Paraquat-induced, dose-dependent conditioned taste aversions and weight loss mediated by the area postrema

TOXICOLOGY

AND

APPLIED

PHARMACOLOGY

87,2 12-22

1 ( 1987)

Paraquat-Induced, Dose-Dependent and Weight Loss Mediated MICHAEL

S. DEY, ROBERT

Conditioned Taste Aversions by the Area Postrema

I. KRIEGER. ‘AND ROBERT C.

RITTER’

W’OI Regional Program in Veterinary Medical Education and Department of Veterinary Science. University ofIdaho, Moscow, Idaho 83843

Received February 10, 1986; accepted September 15. 1986 Paraquat-Induced, Dose-Dependent Conditioned Taste Aversions and Weight Loss Mediated by the Area Postrema. DEY, M. S.. KRIEGER, R. I., AND RITTER, R. C. (1987). Toxicol. Appl. Pharmacol. 87, 2 12-22 I. Paraquat’s (PQ) effect on feeding behavior in the rat was examined using a conditioned taste aversion (CTA) paradigm. CTA is a learned avoidance of tastes closely associated with prior illness. Male Sprague-Dawley rats trained to drink an instant breakfast solution were subsequently offered a novel-flavored solution and consumption was measured over 30 min. Following consumption of the novel solution, PQ (0.48-48.0 pmol/kg) was injected subcutaneously. Peak blood PQ concentrations were measured by serially sampling blood (0.15 ml) from an indwelling jugular cannula between 10 and 35 min after injection. Two days later, the rats were offered the same novel-flavored solution. Paraquat produced dose-dependent avoidance of the novel solution when injected subcutaneously. A PQ dosage of 2.7 rmol/kg or less did not alter consumption. The EDSO for CTA production of 13.0 pmol/kg was determined by log-probit analysis. The minimum effective dosage was 4.2 @m/kg. The doses examined did not produce overt clinical or histological signs of toxicity. Peak blood paraquat concentration was linearly related (r = 0.995) to dosage. Additionally when administered by gavage CTAs occurred only with a much larger PQ dosage (480 pmol/kg). Thermal lesions of a hindbrain circumventricular organ. the area postrema (AP). prevented PQ-induced CTAs despite repeated PQ injections. Additionally, weight loss following PQ exposure was also attenuated by AP lesions. CTAs were induced in these same AP-lesioned rats by oral administration of copper sulfate. This substance conditions taste aversions by activating vagal afferent neurons. The fact that copper sulfate-induced aversions were not blocked by lesions of the area posterema indicates that the lesioned rats are capable of forming CTAs when treated with a toxicant which does not act via the AP. These data indicate that PQ produces CTAs in a dose-dependent manner. Furthermore. PQ-induced CTAs and weight loss are mediated by the AP. The AP may contain receptors which detect xenobiotics, enabling animals to avoid future contact with these compounds. 0 1987 Academic Press, 1%

Paraquat (PQ) is a widely used herbicide which has undergone extensive toxicological examination, focused primarily on its lung toxicity. Effects of PQ on behavior have largely been unstudied. Parenteral exposure of rats and dogs to PQ results in decreased food consumption (Murray and Gibson,

1972; Sharp et al., 1972) and consequent loss of body weight (Gardiner and McAnalley, 1979; Saunier et al., 1982). In fact, loss of body weight following PQ exposures is so dramatic and consistent that it was used recently to predict the presence of PQ-induced lung disease in rats (Akahori and Oehme, 1983). Despite the consistency of PQ-induced body weight loss, there is no evidence that this effect is directly related to lung toxicity. PQ-induced weight loss and reduction of

’ Present address: California Department of Food and Agriculture, Worker Health and Safety Unit, 1220 N, Sacramento, CA 95814. * To whom reprint requests should be addressed. 0041-008X/87

$3.00

Copyright 0 1987 by Academic Press. Inc. All rights of reproduction in any form reserved.

