BEHAVIORALAND NEURALBIOLOGY58, 1--7 (1992)
Scheduled Running Wheel Activity Indexes the Specificity of Pharmocological Anorexia NORI GEARY, JENNIFER FUDGE, AND JOSEPH LE SAUTER1 Department of Psychology, Columbia University, New York, New York 10027
Nondeprived male Sprague-Dawley rats that were given scheduled access to running wheels for 60 min daily ran immediately and energetically. Intraperitoneal inject,ions of 400 t~g/kg pancreatic glucagon and 0.15 ~g/kg cholecystokinin octapeptide had no effect on scheduled running, but significantly inhibited feeding when the rats were offered condensed milk instead of access to the running wheels. This is consistent with the hypothesized fhnction of these peptides as postprandial satiety signals. In contrast, 0.5 mg/kg amphetamine and 75 t~M/kg LiC1, which produced similar degrees of anorexia, inhibited running by about 50%. Amphetamine, but neither peptide, also inhibited water drinking and disrupted the behavioral sequence of postprandial satiety. The distance run during scheduled running tests was inversely related to body weight, but the patterns of the drugs' effects were not altered by baseline running differences. Scheduled wheel running is a robust consummatory behavior that appears to provide a relatively valid, simple, and sensitive test of the behavioral specificity of pharmacological anorexia. ~ 1992 Academic Press, Inc. The physiological controls of consummatory behavior are presumed to be highly selective. Providing convincing demonstrations of behavioral specificity, however, has been a persistent problem in the behavioral neurobiology of motivation. The problem is especially acute in studies of the satiation of hunger. First, many nonspecific effects that can interfere with ingestion, such as sedation or malaise, may superficially mimic postprandial satiety. Second, important components of the neural mechanisms mediating ingestive behavior may also participate in the control of very different responses 1 Please address reprint requests to N. Geary, Bourne Laboratory, New York Hospital-Cornell Medical College, 21 Bloomingdale Road, White Plains, NY 10605. We thank Dr. Anthony Sclafani for the use of the running wheels and Ms. Debra Rosenzweig for help with the experiments. This research was supported by NIH Grant DK-32448 to N. Geary.
(for discussion see Berridge & Valenstein, 1991; Hoebel, 1988; Stricker, 1990). For these reasons, a great deal of effort has been devoted to the development of behavioral criteria to evaluate the specificity of hypothesized postprandial satiety mechanisms in animals. Specificity screens have been proposed based on the microstructure of periprandial behaviors (Blundell, 1985; Crawley, 1983), meal patterning (VanderWeele, Granja, & Deems, 1982), dietary self-selection (VanderWeele, Deems, & Gibbs, 1984), presence of species-typical postprandial behaviors (Antin, Gibbs, Holt, Young, & Smith, 1975; Levine, Kuskowski, Grace, & Billington, 1991; Smith & Gibbs, 1979), operant performance for food reward (Flood, Silver, & Morley, 1990), water intake (Gibbs, Young, & Smith, 1973), reaction to food deprivation (Billington, Levine, & Morley, 1983), and acquisition of conditioned aversions (CTA; Deutsch & Hardy, 1977; Gibbs et al., 1973; Swerdlow, Van der Koy, Koop, & Wenger, 1983). To these behavioral criteria may be added criteria of physiological specificity, such as EEG (Danguir, Nicolaidis, & Gerard, 1979; Mansbach & Lorenz, 1983), body temperature (Gibbs et al., 1973), gastrointestinal motility (Deutsch, Thial, & Greenberg, 1978), and neuroendocrine responses (Verbalis, McCann, McHale, & Stricker, 1986). Unfortunately, the sensitivity and reliability of each of these criteria remains questionable. For this reason, convincing evidence for a satiety mechanism must include converging evidence from several tests of specificity. Here we describe a simple and novel test of behavioral specificity to supplement existing techniques. The test is based on comparison of drug effects on scheduled test meals and on scheduled periods of access to running wheels. Cholecystokinin octapeptide (CCK) and glucagon (PG), which are hypothesized satiety signals (Geary, 1990; Smith & Gibbs, 1979, 1991), were found to inhibit feeding, 0163-1047/92 $5.00 Copyright © 1992 by AcademicPress, Inc. All rights of reproduction in any form reserved
2
GEARY, FUDGE, AND LE SAUTER
but not running. In contrast, LiC1 and amphetamine, drugs with clear nonspecific effects on feeding, inhibited both feeding and running. METHODS
Subjects Sixty male Sprague-Dawley rats (Charles River, Wilmington, MA), 425-750 g, were individually housed in metal cages (37 x 35 x 20 cm) with wood chip bedding. Pelleted rat chow (Purina 5012, St. Louis, MO) and water were presented ad lib, except as described. Rooms were maintained at 20 ± 2°C with a 12:12 light-dark cycle.
Experiment 1 Feeding and wheel running were tested in 2 groups of 12 rats each. One group was tested 2 h after light onset ("early light") and the other 6 h after light onset ("mid-light"). For each group, PG and CCK, then amphetamine, and finally LiC1 was tested. For the peptide tests, groups were randomly subdivided into two squads of six rats each. One squad received feeding tests before wheel running tests, and the other wheel running tests before feeding tests. For feeding tests, rats were intraperitoneally injected with 1 ml/kg 0.9% NaC1, returned to their home cages, and presented evaporated milk (Pet Foods, St. Louis, MO) in graduated tubes for 60 rain. Milk intakes were measured every 10 rain. Pellets were removed during the test. For wheel running tests, rats were similarly injected and placed in running-wheel cages for 60 rain with water, but no food. These cages consisted of activity wheels (radius 18 cm, width 10 cm) with rotation counters attached to small rest cages (25 x 15 x 12.5 cm). Running (wheel rotations) was measured every 10 min. To ensure stable baselines, these procedures were repeated Monday-Friday until the standard deviation of the last 3 days' mean data was less than 30% of the mean for each rat. This procedure took about 2 weeks. Then, on 6 consecutive days, rats were injected (1 ml/kg) in random order with 400 t~g/ml PG (Eli Lilly, Indianapolis, IN), 0.15 t~g/ml CCK (Kinevac sincalide, Squibb, New Brunswick, NJ), 0.15 tLg/ml CCK plus 400 t~g/ml PG, each tested once, or 0.9% NaC1, tested three times. After this, the squad that had received feeding tests was adapted to the activity wheels, the squad that had been running was adapted to the feeding procedure, and the entire procedure was repeated. Amphetamine was then similarly tested in each
group. Separate crossover tests of 1 ml/kg 0.9% NaC1 vs 0.1 or 0.5 mg/ml D-amphetamine (Sigma, St. Louis, MO) were done for both the food intake and the wheel running procedure. Again, one squad was tested on food intake first and the other squad on wheel running first. Finally, LiC1 (5 ml/kg 0.15 M LiC1, Sigma) and 5 ml/kg 0.9% NaC1 was similarly tested.
Experiment 2 To provide a standard of comparison for the wheel running results, amphetamine's effects on feeding, drinking, and open field activity were also tested. A new group of 12 rats was adapted to a daily 18h water (but not food) deprivation schedule. Then, separate crossover tests were done of the effects of 1 ml/kg 0.9% NaC1 vs. 0.1 or 0.5 mg/ml amphetamine on 60 min postdeprivation water intake in the absence of food. Amphetamine's anorexic effect was retested in another group of 12 rats using the procedure described above. Amphetamine's effect on open field activity in a novel environment was tested in 12 other rats. These were injected with 1 ml/kg 0.9% NaC1 and placed in plastic cages (20 x 24 x 20 cm) with wire mesh floors for 60 rain for 6 days. Then, separate crossover tests of 0.9% NaC1 vs. 0.1 or 0.5 mg/kg amphetamine were done. An experimenter blind to the injection condition observed and recorded each rat's behavior once each minute for 40 rain, using a time sampling technique (Hinton, Rosofsky, Granger, & Geary, 1986). Behaviors were grouped into grooming, activity, and resting behaviors. The activity category included locomotion, rearing, and sniffing. All these were early lght tests. Amphetamine's effects on the behavioral sequence of postprandial satiety (Antin et al., 1975; Rosofsky & Geary, 1989) were also measured. The rats and procedure of the novel environment activity test were used, except evaporated milk was offered after injections. This test was done at mid-light.
