The effects of lindane, DDT, and chlordecone on avoidance responding and seizure activity

TOXICOLOGY

AND

APPLIED

PHARMACOLOGY

88,57-65

( 1987)

The Effects of Lindane, DDT, and Chlordecone on Avoidance Responding and Seizure Activity H. A. TILSON,

S. SHAW, AND R. L. MCLAMB

Laboratory of Behaviorul and Neurological To.yicology, National Institute ofEnvironmental Health Sciences. P.U. Box 12233. Research Triangle Park. North Carolina 27709

Received August 4. 1986: accepted November 3, 1986 The Effects of Lindane. DDT, and Chlordecone on Avoidance Responding and Seizure Activity. TILSON, H. A., SHAW, S.. AND MCLAMB. R. L. (1987). Toricol. .4ppl. Pharmacol. 88, 5165. Male adult Fischer-344 rats were given various doses of lindane (0, 15, and 30 mg/kg, PO), chlordecone (0, 25, 50, or 100 mg/kg, ip), or p.p’-dichlorodiphenyltrichloroethane (p,p’-DDT) (0. 25, 50, or 100 mg/kg, po) and tested for their ability to perform a two-way shuttle box task or to learn and retain a step-through passive avoidance response. Administration of p,p’-DDT or chlordecone 3 hr prior to acquisition did not affect the number of shuttle box avoidance responses made during a 60-trial training task, while responses during the intertrial interval (ITI) were decreased. Rats receiving I5 or 30 mg/kg of lindane made fewer avoidance responses, but did not differ from controls in terms of the number of responses during the ITI. When 30 mg/ kg lindane was given 3 hr prior to passive avoidance acquisition. retention was impaired 7 days later; the lower dose of lindane. and all doses of chlordecone or p.p’-DDT had no effect under these conditions. When these chemicals were given immediately after passive avoidance training, animals treated with lindane were not affected. Animals receiving 100 mg/kg ofp.p’-DDT or chlordecone displayed marked signs of toxicity and animals tested 7 days after training showed an impaired retention. Pretreatment with anticonvulsants such as phenobarbital and chlordiazepoxide, which may enhance GABA-mediated responses, blocked the disruptive effecs of lindane (30 mg/kg) on shuttle box avoidance. The seizure-related activity produced by a higher dose of lindane (60 mg/kg) and kainic acid. a hippocampal excitotoxin, was also blocked by phenobarbital and chlordiazepoxide. Pretreatment with phenytoin. which is thought to bind to the inactivation gates of sodium, had no effect on the effects produced by lindane or kainic acid. These data suggest that treatment with nonconvulsant doses of lindane can interfere with the ability to acquire and use new information and that these effects may be associated with 15 1987 Academic PKS. h alterations in GABA.

The neurobiological effects of chlorinated hydrocarbon agents such as dichlorodiphenyltrichloroethane (DDT), chlordecone (Kepone), and r-hexachlorocyclohexane (lindane) have been known for several years. Exposure to DDT and chlordecone produces tremor and hyperresponsiveness to external stimulation (Joy, 1982a; Woolley, 1982), while lindane is primarily a convulsant (Joy, 1982b). Recent research in our laboratory has indicated that the neuropharmacological basis for the tremor and hyperresponsiveness

produced by these chemicals is different (Tilson et al.. 1985; Tilson et al., 1986; Herr et al., 1987).

In spite of the prominent effects of chlorinated hydrocarbons on neurological functioning and their known neurochemical and neurophysiological effects on neuronal processes, there are few studies concerning the effects of these agents on cognitive or associative processes such as learning and memory. Data from the animal literature indicate that acute exposure to DDT can affect acquisition 57

