs.4 Task 4 I vs2 I vs3 I vs4
LEAD
-
-
NS 0.05
-
not performed.
third, while Group 4 was marginally impaired with the same pattern (Fig. 2). On Task 2, when irrelevant cues were introduced, Group 2 was impaired on all three blocks of reversals (Fig. 3). On Task 3, Group 2 was impaired on blocks 1 and 2 as in Task 1, while Group 4 was impaired on blocks 2 and 3 (Fig. 4). When all tasks were combined, there is evidence that Group 2 was impaired over all three blocks of reversals, while Group 4 was impaired only on the first two blocks. Acquisition of the task was retarded only for Group 2 on Task 2, when irrelevant cues were introduced. Analysis of attention to irrelevant cues revealed that there was a tendency for Group 2 to have more total perseverative errors than Group 1 over the three tasks for which the total number of errors differed (Table 4). There was no strong evidence that this was true for Group 4, however (Task 1, p = 0.33, Task 3, p = 0.11). Further comparison of
Group 1 vs Group 2 revealed that on Task 1, treated monkeys had greater ratios of responding incorrectly on the left button but not for avoiding either button. For Task 2, treated monkeys responded on the left and avoided the right button, and attended to the irrelevant color stimuli more than controls. For Task 3, treated monkeys switched position strategies, responding more on the right and avoiding the left relative to controls; there was no strong indication that they attended to the irrelevant form stimuli more than controls. DISCUSSION The decrease in blood lead levels observed after withdrawal of infant formula replicates previous research from this laboratory (Rice et al., 1979; Rice, 1985). The fact that the same pattern of blood lead values was ob-
106
RICE AND GILBERT TASK
1
TASK
2
3500-
v) 2 2
2750-
:
2250-
8 g Y
1750. 1250-
8
7504
-
j
25;$:...
!I.
2
TASK
i
1100
E
900
500 300 100
. 1.
“m
Fe*
3
:
**a ir
!i:
1. . *
: : i:
birth, with peak blood levels of 15 or 25 pg/ dl, and steady state postweaning levels of 1 1 or 13 pg/dl, were impaired on a nonspatial form discrimination reversal task and a color discrimination reversal task that followed immediately. They were not impaired on the third task, a form discrimination reversal task with irrelevant color cues. In the present experiment, Group 4, dosed beginning at 300 days of age, was impaired on two of the tasks, to a lesser degree than Group 2. Group 3, dosed from birth to 400 days, was not impaired. In general, the increased total number of errors over all reversals was the result of more errors toward the beginning rather than at the end of the task. This has been observed by
:
2750
i.
2250
GROUP
FIG. 1. Number oferrors over all 15 reversals for Tasks 1, 2, and 3. Each dot represents an individual monkey. Group 1, dosed with vehicle from birth onward: Group 2, dosed with lead from birth onward; Group 3, dosed with lead from birth to 400 days of age and vehicle thereafter; Group 4, dosed with vehicle from birth to 300 days of age and lead thereafter.
served in Groups 2 and 4, associated with diet but not age, indicates that the decrease in blood lead levels observed routinely in our laboratory after weaning is an effect of diet rather than age. Similarly, the statistically significant increase in FEP levels at blood lead levels above 30 pg/dl, with some individuals exhibiting an increase at 20 pg/dl, replicates previous finding from this laboratory (Rice, 1985) and is consistent with data in children (CavaIleri et al., 1981; Bush et al., 1982). The deficits observed in Group 2 on the first three tasks replicates effects observed in monkeys with lower blood lead levels (Rice, 1985). In that study, monkeys dosed from
REVERSALS
l-5
REVERSALS
6-10
REVERSALS
11-15
1750 1250
900-
900 1 700
1
*
1
-
500 300
FIG. 2. Number oferrors for Reversals l-5,6-10, 1I - 15 on Task 1. Symbols as in Fig. 1.
and
SENSITIVE
107
PERIODS FOR LEAD NEUROTOXICITY
ACQUISITION
REVERSALS
1-5
i;[Z::. .$ :..... ;E”-‘;. 8: 1
2
3
4
GROUP
1
2
3
4
FIG. 3. Number oferrors for the Acquisition, Reversals l-5, 6-10, and 1 l-1 5 on Task 2. Symbols as in Fig. 1.