Data Analysis In Experiment 1, rats were active and fed most during the first 10 min and 30-min data corresponded to the size of the milk meals. Therefore, we report both 0- to 10- and 0- to 30-rain cumulative data. Separate repeated measures analyses of variance (ANOVA) were done for 10- and 30-rain data for both measures (feeding and running) at both circadian times (early light, mid-light). Each of the eight ANOVAs included peptide, amphetamine, and LiC1 data. Because preliminary analyses failed to detect differences among the three NaC1 control in-
SATIETY AND SCHEDULEDRUNNING jections during any of the peptide tests or between the two NaC1 injections during any of the amphetamine tests, these data were collapsed into single peptide control and amphetamine control values. Significant overall effects were followed by post hoc t tests of differences between individual means. The intensity of running after control injections varied through these series of experiments. The mean control running values for the six running subtests (peptides, amphetamine, and LiC1 at each circadian time) ranged from about 35 to 120 rotations/30 min. These values were highly correlated to group body weights (r = -0.95, p < .01). The drug effects reported below, however, were not noticeably affected by these baseline variations. Running, but not feeding, baselines were further reduced by a carryover effect of LiC1. Rats tested with LiC1 on the first day of the LiC1 crossover test ran much less during the next day's control test. This was clearest in the 30-min early light data, where control running values for rats receiving control tests after LiC1 tests were only 45% of the average of their peptide and amphetamine baselines. In contrast, rats receiving control tests before LiC1 ran 90% as much as previously. This carryover effect resulted in significantly depressed LiC1 baselines (Table 1). As described below, LiC1 significantly inhibited running even in comparison to these reduced baselines. In Experiment 2 ANOVAs or t tests were used ibr the feeding and running data, as appropriate. Frequencies of observed behaviors were tested nonparametrically with Friedman's tests and post hoc Wilcoxon's tests. All analyses were done with untransformed data. Changes are also reported as percentages to facilitate across experiment comparisons. RESULTS
Experiment 1 The pattern of results was very similar in early light and mid-light tests. CCK and PG inhibited feeding but not running. In contrast to the peptides' selective effects, amphetamine and LiC1 inhibited both feeding and running (Table 1). This pattern was clear both 10 min into the tests, when the rats were feeding or running enthusiastically, and after 30 min, when meals were over and running was frequently interrupted by pauses. The difference in specificity between the peptides and amphetamine of LiC1 occurred despite the fact that feeding was inhibited to a similar extent by
3
all the treatments. At early light, PG plus CCK inhibited feeding 58 - 5%; 0.5 mg/kg amphetamine, 61 _+ 6%; and LiC1, 60 -+ 9%. At mid-light, CCK inhibited feeding 36 -+ 8%; CCK plus PG, 36 _+ 9%; 0.5 mg/kg amphetamine, 41 _+ 8%; and LiC1 52 _+ 11% (which is not significantly different from the other mid-light values).
Experiment 2 Thirsty rats drank significantly less water after 0.5 mg/kg amphetamine (Table 2). The decrease of 16 _+ 6%, however, was less than the 72 _+ 6% decrease in milk intake caused by this amphetamine dose, t(22) = 5.78, p < .01. In the novel environment, 0.5 mg/kg amphetamine doubled the frequency of activity and almost eliminated resting (Table 3). Finally, in feeding rats, this dose of amphetamine disrupted the behavioral sequence of postprandial satiety by decreasing the frequencies of grooming and resting and increasing the frequency of general activity (Table 4). DISCUSSION We report two new results: first, when running wheels are made available for short, scheduled periods daily, rats run avidly--as much as 75 m in the first 10 min--and, second, scheduled running provides a useful index of the specificity of pharmacological anorexia. We compared the effects of nonspecific anorectics and hypothesized satiety agents on running and feeding to establish the validity of scheduled running as a specificity screen. The nonspecific anorectic amphetamine and LiC1 inhibited running. In contrast, the hypothesized satiety signals PG and CCK (Geary, 1990; Smith & Gibbs, 1979, 1991) did not. This dissociation occurred in both early and mid-light phase tests and was independent of variations in baseline levels of running. The differentiation was not due to potency differences among the agents because running was unaffected by peptide treatments that inhibited feeding as much as 0.5 mg/kg amphetamine. (The LiC1 data are ambiguous on this point. In comparison to its own control days, LiC1 inhibited feeding to a similar extent as did the peptides and amphetamine. But the LiC1 baselines were significantly reduced by a carryover effect in rats receiving control tests the day after LiC1 tests.) The temporal patterns of the anorectics' effects on running and feeding were also similar. Rats ran or fed most in the initial 10 min of tests,
4
GEARY, FUDGE, AND LE SAUTER i
TABLE
1
Differential Effects of Peptides, Amphetamine, and LiC1 on Feeding and Scheduled Wheel Running Early light
Mid-light
10 rain
30 min
10 min
30 min
Peptides Feeding Control PG CCK PG ÷ CCK Activity Control PG CCK PG + CCK
10.2 8.3 8.4 4.8
+-+ -+ -+
1.1 1.4 1.2 1.0"*
65.4 58.3 68.4 56.8
-+ -+ -+ -+
4.0 4.3 5.4 5.8
13.8 10.5 9.8 5.9
-+ -+ -+ _+
119.5 117.6 118.2 115.6
-+ _+ -+ -+
1.0 1.6" 0.9** 0.9** 8.8 9.9 11.4 13.4
12.9 12.3 8.7 7.2
-+ -+ -+ _+
1.3 1.6 1.3" 1.4"*
15.6 13.2 9.8 9.7
-+ -+ -+ -+
34.5 40.7 37.5 33.8
-+ -+ -+ -+
8.0 7.0 8.6 5.6
69.9 68.7 65.9 64.0
-+ -+ -+ +-
1.7 1.7 1.5"* 1.4"* 14.6 13.1 16.2 11.6
Amphetamine Feeding Control 0.1 m g / k g 0.5 m g / k g Activity Control 0.1 m g / k g 0.5 m g / k g
9.6 -+ 1.0 8.8 -+ 1.0 5.3 -+ 1.1"*
15.2 -+ 1.3 11.0 -+ 1.5"* 6.0 _+ 1.1"*
11.8 -+ 1.8 11.5 -+ 1.5 8.3 -+ 1.6
14.2 -+ 1.6 12.3 -+ 1.5 8.7 -+ 1.5"*
59.9 -+ 11.1 55.8 -+ 6.2 43.3 +- 6.8**
112.1 -+ 15.1 93.8 _+ 11.3 75.0 -+ 9.6**
31.4 -+ 7.4 22.5 -+ 5.5 23.5 -+ 7.2*
56.2 -+ 16.4 44.6 -+ 11.8 37.8 -+ 11.8"
LiC1 Feeding Control LiC1 Activity Control LiC1
8.1 -+ 1.2 5.6 -+ 1.3"*
14.1 -+ 1.2 5.8 -+ 1.3"*
13.1 -+ 2.0 7.7 -+ 2.0**
14.5 -+ 2.0 7.9 -+ 2.1"*
49.6 -+ 6.5 21.8 -+ 3.5**
75.4 -+ l l . 2 t 29.9 -+ 5.2**
24.6 -+ 7.6 11.8 -+ 3.4**
35.4 -+ 10.9t 15.8 -+ 5.1"*
Note. Feeding in ml and activity in r u n n i n g wheel rotations, m -+ SE. *p < .05 vs same subtest control, **p < .01; t p < .01 vs. other subtest control values.
and amphetamine and LiC1 each inhibited feeding and activity similarly in this interval. Thus, because putative satiety agents failed to affect scheduled running, whereas feeding and running were similarly affected by nonspecific anorectics, running wheel activity appear a valid test of behavioral specificity. Scheduled wheel running may present other advantages as a specificity screen. It requires minimal instrumentation and lends itself to either manual TABLE
2
Amphetamine Inhibits Both Feeding and Drinking
Saline Amphetamine (0.1 mg/kg) Amphetamine (0.5 mg/kg)
Food intake
Water intake
12.4 -+ 1.0 10.2 -+ 1.2 3.8 -+ 1.2"*
16.8 -+ 0.8 16.4 -+ 1.4 13.8 -+ 0.9*
Note. Data are 30 min intakes, ml, means -+ SE. *p < .05 vs. saline, **p < .01.
or automatic data acquisition. Rats run avidly even when begun on scheduled wheel access late in life(i.e., body w > 500 g), and after a few adaptation sessions both within and between subject variablilTABLE 3 Amphetamine Stimulates Activity in a Novel Environment Grooming
Activity
Resting
Saline 2.7 _+ 0.4 20.5 -+2.9 35.2 -+ 2.8 Amphetamine (0.1 mg/kg) 2.8 -+ 0.5 23.4 -+ 3.1 32.9 -+ 3.1 Amphetamine (0.5 mg/kg) 1.9 +- 0.6 52.0 -+ 2.2* 5.0 -+ 2.5*
Note. Data are frequencies of observed behaviors, one rating each min during a 40-min test, median -+ semi-interquartile range. Grooming is licking, biting, scratching, or rubbing with paws of any part of the body; activity is locomoting, rearing, or licking, biting, scratching, or sniffing any part of the cage; resting is reclining with body on cage floor without observation of any other behavior, eyes open or closed. *p < .01 vs. saline, Wilcoxon test after significant Friedman's test.