0041-008X/87 Copyright All n&s

$3.00

0 1987 by Academic Press, Inc. of reproductron in any form reserved

58

TILSON,

SHAW, AND MC LAMB

and/or retention of active (Uppal et al., 1983a) and passive avoidance responding in rats or mice (Uppal et al., 1983b; Sobotka, 197 I), while performance of a food motivated response in a modified Hebb-Williams closed field test was not affected in rats (Khairy, 1959). Repeated exposure to lindane has been reported to increase the number of errors made by rats tested in a food reinforced T-maze (Desi, 1974). Tilson et al. (1979) reported that chronic exposure to chlordecone decreased retention of a oneway avoidance response in rats. Because of the apparent lack of data in this area, the present series of experiments compared systematically the dose-dependent effects of acutely administered DDT, chlordecone, and lindane on the acquisition of a two-way shuttle box avoidance response and the acquisition and retention of a passive avoidance response in rats. Because of the finding that lindane appeared to interfere with the acquisition of avoidance responding at doses lower than those producing observable neurological impairment, subsequent experiments were conducted using pharmacological probes to determine the possible mechanism responsible for these behavioral effects. In these studies, rats were pretreated with three anticonvulsants, phenytoin, chlordiazepoxide, and phenobarbital, prior to administration of lindane. Phenobarbital and chlordiazepoxide are believed to exert their anticonvulsant effects in part by enhancing GABA-mediated responses with little or no effect on the function of the sodium channel (MacDonald et al., 1985). Phenytoin, at therapeutically relevant concentrations, is believed to bind to the inactivation channels of sodium to prevent repetitive firing of neurons, whereas significant GABA-mediated effects are observed at higher concentrations (MacDonald, 1983). Finally, the effect of these three anticonvulsants on the seizure-related activity of a high dose of lindane (60 mg/kg) was also determined and compared to kainic acid. Pretreatment with phenobarbital and diazepam, but not phenytoin, has been

reported to block the effects of systemically administered kainic acid, an excitotoxicant causing damage to the limbic forebrain, particularly in the hippocampus (Fuller and Olney, 198 1). It has been reported that the seizure activity produced by lindane may be mediated via the limbic system (Woolley et al.. 1984). METHODS Subjects. Ten- to twelve-week-old male Fischer-344 rats (Charles River Breeding Co., Wilmington, MA) were housed four per cage in a room with a constant temperature (2 I f 2°C) relative humidity (50 + lo%), and a l2hr light/dark cycle (dark, 1900 hr). Feed (NIH No. 3 1) and water were freely available in the home cages throughout the studies, unless noted otherwise. Dosing. The effects of organochlorines on the acquisition and performance oftwo-way shuttle box responding were studied in three separate experiments. In the first experiment, rats were dosed by gavage (PO) with 15 or 30 mg/kg of r-hexachlorocyclohexane (Sigma Chemical Co., St. Louis, MO) dissolved in corn oil or I mg/kg of the corn oil vehicle. A higher dose of lindane (60 mg/kg) produced seizures and death in over three-fourths of the animals within 24 hr. Another group of rats received either 2550, or 100 mg/kg ofp,p’-DDT (Aldrich Chemical Co., Milwaukee, WI) or 1 mg/kg of corn oil vehicle by gavage. A third group received either 0. 25, 50, or 100 mg/kg of chlordecone ip (Radian Corporation, Austin, TX). The highest dose of DDT and chlordecone produced mortalities in approximately 30% of the rats 1 week after dosing. Administration of the organochlorines occurred 3 hr prior to testing in the shuttle box. Routes of administration and times of testing are based on work reported elsewhere (Tilson et af.. 1984). The effects of the organochlorines on passive avoidance responding were studied in two separate experiments. In the first experiment, rats were given one of three doses of p,p’-DDT. chlordecone, lindane, or corn oil vehicle 3 hr prior to acquisition of the passive avoidance task. In the second experiment, the organochlorines were given immediately after the passive avoidance training. Retention of avoidance responding was tested 1 week later. In experiments involving pharmacological pretreatment prior to lindane, phenytoin (Sigma Chemical Co.) was suspended in two drops of Tween 80 per 10 ml of distilled water and given by gavage in a volume of 1 ml/ kg 30 min prior to lindane. Phenobarbital (Winthrop) was dissolved in distilled water and given by gavage in a dose of 50 mg/kg 30 min prior to lindane, while chlordiazepoxide hydrochloride (Roche) was dissolved in dis-

