- Email: [email protected]
Effect of chronic administration of dexfenfluramine on stress- and palatability-induced food intake in rats
Physiology& Behavior,Vol. 46, pp. 145-149. ©Pergamon Press plc, 1989. Printed in the U.S.A.
0031-9384/89 $3.00 + .00
Effect of Chronic Administration of Dexfenfluramine on Stress- and Palatability-Induced Food Intake in Rats A N N E - M A R I E S O U Q U E T A N D N E I L E. R O W L A N D 1
Department o f Psychology and Centers for Neurobiological and Nutritional Sciences University of Florida, Gainesville, FL R e c e i v e d 25 M a y 1989
SOUQUET, A.-M. AND N. E. ROWLAND. Effect of chronic administration of dexfenfluramine on stress- and palatability-induced food intake in rats. PHYSIOL BEHAV 46(2) 145-149, 1989.--Dexfenfluramine, administered chronically either SC via osmotic minipump (3 mg/kg/day) or as dally IP injections (1.5 mg/kg) to rats, suppressed tail pinch-induced eating for up to 14 days. The minipump regimen also caused a reduction of intake of a palatable "dessert" for up to 9 days in heavy, adult rats, but was only transiently effective in young, rapidly growing rats. Nonetheless, the regimen produced decreased weight gain and insensitivity to a subsequent acute test dose of the agent. These results illustrate the sustained anorectic efficacy of dexfenfluramine in paradigms that do not involve restriction of diet, and these may be animal models of snacking or binging. The data also corroborate previous suggestions that the agent is more effective in older or more obese populations. Dexfenfluramine
Chronic drug administration
Anorexia
STUDIES from this and several other laboratories have indicated that the anorectic effect of d,l-fenfluramine decreases across repeated daily treatments in rats maintained on schedules of restricted food access (8,16). On such schedules, there is no opportunity for the animals to make up the food not eaten on one day, and so some weight loss (from an already reduced level due to the schedule) occurs. Levitsky and colleagues (10) have argued that this apparent tolerance to the anorexia is because of the weight loss and that when the rats start below some "set point" the anorectic agent is less effective. Subsequent theoretical (19) and empirical (5) accounts support the idea that fenfluramine lowers a set point for body weight. We have presented two empirical counterarguments. In the first, we showed that injections of fenfluramine following scheduled feeds also produced tolerance despite minimal weight loss (14), and in the second we found that weight loss by food restriction prior to fenfluramine treatment failed to greatly diminish the anorectic effect of the drug (2). Increased brain serotonin (5HT) transmission is the likely mechanism by which fenfluramine causes anorexia (6, 7, 15). However, large doses of fenfluramine deplete brain 5HT in rats (15), and Kleven et al. (8) have suggested that anorectic tolerance occurs in chronic regimens because subsequent injections are unable to release as much 5HT from a depleted nerve terminal. The present experiments are designed to overcome what we see
Palatability
Stress
Eating
as several major shortcomings that complicate the interpretation of the above studies. First, we use the active d(+) enantiomer of fenfluramine (DFEN), rather than the racemate used in all of the above work. The 1( - ) enantiomer, although not very potent as an anorectic, has neurochemical effects that might modify the longterm efficacy of its active partner (15). Second, we have used low doses of DFEN that produce little or no depletion of brain 5HT, and so release might remain uncompromised during a chronic regimen. As in previous studies (3, 12, 17), we use osmotic minipumps to deliver the agent continuously rather than by daily injections: because the tissue half-lives of DFEN and its active metabolite, d-norfenfluramine are a few hours in rats, tissue drug levels will fluctuate dramatically across the day with injection paradigms. Third, we use two acute models of overeating that do not employ food deprivation and/or weight loss. The first is tail pinch-induced eating and the second is a "dessert" test in which rats fed chow ad lib also receive brief daily access to a sweet fluid. Acute injection of DFEN is known to inhibit eating in these and other paradigms (13,17). EXPERIMENT 1: TAIL PINCH STUDIES WITH DFEN METHOD
Animals and Housing Male Sprague-Dawley (Zivic Miller) rats, 3-6 months old and
tRequests for reprints should be addressed to Dr. Neil Rowland, Department of Psychology, University of Florida, Gainesville, FL 32611.
145
146
SOUQUET AND ROWLAND
weighing 290--600 g, were housed individually in hanging stainless steel wire cages with tap water and Purina rodent chow pellets available ad lib. The vivarium was light (on: 0700-1900) and temperature (23°C) controlled. All tests were performed in this vivarium during the morning.