SATIETY AND SCHEDULED RUNNING
TABLE 4 Amphetamine Disrupts the Behavioral Sequence of Postprandial Satiety Feeding
Grooming
Activity
Resting
Saline 9.1 ± 0.8 3.2 ± 0.7 18.6 ± 1.8 21.2 ± 2.7 0.1 m g / k g AM 7.8 ± 0.9 4.5 + 1.2 25.1 ± 2.5 13.3 ± 2.3 0.5 m g / k g AM 2.9 ± 0.6* 2.1 ± 0.7* 37.8 ± 4.3* 6.3 ± 3.3*
Note. D a t a a r e f r e q u e n c i e s of o b s e r v e d b e h a v i o r s , one r a t i n g e a c h m i n d u r i n g a 6 0 - m i n t e s t , m e d i a n +- s e m i - i n t e r q u a r t i l e r a n g e . A M = a m p h e t a m i n e . B e h a v i o r s a r e defined in T a b l e 3. *P < .01 vs. s a l i n e , Wilcoxon t e s t a f t e r s i g n i f i c a n t F r i e d m a n ' s tests.
ity is modest (especially in comparison to the large between-subject variability of rats given ad libitum access to running wheels (Iverson & Iverson, 1975)). Scheduled wheel running appears to be similar in sensitivity to the other specificity screens we tested. Amphetamine's thresholds for inhibition of feeding palatable food, inhibition of scheduled running, inhibition of drinking after water deprivation, disruption of the behavioral sequence of postprandial satiety, and stimulation of general activity in a novel cage were all between 0.1 and 0.5 mg/kg. These are small doses in comparison to the typical ED of about 1-2 m g / k g in food deprived rats (Blundell, 1985; Carruba & Mantegazza, 1981; Cox & Maickel, 1972). Whether small doses of amphetamine have a specific inhibitory effect on appetite (Blundell, 1985; Carruba & Mantegazza, 1981) remains controversial (Rosofsky & Geary, 1989). Intraperitoneally injected CCK and PG reduced meal size 25-60% in these tests without producing signs of behavioral nonspecificity. This result is consistent with the hypothesis that these peptides inhibit feeding by facilitating postprandial satiety (Geary, 1990; Smith & Gibbs, 1991). We did not, however, compare the peptides' effects on feeding and running over the full range of their behavioral actions. Larger doses of these peptides certainly can elicit nonspecific feeding effects, but this does not obviate the significance of the specific actions of more moderate doses, such as we report. An additional test of behavioral specificity is useful because there is no single unambiguously valid measure of it. Conditioned taste aversion (CTA) tests have been advanced as crucial test of specificity (Deutsch, 1983). But CTAs appear to produce both false positives and false negatives (Goudie, 1979, 1985). For example, drugs which maintain self-administration reward (Berger, 1972) or which stimulate feeding, such as 2-deoxy-D-glucose (Thompson & Zzagon, 1981) have produced CTAs. Thus, CTAs
5
are not always sufficient to inhibit either appetitive or consummatory responding. Several drugs have also simultaneously elicited both conditioned aversions and preferences (Steward & Grupp, 1989). Comparison of liquid food intake and drinking has a great deal of face validity as a specificity screen because the two behaviors require similar ingestive responses. But this test also has complications. Water deprived rats usually eat less, so animals may be hungry as well as thirsty during tests. Further, both drinking and CTAs are usually tested under different conditions than feeding is, which may affect the actions of the drugs under examination. The use of running wheel activity as a test of behavioral specificity assumes that running and feeding are mediated by different neural mechanisms. This is indicated by several observations in addition to the dissociations between inhibitions of running and feeding reported here. For example, despite the fact that postprandial satiety normally includes behavioral quiescence, EEG synchronization, and resting (Mansbach & Lorenz, 1983; Smith & Gibbs, 1979), and that large doses of CCK (5100 t~g/kg) have been shown to reduce exploratory behavior in a nonfeeding situation (Crawley, 1983; Soar, Hewson, Leighton, Hill, & Hughes, 1989), meal-contingent peptide treatments have often dramatically accelerated postprandial resting without affecting the amount of general activity prior to resting (Antin et al., 1975; Le Sauter & Geary, 1987). Similarly, in free-feeding rats, meal size was positively correlated with the amount of postprandial rest, but was not correlated with the amount of postprandial activity (Bernstein, 1975). Feeding and activity also display independent circadian rhythms (Boulos & Terman, 1980). Finally, recent evidence suggests that striatal dopaminergic controls of feeding and activity are localized separately (Kelley, Gauthier, & Lang, 1989). Interestingly, amphetamine both inhibited activity in running wheels and stimulated activity in a novel environment. This difference may be related to a functional difference in the behaviors. Psychomotor stimulants' tend to facilitate behaviors that are simple or that occur at lower rates, such as open field activity, and tend to interfere with more complex or higher rate behaviors, presumably such as scheduled running (Harvey, 1987). Another possibility is that the apparently different neural organizations of exploratory locomotion and of wheel running are differentially sensitive to amphetamine's synaptic effects (Iverson & Iverson, 1975). If scheduled running is more similar to feeding in
6
GEARY, FUDGE, AND LE SAUTER
functional or neural organization that is open field activity, then running probably also provides a better test of the behavioral specificity of anorexia. We failed to replicate our finding that simultaneous injections of PG and CCK elicit a functionally synergistic satiety effect (Hinton et al., 1986). This may have occurred because CCK alone inhibited feeding relatively potently here. PG and CCK synergized only when the effect of CCK alone was small in some (Hinton et al., 1986; Le Sauter & Geary, 1987) but not all (Geary, Kissileff, Pi-Sunyer, & Hinton, 1991; Kalogeris, Reidelberger, Mendel, & Solomon, 1991) previous tests. REFERENCES Antin, J., Gibbs, J., Holt, J., Young. R. C., & Smith, G. P. (1975). Cholecystokinin elicits the complete behavioral sequence of satiety in rats. Journal of Comparative and Physiological Psychology, 89, 784-790. Ashe, J. H., & Nackman, M. (1980). Neural mechanisms in taste aversion learning. In J. M. Sprague & A. N. Epstein (Eds.), Progress in psychobiology and physiological psychology (pp. 233-262). New York: Academic Press. Berger, B. D. (1972). Conditioning of food aversions by injections of psychoactive drugs. Journal of Comparative and Physiological Psychology, 81, 21-26. Bernstein, I. L. (1975). Relationship between activity, rest, and free feeding in rats. Journal of Comparative and Physiological Psychology, 89, 253-257. Berridge, K. C., & Valenstein, E. S. (1991) What psychological process mediates feeding evoked by electrical stimulation of the lateral hypothalamus? Behavioral Neuroscience, 105, 3 14. Billington, C. J., Levine, A. S., & Morely, J. E. (1983). Are peptides truly satiety agents? A method of testing for neurohumoral satiety effects. American Journal of Physiology, 245, R920-R926. Blundell, J. E. (1985). Psychopharmacology of centrally acting anorectic agents. In M. Sandler & T. Silverstone (Eds.), Psychopharmacology and Food (pp. 71-109). London: Oxford Univ. Press. Booth, D. A. (1972). Conditioned satiety in the rat. Journal of Comparative and Physiological Psychology, 81, 457-471. Boulos, Z., & Terman, M. (1980). Food availability and daily biological rhythms. Neuroscience and Biobehavioral Reviews, 4, 119-131. Carruba, M. O., & Mantegazza, P. (1981). Drugs with effects on food intake and energy expenditure. In L. A. Cioffi (Ed.),