EFFECTS OF ORGANOCHLORINES tilled water and given subcutaneously (SC)30 min prior to lindane. In experiments concerning two-way shuttle box performance, a dose of 30 mg/kg of lindane was used and behavioral testing occurred 3 hr after administration of lindane. In the comparison between lindane and kainit acid, the anticonvulsants were administered 30 min prior to 60 mg/kg of lindane (PO) or 8 mg/kg of kainic acid (SC)(Sigma Chemical Co.). Kainic acid was prepared prior to the experiment by dissolving it in distilled water. Shuttle box acquisition. Acquisition of discriminated avoidance responding was measured in a shuttle box comprised of two different sized compartments (Harry and Tilson, 198 I). The larger compartment was 24.4 cm wide X 26.7 cm high X 29.21 cm long, while the smaller one measured 11.43 cm X 13.3 cm X 19.1 cm. The chambers were constructed of 0.95-cm black Plexiglas and had stainless-steel grids positioned 1.75 cm apart. Two white lights (GE PR 17) and a Sonalert (Model 5C628) were positioned at the rear of each chamber. Electric foot shock was applied to the grids by a solidstate shocker scrambler (Coulboum Instruments, Lehigh valley, PA; Model E- I3- 16). The two Plexiglas compartments were separated by an automated guillotine door (Lafayette Instruments, Lafayette, IL; Model 85013). The position of the animal in the shuttle box was determined by a bank of photocells located 0.5 cm above the grids. Programming of a PDP-IA minicomputer allowed for automatic activation of stimuli and recording of responses. Each session began by placing the rat in the smaller compartment, closing the lid, and raising the guillotine door. Each trial consisted of the presentation of a combined light- and tone-conditioned stimulus occurring in the compartment containing the animal. If the animal did not leave the compartment having the light and tone stimulus within IO set (avoidance response) electric foot shock (0.4 mA) was applied to thegrids ofthe floor. Shuttling to the safe (unshocked side) was recorded as an escape response. The occurrence of an avoidance and escape response or 10 set of foot shock without making an escape response (escape loss) terminated the light-tone cues and/or shock and initiated a IO- to 20-set variable intertrial interval (ITI). Responses during the IT1 were recorded. Latencies to make an avoidance or escape response were also recorded. All animals were given 60 training trials. Rats were dosed with one of the organochlorine chemicals and testing occurred 3 hr later. Passive avoidance acquisition and retention. A twocompartment trough-shaped alley [51 L X 30/6.8 W (top/bottom) X 15 cm H] similar in design to that described by Jarvik and Kopp ( 1967) was used for passive avoidance training and retention testing. The alley was constructed of Plexiglas and consisted of two compartments, I7 and 34 cm in length. The compartments were separated by a door that could be raised and lowered by a pullchain. A single 15 W incandescent bulb was suspended 10 cm above the lid of the smaller compartment.

ON LEARNING

AND MEMORY

59

The floor of each compartment consisted of parallel stainless-steel plates, separated along their length by 2 cm. Constant current electric footshock (0.4 mA) from a Coulboum solid-state shocker (Model E 13- 16) was administered through the floor plates ofthe larger companment. The passive avoidance apparatus was housed in a larger sound- and light-attenuated chamber with background illumination and masking noise provided by the chamber exhaust fan. On the day of testing, rats were placed facing the door into the illuminated compartment of the passive avoidance apparatus. Ten seconds later, the guillotine door was raised and the latency to cross from the smaller to the larger compartment was recorded with a stopwatch. Contingent upon crossing into the darkened compartment (up to 300 set), the door was closed and footshock was delivered for 1 sec. Retention was assessed7 days later using a procedure similar to that of training, except that footshock was not delivered. For up to 300 set, the latency to cross from the small to the large compartment was recorded. In addition, as a measure of vacillatory responding (Mactutus et al.. 1982). the total number ofheadpokes (head including ears) and halfcrosses (head and two forepaws) was recorded. In one experiment, rats were given the organochlorine insecticides 3 hr prior to acquisition, while in the second experiment, dosing occurred immediately after being removed from the chamber. Seizure activity. Rats were pretreated with vehicle or one of the anticonvulsants 30 min prior to receiving 60 mg/kg oflindane or 8 mg/kg of kainic acid. The rats were placed individually into domiciliary cages containing corncob bedding and were observed for the presence or absence of the lindane- or kainic acid-induced seizure syndrome. Kainic acid seizures were characterized by facial twitching and myoclonic motions ofthe jaws, rearing on hindlegs with head bobbing, and forelimb clonus. The seizure pattern following lindane was similar to kainic acid, consisting primarily of forelimb clonus and whole body jerks. Statistical ana/ysis. Differences between groups were assessedwith Fisher’s least significant difference test after significant analysis of variance (Winer, 197 I).

RESULTS Two-way shuttle box acquisition. Both 15 and 30 mg of lindane produced a significant decrease in the number of correct avoidance responses made during the 60-trial acquisition session (Fig. 1). These effects on avoidance responding were seen in the absence of significant changes in the number of responses during the intertrial interval. Table 1 shows that lindane significantly increased

60

TILSON, LINDANE

0

15 30

CHLORDECONE

0 25 50 100 DOSAGE

SHAW

P,P1-DDT

0 25

50

100

(MG/KG)

FIG. 1. Dose-related effects of lindane, chlordecone. andp.p’-DDT on the number ofavoidance responses and responses during ITI of a 60-trial training session. There were 16.26, and 24 rats in the control groups for lindane. chlordecone. andp,p’-DDT, respectively. All other treatment groups had eight rats per group, except 50 mg/kg of chlordecone, which had 16 animals. Asterisks indicate a significant difference from vehicle control (p i 0.05, Fisher’s least significant difference after significant analysis of variance).