~, aZ
180
160
120, o Z
Tail Pinch Procedure
T
T~<
]
1
O ~ 0
140,
,~0[
-
ContrOIr 0
1
100,
T
80.
The rats were placed individually into rectangular plastic washing-up bowls (Rubbermaid) with chow pellets scattered on the floor. In an initial screening test, each rat was administered four 120 sec tail pinches, each 10-15 min apart, using hand-held, foam-padded sponge forceps in a procedure similar to that of Antelman et al. (1). The time spent eating or gnawing food during each pinch was recorded using a stopwatch. Because rats don't eat between trials in these daytime tests, duration is correlated with food consumed and/or gnawed, especially using a within-subjects design. Rats that did not exhibit sustained eating during these screening trials were not included in the main studies.
h
o Z
o
60.
40-
~ / 1
'
i
DFEN
20a
0,
' 1
0
: 2
I 3
.' 4
: 5
: 6
: 7
: 8
~ : : : ; I I 9 10 11 12 13 14 15
DAYS OF TREATMENT
Study A: Minipump DFEN
FIG. 1. Tail pinch-induced food intake, expressed as Mean --+ SE of baselines summed over four 120 sec trials on days indicated, in male rats receiving no drug (Control; open circles) or d-fenfluramine (DFEN; 3 mg/kg/day SC via minipump). Intakes in the DFEN group are smaller than Control (p<0.05) on all days.
Three daily baseline tests were conducted to ensure that the durations of tail pinch-induced eating were stable. Seven rats then received, under brief ether anesthesia, a subcutaneously-implanted osmotic minipump (Alzet No. 2002, 14 days) preloaded with DFEN (hydrochloride salt, from Servier Labs) dissolved in distilled water at a concentration to deliver 3 mg/kg/day. An additional 7 rats received no drug. Measures of tail pinch-induced eating were made 1, 4, 7, 10 and 14 days later.
end of the study, which gave rise to a drug x days interaction term for d,l-fenfluramine (p<0.01). However, it is clear from Fig. 2 that d,l-fenfluramine- and DFEN-treated rats performed similarly, and showed no marked change in the initial (day 1) effect of the drug with repeated treatment. As for the minipump study A, the general other behaviors of drug-treated rats during tail pinch were similar to controls.
Study B: Injections of DFEN After baselines were established as above, rats received daily injections of either DFEN (1.5 mg/kg IP, N = 6), d,l-fenfluramine (3 mg/kg IP, N = 6 ) , or 0.15 M saline as control ( N = 5 ) . Tail pinch tests (days 1, 2, 4, 7, 10 and 14) started 30 min after injections.
Data Analysis All durations of eating were converted to % of individual baselines, and were then compared between treatment groups and days by analysis of variance, and post hoc two-tailed t-tests. RESULTS
Study A The mean baseline duration of tail pinch-induced eating was 62 sec/120 sec trial. The control (no drug) rats improved above this baseline on each of the five test days (Fig. 1). DFEN suppressed tail pinch-induced eating, F(1,12)= 12.5, p<0.01. There was no main or interactive effect of days, indicating that DFEN remained effective throughout the 14 day period. The effect on day 1 was somewhat greater than on subsequent days. In addition to eating and gnawing, control rats also showed locomotion, grooming, and occasional escape attempts during the pinch; DFEN-treated rats were indistinguishable from controls in this respect.
Study B The mean baseline duration of tail pinch-induced eating was 75 sec/120 sec trial and this was significantly reduced by both d,l-fenfluramine, F(1,9)= 16.8, p<0.01, and DFEN, F(1,9)= 6.3, p<0.05, relative to vehicle-injected controls (Fig. 2). Atypically, the eating durations of the control rats decreased toward the
DISCUSSION
In study B, we have replicated previous reports that racemic fenfluramine (1) and DFEN (6) suppress tail pinch-induced eating in acute tests. We have now extended these findings by showing that the drugs do not produce tolerance in these paradigms with either infusion (Study A) or injection (Study B) modes of administration. Further, the major effect is due to the d(+)enantiomer, since an equal dose of DFEN mixed with 1-fenfluramine (viz. 3 mg racemate/kg) had a similar effect to 1.5 mg DFEN/kg alone (Fig. 2). The suppression of stress-elicited behav-
lid Z 1.1.1
v 0 Z
tl. 0 Z 0
120-
T / ~
1 DO 80-
1-?
T
1
1~
Controlol
'
D,L-FEN "
60 4.0 20
&~'?