212

PARAQUAT

TASTE

food intake may result from direct interaction of PQ with physiological systems which normally control feeding. Alternatively, these signs of PQ toxicity may be due to PQ-induced malaise rather than to a direct and specific effect on food intake control systems. In either case, it is likely that PQ exerts its behavioral effects by interacting directly with the brain or by stimulating visceral afferent nerves. e.g., the vagus nerve. Mediation of behavioral effects of ingested PQ by vagal afferents is plausible. since PQ ingestion has been reported to cause oral, esophageal, and gastric ulceration (Matthew et al., 1968; Conning et ~1.. 1969; Beebeejaun et ~1.. 197 1) and is capable of inducing emesis in humans after ingestion (Beebeejaun et al., 197 1). Since dogs vomit after intravenous (iv) PQ injection (Giri cf al.. 1982), PQ may also act directly on brain structures which mediate emesis. Although brain slices accumulate PQ by an energy-dependent process (Rose er al., 1976), PQ is a highly charged cation and should not easily cross the blood brain barrier (BBB) in riro. Therefore, if PQ does act directly on the brain. it most likely acts via a brain structure which lacks typical BBB characteristics. One likely brain structure through which PQ might exert its behavioral effects is the area postrema (AP), located in the dorso caudal medulla oblongata. The AP lacks a BBB and appears to contain chemoreceptors which mediate emesis in response to some bloodborne chemicals (Borrison and Wang, 1953: Borrison. 1974). Additionally, the AP has been implicated in the conditioning of taste aversions in response to parenterally injected lithium chloride or scopolamine methyl nitrate (Berger et ~1.. 1973; Ritter e/ al., 1980). Conditioned taste aversions (CTAs) are learned responses which reduce the likelihood of repeated ingestion of a food associated with illness. The CTA paradigm involves allowing rats to eat a novel test food (conditioned stimulus). Immediately after they have ingested the novel-flavored food, the rats are injected with a test chemical (unconditioned stimulus). If the administered chem-

213

AVERSIONS

ical causes malaise, then on subsequent presentations the rats will reject (conditioned response) the food which is associated with the previous toxicity. This form of conditioning is believed to be responsible for “bait shyness” in wild animals (Berger et al., 1973). This study was undertaken (1) to determine whether PQ produces dose-dependent conditioning of taste aversions, (2) to confirm the occurrence of PQ-induced weight loss, (3) to determine whether the intact AP participates in this loss of weight, (4) to determine whether PQ can serve as the unconditioned stimulus for a learned change in ingestive behavior (conditioned taste aversion), and (5) to evaluate the AP’s role in taste aversions conditioned by PQ. METHODS Paraquat (methyl viologen) was obtained from Sigma Chemical Co. (St. Louis. MO). Methyl-labeled [‘“Clparaquat (14CH,-PQ. I1 I mCi/mmol) was purchased from Amersham Radiopharmaceutical Co. (Arlington Heights, IL). :lnirnuls Adult male Sprague-Dawley derived rats (Washington State University Laboratory Animal Facility) weighing 325 ? 10 g (initial weights) were used. The rats were maintained in an AALAC-approved facility on a I?: I ?-hr 1ight:dark cycle (lights on at 7:00 AM). They were housed individually in hanging wire cages and had free access to food (Purina Rat Chow) and water except during training and testing sessions. The rats were acclimated for 5 days before any studies were initiated, Training and testing were performed in the rats’ home cages beginning at lo:30 AM. Surgeries on the AP were performed under methoxyfluorane anesthesia. The foramen magnum was exposed by blunt dissection of the muscles overlying the atlanto occipital junction. Following incision of the meninges, the foramen magnum was enlarged slightly by removing a few chips of occipital bone with microrongeurs. Subsequently, the vermis of the cerebellum was elevated. and the AP was exposed and thermally lesioned (M&one ef al., 1980). Sham-operated control rats (shams) were treated identically. except that the AP was touched by a cotton-tipped swab instead of by the thermocautery tip. All rats were allowed a 15 to 30day recovery period following surgery before the behavioral studies were initiated. After completion of the behavior studies, both shams and lesioned rats were deeply anesthetized with sodium pentobarbital (75 mg/kg) and exsanguinated. The rats were perfused via the left heart with saline followed by 10% buffered formalin using a 25-cm pressure head. The

214

DEY,

KRIEGER.