latencies to make an avoidance response, while latencies to make an escape response were not affected. These data suggest lindane decreased the frequency of avoidance responding without changing activity during the IT1 or responsiveness to the unconditioned stimulus. Chlordecone and p,p’-DDT had no significant effect on the frequency of avoidance responses up to 100 mg/kg (Fig. 1). However, rats receiving 50 or 100 mg/kg of either compound were observed to be tremulous and hyperirritable when removed from the test apparatus. That chlordecone had effects on behavior during the test is indicated by the observation that 50 and 100 mg/kg decreased the number of IT1 responses and increased latencies to escape the shock. All doses of chlordecone increased latencies to make an avoidance response (Table I). Likewise, 100 mg/kg of p,p’-DDT significantly decreased responses during the IT1 and increased latenties to make an escape response. These data indicate that although some doses of chlorde-

AND MC LAMB

cone and p,p’-DDT may have produced neurological signs such as tremor potentially interfering with responding during the IT1 or the latency to respond to the conditioned or unconditioned stimuli, these rats were capable of acquiring the conditioned response. Eflects of passive avoidance. Administration of lindane, chlordecone, or p.p’-DDT prior to training had no significant effects on the latency to cross from the small, lighted compartment to the larger chamber during the acquisition session (Table 2), indicating that the animals were not significantly hyperactive or impaired at the time of training. When retested 1 week later, rats having received 30 mg/kg of lindane made significantly more crossovers and spent less time in the lighted, previously unshocked side than controls. Chlordecone or p,p’-DDT had no significant effect on retention of passive avoidance when training occurred under the influence of chemicals.

TABLE 1 EFFXTSOFLINDANE.&-DDT,ORCHLORDECONE ON AVOIDANCEAND ESCAPE LATENCIESINTHE WAYSHUTTLE Box

Treatment and dose M-&k) Lindane 0 15 30

Chlordecone 0 25

50 100 p,p’-DDT 0

Two-

Response latency (set)” N

Avoidances

16 8 8

3.7 +- 0.2 6.0?0.3* 5.1 k 0.5*

0.6 0.7 0.9

26

3.4 -t 0.2 4.6 -t- 0.8* 4.5,0.3* 4.7 ?z0.2*

1.0 f 0.04

3.4 i 0.2 3.6 ? 0.4 4.1 f 0.5 3.4 10.4

0.8 aO.04 0.7 kO.05 0.8 f 0.06 2.2 +0.24*

8 15 8 24

25 50

8 8

100

8

Escapes f 0.03 kO.12 + 0.12

1.o f 0.08 1.2kO.14’ 1.7 + 0.22*

’ Mean f SE. * p < 0.05, Fisher’s least significant difference after significant analysis of variance.

EFFECTS

OF

ORGANOCHLORINES

ON TABLE

LEARNING

AND

Treatment and dose (mg/kg) Control Lindane IS 30 Chlordecone 25 50 100 [I.]“-DDI2s 50 IO0

Iv

Acquisition” latencies (xc)

61

2

EFFECTS OF PRETRAINING ADMINISTRATION OF LINDANE, CHLORDECONE, ON ACQUISITION AND RETENTION OF PASSIVE AVOIDANCE Average

MEMORY

response

f SE

Partial

Complete

Time spent on lighted (unshocked) side (set)’

Retention Nosepokes”

AND p.p’-DDT

crossoversh

32

24.5 + 5.9

2.7 + 0.4

1.4kO.2

0.8 t 0.2

233.4 f 17.6

8 8

25.9 f 7. I 22.8 f 7.2

4.1 + 1.1 4.0 * 1.0

2.3 r 0.7 1 .o f 0.5

1.5 kO.8 3.1 * 1.2*

211.3 i 39.9 136.0+41.0*

8 8 5

15.8 + 7.1 24.1 + 7.9 19.6 + 4.7

2.4 r 0.8 2.9 k 0.2 2.2 t 0.6

I .o t 0.4 1. I f 0.4 0.8 t 0.6

1.5 r 0.7 1.6kO.8 0.8 t 0.4

166.8 k 46.1 193. I f 43.8 191.6 + 53.0

8 8 7

16.1 + 2.4 15.8 iz 3.1 21.6 + 6.0

2.4 zi- 0.4 2.9 t 0.6 2.3 t 0.5

0.9 f 0.3 1.6 f 0.7 0.9 -+ 0.3

0.3 t 0.2 0.3 k 0.6 I .o * 0.5

255.0 i 42.6 176.5 f 28.8 188.3 f 53.1

‘I Latency (set) to cross from lighted side to dark side during training. ’ Frequency of response occurring during 300-set retest period 1 week after training. Time spent on previously unshocked side during a 300-set retest period 1 week after training. * 17 < 0.05. from control (Mann-Whitney Utest).