O,
0
:
;
;
;
I
I
i ,
i ,
i ,
1
2
3
4
5
6
7
8
9
i .
.
i .
.
i
i
, ,
, ,
10 11 12 13 14 15
DAYSOFTREATMENT FIG. 2. Tail pinch-induced food intake, expressed as in Fig. 1, in rats given daily injections of either saline (Control), d-fenfluramine (DFEN, 1.5 mg/kg IP), or d,l-fenfluramine (D,L-FEN, 3 mg/kg IP) 30 rain before the start of testing each day. Intakes of the DFEN group were lower than Control (p<0.05) on days 1, 2, 4 and 7. Intakes of the D,L-FEN group were lower than control (p<0.0l) on days 1, 2, 4, 7 and 14.
CHRONIC DEXFENFLURAMINE AND FEEDING
147
TABLE 1
14o'
1
12olOO.
/,
1
80. e
60.
/
ANORECTIC EFFECT OF AN ACUTE DOSE OF DFEN (1 mg/kg) ON MILK DESSERT INTAKE IN RATS PREVIOUSLY TREATED FOR 14 DAYS EITHER WITH MINIPUMP DFEN (3 mg/kg/day) OR NOTHING (CONTROLS)
Control
~
Intake (ml) After Acute Injection Sex Chronic Pretreatment
DFEN
40, u1
20,
C~
0
0
: I
2
I
I
3
4 5 6 7 8 DAYS OF TREATMENT
I
;
I
I
I
9
I
I
I
10 11
12
FIG. 3. Dessert test intake, expressed as Mean --- SE of 30 min baseline tests, in female rats given either no drug (Control) or d-fenfluramine (DFEN, 3 mg/kg/day SC via minipump). Intakes in the DFEN group are smaller than control (/9<0.01) on days 1, 3, 5 and 9.
Vehicle
DFEN
Female Control DFEN
25.3 - 3.7 15.8 - 1.5
5.3 - 2.0* 14.8 - 1.5
Male Control DFEN
15.0 +-- 0.7 16.8 +-- 4.5
3.5 -+ 2.5* 11.0 --- 3.2
Shown are Mean _+ SE for groups of 4 rats. *p<0.05 DFEN
Procedure iors is restricted to feeding/gnawing because noneating behaviors were apparently normal (in fact, as less time was spent eating after drug treatment, correspondingly more time was in principle available for other behaviors). EXPERIMENT 2: DESSERT TEST AND FENFLURAMINE METHOD
Animals and Housing This experiment was performed as two studies. In the first, female Sprague-Dawley rats (4-8 months old, weighing 310-450 g) were used, and were tested in the middle part of the light phase of a 12:12 light cycle. In the second, age-matched (2-3 months) male and female Sprague-Dawley rats (mean body weights 350 and 260 g, respectively, at the start of the study) were housed in a 12:12 cycle with lights on 0200-1400 hr. The rats were tested at 1300 hr, during the last hour of the artificial daytime. All other details of housing were as in Experiment 1.
Rats were trained to drink a sweetened solution, each liter of which contained 125 g commercial sugar and 125 g fat-free powdered milk made in tap water, for 30 min per day. Chow was available at all times, but water was removed during the 30 min dessert test. Milk was presented in 50 ml graduated tubes fitted with sipper tubes. After about one week, the baseline intakes were stable. In the first study, 5 rats received an osmotic minipump, as described in Experiment 1, to deliver DFEN at 3 mg/kg/day for 14 days. Another 5 rats served as untreated controls. In the second study, 7 males and 7 females received DFEN, and another 7 of each sex served as untreated controls. In addition to the time of day, two more features differed in this compared with the preceding study. First, powdered Purina chow was available ad lib from jars inside the cages, and daily intakes were recorded (as well as the dessert intake). Second, an acute tolerance test was performed on day 16, 48 hr after the minipumps nominally expired. For the acute test, half of the rats were injected with DFEN (1 mg/kg, IP) and the other half with saline 30 min before presentation of the dessert.
Data Analysis 120' O~O I00.
80.
The milk intakes for each individual were expressed as % of the baseline, and were treated by analysis of variance with post hoc t-tests. For the acute tolerance test, a Mann-Whitney test was used for comparison of small groups.
male control
I o : l O ~alm~leD¢l~ontroI e~4, f e m
/./ ~
"
60.
~S~TS
Study A
20.