other investigators at higher blood lead levels (Bushnell and Bowman, 1979) as well as in the group of monkeys discussed previously (Rice, 1985). Group 2 was impaired on the acquisition of Task 2, when irrelevant cues were introduced. Such impairment in acquisition was not observed at lower doses on nonspatial discrimination reversal, but was observed in those same monkeys on a series of spatial discrimination reversal tasks when irrelevant cues were introduced (Gilbert and Rice, 1987). It is apparent from the figures that there are individual monkeys in treated Groups 2 and 4 that are severely impaired, while other individuals are not different from controls. This results in increased variability within these treated groups. In fact, the greater statistical significance in Group 2 than Group 4 appears to result from the increased number of individuals with aberrant performance in the former group, rather than a difference in the magnitude of the deficit within impaired individuals. This phenomenon has been observed repeatedly in our laboratory (Rice, 1985; Gilbert and Rice, 1987; Rice and Kar-
pinski, 1988; Rice et al., 1979) as well as in lead-exposed rats (Cory-Slechta et al., 1985, 1983). In the present study, the aberrant performances across tasks were typically due to different individuals, although this was less apt to be true within tasks. Impaired performance in the presence of distracting stimuli or changes in environmental contingencies (as in the requirement for reversal of the previously correct response) appears to be a hallmark of lead toxicity in the monkey (see Rice, in press, for review). Such procedures have proved more sensitive than the acquisition of a behavioral task, except perhaps where contingencies change, as in the deficit observed on Task 2 by Group 2. In the present experiment, Group 2 monkeys attended to position as an irrelevant stimulus over all three tasks for which they showed an increase in total number of errors. In addition, they attended to irrelevant colors when irrelevant stimuli were introduced in Task 2, but not to irrelevant forms in Task 3, when irrelevant cues were familiar. These results are consistent with previous results in monkeys with lower blood levels (Rice, 1985; Gil-
108
RICE AND GILBERT REVERSALS
Data from the present experiment suggest that lead exposure very early in life is not necessary for lead-induced behavioral impairment. Three hundred days of age in the monkey (when dosing of Group 4 commenced) corresponds roughly to 3.0-3.5 years of age in the child in terms of development. In the human, blood lead values peak at approximately 2 years of age (Mahaffey et al., 1982) so that exposure in the present study started after the equivalent of the usual peak in children. It is also evident, however, that exposure during this early period as well exacerbated the impairment observed in the present experiment, as is obvious from a comparison of Group 2 to Group 4. The data from the discrimination reversal task suggest that exposure early in life only, up to an equivalent brain age of 4.0-4.5 years in the human (Garey and de Courten, 1983; Booth et al., 1988; Teller et al., 1978) does not result in lasting impairment in the monkey. However, there is some indication from performance on intermittent schedules in these monkeys
l-5
350 .
it
0 g
w
REVERSALS
6-10 .
350
REVERSALS 3304 270
11-15
. i
i FIG. 4. Number 1 1- 15 on Task 3.
i
GROUP
’
3
i
of errors for Reversals l-5.6-10, Symbols as in Fig. 1.
TABLE 4 and
P-VALUES ATTENTION
(GOOD’S
RANDOMIZATION
TO IRRELEVANT
TEST)
STIMULI-GROUPS
Task
1
Task 2
bert and Rice, 1987). In those studies, treated Total perseveration 0.090 0.006 monkeys attended to irrelevant cues when Right incorrect 0.067 0.183 0.018 0.004 they were first introduced on a series of non- Left incorrect Right avoid 0.329 0.065 spatial discrimination reversal tasks: on a seLeft avoid 0.205 0.365 ries of spatial discrimination tasks, deficits Red incorrect N/A” 0.005 were observed in the presence but not the abGreen incorrect 0.014 N/A sence of irrelevant cues. In the previous study Red avoid 0.089 N/A of nonspatial discrimination reversal, unlike Green avoid 0.145 N/A the present study, the color discrimination Square incorrect N/A N/A with irrelevant form cues was tested before Triangle incorrect N/A N/A Square avoid the form discrimination with irrelevant color N/A N/A N/A N/A cues, while in the spatial discrimination re- Triangle avoid Position incorrect 0.017 0.011 versal study irrelevant form cues were also in0.195 0.011 troduced before irrelevant color cues. Thus it Position avoid Stimulus incorrect 0.005 N/A is unlikely that the results of the present study Stimulus avoid 0.032 N/A are the result of color cues being more distracting. ” N/A. analysis not applicable for the task.
FOR 1 vs 2
Task 3 0.090 0.044 0.175 0.105 0.034 N/A N/A N/A N/A 0.148 0.088 0.134 0.238 0.133 0.068 0.143 0.122
SENSITIVE
that early-exposed monkeys may behave differently from controls (Rice and Gilbert, 1987). It is clear that exposure to higher levels of lead than those in the present study during the first year of life only can result in later impairment on discrimination reversal performance (Bushnell and Bowman, 1979) as well as other tasks (Levin and Bowman, 1986, 1983) in the rhesus monkey. Whether the present group, exposed to lower doses of lead, has permanent behavioral sequelae as a result of early lead exposure awaits further research.
GILBERT, S. G., AND RICE, D. C. (1987). Low-level lifetime lead exposure produces behavioral toxicity (spatial discrimination reversal) in adult monkeys. Toxicol. Appl. Pharmacol.