brain was removed and postfixed and 20-pm sections were stained using Weil and cresyl violet procedures. The extent of the dorsal brain stem lesions was mapped and reconstructions were made using anatomical landmarks provided by the atlas of Pellegrino et ul. (1979). Statistical analyses of the data were performed using paired t tests. Procedures. To examine the dose-response relationship for PQ-induced CTA formation rats were trained to drink a solution of vanilla instant breakfast (1 g:6.6 g water: Western Family Foods, Portland, OR) from 50-ml calibrated drinking tubes. This solution was presented every other day for 30 mm for approximately 2 weeks until the rats consistently drank the same amount (*IO%) during each trial. Seven to 10 training sessions were required. Seventy-two hours prior to dosing with PQ. the rats were implanted with indwelling jugular cannulas. On the day of PQ dosing (Day I), rats were offered an instant breakfast flavor (i.e., strawberry. banana, etc.) to which they had not been exposed previously (novelflavored). The rats had access to the novel-flavored instant breakfast for 30 min. The volume consumed by each rat was recorded and the drinking tubes were removed immediately following the 30-min period. Within 10 min of removing the drinking tubes, each rat that drank 3.0 ml or more of the novel solution received a subcutaneous (SC) injection of 14CH3-PQ in the thigh. Dosages of 14CH3-PQ. dissolved in distilled water. ranged from 0.48 to 48.0 pmol/kg. To measure peak blood PQ concentrations, blood (0.15 ml) was sampled serially via the indwelling cannulas between 10 and 35 min after injection (Dey ef al.. 1986). Forty-eight hours later (Day 3) these same rats were again offered the novel instant breakfast solution and their 30-min consumption was measured. The Day 3 effective dose (ED50) which resulted in a mean decrease of half the Day I consumption was determined using log-probit graphic analysis (Litchfield and Wilcoxon. 1949) based on administered dosage. The ED50 and slope of the curve for mean decrease in consumption were also determined from mean peak blood PQ concentrations. Correlation between dosage and peak blood PQ concentrations was examined by graphic analysis. The slope of the curve was determined by linear regression analysis. The first experiment examined the effects of PQ on body weight in sham and AP-lesioned (APL) rats using a dosage which did not produce lung or renal damage when injected subcutaneously. Based on earlier unpublished studies, a dosage of 48 pmol/kg was used. Ten APL rats and nine shams were injected SC with PQ. and body weight was recorded 24 and 48 hr postinjection. The role ofthe AP in mediating PQ-induced taste aversions was also investigated. Five APL rats and five shams were trained to drink from calibrated tubes. Once trained, the rats were offered a novel-flavored instant breakfast solution (strawberry) for 30 min in place of the familiar vanilla solution (Day 1). The volume consumed

AND

RITTER

by each rat was recorded. Immediately after (< 10 min) the drinking period, each animal that drank at least 3.0 ml of the novel solution received an SC injection of PQ (48.0 rmol/kg). Forty-eight and 96 hr later (Days 3 and 5). these same rats were again presented with strawberry instant breakfast and their 30-min consumption was measured. To establish whether the AP lesion or incidental damage to the adjacent medullary vagal projections was responsible for the impaired ability to acquire PQ-induced taste aversions, consumption ofa novel-flavored solution was paired to intragastric cupric sulfate administration (Coil et ~1.. 1978). Therefore, rats from the first experiment were retrained on the dilute solution of vanilla instant breakfast. Once retrained, the rats were tested as before. Each rat drinking 3.0 ml or more of the novelflavored instant breakfast received a solution of cupric sulfate (5 mg/kg; 0.68 mg/ml) by gavage using a French urethral catheter (No. 8. Darvol. Inc.). Forty-eight (Day 3) and 96 hr (Day 5) later, the animals were again offered the novel-flavored solution for 30 min and consumption was measured. To determine whether PQ absorbed from the gastrointestinal tract induced CTAs. the effect of intragastrically administered PQ was examined. Additionally. the importance of the AP for the formation of CTAs conditioned by intragastric PQ was examined. Five APL rats and five shams were trained and tested as before. using vanilla- and anise-flavored instant breakfast. All animals consuming 3.0 ml or more were intragastrically intubated with PQ (48 bmol/kg). Forty-eight hours later, consumption of the anise-flavored solution was measured again. Subsequently. this experiment was repeated using 480 pmol/kg and orange-flavored instant breakfast. The effect of hunger induced by food deprivation on PQ-induced CTA formation was examined. Food deprivation effectively increased consumption of instant breakfast by all experimental animals. Thus, conditioning of an aversion must overcome not only the rats’ drive to consume a highly palatable food. but also their hunger due to prior restriction of intake. Five APL rats and four shams were trained as described above. All nine rats were fasted for 24 hr prior to being offered a banana-flavored solution. Consumption was recorded every 30 min during a 120-min testing period. All animals consuming 3.0 ml or more in the first 30 min were injected SC with PQ (48 gmol/kg) immediately (< I5 min) after the end of the 120.min testing period. Weights were recorded daily. When the rats’ body weights had stabilized following the injection, they were again fasted for 24 hr. The bananaflavored solution was offered again and consumption was recorded. Within I5 min after the 120-min test, the animals were again injected with PQ. This regimen was repeated two more times until there was no consumption of the flavored solution by the shams. Nochlin and Levine (1982) demonstrated that nonnecrotizing lesions in the area of the AP were produced