When exposure to the chemicals occurred immediately after training, lindane had no effects on retention 1 week later (Table 3). Rats given the highest dose of chlordecone or p,p’-DDT spent less time in the lighted compartment during the retention test; 100 mg/ kg of chlordecone also increased significantly the number of complete crossovers during retention testing. In both passive avoidance experiments, the 100 mg/kg dose of p,g’-DDT or chlordecone resulted in mortality during the week after the exposure. Pharmacological interactions with tindane. Table 4 shows that none ofthe pharmacological probes had significant effects on the number of avoidance responses at the time of testing (3.5 hr after receiving the probe). Lindane significantly decreased the number of avoidance responses during the 60-trial session. Pretreatment with phenobarbital or chlordiazepoxide antagonized this disruptive effect of lindane, while phenytoin did not.

In the study concerning the interactions of the pharmacological probes with the convulsant effects of lindane and kainic acid, five of six animals given 60 mg/kg of lindane or 8 mg/kg of kainic acid displayed the full seizure pattern characteristic of each agent. This was completely blocked in animals pretreated with chlordiazepoxide or phenobarbital (none of the six rats in either group showed seizures). Pretreatment with phenytoin did not protect against seizures produced by kainit acid (five of six animals showed seizures) or lindane (six of six rats showed seizures).

DISCUSSION The results of the present study indicate that a single dose of lindane can interfere with the ability of rats to acquire a two-way shuttle box avoidance task and to retain a stepthrough passive avoidance task when trained

62

TILSON,

SHAW, TABLE

EFFECTS OF POST-TRAINING ON RETENTION

AND

MC

LAMB

3

ADMINISTRATION OF LINDANE, OF PASSIVE AVOIDANCE 1 WEEK

CHLORDECONE, AFTER TRAINING

AND

p,~‘-DDT

Time Treatment

and dose (m/k) Control Lindane 15 30 Chlordecone 25 50 100 p,p’-DDT 25 50 100

Crossovers”

spent

on lighted” (unshocked) side (set)

Iv

Nosepokes

Partial

Complete

33

1.7kO.4

0.8 * 0.2

0.2 f 0.9

289.5 2

7 7

2.7 f 1.0 2.4 f 0.8

0.7 -+ 0.4 1.6 t 0.5

0.9 f 0.6 0.1 f 0.1

264.9 + 22.X 276.4 t 23.6

8 8 8

1.3io.5 1.6 f 0.6 1.8kO.7

0.9 zk 0.3 0.4 f 0.3 0.9 * 0.4

0.4 f 0.3 0.4 f 0.3 1.8+0.5*

289.0 k 1 1.U 277.6 +- 14.8 144. I f 43.3*

8 8 5

2.7 + 1.2 1.9 ? 0.4 2.0 f 0.7

0.2 + 0.2 0.3 f 0.4 1.o f 0.4

0.7 * 0.7 0.2 f 0.5 1.3 kO.2

283.0 k 17.0 300.0 t 0 213.4 2 26.8*

5.7

’ Frequency of responses occurring during 300-set retest period 1 week after training. ” Time (set) spent on previously unshocked side during a 300-set retest period 1 week after training. *p < 0.05, from control (Mann-Whitney Utest).

under the influence of lindane. Post-training administration of lindane did not affect retention 7 days after learning. In general, these data are consistent with those of Desi (1974) who found that repeated exposure to lindane increased the number of errors made in a food-reinforced maze. One interpretation of these results is that lindane may interfere directly with learning. Woolley et al. ( 1984) reported that lindane has effects on long-term potentiation in the hippocampus and it is possible that this effect may compete or interfere with the utilization of new information. That post-training administration of lindane did not affect retention 7 days later suggests that the process of memory consolidation is not altered. However, it should be noted that the onset of pharmacological activity after oral administration of lindane is approximately 45-90 min, which may be too far removed from the consolidation process. Although it is possible that the impaired retention observed in rats trained in the passive avoidance task under the influence of lindane may be associated with state-dependent cues,

disruptive effects were not observed in rats trained under the influence of chlordecone or p,p’-DDT. Our results also indicate that chlordecone and p,p’-DDT have few consistent effects on the ability of rats to learn a conditioned twoway shuttle box response. Since it was observed that rats given higher doses of chlordecone or p,p’-DDT were hyperresponsive to external stimulation, it is possible that the avoidance responses obtained in these rats represent an enhanced reactivity to the light and tone stimuli. A pilot study, in which electric foot shock was not presented following the light and tone combination, found that vehicle control rats made 7 + 2 shuttle box responses in 60 trials, while rats receiving 100 mg/kg of either chlordecone or p,p’-DDT averaged 2 + 2 responses under these conditions (Shaw and Tilson, unpublished observations). The finding that p,p’-DDT had no effect on the ability of rats to perform a two-way conditioned avoidance response is in general accord with the data of Uppal et al. (1983a).