ACKNOWLEDGMENTS thank Bruce Martin for testing the monanalysis, Stephen Hayward for statistical Tanner for lead and FEP analysis, and and Fred Bryce for review of the manu-
REFERENCES BOOTH,
R. G.,
KIORPES.
L., WILLIAMS,
R. A.. AND
TELLER, D. Y. ( 1988). Operant measurement of contrast sensitivity in infant macaque monkeys during normal development. Vision Res. 28,387-396. B~ISH, B., DORAN. D., AND JACKSON, K. ( 1982). Evaluation of erythrocyte protoporphyrin and zinc protoporphyrin as micro screening procedures for lead poisoning detection. .4nn. Clin. Biochem. 19,7 I-76. BUSHNELL. P. J., AND BOWMAN, R. E. (1979). Persistance of impaired reversal learning in young monkeys exposed to low levels of dietary lead. J. Toxicol. Environ. Health 5, 1015-1023. CAVALLERI, A., BARUFFINI, A., MINOIA. C., AND BIANCO, L. (198 1). Biologic response of children to low levels of inorganic lead. Environ. Res. 25,4 15-423. CORY-SLECHTA, D. A., WEISS, B., AND Cox, C. (1983). Delayed behavioral toxicity of lead with increasing exposure concentration. Toxicol. Appl. Pharmacol. 71, 342-352. CORY-SLECHTA. D. A., WEISS, B., AND Cox, C. (1985). Performance and exposure indices of rats exposed to low concentrations of lead. Toxicol. Appl. Pharmacol. 78.29 l-299. EDGINGTON, E. (1980). Randomization Tests, pp. 182184. Dekker, New York. GAREY, L. J., AND DE COURTEN, C. (1983). Structural development of the lateral geniculate nucleus and visual cortex in monkey and man. Behav. Brain Res. 10, 3-13.
19,484-490.
GOOD, P. (1979). Detection of a treatment effect when not all subjects will respond to treatment. Biometrics 35,483-489. LEVIN, E. D., AND BOWMAN. R. E. (1983). The effect of pre- or post-natal lead exposure on Hamilton search task in monkeys. Neurobehav. Toxicol. Teratol. 5, 391-394. LEVIN, E. D., AND BOWMAN, R. E. (1986). Long-term lead effects on the Hamilton search task and delayed spatial alternation in monkeys. Neurobehav. Toxicol. Teratol.
The authors keys and data analysis. Roy John Truelove script.
109
PERIODS FOR LEAD NEUROTOXICITY
8,2 19-224.
MAHAFFEY, K., ANNEST. J., ROBERTS, J., AND MURPHY, R. (1982). National estimates ofblood lead levels: United States, 1976- 1980. N. Engl. J. Med. 307,573579. RICE, D. C. ( 1984). Behavioral deficits (delayed matching to sample) in monkeys exposed from birth to low levels of lead. Toxicol. Appl. Pharmacol. 75,337-345. RICE, D. C. (1985). Chronic low-lead exposure from birth produces deficits in discrimination reversal in monkeys. Toxicol. Appl. Pharmacol. 77,20 l-2 10. RICE, D. C. (1988). Schedule-controlled behavior in infant and juvenile monkeys exposed to lead from birth. Neurotoxicology9,75-88.
RICE, D. C. Behavioral impairment produced by developmental lead exposure: Evidence from primate research. In Human Lead Exposure (H. L. Needleman. Ed.). CRC Press, Boca Raton. In press. RICE, D. C., AND GILBERT. S. G. (1985). Chronic lowlevel lead exposure from birth produces behavioral toxicity (DRL) in monkeys. To.~icol. .4ppl. Pharmacol. 80,42 l-426. RICE, D. C.. AND GILBERT, S. G. (1987). Sensitive periods for lead behavioral toxicity in the monkey. The Toxicologist.
7, 249.
RICE, D. C.. GILBERT, S. G., AND WILLES, R. F. (1979). Neonatal low-level lead exposure in monkeys (Macaca fascicularis): Locomotor activity, schedule-controlled behavior, and the effects of amphetamine. Toxicol. Appl. Pharmacol.
51,503-5
13.
RICE, D. C., AND KARPINSKI, K. F. (1988). Lifetime lowlevel lead exposure produces deficits in delayed alternation in adult monkeys. Nezcrotoxicol. Teratol. 10, 207-2 14. RICE, D. C.. AND WILLES. R. F. (1979). Neonatal lowlevel lead exposure in monkeys (Macaca fascicularis): Effects on two-choice non-spatial form discrimination. J. Environ. Pathol. Toxicol. 2, 1195-1203. TELLER, D. Y.. REGAL, D. M., VIDEEN, T. 0.. AND PULOS, E. (1978). Development of visual acuity in infant monkeys (Macaca nemistrina) during the early postnatal weeks. Vision Res. l&561-566.