PARAQUAT

TASTE TABLE

DOSE-RESPONSE RELATIONSHIP ANI) MEAN PERCENTAGE DECREASE

BETWEEN MEAN IN CONSUMPTION

215

AVERSIONS 1

PEAK BLOOD PARAQUAT OF NOVEL-FLAVORED

CONCENTRATIONS

INSTANT

(AND

BREAKFAST

DOSAGE)

(AND NUMBER

OF RUTS AVERTING) Mean peak blood paraquat concentrations (nmol/ml _t SD)

Dosage (pmohz) 48.0 18.3 10.7 8.2 4.8 4.2 2.7 1.5 0.48 ‘I Aversion sumption. ” Decrease

52 16 14 9.4 3.2 5.4 3.5 2.2 0.54 determined

f 7.8 It 5.6 f 1.2 f 1.5 * 1.9 kO.9 kO.02 f0.17 k 0.07

as a decrease

in Day 3 consumption

No. of rats averting”/ No. of rats tested 515 414 415 515 3/5

100 64.3 44.9 29.2 18.1 10.8 0 0 0

215 015 O/5 015

in Day 3 consumption

during

Mean % decrease in consumption b

a 30-min

feeding

by an aliphatic triamine. Therefore. the possibility that PQ itself produced lesions of the AP region was examined. Four rats were injected SC with PQ (48 smol/kg). Twenty-four hours later, the rats were anesthetized with sodium pentobarbital and perfused via the left heart with saline followed by 10% buffered formalin. The brains were sectioned in 20-pm sections. stained with hematoxylin and eosin. and examined for neuropathology.

RESULTS A dose-dependent relationship was observed between PQ dosage and consumption. With increasing dosages, an increase in the number of rats forming aversions was observed. Additionally, the volume of novelflavored instant breakfast that was consumed decreased with increasing PQ dosages (Table 1). Decreases (lo- 15%) of Day 1 consumption were considered significant. Peak blood PQ, measured 10 to 20 min after SCinjection, increased linearly (r = 0.995) with dosage. Increases in peak blood PQ concentrations correlated with the decrease in consumption of novel-flavored solution and the number of rats forming CTAs (Table 1). Peak blood PQ concentrations of 3.5 nmol/ml (dosage = 2.7 pmol/kg body wt) or less did not alter consumption by any of the rats.

of novel-flavored trial expressed

solution as a % of Day

compared

with Day

I con-

1 consumption.

An ED50 of 13.0 pmol/kg and slope of 2.6%/pmol/kg were determined based on a log-probit plot of the mean percentage decrease in consumption vs dosage. Similar graphic analysis of the mean percentage decrease vs mean peak blood PQ concentration yielded an ED50 equal to 13.9 nmol/ml and a slope of 2.3%/nmol/ml. The slopes of the lines were not significantly different (p > 0.05). No damage was evident in the coronal sections through the region of the AP in the sham-operated rats (Fig. 1). The AP was totally absent in the APL rats, except in one animal in which only 70% of the AP was destroyed. Damage to tissue surrounding the AP was variable. Nevertheless, the behavior of the rats with more extensive damage to adjacent structures was not different from that of the rats with less extensive lesions. Typical lesions totally destroyed the AP and also caused moderate damage to the immediately adjacent nucleus of the solitary tract. in some rats, the lesion impinged on the dorsal motor nucleus of the vagus. Representative photomicrographs are shown in Fig. 1. Histological examination of coronal sections through the

216

DEY,

KRIEGER,

AND

RITTER

FIG. I. Photomicrographs of the AP region from sham-operated controls AP = area postrema; DMN = dorsal motor nucleus of the X cranial nerve: CR = nucleus gracilis: TSL = solitary tract: XII = nucleus of the XII cranial

AP region from rats killed 24 hr after an SC injection of PQ revealed no damage to the region. Subcutaneously injected PQ (48 pmol/kg) caused a significant (p < 0.005) weight loss in shams both 24 and 48 hr later. Body weight returned to preinjection levels by 9 days. APlesioned rats did not lose significant body weight (Table 2). A single SC injection of PQ into shams following exposure to a novel-flavored diet (Day 1) abolished consumption when this solution was offered again 48 and 96 hr later (Day 3). APL rats, however, did not reduce their Day 3 or Day 5 intake of the instant breakfast after PQ treatment (Table 3).

(A) and AP-lesioned rats (B). SOL = solitary tract nucleus: nerve.