O. -20.
0
:
:
:
:
:
:
:
1
2
3
4
5
6
7
I
I
I
I
| •
| ,
i ,
i °
8 9 10 11 12 13 14 15 DAYS OF TREATMENT
FIG. 4. Mean body weight change of female and male rats during 14 day regimens of either no drug (Control) or DFEN (3 mg/kg/day via minipump). These rats had chow ad lib in addition to daily 30 rain dessert tests. Control vs. DFEN groups differ (p<0.05) on days 1, 2, 5 and 14 (females) and on days 1, 3, 5 and 8 (males).
The dessert intake (mean baseline = 16.0 ml) shown in Fig. 3 remained stable in the controls, and was suppressed by DFEN, F(1,8) = 21.9, p<0.001. There was also an interaction between drug and days, F(5,40)=2.9, p < 0 . 0 5 , because the intake in the DFEN-treated rats increased during the treatment and, by day 11 was not different from controls. The mean body weight of the rats also was decreased by DFEN, F(1,8)=9.99, p = 0 . 0 1 , with a net loss of 20 g on day 1, and a slow regain to 5 g below starting weight by day 11. This sustained, modest weight loss is comparable to previous findings with DFEN (3,12).
148
SOUQUET AND ROWLAND
Study B In contrast to the above result, the effects of DFEN were small in study B. In females, the intake of milk was reduced to about 70% of controls on days 1-3, and thereafter was at control values. The overall drug effect, F(1,12)=3.5, p = 0 . 0 8 , did not reach customary levels of significance. In the males, no drug effect on milk intake was observed, F(1,12)=0.0. The DFEN evidently had some effects in these animals. The 24 hr intakes of chow were significantly decreased by 60 and 30% on days 1 and 2, respectively, in females and by 30% on day 1 in males, although intakes were at control levels for the rest of the 14 days. Thus, overall ANOVAs for food intake were not significant. Body weight, expressed as weight gain from day 0, was decreased by DFEN [females: F(1,14)=7.12, p<0.02; males: F(1,14)= 5.63, p<0.04], as shown in Fig. 4. In the acute tolerance test, a combining data from males and females, the milk intake was comparable in all vehicle-treated groups, regardless of prior chronic treatment. As expected, the original controls receiving acute DFEN were profoundly anorexic (intake 20% of vehicle group), while the chronic DFEN rats showed marked tolerance to the anorectic action of the acute dose (intake about 80% of vehicle group), see Table 1. DISCUSSION In apparent contrast to the tail pinch results in Experiment 1, the effects of DFEN on dessert intake are less marked and/or show considerable tolerance. The effects are clearly greater in the older and heavier females of Study A, than in the more rapidly growing, younger rats of Study B. We have previously noted that the weight reducing effect of DFEN is greatest in mature rats that either are overweight and/or are not gaining weight rapidly (3, 12, 15, t7). The present result suggests this is true with respect to dessert intake, although we cannot rule out two other possibilities. First, the time of day differed between studies A and B, and the rats tested in the middle of the day (A) showed a larger effect than at the end of the day (B). However, Leibowitz (9) has argued that the anorectic action of DFEN on carbohydrate intake (our milk dessert is 80% carbohydrate and 20% protein in calories) is maximal at the
end of day/early night period. Based on this, we anticipated larger effects of DFEN in study B, which is clearly at variance with our findings. A second possibility is that the dose and/or duration of the minipump regimen in Study B was inadvertently lower than in Study A. We have no direct plasma measures of DFEN to address this issue, but the fact that body weight reduction relative to controls was evident in Study B, and that there was substantial tolerance to the acute test dose [as reported before (12)] argues that drug was delivered by the pumps. It is noteworthy that body weight was consistently lower during chronic DFEN, despite little overall anorexia. This presumably reflects decreased feed efficiency in the DFEN-treated rats, as has been suggested in acute studies in rats (4,11). CONCLUSIONS We have previously shown partial tolerance to the effect of d,l-fenfluramine (daily injections) on both dessert intake and tail pinch-induced eating (13,16). We did not observe full tolerance in the present tail pinch studies, using either d,1- or d-fenfluramine. The effects of DFEN on the milk intake were less marked than on stress-related eating, and some tolerance was evident. This result is consistent with findings that the 50% inhibitory dose of acute DFEN is lower in tail pinch than dessert paradigms (17). Even when behavioral tolerance did occur, sustained effects of DFEN on body weight were evident. The chronic mode of administration may be less conducive to development of tolerance than daily injections. We did not measure brain levels of 5HT in the present work, although we generally find little or no depletion of 5HT with either chronic or acute DFEN at these doses (3, 6, 12, 17). Thus, while at higher and/or intermittent doses, DFEN may reduce brain 5HT and so contribute to anorectic tolerance (8), this is unlikely to be an important mechanism at low clinical doses. The DFEN regimens used here produce plasma levels of DFEN (and norfenfluramine) that are 5 - 1 0 × those reported as therapeutic doses in humans (17,18). ACKNOWLEDGEMENT We thank L'Institut des Recherches Servier for support of this paper.