The Body Weight Regulatory System: Normal and Disturbed Mechanisms (pp. 279-287). New York: Raven Press. Cox, R. H., & Maickel, R. P. (1972). Comparison of anorexigenic and behavioral potency of phenylethylamines. Journal of Pharmacology and Experimental Therapeutics, 81, 1-9. Crawley, J. N. (1983). Divergent effects of cholecystokinin, bombesin, and lithium on rat exploratory behaviors. Peptides, 4, 405-410. Danguir, J., Nicolaidis, S., & Gerard, H. (1979). Relations be-
tween feeding and sleep patterns in the rat. Journal of Comparative and Physiological Psychology, 93, 820-830. Deutsch, J. A. (1983). Dietary control and the stomach. Progressive Neurobiology, 20, 313-332. Deutsch, J. A., & Gonzalez, M. F. (1978). Food intake reduction: Satiation or aversion? Behavioral Biology, 24, 317-327. Deutsch, J. A., & Hardy, W. T. (1977). Cholecystokinin produces bait shyness in rats. Nature (London,) 266, 196. Deutsch, J. A., Melina, F., & Peurto, A. (1976). Conditioned taste aversion caused by palatable nontoxic nutrients. Behavioral Biology, 16, 161-174. Deutsch, J. A., Thial, T. R., & Greenberg, L. H. (1978). Duodenal motility after cholecystokinin injection or satiety. Behavioral Biology, 24, 393-399. Eikelboom, R., & Mills, R. (1988). A microanalysis of wheel running in male and female rats. Physiology and Behavior, 43, 625-630. Flood, J. F., Silver, A. J., & Morley, J. E. (1990). Do peptideinduced changes in feeding occur because of changes in motivation to eat? Peptides, 11, 265-270. Geary, N. (1990). Pancreatic glucagon signals postprandial satiety. Neuroscience and Biobehavioral Reviews, 14, 323-338. Geary, N. Kissileff, H. R., Pi-Sunyer, F. X., & Hinton, V. (1991). Individual, but not simultaneous, glucagon and cholecystokinin infusions inhibit feeding in rats. American Journal of Physiology, in press. Gibbs, J., Young, R. C., & Smith, G. P. (1973). Cholecystokinin decreases food intake in rats. Journal of Comparative and Physiological Psychology, 3, 488-495. Goudie, A. J. (1979). Aversion stimulus properties of drugs. Neuropharmacology, 18, 971-979. Goudie, A. J. (1985). Aversive stimulus properties of drugs: The conditioned taste aversion paradigm. In A. Greenshaw & C. Dourish (Eds.), Experimental approach in psychopharmacology. Clifton, NJ: Humana Press. Harvey, J. A. (1987). Behavioral pharmacology of central nervous system stimulants. Neuropharmacology, 26, 887-892. Hinton, V., Rosofsky, M., Granger, J., & Geary, N. (1986). Combined injection potentiates the satiety effects of pancreatic glucagon, cholecystokinin, and bombesin. Brain Research Bulletin, 17, 615-619. Hoebel, B. G. (1988). Neuroscience and motivation: pathways and peptides that define motivational systems. In R. C. Atkinson, R. J. Herrnstein, G. Lindzey & R. D. Luce (Eds.),
Stevens' handbook of experimental psychology: Perception and motivation, 2nd ed., Vol. 1 (pp. 547-625). New York: Wiley. Iverson, S. P., & Iverson, L. L. (1975). Behavioral pharmacology. New York: Oxford Univ. Press. Kalogeris, T. J., Reidelberger, R. D., Mendel, V. E., & Solomon, T. J. (1991). Interaction of CCK-8 and pancreatic glucagon in control of food intake in dogs. American Journal of Physiology, 260, R688-R693. Kelley, A., E., Gauthier, A. M., & Lang, C. G. (1989). Amphetamine microinjections into distinct striatal subregions cause dissociable effects on motor and ingestive behavior. Behavioral Brain Research, 5, 163-174. Kushner, L. R., & Mook, D. G. (1984). Behavioral correlates of oral and postingestive satiety in the rat. Physiology and Behavior, 33, 713-718.
SATIETY AND SCHEDULED RUNNING Le Sauter, J., & Geary, N. (1987). Pancreatic glucagon and cholecystokinin synergistically inhibit sham feeding in rats. American Journal of Physiology, 253, R719-R725. Levine, A. S., Kuskowski, Grace, M., & Billington, C. J. [1991). Food deprivation-induced vs. drug-induced feeding: A behavioral evaluation. American Journal of Physiology, 260, R546-R552. Mansbach, R. S., & Lorenz, D. N. (1983). Cholecystokinin (CCK8) elicits prandial sleep in rats. Physiology and Behavior, 30, 179-183. Rosofsky, M., & Geary, N. (1989). Phenylpropanolamine and amphetamine disrupt postprandial satiety in rats. Pharmacology, Biochemistry, and Behavior, 34, 797-803. Smith, G. P., & Gibbs, J. (1979). Postprandial satiety. In J. Sprague & A. E. Epstein (Eds.}, Progress in psychobiology and physiological psychology, Vol 8 (pp. 179-242). New York: Academic Press. Smith, G. P., & Gibbs, J. (1992). The development and proof of the cholecystokinin satiety hypothesis. In C. T. Dourish, S. J. Cooper, S. P. Iverson, & L. L. Iverson (Eds.), Multiple cholecystokinn receptors in the CNS. London: Oxford Univ. Press, in press. Soar, J., Hewson, G., Leighton, G. E., Hill, R. G., & Hughes, J. (1989). L364, 718 antagonizes the cholecystokinn-induced suppression of locomotor activity. Pharmacology Biochemistry and Behavior, 33, 637-640. Steward, R. B., & Grupp, L. A. (1989). Conditioned place aversion mediated by self-administered ethanol in the rat: A consideration of blood ethanol levels. Pharmacology, Biochemistry, and Behavior, 32, 431-437. Stricker, E. M. (1990). Homeostatic origins of ingestive behavior.
7
In E. M. Stricker lEd.), Handbook of Behavioral Neurobwlogy, Vol. 10, Neurobiology of Food and Fluid Intake (pp. 4560). New York: Plenum Press. Swerdlow, N. R., Van der Koy, D., Koop, G. F., & Wenger, J. R. (1983). Cholecystokinin produces conditioned place-aversion, not place-preference, in food-deprived rats: Evidence against involvement in satiety. Life Science, 32, 2087-2093. Thompson, C. I., & Zzagon, I. A. (1981). 2-Deoxy-D-glucose produces delayed hypophagia and conditioned taste aversion in rats. Physiology and Behavior, 27, 1001-1004. VanderWeele, D. A., Deems, D. A., & Gibbs, J. ~1984). Cholecystokinin, lithium, and diet self-selection in the rat: Lithium chloride decreases protein, while cholecystokinin lowers fat and carbohydrate ingestion. Nutrition and Behavior, 2, 127-135. VanderWeele, D. A., Geiselman, P. J., & Novin, D. (1979). Pancreatic glucagon, food deprivation and feeding in intact and vagotomized rabbits. Physiology and Behavior, 23, 155-158. VanderWeele, D. A., Granja, J. A., & Deems, D. A. (1982). Discomfort or satiety: The spontaneous meal pattern may serve as a predictor. In B. G. Hoebel & D. Novin (Eds.), The neural basis of feeding and reward (pp. 167-173). Brunswick, ME: Haer Institute. Verbalis, J. G., McCann, M. J., McHale, C. M., & Stricker, E. M. (1986). Oxytocin secretion in response to cholecystokinin and food: Differentiation of nausea from satiety. Science, 232, 1417-1419. Verbalis, J. G., Richardson, D. W., & Stricker, E. M. (1987). Vasopressin release in response to nausea-producing agents and cholecystokinin in monkeys. American Journal of Physwlogy, 252, R749-R753.