In contrast to parenteral PQ, intragastric administration of cupric sulfate virtually abolished Day 3 consumption of a novel-flavored solution by both sham and APL rats. Day 5 consumption of the novel instant breakfast was also significantly reduced in both groups (Table 3). Unlike intragastric cupric sulfate, intragastric PQ (480 pmol/kg) conditioned taste aversions only in shams, and not in APL rats. When administered intragastrically, 48 pmol/kg of PQ failed to produce CTAs in either group (Table 3). No change in consumption of the novelflavored instant breakfast was observed following repeated injections of PQ in 24-hr

PARAQUAT

TASTE TABLE

217

AVERSIONS 2

BODY WEIGHT LOSSES”OF SHAM AND APL RATS 1 AND 2 DAYS AVER A SUBCUTANEOUS INJECTIONOF PQ (48 pmol/kg) Day I

Day 2

9

Yobody wt

g

90body wt

14 *9.9*

3.4 f 2.2+

16 k 12.7*

3.6 k 2.3*

4.8 f 8.6

1.3 + 1.9

4.3 + 8.9

1.2 f 1.9

Sham-operated controls n=9

AP-lesioned rats II = 10

* Significantly different from preinjection weights (p < 0.005). o Values expressed are the means i SD.

food-deprived APL rats (Table 4). However, food-deprived shams decreased their consumption of the novel-flavored solution after

TABLE

each PQ injection. Reduced consumption was significant (p < 0.05) after the first PQ injection. However, four injections were re-

3

NOVEL-FLAVORED INSTANT BREAKFAST CONSUMPTION’ IN SHAM AND APL RATS PQ injected 48 pmol/kg Sham-operated conrols n=5 AP-lesioned n=5 Cupric sulfate gavage 5.0 mg/kg Sham-operated controls n=4 AP-lesioned n=5 PQ gavage 48 rmol/kg Sham-operated controls 12= 4 AP-lesioned n=4 480 pmol/kg Sham-operated controls n=3 AP-lesioned n=5

Day 1

Day 3

Day 5

6.1 f 1.8

o.o**

o.o**

18 + 3.4

19 +- 1.9

6.7 + 2.7

o.o*

16 ? 7.2

1.1 i- 2.3*

5.0 + 0.9

6.2 of 1.8

17 f3.2 7.2 It 2.4 23 f 5.3

18 k4.6

0. I f 0.04* 2.7 k 5.1*

18 k-2.0 o.o* 21 +5.7

&‘o/e. Day I represents predosing consumption. Consumption on Days 3 and 5 is after administration intragastric cupric sulfate. * Significantly different from Day I consumption (p < 0.05). ** Significantly different from Day I consumption (p < 0.005). ” Values are expressed in milliliters as the means -+ SD.

of PQ or

218

DEY, KRIEGER,

AND RlTTER

TABLE 4 PREFASTED

BODY WEIGHTS(~)

Sham-operated controls n=4 NFlBconsumed (120 min) Weight AP-lesioned n=5 NFIB consumed ( 120 min) Weight

AND NOVEL-FLAVORED INSTANT BREAKFAST (NFIB) CONSUMPTION (ml) AFTER 24-hr FAST IN SHAM AND APL RATS Day 1

Day IO

Day 19

Day 25

2’2 5.2” 5OOk23”

I I + 10.2* 490 f 30

8.6 k 10.2* 480 * 30

4.4_t 5.1* 480 +28

o.o* 490 +- 2x

33 f 7.0” 460 f 29”

33 k 3.4 460 f 27

32 +- 4.5 470 k 28

30 -t 5.9 470 -t 27

33 f 4.1 470 +- 28

Day 34

Note. Day 1 represents preinjection values. Body weights and consumptions on Days IO, 19. 25, and 34 are after repeated subcutaneous injections of PQ (48 pmol/kg). * Significantly different from Day I consumption (p < 0.05). 0 Values are expressed as the means + SD

quired to totally shams.