REFERENCES 1. Antelman, S. M.; Caggiula, A. R.; Eichler, A. J.; Lucik, R. R. The importance of stress in assessing the effects of anorectic drugs. Curr. Med. Res. Opin. 6(Suppl. 1):73-82; 1979. 2. Carlton, J.; Rowland, N. E. Effects of initial body weight on anorexia and tolerance to fenfluramine in rats. Pharmacol. Biochem. Behav. 23:551-554; 1985. 3. Carlton, J.; Rowland, N. E. Long term actions of d-fenfluramine in two rat models of obesity: sustained reductions in body weight and adiposity without depletion of brain serotonin. Int. J. Obes.; in press. 4. Even, P.; Coulaud, H.; Nicolaidis, S. Lipostatic and ischymetric mechanisms originate dexfenfluramine-induced anorexia. Pharmacol. Biochem. Behav. 30:89-99; 1988. 5. Fantino, M.; Faion, F.; Rolland, Y. Effect of dexfenfluramine on body weight set point: study in the rat with hoarding behaviour. In: Nicolaidis, S., ed. Serotoninergic system, feeding and body weight regulation. London: Academic Press; 1986:115-126. 6. Garattini, S. Effects of d-fenfluramine on eating disorders. In: Ferrari, E., ed. Disorders of eating behaviour: a psychoneuroendocrine approach. Oxford: Pergamon; 1986:327-341. 7. Garattini, S.; Bizzi, A.; Caccia, S.; Mennini, T.; Samanin, R. Progress in assessing the role of serotonin in the control of food intake. Clin. Neuropharmacol. 1l(Suppl. 1):$8-$32; 1988. 8. Kleven, M. S.; Schuster, C, R.; Seiden, L. S. Effect of depletion of brain serotonin by repeated fenfluramine on neurochemical and
9.
10. 11. 12. 13. 14. 15. 16.
anorectic effects of acute fenfluramine. J. Pharmacol. Exp. Ther. 246:822-828; 1988. Leibowitz, S. F.; Weiss, G. F.; Shor-Posner, G. Hypothalamic serotonin: pharmacological, biochemical, and behavioral analyses of its feeding-suppressive action. Clin. Neuropharmacol. l l(Suppl. 1): $51-$71; 1988. Levitsky, D. A.; Strupp, B. J.; Lupoli, J. Tolerance to anorectic drugs: Pharmacological or artifactual. Phannacol. Biochem. Behav. 14:661-667; 1981. Levitsky, D. A.; Schuster, J. A.; Stallone, D.; Strupp, B. J. Modulation of the thermic effect of food by fenfluramine. Int. J. Obes. 10:169-173; 1986. Rowland, N. E. Effect of continuous infusions of dexfenfluramine on food intake, body weight and brain amines in rats. Life Sci. 39: 2581-2586; 1986. Rowland, N. E.; Antelman, S. M.; Bartness, T. J. Comparison of the effects of fenfluramine and other anorectic agents in different feeding and drinking paradigms in rats. Life Sci. 36:2295-2300; 1985. Rowland, N. E.; Carlton, J. Different behavioral mechanisms underlie tolerance to the anorectic effects of fenfluramine and quipazine. Psychopharmacology (Berlin) 81:155-157; 1983. Rowland, N. E.; Carlton, J. Neurobiology of an anorectic drug: fenfluramine. Prog. Neurobiol. 27:13-62; 1986. Rowland, N. E.; Carlton, J. Tolerance to fenfluramine anorexia: fact
CHRONIC DEXFENFLURAMINE AND FEEDING
or fiction? In: Nicolaidis, S., ed. Serotoninergic system, feeding and body weight regulation. London: Academic Press; 1987:71-83. 17. Rowland, N. E.; Carlton, J. Dexfenfiuramine: effects on food intake in various animal models. Clin. Neuropharmacol. ll(Suppl. 1): $33-$50; 1988. 18. Rowland, N. E.; Souquet, A.-M.; Edwards, D. J. Long term actions
149
of dexfenfluramine on food intake, body weight and brain serotonin in rodents. In: Paoletti, R.; Vanhoutte, P. M., eds. Serotonin: From cell biology to pharmacology and therapeutics. Dordrecht: Kluwer Academic Publishers; in press. 19. Stunkard, A. J. Anorectic agents lower a body weight set point. Life Sci. 30:2043-2055; 1982.