Scheduled Running Wheel Activity Indexes the Specificity of Pharmocological Anorexia NORI GEARY, JENNIFER FUDGE, AND JOSEPH LE SAUTER1 Department of Psychology, Columbia University, New York, New York 10027
Nondeprived male Sprague-Dawley rats that were given scheduled access to running wheels for 60 min daily ran immediately and energetically. Intraperitoneal inject,ions of 400 t~g/kg pancreatic glucagon and 0.15 ~g/kg cholecystokinin octapeptide had no effect on scheduled running, but significantly inhibited feeding when the rats were offered condensed milk instead of access to the running wheels. This is consistent with the hypothesized fhnction of these peptides as postprandial satiety signals. In contrast, 0.5 mg/kg amphetamine and 75 t~M/kg LiC1, which produced similar degrees of anorexia, inhibited running by about 50%. Amphetamine, but neither peptide, also inhibited water drinking and disrupted the behavioral sequence of postprandial satiety. The distance run during scheduled running tests was inversely related to body weight, but the patterns of the drugs' effects were not altered by baseline running differences. Scheduled wheel running is a robust consummatory behavior that appears to provide a relatively valid, simple, and sensitive test of the behavioral specificity of pharmacological anorexia. ~ 1992 Academic Press, Inc. The physiological controls of consummatory behavior are presumed to be highly selective. Providing convincing demonstrations of behavioral specificity, however, has been a persistent problem in the behavioral neurobiology of motivation. The problem is especially acute in studies of the satiation of hunger. First, many nonspecific effects that can interfere with ingestion, such as sedation or malaise, may superficially mimic postprandial satiety. Second, important components of the neural mechanisms mediating ingestive behavior may also participate in the control of very different responses 1 Please address reprint requests to N. Geary, Bourne Laboratory, New York Hospital-Cornell Medical College, 21 Bloomingdale Road, White Plains, NY 10605. We thank Dr. Anthony Sclafani for the use of the running wheels and Ms. Debra Rosenzweig for help with the experiments. This research was supported by NIH Grant DK-32448 to N. Geary.
(for discussion see Berridge & Valenstein, 1991; Hoebel, 1988; Stricker, 1990). For these reasons, a great deal of effort has been devoted to the development of behavioral criteria to evaluate the specificity of hypothesized postprandial satiety mechanisms in animals. Specificity screens have been proposed based on the microstructure of periprandial behaviors (Blundell, 1985; Crawley, 1983), meal patterning (VanderWeele, Granja, & Deems, 1982), dietary self-selection (VanderWeele, Deems, & Gibbs, 1984), presence of species-typical postprandial behaviors (Antin, Gibbs, Holt, Young, & Smith, 1975; Levine, Kuskowski, Grace, & Billington, 1991; Smith & Gibbs, 1979), operant performance for food reward (Flood, Silver, & Morley, 1990), water intake (Gibbs, Young, & Smith, 1973), reaction to food deprivation (Billington, Levine, & Morley, 1983), and acquisition of conditioned aversions (CTA; Deutsch & Hardy, 1977; Gibbs et al., 1973; Swerdlow, Van der Koy, Koop, & Wenger, 1983). To these behavioral criteria may be added criteria of physiological specificity, such as EEG (Danguir, Nicolaidis, & Gerard, 1979; Mansbach & Lorenz, 1983), body temperature (Gibbs et al., 1973), gastrointestinal motility (Deutsch, Thial, & Greenberg, 1978), and neuroendocrine responses (Verbalis, McCann, McHale, & Stricker, 1986). Unfortunately, the sensitivity and reliability of each of these criteria remains questionable. For this reason, convincing evidence for a satiety mechanism must include converging evidence from several tests of specificity. Here we describe a simple and novel test of behavioral specificity to supplement existing techniques. The test is based on comparison of drug effects on scheduled test meals and on scheduled periods of access to running wheels. Cholecystokinin octapeptide (CCK) and glucagon (PG), which are hypothesized satiety signals (Geary, 1990; Smith & Gibbs, 1979, 1991), were found to inhibit feeding, 0163-1047/92 $5.00 Copyright © 1992 by AcademicPress, Inc. All rights of reproduction in any form reserved
2
GEARY, FUDGE, AND LE SAUTER
but not running. In contrast, LiC1 and amphetamine, drugs with clear nonspecific effects on feeding, inhibited both feeding and running. METHODS
Subjects Sixty male Sprague-Dawley rats (Charles River, Wilmington, MA), 425-750 g, were individually housed in metal cages (37 x 35 x 20 cm) with wood chip bedding. Pelleted rat chow (Purina 5012, St. Louis, MO) and water were presented ad lib, except as described. Rooms were maintained at 20 ± 2°C with a 12:12 light-dark cycle.
Experiment 1 Feeding and wheel running were tested in 2 groups of 12 rats each. One group was tested 2 h after light onset ("early light") and the other 6 h after light onset ("mid-light"). For each group, PG and CCK, then amphetamine, and finally LiC1 was tested. For the peptide tests, groups were randomly subdivided into two squads of six rats each. One squad received feeding tests before wheel running tests, and the other wheel running tests before feeding tests. For feeding tests, rats were intraperitoneally injected with 1 ml/kg 0.9% NaC1, returned to their home cages, and presented evaporated milk (Pet Foods, St. Louis, MO) in graduated tubes for 60 rain. Milk intakes were measured every 10 rain. Pellets were removed during the test. For wheel running tests, rats were similarly injected and placed in running-wheel cages for 60 rain with water, but no food. These cages consisted of activity wheels (radius 18 cm, width 10 cm) with rotation counters attached to small rest cages (25 x 15 x 12.5 cm). Running (wheel rotations) was measured every 10 min. To ensure stable baselines, these procedures were repeated Monday-Friday until the standard deviation of the last 3 days' mean data was less than 30% of the mean for each rat. This procedure took about 2 weeks. Then, on 6 consecutive days, rats were injected (1 ml/kg) in random order with 400 t~g/ml PG (Eli Lilly, Indianapolis, IN), 0.15 t~g/ml CCK (Kinevac sincalide, Squibb, New Brunswick, NJ), 0.15 tLg/ml CCK plus 400 t~g/ml PG, each tested once, or 0.9% NaC1, tested three times. After this, the squad that had received feeding tests was adapted to the activity wheels, the squad that had been running was adapted to the feeding procedure, and the entire procedure was repeated. Amphetamine was then similarly tested in each
group. Separate crossover tests of 1 ml/kg 0.9% NaC1 vs 0.1 or 0.5 mg/ml D-amphetamine (Sigma, St. Louis, MO) were done for both the food intake and the wheel running procedure. Again, one squad was tested on food intake first and the other squad on wheel running first. Finally, LiC1 (5 ml/kg 0.15 M LiC1, Sigma) and 5 ml/kg 0.9% NaC1 was similarly tested.
Experiment 2 To provide a standard of comparison for the wheel running results, amphetamine's effects on feeding, drinking, and open field activity were also tested. A new group of 12 rats was adapted to a daily 18h water (but not food) deprivation schedule. Then, separate crossover tests were done of the effects of 1 ml/kg 0.9% NaC1 vs. 0.1 or 0.5 mg/ml amphetamine on 60 min postdeprivation water intake in the absence of food. Amphetamine's anorexic effect was retested in another group of 12 rats using the procedure described above. Amphetamine's effect on open field activity in a novel environment was tested in 12 other rats. These were injected with 1 ml/kg 0.9% NaC1 and placed in plastic cages (20 x 24 x 20 cm) with wire mesh floors for 60 rain for 6 days. Then, separate crossover tests of 0.9% NaC1 vs. 0.1 or 0.5 mg/kg amphetamine were done. An experimenter blind to the injection condition observed and recorded each rat's behavior once each minute for 40 rain, using a time sampling technique (Hinton, Rosofsky, Granger, & Geary, 1986). Behaviors were grouped into grooming, activity, and resting behaviors. The activity category included locomotion, rearing, and sniffing. All these were early lght tests. Amphetamine's effects on the behavioral sequence of postprandial satiety (Antin et al., 1975; Rosofsky & Geary, 1989) were also measured. The rats and procedure of the novel environment activity test were used, except evaporated milk was offered after injections. This test was done at mid-light.