EFFECTS

OF

TABLE

ORGANOCHLORINES

ON

4

EFFECTS OF PHARMACOLOGICAL PRETREATMENT ON DISRUPTIVE EFFECTS OF LINDANEON TWO-WAY SHUTTI.E Box RESPONDING Average Treatment and dose (m/kg) Vehicle Phenobarbital 50 Phenytoin 15 Chlordiazepoxide IO Lindane 30 Lindane + phenobarbital Lindane + phenytoin Llndane + chlordiazepoxide

response

f SE

N

Avoidances”

Responding during the IT1

50

29.1+

79.2 +

8

1.7

4.6

31.1 t4.4

55.1 f 13.2*

12

29.9 + 3.0

69.0 k

9.0

8

34.1 k5.1

90.5 f

6.8

50

18.9 k 1.5*

73.5 f

3.7

12

28.2 f 3.0**

65.2 +

4.2

12

15.2 k 2.8*

61.0 t

6.9

8

27.3 f 3.8**

55.3 -+ 9.0*

li Correct avoidances or responses made during the intertrial interval during a 60-trial acquisition test. * p < 0.05, from vehicle control. ** p < 0.05. differs significantly from lindane, indicat~ng pharmacological antagonism ofdisruptive effect.

These investigators reported that acute exposure to DDT did not affect the ability of rats already trained to respond to a pole-climb response to avoid electric shock, although avoidance latencies were increased. However. Uppal et al. (1983a) also reported that the acquisition of the pole-climb response over days was inhibited in rats receiving DDT prior to each training session. Furthermore, DDT treatment facilitated the extinction of this response. Thus, although our data indicate that p,p’-DDT did not interfere with the acquisition of a shock-motivated shuttle response, it is possible that DDT might interfere with acquisition of responses having different topographies, or maintained by different reinforcers or training regimens (i.e.,

LEARNING

AND

MEMORY

63

acquisition over days, instead of within a single session). Finally, the effects of repeated dosing with DDT may be different from those following a single exposure. Uppal et al. (1983b) reported that rats trained in a step-through passive avoidance task under the influence of DDT were impaired when tested 24 or 48 hr later; when given in a high dose (95.0 mg/kg) 24 hr after training, retention was impaired 48 hr later. Although the latter observation is consistent with our results, the finding that pretraining doses of DDT (47.5 or 95 mg/kg) interfered with subsequent retention is not. Differences in predosing time, strain of rat, configuration of the apparatus, and other procedural differences might account for these disparate observations. Acute exposure to chlordecone had no significant effect on the acquisition of a two-way shuttle box task. Previous work in this laboratory (Tilson et al., 1979) indicated that chronic exposure to chlordecone via the diet did not impair the acquisition of a oneway avoidance task. However, retention of this task was impaired. Our data showing that chlordiazepoxide and phenobarbital block the behavioral effects of lindane are consistent with the interpretation that lindane may act by interfering with GABA. There is evidence that benzodiazepines exhibit anticonvulsant properties by a GABAergic mechanism (Killam and Suria, 1980), possibly acting on GABAergic containing basket cells to enhance recurrent inhibitions over the firing of pyramidal cells in the hippocampus (Lee et al. 1979). Furthermore, phenobarbital is also known to enhance GABAergic pre- and postsynaptic inhibition (Haeffely, 1980). Phenytoin, which at therapeutic concentrations does not appear to facilitate postsynaptic responses to GABA in rat hippocampal neurons (Herschkowitz and Ayala, 198 l), had no significant effect in lindane-treated rats. Matsumura and associates have reported that lindane can act as an agonist on the picrotoxinin receptor, inhibiting GABA-stimulated chloride uptake (Ghiasuddin and