0031-9384/89 $3.00 + .00
Effect of Chronic Administration of Dexfenfluramine on Stress- and Palatability-Induced Food Intake in Rats A N N E - M A R I E S O U Q U E T A N D N E I L E. R O W L A N D 1
Department o f Psychology and Centers for Neurobiological and Nutritional Sciences University of Florida, Gainesville, FL R e c e i v e d 25 M a y 1989
SOUQUET, A.-M. AND N. E. ROWLAND. Effect of chronic administration of dexfenfluramine on stress- and palatability-induced food intake in rats. PHYSIOL BEHAV 46(2) 145-149, 1989.--Dexfenfluramine, administered chronically either SC via osmotic minipump (3 mg/kg/day) or as dally IP injections (1.5 mg/kg) to rats, suppressed tail pinch-induced eating for up to 14 days. The minipump regimen also caused a reduction of intake of a palatable "dessert" for up to 9 days in heavy, adult rats, but was only transiently effective in young, rapidly growing rats. Nonetheless, the regimen produced decreased weight gain and insensitivity to a subsequent acute test dose of the agent. These results illustrate the sustained anorectic efficacy of dexfenfluramine in paradigms that do not involve restriction of diet, and these may be animal models of snacking or binging. The data also corroborate previous suggestions that the agent is more effective in older or more obese populations. Dexfenfluramine
Chronic drug administration
Anorexia
STUDIES from this and several other laboratories have indicated that the anorectic effect of d,l-fenfluramine decreases across repeated daily treatments in rats maintained on schedules of restricted food access (8,16). On such schedules, there is no opportunity for the animals to make up the food not eaten on one day, and so some weight loss (from an already reduced level due to the schedule) occurs. Levitsky and colleagues (10) have argued that this apparent tolerance to the anorexia is because of the weight loss and that when the rats start below some "set point" the anorectic agent is less effective. Subsequent theoretical (19) and empirical (5) accounts support the idea that fenfluramine lowers a set point for body weight. We have presented two empirical counterarguments. In the first, we showed that injections of fenfluramine following scheduled feeds also produced tolerance despite minimal weight loss (14), and in the second we found that weight loss by food restriction prior to fenfluramine treatment failed to greatly diminish the anorectic effect of the drug (2). Increased brain serotonin (5HT) transmission is the likely mechanism by which fenfluramine causes anorexia (6, 7, 15). However, large doses of fenfluramine deplete brain 5HT in rats (15), and Kleven et al. (8) have suggested that anorectic tolerance occurs in chronic regimens because subsequent injections are unable to release as much 5HT from a depleted nerve terminal. The present experiments are designed to overcome what we see
Palatability
Stress
Eating
as several major shortcomings that complicate the interpretation of the above studies. First, we use the active d(+) enantiomer of fenfluramine (DFEN), rather than the racemate used in all of the above work. The 1( - ) enantiomer, although not very potent as an anorectic, has neurochemical effects that might modify the longterm efficacy of its active partner (15). Second, we have used low doses of DFEN that produce little or no depletion of brain 5HT, and so release might remain uncompromised during a chronic regimen. As in previous studies (3, 12, 17), we use osmotic minipumps to deliver the agent continuously rather than by daily injections: because the tissue half-lives of DFEN and its active metabolite, d-norfenfluramine are a few hours in rats, tissue drug levels will fluctuate dramatically across the day with injection paradigms. Third, we use two acute models of overeating that do not employ food deprivation and/or weight loss. The first is tail pinch-induced eating and the second is a "dessert" test in which rats fed chow ad lib also receive brief daily access to a sweet fluid. Acute injection of DFEN is known to inhibit eating in these and other paradigms (13,17). EXPERIMENT 1: TAIL PINCH STUDIES WITH DFEN METHOD
Animals and Housing Male Sprague-Dawley (Zivic Miller) rats, 3-6 months old and
tRequests for reprints should be addressed to Dr. Neil Rowland, Department of Psychology, University of Florida, Gainesville, FL 32611.
145
146
SOUQUET AND ROWLAND
weighing 290--600 g, were housed individually in hanging stainless steel wire cages with tap water and Purina rodent chow pellets available ad lib. The vivarium was light (on: 0700-1900) and temperature (23°C) controlled. All tests were performed in this vivarium during the morning.