Data Analysis In Experiment 1, rats were active and fed most during the first 10 min and 30-min data corresponded to the size of the milk meals. Therefore, we report both 0- to 10- and 0- to 30-rain cumulative data. Separate repeated measures analyses of variance (ANOVA) were done for 10- and 30-rain data for both measures (feeding and running) at both circadian times (early light, mid-light). Each of the eight ANOVAs included peptide, amphetamine, and LiC1 data. Because preliminary analyses failed to detect differences among the three NaC1 control in-
SATIETY AND SCHEDULEDRUNNING jections during any of the peptide tests or between the two NaC1 injections during any of the amphetamine tests, these data were collapsed into single peptide control and amphetamine control values. Significant overall effects were followed by post hoc t tests of differences between individual means. The intensity of running after control injections varied through these series of experiments. The mean control running values for the six running subtests (peptides, amphetamine, and LiC1 at each circadian time) ranged from about 35 to 120 rotations/30 min. These values were highly correlated to group body weights (r = -0.95, p < .01). The drug effects reported below, however, were not noticeably affected by these baseline variations. Running, but not feeding, baselines were further reduced by a carryover effect of LiC1. Rats tested with LiC1 on the first day of the LiC1 crossover test ran much less during the next day's control test. This was clearest in the 30-min early light data, where control running values for rats receiving control tests after LiC1 tests were only 45% of the average of their peptide and amphetamine baselines. In contrast, rats receiving control tests before LiC1 ran 90% as much as previously. This carryover effect resulted in significantly depressed LiC1 baselines (Table 1). As described below, LiC1 significantly inhibited running even in comparison to these reduced baselines. In Experiment 2 ANOVAs or t tests were used ibr the feeding and running data, as appropriate. Frequencies of observed behaviors were tested nonparametrically with Friedman's tests and post hoc Wilcoxon's tests. All analyses were done with untransformed data. Changes are also reported as percentages to facilitate across experiment comparisons. RESULTS
Experiment 1 The pattern of results was very similar in early light and mid-light tests. CCK and PG inhibited feeding but not running. In contrast to the peptides' selective effects, amphetamine and LiC1 inhibited both feeding and running (Table 1). This pattern was clear both 10 min into the tests, when the rats were feeding or running enthusiastically, and after 30 min, when meals were over and running was frequently interrupted by pauses. The difference in specificity between the peptides and amphetamine of LiC1 occurred despite the fact that feeding was inhibited to a similar extent by
3
all the treatments. At early light, PG plus CCK inhibited feeding 58 - 5%; 0.5 mg/kg amphetamine, 61 _+ 6%; and LiC1, 60 -+ 9%. At mid-light, CCK inhibited feeding 36 -+ 8%; CCK plus PG, 36 _+ 9%; 0.5 mg/kg amphetamine, 41 _+ 8%; and LiC1 52 _+ 11% (which is not significantly different from the other mid-light values).
Experiment 2 Thirsty rats drank significantly less water after 0.5 mg/kg amphetamine (Table 2). The decrease of 16 _+ 6%, however, was less than the 72 _+ 6% decrease in milk intake caused by this amphetamine dose, t(22) = 5.78, p < .01. In the novel environment, 0.5 mg/kg amphetamine doubled the frequency of activity and almost eliminated resting (Table 3). Finally, in feeding rats, this dose of amphetamine disrupted the behavioral sequence of postprandial satiety by decreasing the frequencies of grooming and resting and increasing the frequency of general activity (Table 4). DISCUSSION We report two new results: first, when running wheels are made available for short, scheduled periods daily, rats run avidly--as much as 75 m in the first 10 min--and, second, scheduled running provides a useful index of the specificity of pharmacological anorexia. We compared the effects of nonspecific anorectics and hypothesized satiety agents on running and feeding to establish the validity of scheduled running as a specificity screen. The nonspecific anorectic amphetamine and LiC1 inhibited running. In contrast, the hypothesized satiety signals PG and CCK (Geary, 1990; Smith & Gibbs, 1979, 1991) did not. This dissociation occurred in both early and mid-light phase tests and was independent of variations in baseline levels of running. The differentiation was not due to potency differences among the agents because running was unaffected by peptide treatments that inhibited feeding as much as 0.5 mg/kg amphetamine. (The LiC1 data are ambiguous on this point. In comparison to its own control days, LiC1 inhibited feeding to a similar extent as did the peptides and amphetamine. But the LiC1 baselines were significantly reduced by a carryover effect in rats receiving control tests the day after LiC1 tests.) The temporal patterns of the anorectics' effects on running and feeding were also similar. Rats ran or fed most in the initial 10 min of tests,
4
GEARY, FUDGE, AND LE SAUTER i
TABLE
1
Differential Effects of Peptides, Amphetamine, and LiC1 on Feeding and Scheduled Wheel Running Early light
Mid-light
10 rain
30 min
10 min
30 min
Peptides Feeding Control PG CCK PG ÷ CCK Activity Control PG CCK PG + CCK
10.2 8.3 8.4 4.8
+-+ -+ -+
1.1 1.4 1.2 1.0"*
65.4 58.3 68.4 56.8
-+ -+ -+ -+
4.0 4.3 5.4 5.8
13.8 10.5 9.8 5.9
-+ -+ -+ _+
119.5 117.6 118.2 115.6
-+ _+ -+ -+
1.0 1.6" 0.9** 0.9** 8.8 9.9 11.4 13.4
12.9 12.3 8.7 7.2
-+ -+ -+ _+
1.3 1.6 1.3" 1.4"*
15.6 13.2 9.8 9.7
-+ -+ -+ -+
34.5 40.7 37.5 33.8
-+ -+ -+ -+
8.0 7.0 8.6 5.6
69.9 68.7 65.9 64.0
-+ -+ -+ +-
1.7 1.7 1.5"* 1.4"* 14.6 13.1 16.2 11.6
Amphetamine Feeding Control 0.1 m g / k g 0.5 m g / k g Activity Control 0.1 m g / k g 0.5 m g / k g
9.6 -+ 1.0 8.8 -+ 1.0 5.3 -+ 1.1"*
15.2 -+ 1.3 11.0 -+ 1.5"* 6.0 _+ 1.1"*
11.8 -+ 1.8 11.5 -+ 1.5 8.3 -+ 1.6
14.2 -+ 1.6 12.3 -+ 1.5 8.7 -+ 1.5"*
59.9 -+ 11.1 55.8 -+ 6.2 43.3 +- 6.8**
112.1 -+ 15.1 93.8 _+ 11.3 75.0 -+ 9.6**
31.4 -+ 7.4 22.5 -+ 5.5 23.5 -+ 7.2*
56.2 -+ 16.4 44.6 -+ 11.8 37.8 -+ 11.8"
LiC1 Feeding Control LiC1 Activity Control LiC1
8.1 -+ 1.2 5.6 -+ 1.3"*
14.1 -+ 1.2 5.8 -+ 1.3"*
13.1 -+ 2.0 7.7 -+ 2.0**
14.5 -+ 2.0 7.9 -+ 2.1"*
49.6 -+ 6.5 21.8 -+ 3.5**
75.4 -+ l l . 2 t 29.9 -+ 5.2**
24.6 -+ 7.6 11.8 -+ 3.4**
35.4 -+ 10.9t 15.8 -+ 5.1"*
Note. Feeding in ml and activity in r u n n i n g wheel rotations, m -+ SE. *p < .05 vs same subtest control, **p < .01; t p < .01 vs. other subtest control values.
and amphetamine and LiC1 each inhibited feeding and activity similarly in this interval. Thus, because putative satiety agents failed to affect scheduled running, whereas feeding and running were similarly affected by nonspecific anorectics, running wheel activity appear a valid test of behavioral specificity. Scheduled wheel running may present other advantages as a specificity screen. It requires minimal instrumentation and lends itself to either manual TABLE
2
Amphetamine Inhibits Both Feeding and Drinking
Saline Amphetamine (0.1 mg/kg) Amphetamine (0.5 mg/kg)
Food intake
Water intake
12.4 -+ 1.0 10.2 -+ 1.2 3.8 -+ 1.2"*
16.8 -+ 0.8 16.4 -+ 1.4 13.8 -+ 0.9*
Note. Data are 30 min intakes, ml, means -+ SE. *p < .05 vs. saline, **p < .01.
or automatic data acquisition. Rats run avidly even when begun on scheduled wheel access late in life(i.e., body w > 500 g), and after a few adaptation sessions both within and between subject variablilTABLE 3 Amphetamine Stimulates Activity in a Novel Environment Grooming
Activity
Resting
Saline 2.7 _+ 0.4 20.5 -+2.9 35.2 -+ 2.8 Amphetamine (0.1 mg/kg) 2.8 -+ 0.5 23.4 -+ 3.1 32.9 -+ 3.1 Amphetamine (0.5 mg/kg) 1.9 +- 0.6 52.0 -+ 2.2* 5.0 -+ 2.5*
Note. Data are frequencies of observed behaviors, one rating each min during a 40-min test, median -+ semi-interquartile range. Grooming is licking, biting, scratching, or rubbing with paws of any part of the body; activity is locomoting, rearing, or licking, biting, scratching, or sniffing any part of the cage; resting is reclining with body on cage floor without observation of any other behavior, eyes open or closed. *p < .01 vs. saline, Wilcoxon test after significant Friedman's test.