abolish

consumption

by

CTA induction by low doses of PQ was dose-dependent. Further, a linear relationship that was observed between dosage and peak blood PQ concentrations, over the dosDISCUSSION age range tested, may be used for predicting PQ-induced CTAs. Taste aversion conditioning is a sensitive Loss of body weight following PQ adminismeasure of response to xenobiotic exposure. tration was prevented by lesions which deRats acquire taste aversions without overt or stroyed the AP. PQ-induced conditioned even histological signs of toxicity. PQ retaste aversions were also abolished by such leduced feeding by causing illness at a dosage sions. The importance of an intact AP has one-tenth that required to produce lung disease. The ED50 for PQ-induced CTAs was been demonstrated in the formation of CTAs scopolamine 13.0 gmol/kg with a minimum effective dose to parenterally administered methyl nitrate (Berger et al., 1973) and lithof 4.2 pmol/kg. A dose greater than 48 pmol/ kg is required to produce lung disease (un- ium chloride (Ritter et al., 1980). Since the published observations). PQ’s induction of AP is outside the BBB and, therefore, is CTAs, at a blood PQ concentration as low as highly accessible to blood-borne chemicals, it 5.4 nmol/ml, demonstrates the sensitivity of is probable that PQ reduces food intake and aversions by acting directly this behavioral response at a blood PQ con- conditioned centration which fails to produce lung dis- through this neural structure. Alternatively, the AP and adjacent nucleus of the solitary ease. tract are sites of vaga1 afferent termination Blood PQ concentrations of 3.5 nmol/ml or less do not produce CTAs and, therefore, and, consequently, AP lesions may impair are not expected to alter food intake or reduce formation of PQ-induced CTAs because of body weight. This is consistent with studies in damage to vagal afferent projections. This inwhich infused doses of PQ resulted in blood terpretation is unlikely since 70% destruction PQ concentrations of 0.8 nmol/ml after 21 of the AP, with minimal damage to adjacent days and did not alter food intake or body areas, prevented conditioning of taste averweight gain (unpublished observations). sions by PQ. Thus, rats with little or no vagal

PARAQUAT

TASTE

afferent damage behaved similarly to rats having complete destruction of the AP and more extensive damage to surrounding tissue. Additionally, intragastric cupric sulfate produced CTAs in both APL and sham-lesioned rats (Table 2). Since Coil d al. ( 1978) demonstrated that intragastrically administered cupric sulfate solution produces CTAs which are principally mediated by vagal afferents from the gut, persistence of cupric sulfate-induced aversions in APL rats suggests that vagal afferents necessary for conditioning of some aversions were intact in lesioned rats. Therefore, the formation of conditioned taste aversions following PQ exposure is reliant on the intact AP. Finally, the fact that cupric sulfate did condition aversions in APL rats demonstrates that the ability to learn is not impaired by the lesion. Intragastric administration of PQ (480 pmol/kg) also caused the formation of CTAs in intact rats. Therefore, the ability of PQ to induce formation of CTAs is not limited to the subcutaneous route of administration. Furthermore, these CTAs conditioned with intragastric PQ were abolished by AP lesions and do not depend on stimulation of gastrointestinal vagal sensory neurons. Therefore, impaired food intake after oral PQ does not depend on gastrointestinal irritation with consequent vagal stimulation. Since the formation of conditioned taste aversions was dose-dependent, the failure of 48 pmol/kg of PQ administered by gavage to induce CTAs was probably due to the poor gastrointestinal absorption of PQ (Daniel and Gage, 1966; Murray and Gibson, 197 1). AP-lesioned rats consistently consumed at least two times more instant breakfast solution than the shams, giving a consumption ratio (milliliters consumed by shams/milliliters consumed by APL rats) of less than 0.5 (Table 2). The effect of the lesion has been reported previously (Edwards and Ritter, 198 1) and appears to be due to interference with the hindbrains role in control of appetite. Fasting the rats for 24 hrjust prior to testing increased the consumption ratio of shams

AVERSIONS

219

to AP-lesioned rats to greater than 0.7. However, despite the increased consumption of the two groups, the shams formed CTAs while the APL rats failed to form aversions even after repeated PQ injections. Food deprivation is known to increase resistance to conditioning of taste aversions. Fasting increased resistance to formation of CTAs in response to PQ. A single injection of PQ significantly (p < 0.05) decreased future consumption of a novel food by sham-operated rats even when they were hungry following a 24-hr fast. Four PQ injections totally abolished consumption by shams (Table 3). Nevertheless, the fact that APL rats did not reduce intake of the novel solution even after four PQ injections suggests that they were indeed refractory to PQ-induced malaise. During repeated fasting, the body weights of the shams decreased, as did consumption of the novel instant breakfast for the first two PQ injections (Table 3). However, after the third injection the body weights began to increase, despite a continued decrease in the consumption of the flavored solution. One explanation of the transient weight loss by the shams in the present study would be the formation of a transient aversion to the animals’ familiar diet. Similar observations have been made by Bernstein et al. ( 1980) and Bernstein and Goehler ( 1983) in which food intake was found to parallel aversion formation. Furthermore, these authors demonstrated that the declines in food intake were transient for familiar diets but persisted in animals on novel diets. While PQ-induced malaise may be responsible for the transient decrease in food consumption and subsequent weight loss, the increase in body weight of shams after the third and fourth injections suggests that malaise alone was not responsible. Regardless, the PQ-induced weight loss was blocked when the taste aversions were eliminated by AP lesions (Table 3). Decreased food consumption is often used as an indication of toxicity. Clearly, behavioral changes which affect feeding can occur at doses less that those which produce tissue