~, aZ
180
160
120, o Z
Tail Pinch Procedure
T
T~<
]
1
O ~ 0
140,
,~0[
-
ContrOIr 0
1
100,
T
80.
The rats were placed individually into rectangular plastic washing-up bowls (Rubbermaid) with chow pellets scattered on the floor. In an initial screening test, each rat was administered four 120 sec tail pinches, each 10-15 min apart, using hand-held, foam-padded sponge forceps in a procedure similar to that of Antelman et al. (1). The time spent eating or gnawing food during each pinch was recorded using a stopwatch. Because rats don't eat between trials in these daytime tests, duration is correlated with food consumed and/or gnawed, especially using a within-subjects design. Rats that did not exhibit sustained eating during these screening trials were not included in the main studies.
h
o Z
o
60.
40-
~ / 1
'
i
DFEN
20a
0,
' 1
0
: 2
I 3
.' 4
: 5
: 6
: 7
: 8
~ : : : ; I I 9 10 11 12 13 14 15
DAYS OF TREATMENT
Study A: Minipump DFEN
FIG. 1. Tail pinch-induced food intake, expressed as Mean --+ SE of baselines summed over four 120 sec trials on days indicated, in male rats receiving no drug (Control; open circles) or d-fenfluramine (DFEN; 3 mg/kg/day SC via minipump). Intakes in the DFEN group are smaller than Control (p<0.05) on all days.
Three daily baseline tests were conducted to ensure that the durations of tail pinch-induced eating were stable. Seven rats then received, under brief ether anesthesia, a subcutaneously-implanted osmotic minipump (Alzet No. 2002, 14 days) preloaded with DFEN (hydrochloride salt, from Servier Labs) dissolved in distilled water at a concentration to deliver 3 mg/kg/day. An additional 7 rats received no drug. Measures of tail pinch-induced eating were made 1, 4, 7, 10 and 14 days later.
end of the study, which gave rise to a drug x days interaction term for d,l-fenfluramine (p<0.01). However, it is clear from Fig. 2 that d,l-fenfluramine- and DFEN-treated rats performed similarly, and showed no marked change in the initial (day 1) effect of the drug with repeated treatment. As for the minipump study A, the general other behaviors of drug-treated rats during tail pinch were similar to controls.
Study B: Injections of DFEN After baselines were established as above, rats received daily injections of either DFEN (1.5 mg/kg IP, N = 6), d,l-fenfluramine (3 mg/kg IP, N = 6 ) , or 0.15 M saline as control ( N = 5 ) . Tail pinch tests (days 1, 2, 4, 7, 10 and 14) started 30 min after injections.
Data Analysis All durations of eating were converted to % of individual baselines, and were then compared between treatment groups and days by analysis of variance, and post hoc two-tailed t-tests. RESULTS
Study A The mean baseline duration of tail pinch-induced eating was 62 sec/120 sec trial. The control (no drug) rats improved above this baseline on each of the five test days (Fig. 1). DFEN suppressed tail pinch-induced eating, F(1,12)= 12.5, p<0.01. There was no main or interactive effect of days, indicating that DFEN remained effective throughout the 14 day period. The effect on day 1 was somewhat greater than on subsequent days. In addition to eating and gnawing, control rats also showed locomotion, grooming, and occasional escape attempts during the pinch; DFEN-treated rats were indistinguishable from controls in this respect.
Study B The mean baseline duration of tail pinch-induced eating was 75 sec/120 sec trial and this was significantly reduced by both d,l-fenfluramine, F(1,9)= 16.8, p<0.01, and DFEN, F(1,9)= 6.3, p<0.05, relative to vehicle-injected controls (Fig. 2). Atypically, the eating durations of the control rats decreased toward the
DISCUSSION
In study B, we have replicated previous reports that racemic fenfluramine (1) and DFEN (6) suppress tail pinch-induced eating in acute tests. We have now extended these findings by showing that the drugs do not produce tolerance in these paradigms with either infusion (Study A) or injection (Study B) modes of administration. Further, the major effect is due to the d(+)enantiomer, since an equal dose of DFEN mixed with 1-fenfluramine (viz. 3 mg racemate/kg) had a similar effect to 1.5 mg DFEN/kg alone (Fig. 2). The suppression of stress-elicited behav-
lid Z 1.1.1
v 0 Z
tl. 0 Z 0
120-
T / ~
1 DO 80-
1-?