64

TILSON,

SHAW,

Matsumura, 1982; Kadous et al., 1983; Matsumura and Ghiasuddin, 1983). This conclusion is supported by other studies indicating that lindane is a competitive inhibitor at the picrotoxin binding site within the GABA receptor-ionophore complex (Lawrence and Casida, 1984; Abalis et al., 1985). Woolley et al. ( 1984) have suggested that lindane may reduce recurrent inhibition in the hippocampus, using a paired pulse stimulation technique. This observation is important since the first stimulus is believed to activate recurrent collaterals that activate inhibitory GABAergic basket cells (Storm-Mathisen, 1977); GABA released from these cells is thought to inhibit responsiveness to the second stimulus. These data suggest that lindane interferes with GABAergic inhibition. Our data implicating GABA in some of the effects of lindane differentiates it from chlordecone and P,P’-DDT, which appear to have different neural mechanisms of action. DDT appears to slow the falling phase of the sodium current by binding to sodium channels in the open position, resulting in prolonged depolarization causing repetitive discharges of nerve fibers (Narahashi and Hass, 1967, 1968). Other data have indicated that the mechanism of chlordecone differs from that of DDT (Tilson et al., 1985; Tilson et al.. 1986). possibly by interfering with intracellular calcium (End et al., 1979). Phenytoin, which at therapeutic doses binds to the inactivation gates of sodium to block repetitive firing (MacDonald, 1983), blocks and exacerbates the effects of p,p’-DDT and chlordecone, respectively (Tilson et al., 1985; Tilson et al., 1986). In the present experiment, phenytoin had no effect on the behavioral effects of lindane. If phenytoin had effects on GABA responses, such as those that could occur with higher concentrations, then it should have had some effect on the lindane-induced alterations in behavior. The results of the present experiments demonstrate that lindane interferes with avoidance responding at lower doses (15 to 30 mg/kg) and produces seizures at higher

AND

MC

LAMB

doses (60 mg/kg). Although pharmacological probes suggest that these effects may be mediated by a GABAergic mechanism, additional studies will be required to determine more precisely the interaction between lindane and GABAergic inhibitory systems in the hippocampus. ACKNOWLEDGMENTS The authors are grateful for the assistance of Mrs. Loretta Moore in the preparation of this manuscript, Mr. Robert Hall in the construction ofthe behavioral apparatus used in these experiments, and the technical assistance of Mr. Charles Hamm.

REFERENCES ABALIS. I., ELDEFRAWI, M. E.. AND ELDEFRAWI, H. T. (1985). High-affinity stereospecific binding of cyclodiene insecticides and r-hexachlorocyclohexane to yaminobutyric acid receptor of rat brain. Pest. Biothem. Physiol. 24,95-102. DESI, I. (1974). Neurotoxicological effect ofsmall quantities of lindane. Int. Arch. Arbeitsmed. 33, I53- 162. END, D. W.. CARCHMAN. R. A.. AMEEN, R., AND DEWEY, W. L. (1979). Inhibition of rat brain mitochondrial calcium transport by chlordecone. Toxicol Appl. Pharmacol. 51, I89- 196. FULLER, T. A.. AND OLNEY, J. W. ( 1981). Only certain anticonvulsants protect against kainate neurotoxicity. Neurobehav. To-xicol. Teratol. 3,355-36 I. GHIASUDDIN, S. M., AND MATSLJMURA. F. (1982). Inhibition of y-aminobutyric acid (GABA)-induced chloride uptake by r-BHC and heptachlorepoxide. C.tw?p. Biochem. Physiol. c 73, I4 1- 144. HAEFFELY, W. E. (1980). GABA and the anticonvulsant action of benzodiazepines and barbiturates. Brain Re.\. BUN. S(supp1. 2), 873-878. HARRY, G. J., AND TILSON, H. A. ( 1981). The effects of postpartum exposure to triethyltin on the neurobehavioral functioning of rats. Neurotosicologv 2,283-196. HERSHKOWITZ, N.. AND A~ALA, G. F. ( I98 I ). Effects of phenytoin on pyramidal neurons of the rat hippocampus. Brain Rex 208,487-492. HERR, D. W., GALLUS. J. A., AND TILSON, H. A. ( 1987). Pharmacological modification of tremor and enhanced acoustic startle by chlordecone and p,p’-DDT. P.~ychopharmacology. in press. JARVIK, M. E.. AND KOPP, R. (1967). An improved onrtrial passive avoidance learning situation. P.syc~hol. Rep. 21,22 1-224. J OY, R. M. (1982a). Chlorinated hydrocarbon insectitides. In Pesticides and Neurological Diseases (D. J.