220

DEY,

KRIEGER,

AND

RITTER

damage. Conditioned taste aversions are sen- AULERICH, R. J., AND RINGER, R. K. (1977). Current status of PCB toxicity to mink, and effect on their resitive changes in behavior which may indiproduction. Arch. Environ. Contam. Toxicol. 6,27B9cate exposure to xenobiotics. Other chemi292. cals also cause weight loss and decreased food BEEBEEJAUN, A. R., BEEVERS, G., AND ROGERS. W. N. consumption in animals (Montgomery and (197 1). Paraquat poisoning: Prolonged excretion. Clin. Toxicol. 4, 397-407. Holtzman. 1975; Aulerich and Ringer, 1977; BERGER, B. D., WISE. C. D.. AND STEIN, L. (1973). Area Tjarka et al., 1981: Hietamen et al., 1982; postrema damage and bait shyness. J. Camp. Phwiol. Gupta et al., 1983; Ring et al., 1983). The PsychoI. 82,475-479. possibility that weight loss may be linked to BERNSTEIN, I. L.. AND GOEHLER, L. E. (1983). Chronic taste aversions may have important implicalithium chloride infusions: Conditioned suppression of food intake and preference. Behov. Meuro.rci. 97, tions for toxicity testing, since changes in 290-298. weight are a frequently used indicator of animal health. The possibility exists that weight BERNSTEIN, 1. L., VITIELLO. M. V., AND SIGMUNDI, R. A. (1980). Effects of interference stimuli on the acloss may be due to developing aversions to quisition of learned aversions to foods in the rat. J. familiar laboratory diets or to diets where the Comp. Physiol. Psychol. 94,92 l-93 1. flavor varies sufficiently to be perceived as a BORRISON,H. L. ( 1974). Area postrema: Chemoreceptor trigger zone for vomiting-Is that Ally L(fe Sci. 14, novel diet. Furthermore, decreased food and 1807-1817. water consumption results in biochemical BORRISON, H. L.. AND WANG, S. C. (1953). Physiology changes, e.g., decreased lung and liver desatand pharmacology of vomiting. Pharmacol. Rev. 5. urase activity (Montgomery and Holtzman, 193-230. 1975; Montgomery, 1976) and pathological COIL, J. D.. ROGERS, R. C., GARCIA. J., AND NOVIN. D. ( 1978). Conditioned taste aversions: Vagal and circuchanges, e.g., kidney, which may inadverlatory mediation of the toxic unconditioned stimulus. tently be mistaken for direct effects of the Behov. Biol. 24, 509-5 19. chemical under study. Pair feeding may be CONNING. D. M.. FLETCHER, K., AND SWAN, A. A. B. necessary to separate these indirect from di( 1969). Paraquat and related bipyridyls. Brit. Med. J. rect effects. 25,245-249. DANIEL, J. W., AND GAGE, J. C. ( 1966). Absorption and In summary, PQ induces dose-dependent excretion of diquat and paraquat in rats. Brit. J. Ind. conditioned taste aversions. The AP is an imMed. 23, 133- 136. portant neural site which participates in the DEY, M. S., BREEZE, R. E., HAYTON, W. L., KARARA, detection of blood-borne chemicals and has A. H.. KRIEGER, R. I., NASER. L. J., AND RENZI, B. E. an important role in the control of PQ-in(1987). Paraquat pharmacokinetics using a subcutaneduced taste aversion formation and weight ous toxic. low dose in the rat. Fundam. Appl. To.uicol.. in press. loss. PQ-induced weight loss and CTA formaEDWARDS, G. L.. AND RITTER, R. C. ( 198 1). Ablation of tion are blocked by destruction of the AP. the area postrema causes exaggerated consumption of The inability of PQ to produce CTAs and preferred foods in the rat. Brain Res. 216,265-276. weight loss in APL rats is due to loss of the GARDINER. T. H.. AND MCANALLEY, B. H. (1979). AP and not to the inability to acquire taste Changes in the pulmonary uptake and binding of drugs in an experimental model of lung fibrosis. Toxiaversions, since intragastric cupric sulfate, col. Appi. Pharmacol. 49,487-496. which acts primarily via the gastrointestinal GIRI. S. N., PARKER, H. R., SPANGLER, W. L., MISRA, vagal afferents, produces CTAs in both sham H. P., ISHIZAKI, G., SCHIEDT, M. J., AND CHANDLER, and APL rats. The AP may be a site which is D. B. ( 1982). Pharmacokinetics of [‘4C]-paraquat and specifically adapted to detect xenobiotics and associated biochemical and pathologic changes in beagle dogs following intravenous administration. Funinitiate behavioral and physiological action dam. Appl. Toxicol. 2,26 1-269. to reduce the hazards of poisoning. GUPTA, B. N., MCCONNELL, E. E., GOLDSTEIN, J. A., REFERENCES HARRIS, M. W., AND MOORE, J. A. (1983). Effects of AKAHORI, F., AND OEHME, F. (1983). Inhibition ofcollagen synthesis as a treatment for paraquat poisoning. Vet. Hum.