SATIETY AND SCHEDULED RUNNING
TABLE 4 Amphetamine Disrupts the Behavioral Sequence of Postprandial Satiety Feeding
Grooming
Activity
Resting
Saline 9.1 ± 0.8 3.2 ± 0.7 18.6 ± 1.8 21.2 ± 2.7 0.1 m g / k g AM 7.8 ± 0.9 4.5 + 1.2 25.1 ± 2.5 13.3 ± 2.3 0.5 m g / k g AM 2.9 ± 0.6* 2.1 ± 0.7* 37.8 ± 4.3* 6.3 ± 3.3*
Note. D a t a a r e f r e q u e n c i e s of o b s e r v e d b e h a v i o r s , one r a t i n g e a c h m i n d u r i n g a 6 0 - m i n t e s t , m e d i a n +- s e m i - i n t e r q u a r t i l e r a n g e . A M = a m p h e t a m i n e . B e h a v i o r s a r e defined in T a b l e 3. *P < .01 vs. s a l i n e , Wilcoxon t e s t a f t e r s i g n i f i c a n t F r i e d m a n ' s tests.
ity is modest (especially in comparison to the large between-subject variability of rats given ad libitum access to running wheels (Iverson & Iverson, 1975)). Scheduled wheel running appears to be similar in sensitivity to the other specificity screens we tested. Amphetamine's thresholds for inhibition of feeding palatable food, inhibition of scheduled running, inhibition of drinking after water deprivation, disruption of the behavioral sequence of postprandial satiety, and stimulation of general activity in a novel cage were all between 0.1 and 0.5 mg/kg. These are small doses in comparison to the typical ED of about 1-2 m g / k g in food deprived rats (Blundell, 1985; Carruba & Mantegazza, 1981; Cox & Maickel, 1972). Whether small doses of amphetamine have a specific inhibitory effect on appetite (Blundell, 1985; Carruba & Mantegazza, 1981) remains controversial (Rosofsky & Geary, 1989). Intraperitoneally injected CCK and PG reduced meal size 25-60% in these tests without producing signs of behavioral nonspecificity. This result is consistent with the hypothesis that these peptides inhibit feeding by facilitating postprandial satiety (Geary, 1990; Smith & Gibbs, 1991). We did not, however, compare the peptides' effects on feeding and running over the full range of their behavioral actions. Larger doses of these peptides certainly can elicit nonspecific feeding effects, but this does not obviate the significance of the specific actions of more moderate doses, such as we report. An additional test of behavioral specificity is useful because there is no single unambiguously valid measure of it. Conditioned taste aversion (CTA) tests have been advanced as crucial test of specificity (Deutsch, 1983). But CTAs appear to produce both false positives and false negatives (Goudie, 1979, 1985). For example, drugs which maintain self-administration reward (Berger, 1972) or which stimulate feeding, such as 2-deoxy-D-glucose (Thompson & Zzagon, 1981) have produced CTAs. Thus, CTAs
5
are not always sufficient to inhibit either appetitive or consummatory responding. Several drugs have also simultaneously elicited both conditioned aversions and preferences (Steward & Grupp, 1989). Comparison of liquid food intake and drinking has a great deal of face validity as a specificity screen because the two behaviors require similar ingestive responses. But this test also has complications. Water deprived rats usually eat less, so animals may be hungry as well as thirsty during tests. Further, both drinking and CTAs are usually tested under different conditions than feeding is, which may affect the actions of the drugs under examination. The use of running wheel activity as a test of behavioral specificity assumes that running and feeding are mediated by different neural mechanisms. This is indicated by several observations in addition to the dissociations between inhibitions of running and feeding reported here. For example, despite the fact that postprandial satiety normally includes behavioral quiescence, EEG synchronization, and resting (Mansbach & Lorenz, 1983; Smith & Gibbs, 1979), and that large doses of CCK (5100 t~g/kg) have been shown to reduce exploratory behavior in a nonfeeding situation (Crawley, 1983; Soar, Hewson, Leighton, Hill, & Hughes, 1989), meal-contingent peptide treatments have often dramatically accelerated postprandial resting without affecting the amount of general activity prior to resting (Antin et al., 1975; Le Sauter & Geary, 1987). Similarly, in free-feeding rats, meal size was positively correlated with the amount of postprandial rest, but was not correlated with the amount of postprandial activity (Bernstein, 1975). Feeding and activity also display independent circadian rhythms (Boulos & Terman, 1980). Finally, recent evidence suggests that striatal dopaminergic controls of feeding and activity are localized separately (Kelley, Gauthier, & Lang, 1989). Interestingly, amphetamine both inhibited activity in running wheels and stimulated activity in a novel environment. This difference may be related to a functional difference in the behaviors. Psychomotor stimulants' tend to facilitate behaviors that are simple or that occur at lower rates, such as open field activity, and tend to interfere with more complex or higher rate behaviors, presumably such as scheduled running (Harvey, 1987). Another possibility is that the apparently different neural organizations of exploratory locomotion and of wheel running are differentially sensitive to amphetamine's synaptic effects (Iverson & Iverson, 1975). If scheduled running is more similar to feeding in
6
GEARY, FUDGE, AND LE SAUTER
functional or neural organization that is open field activity, then running probably also provides a better test of the behavioral specificity of anorexia. We failed to replicate our finding that simultaneous injections of PG and CCK elicit a functionally synergistic satiety effect (Hinton et al., 1986). This may have occurred because CCK alone inhibited feeding relatively potently here. PG and CCK synergized only when the effect of CCK alone was small in some (Hinton et al., 1986; Le Sauter & Geary, 1987) but not all (Geary, Kissileff, Pi-Sunyer, & Hinton, 1991; Kalogeris, Reidelberger, Mendel, & Solomon, 1991) previous tests. REFERENCES Antin, J., Gibbs, J., Holt, J., Young. R. C., & Smith, G. P. (1975). Cholecystokinin elicits the complete behavioral sequence of satiety in rats. Journal of Comparative and Physiological Psychology, 89, 784-790. Ashe, J. H., & Nackman, M. (1980). Neural mechanisms in taste aversion learning. In J. M. Sprague & A. N. Epstein (Eds.), Progress in psychobiology and physiological psychology (pp. 233-262). New York: Academic Press. Berger, B. D. (1972). Conditioning of food aversions by injections of psychoactive drugs. Journal of Comparative and Physiological Psychology, 81, 21-26. Bernstein, I. L. (1975). Relationship between activity, rest, and free feeding in rats. Journal of Comparative and Physiological Psychology, 89, 253-257. Berridge, K. C., & Valenstein, E. S. (1991) What psychological process mediates feeding evoked by electrical stimulation of the lateral hypothalamus? Behavioral Neuroscience, 105, 3 14. Billington, C. J., Levine, A. S., & Morely, J. E. (1983). Are peptides truly satiety agents? A method of testing for neurohumoral satiety effects. American Journal of Physiology, 245, R920-R926. Blundell, J. E. (1985). Psychopharmacology of centrally acting anorectic agents. In M. Sandler & T. Silverstone (Eds.), Psychopharmacology and Food (pp. 71-109). London: Oxford Univ. Press. Booth, D. A. (1972). Conditioned satiety in the rat. Journal of Comparative and Physiological Psychology, 81, 457-471. Boulos, Z., & Terman, M. (1980). Food availability and daily biological rhythms. Neuroscience and Biobehavioral Reviews, 4, 119-131. Carruba, M. O., & Mantegazza, P. (1981). Drugs with effects on food intake and energy expenditure. In L. A. Cioffi (Ed.),