EFFECTS OF ORGANOCHLORINES Ecobichon and R. M. Joy, Eds.), pp. 91-150. CRC Press, Boca Raton, FL. JOY, R. M. (1982b). Mode of action of lindane, dieldrin and related insecticides in the central nervous system. Neurobehav. Toxicol. Teratol. 4,8 13-823. KADOUS, A. A., GHIASUDDIN, S. M., MATSUMURA, F., SCOTT. J. G., AND TANAKA, K. (1983). Difference in the picrotoxinin receptor between the cyclodiene-resistant and susceptible strains of the German cockroach. Pest. Biochem. Physiol. 19, 151-166. I&AIRY, M. (1959). Changes in behavior associated with a nervus system poison (DDT). Q. J. Exp. Psychol. 11, 84-94. KILLAM, E. K., AND SURIA, A. (1980). Benzodiazepines. In Antiepileptic Drugs. Mechanisms of Action (G. H. Glaser, J. K. Penry. and D. M. Woodbury, Eds.), pp. 597-616. Raven Press, New York. LAWRENCE,L. J., AND CASIDA, J. E. (1984). Interactions of lindane, toxaphene and cyclodienes with brain-specific r-butyl bicyclophosphorothionate receptor. Liff .‘;n. 35, 171-178. LEE, H. K., DUNWIDDIE, T. V., AND HOFFER. B. J. ( 1979). Interaction of diazepam with synaptic transmission in the in vitro rat hippocampus. NaunynSchmiedebergs Arch. Pharmacol. 309, 13 l- 136. MACDONALD, R. L. (1983). Barbiturate and hydantoin anticonvulsant mechanisms of action. In Basic Mechanisms of Neuronal Hyperexcitability (H. H. Jasper and N. M. van Gelder. Eds.). pp. 361-387. A. R. Liss. New York. MACDONALD, R. L., MCLEAN. M. J., AND SKERRITT, J. H. (1985). Anticonvulsant drug mechanisms of action Fed. Proc. 44,2634-2639. MACT~TUS, C. F.. UNGER, K. L., AND TILSON, H. A. ( 1982). Neonatal chlordecone exposure impairs early learning and memory in the rat on a multiple measure passive avoidance task. Neuroto.xicology 3,37-44. MA.TSUMURA. F., AND GHIASUDDIN. S. M. (1983). Evidence for similarities between cyclodiene type insecticides and picrotoxinin in their action mechanism, J. Environ. Sci. Health 318, l- 14. NARAHASHI, T.. AND HAAS, H. G. (1967). DDT: Interaction with nerve membrane conductance changes. Stiti~nc~~157,1438-1440.

ON LEARNING

AND MEMORY

65

NARAHASHI. T., AND HAAS, H. G. (1968). Interaction of DDT with the components of lobster nerve membrane conductance. J. Gen. Physiof. 51, 177- 198. SOBOTKA. T. J. (197 1). Behavioral effects of low doses of DDT. Proc. Sot. Exp. Biof. Med. 137,952-955. STORM-MATHISEN, J. (I 977). Localization of transmitter candidates in the brain: The hippocampal formation as a model. Prog. Neurobiol. 8, 119- 18 1. TILSON, H. A., BYRD, N., AND RILEY, M. (1979). Neurobehavioral effects of exposing rats to Kepone via the diet. Environ. Health Perspect. 33,32 1 [abstract]. TILSON, H. A.. HONG, J. S., AND MACTUTUS. C. F. (1985). Effects of 5,5-diphenylhydantoin (phenytoin) on neurobehavioral toxicity of organochlorine insecticides and permethrin. J. Pharmacol. Exp. Ther. 233, 285-289. TILSON, H. A., HUDSON, P. M.. AND HONG, J. S. (1986). 5.5-Diphenylhydantoin antagonizes neurochemical and behavioral effects of p.p’-DDT. but not chlordecone. J. Neurochem.. 47, 1870-1878. TILSON, H. A., PETERSON, N. J., AND MACTUTUS, C. F. (1984). Comparative effects of three organochlorines on hyperreactivity and tremor in rats. Toxicologist 4, 24. UPPAL, R. P., GARG, B. D., AND AHMAD, A. (1983a). Effect of malathion and DDT on the action of chlorpromazine and diazepam with reference to conditioned avoidance response in rats. Ind. J. Exp. Biol. 21,254-257. UPPAL. R. P.. GARG, B. D.. AND AHMAD, A. (1983b). Effect of malathion and DDT on the action of some tranquilizers on learning and memory traces in rats. Ind. J. E.xp. Biof. Z&617-619. WINER, B. J. ( 197 1). Statistical Principles in Experimental Design. McGraw-Hill. New York. WOOLLEY, D. E. ( 1982). Neurotoxicity of DDT and possible mechanism of action. In Mechanisms @.4ction OfNeurotoxic Substances (K. N. Prasad and A. Vemadakis, Eds.). pp. 95- 138. Raven Press, New York. WOOLLEY, D.. ZIMMER, L., HASAN, Z.. AND SWANSON. K. (1984). Do some insecticides and heavy metals produce long-term potentiation in the limbic system? In Ceflular and Molecular Neuroto.xicology (T. Narahashi, Ed.), pp. 45-69. Raven Press, New York.