Toxicol.

25, 32 1-327.

a polybrominated biphenyl mixture in the rat and mouse. I. Six month study. Touicol. Appl. Pharmucol. 68, l-18.

PARAQUAT

TASTE

HIETAM~N. E.. AITIO, A., KOIVUSAARI. U., K~LPIO. J., NEVALAINEN. T., NARHI, M.. SAVOLAINEN, H.. AND VAINIO. H. (1982). Tissue concentrations and interaction of zinc with lead toxicity in rabbits. To.~ico/ogr 25,113-127. KING, F. G.. DEDRICK. R. L., COLLINS. J. M., MATTHEWS, H. B.. AND BERNBAUM, L. S. (I 983). Physiological model for the phannacokinetics of 2.3,7,8tetrachlorodibenzofuran in several species. Tosicol. .4ppl. Pharmacol.

67.390-400.

MATTHEW, H.. LOC;AN. A., WOODRUFF, M. F. A.. AND HEARD, B. (1968). Paraquat poisoning: Lung transplantation. Brit. Med. J. 3,759-763. MCGLONE, J. J., RITTER. S., AND KELLEY, K. (1980). The antiaggressive effect of lithium is abolished by area postrema lesion. Physiol. Behav. 24, 1095- 1100. MONTGOMERY, M. R. (1976). Interaction of paraquat with the pulmonary microsomal fatty acid desaturase system. T~ruicol. .4ppl. Pharmacol. 36, 543-554. MONTGOMERY. M. R.. AND HOLTZMAN, J. L. (1975). Drug-induced alterations in hepatic fatty acid desaturase activity. Bioc,hem. Pharmarol. 24, 1343- 1347. MURRAY, R. E.. ANDGIBSON, J. E. (1971). Lethality and pharmacokinetics of paraquat in rats. Tosicol. Appl. Pharmacol.

19,405.

R. E.. AND GIBSON, J. E. (1972). A comparative study of paraquat intoxication in rats, guinea pigs and monkeys, E,vp. .Mol. Pathol. 17, 3 I l-325.

MURRAY.

221

AVERSIONS

NOCHLIN, D., AND LEVINE, S. L. (1982). Ultrastructure of hypothalamic and medullary lesions caused by an aliphatic triamine. J. Neuropathol. Exp. Neurol. 41, 233-240. PELLEQRINO, L. J.. PELLEGRINO. A. S., AND CUSHMAN. A. J. (1979). A Stereotaxic Atlas of‘the Rat Brain. 2nd ed. Plenum, New York. RITTER, S.. MCGLONE, J., AND KELLEY, K. (I 980). Absence of lithium-induced taste aversion after area postrema lesion. Brain Rex 201,501-506. ROSE. M. S.. LOCK, E. A., SMITH, L. L.. AND WYATT. I. (1976). Paraquat accumulation: Tissue and species specificity. Biochem. Pharmacol. 25,4 19-423. SAUNIER, C.. SCHRIJEN, F., GILLE. J. P., HORSKY. P., PESLIN, R.. HARTEMANN, D., FOLIGUET. B.. AND LAMBERT, H. (1982). Experimental chronic paraquat poisoning. Pulmonary functional and pathological modifications. Bull. Eur. Physiopathol. Respir. 18, 863-876. SHARP, C. W., OTTOLENCHI, A., AND POSNER, H. S. (1972). Correlation of paraquat toxicity with tissue concentrations and weight loss of the rat. Toxicol. Appl. Pharmacol. 22, 24 l-25 1. TARKJA,

S. M., JR, ZOUMAS,

B. L.,

AND

GANS,

J. H.

(198 1). Effects ofcontinuous administration of dietary theobromine on rat testicular weight and morphology. Towicol.

Appl. Pharmacol.

58, 76-82.