The Body Weight Regulatory System: Normal and Disturbed Mechanisms (pp. 279-287). New York: Raven Press. Cox, R. H., & Maickel, R. P. (1972). Comparison of anorexigenic and behavioral potency of phenylethylamines. Journal of Pharmacology and Experimental Therapeutics, 81, 1-9. Crawley, J. N. (1983). Divergent effects of cholecystokinin, bombesin, and lithium on rat exploratory behaviors. Peptides, 4, 405-410. Danguir, J., Nicolaidis, S., & Gerard, H. (1979). Relations be-
tween feeding and sleep patterns in the rat. Journal of Comparative and Physiological Psychology, 93, 820-830. Deutsch, J. A. (1983). Dietary control and the stomach. Progressive Neurobiology, 20, 313-332. Deutsch, J. A., & Gonzalez, M. F. (1978). Food intake reduction: Satiation or aversion? Behavioral Biology, 24, 317-327. Deutsch, J. A., & Hardy, W. T. (1977). Cholecystokinin produces bait shyness in rats. Nature (London,) 266, 196. Deutsch, J. A., Melina, F., & Peurto, A. (1976). Conditioned taste aversion caused by palatable nontoxic nutrients. Behavioral Biology, 16, 161-174. Deutsch, J. A., Thial, T. R., & Greenberg, L. H. (1978). Duodenal motility after cholecystokinin injection or satiety. Behavioral Biology, 24, 393-399. Eikelboom, R., & Mills, R. (1988). A microanalysis of wheel running in male and female rats. Physiology and Behavior, 43, 625-630. Flood, J. F., Silver, A. J., & Morley, J. E. (1990). Do peptideinduced changes in feeding occur because of changes in motivation to eat? Peptides, 11, 265-270. Geary, N. (1990). Pancreatic glucagon signals postprandial satiety. Neuroscience and Biobehavioral Reviews, 14, 323-338. Geary, N. Kissileff, H. R., Pi-Sunyer, F. X., & Hinton, V. (1991). Individual, but not simultaneous, glucagon and cholecystokinin infusions inhibit feeding in rats. American Journal of Physiology, in press. Gibbs, J., Young, R. C., & Smith, G. P. (1973). Cholecystokinin decreases food intake in rats. Journal of Comparative and Physiological Psychology, 3, 488-495. Goudie, A. J. (1979). Aversion stimulus properties of drugs. Neuropharmacology, 18, 971-979. Goudie, A. J. (1985). Aversive stimulus properties of drugs: The conditioned taste aversion paradigm. In A. Greenshaw & C. Dourish (Eds.), Experimental approach in psychopharmacology. Clifton, NJ: Humana Press. Harvey, J. A. (1987). Behavioral pharmacology of central nervous system stimulants. Neuropharmacology, 26, 887-892. Hinton, V., Rosofsky, M., Granger, J., & Geary, N. (1986). Combined injection potentiates the satiety effects of pancreatic glucagon, cholecystokinin, and bombesin. Brain Research Bulletin, 17, 615-619. Hoebel, B. G. (1988). Neuroscience and motivation: pathways and peptides that define motivational systems. In R. C. Atkinson, R. J. Herrnstein, G. Lindzey & R. D. Luce (Eds.),
Stevens' handbook of experimental psychology: Perception and motivation, 2nd ed., Vol. 1 (pp. 547-625). New York: Wiley. Iverson, S. P., & Iverson, L. L. (1975). Behavioral pharmacology. New York: Oxford Univ. Press. Kalogeris, T. J., Reidelberger, R. D., Mendel, V. E., & Solomon, T. J. (1991). Interaction of CCK-8 and pancreatic glucagon in control of food intake in dogs. American Journal of Physiology, 260, R688-R693. Kelley, A., E., Gauthier, A. M., & Lang, C. G. (1989). Amphetamine microinjections into distinct striatal subregions cause dissociable effects on motor and ingestive behavior. Behavioral Brain Research, 5, 163-174. Kushner, L. R., & Mook, D. G. (1984). Behavioral correlates of oral and postingestive satiety in the rat. Physiology and Behavior, 33, 713-718.
SATIETY AND SCHEDULED RUNNING Le Sauter, J., & Geary, N. (1987). Pancreatic glucagon and cholecystokinin synergistically inhibit sham feeding in rats. American Journal of Physiology, 253, R719-R725. Levine, A. S., Kuskowski, Grace, M., & Billington, C. J. [1991). Food deprivation-induced vs. drug-induced feeding: A behavioral evaluation. American Journal of Physiology, 260, R546-R552. Mansbach, R. S., & Lorenz, D. N. (1983). Cholecystokinin (CCK8) elicits prandial sleep in rats. Physiology and Behavior, 30, 179-183. Rosofsky, M., & Geary, N. (1989). Phenylpropanolamine and amphetamine disrupt postprandial satiety in rats. Pharmacology, Biochemistry, and Behavior, 34, 797-803. Smith, G. P., & Gibbs, J. (1979). Postprandial satiety. In J. Sprague & A. E. Epstein (Eds.}, Progress in psychobiology and physiological psychology, Vol 8 (pp. 179-242). New York: Academic Press. Smith, G. P., & Gibbs, J. (1992). The development and proof of the cholecystokinin satiety hypothesis. In C. T. Dourish, S. J. Cooper, S. P. Iverson, & L. L. Iverson (Eds.), Multiple cholecystokinn receptors in the CNS. London: Oxford Univ. Press, in press. Soar, J., Hewson, G., Leighton, G. E., Hill, R. G., & Hughes, J. (1989). L364, 718 antagonizes the cholecystokinn-induced suppression of locomotor activity. Pharmacology Biochemistry and Behavior, 33, 637-640. Steward, R. B., & Grupp, L. A. (1989). Conditioned place aversion mediated by self-administered ethanol in the rat: A consideration of blood ethanol levels. Pharmacology, Biochemistry, and Behavior, 32, 431-437. Stricker, E. M. (1990). Homeostatic origins of ingestive behavior.
7
In E. M. Stricker lEd.), Handbook of Behavioral Neurobwlogy, Vol. 10, Neurobiology of Food and Fluid Intake (pp. 4560). New York: Plenum Press. Swerdlow, N. R., Van der Koy, D., Koop, G. F., & Wenger, J. R. (1983). Cholecystokinin produces conditioned place-aversion, not place-preference, in food-deprived rats: Evidence against involvement in satiety. Life Science, 32, 2087-2093. Thompson, C. I., & Zzagon, I. A. (1981). 2-Deoxy-D-glucose produces delayed hypophagia and conditioned taste aversion in rats. Physiology and Behavior, 27, 1001-1004. VanderWeele, D. A., Deems, D. A., & Gibbs, J. ~1984). Cholecystokinin, lithium, and diet self-selection in the rat: Lithium chloride decreases protein, while cholecystokinin lowers fat and carbohydrate ingestion. Nutrition and Behavior, 2, 127-135. VanderWeele, D. A., Geiselman, P. J., & Novin, D. (1979). Pancreatic glucagon, food deprivation and feeding in intact and vagotomized rabbits. Physiology and Behavior, 23, 155-158. VanderWeele, D. A., Granja, J. A., & Deems, D. A. (1982). Discomfort or satiety: The spontaneous meal pattern may serve as a predictor. In B. G. Hoebel & D. Novin (Eds.), The neural basis of feeding and reward (pp. 167-173). Brunswick, ME: Haer Institute. Verbalis, J. G., McCann, M. J., McHale, C. M., & Stricker, E. M. (1986). Oxytocin secretion in response to cholecystokinin and food: Differentiation of nausea from satiety. Science, 232, 1417-1419. Verbalis, J. G., Richardson, D. W., & Stricker, E. M. (1987). Vasopressin release in response to nausea-producing agents and cholecystokinin in monkeys. American Journal of Physwlogy, 252, R749-R753.