T
1
1~
Controlol
'
D,L-FEN "
60 4.0 20
&~'?
O,
0
:
;
;
;
I
I
i ,
i ,
i ,
1
2
3
4
5
6
7
8
9
i .
.
i .
.
i
i
, ,
, ,
10 11 12 13 14 15
DAYSOFTREATMENT FIG. 2. Tail pinch-induced food intake, expressed as in Fig. 1, in rats given daily injections of either saline (Control), d-fenfluramine (DFEN, 1.5 mg/kg IP), or d,l-fenfluramine (D,L-FEN, 3 mg/kg IP) 30 rain before the start of testing each day. Intakes of the DFEN group were lower than Control (p<0.05) on days 1, 2, 4 and 7. Intakes of the D,L-FEN group were lower than control (p<0.0l) on days 1, 2, 4, 7 and 14.
CHRONIC DEXFENFLURAMINE AND FEEDING
147
TABLE 1
14o'
1
12olOO.
/,
1
80. e
60.
/
ANORECTIC EFFECT OF AN ACUTE DOSE OF DFEN (1 mg/kg) ON MILK DESSERT INTAKE IN RATS PREVIOUSLY TREATED FOR 14 DAYS EITHER WITH MINIPUMP DFEN (3 mg/kg/day) OR NOTHING (CONTROLS)
Control
~
Intake (ml) After Acute Injection Sex Chronic Pretreatment
DFEN
40, u1
20,
C~
0
0
: I
2
I
I
3
4 5 6 7 8 DAYS OF TREATMENT
I
;
I
I
I
9
I
I
I
10 11
12
FIG. 3. Dessert test intake, expressed as Mean --- SE of 30 min baseline tests, in female rats given either no drug (Control) or d-fenfluramine (DFEN, 3 mg/kg/day SC via minipump). Intakes in the DFEN group are smaller than control (/9<0.01) on days 1, 3, 5 and 9.
Vehicle
DFEN
Female Control DFEN
25.3 - 3.7 15.8 - 1.5
5.3 - 2.0* 14.8 - 1.5
Male Control DFEN
15.0 +-- 0.7 16.8 +-- 4.5
3.5 -+ 2.5* 11.0 --- 3.2
Shown are Mean _+ SE for groups of 4 rats. *p<0.05 DFEN
Procedure iors is restricted to feeding/gnawing because noneating behaviors were apparently normal (in fact, as less time was spent eating after drug treatment, correspondingly more time was in principle available for other behaviors). EXPERIMENT 2: DESSERT TEST AND FENFLURAMINE METHOD
Animals and Housing This experiment was performed as two studies. In the first, female Sprague-Dawley rats (4-8 months old, weighing 310-450 g) were used, and were tested in the middle part of the light phase of a 12:12 light cycle. In the second, age-matched (2-3 months) male and female Sprague-Dawley rats (mean body weights 350 and 260 g, respectively, at the start of the study) were housed in a 12:12 cycle with lights on 0200-1400 hr. The rats were tested at 1300 hr, during the last hour of the artificial daytime. All other details of housing were as in Experiment 1.
Rats were trained to drink a sweetened solution, each liter of which contained 125 g commercial sugar and 125 g fat-free powdered milk made in tap water, for 30 min per day. Chow was available at all times, but water was removed during the 30 min dessert test. Milk was presented in 50 ml graduated tubes fitted with sipper tubes. After about one week, the baseline intakes were stable. In the first study, 5 rats received an osmotic minipump, as described in Experiment 1, to deliver DFEN at 3 mg/kg/day for 14 days. Another 5 rats served as untreated controls. In the second study, 7 males and 7 females received DFEN, and another 7 of each sex served as untreated controls. In addition to the time of day, two more features differed in this compared with the preceding study. First, powdered Purina chow was available ad lib from jars inside the cages, and daily intakes were recorded (as well as the dessert intake). Second, an acute tolerance test was performed on day 16, 48 hr after the minipumps nominally expired. For the acute test, half of the rats were injected with DFEN (1 mg/kg, IP) and the other half with saline 30 min before presentation of the dessert.
Data Analysis 120' O~O I00.
80.
The milk intakes for each individual were expressed as % of the baseline, and were treated by analysis of variance with post hoc t-tests. For the acute tolerance test, a Mann-Whitney test was used for comparison of small groups.
male control
I o : l O ~alm~leD¢l~ontroI e~4, f e m
/./ ~
"
60.
~S~TS
Study A
20.
O. -20.
0
:
:
:
:
:
:
:
1
2
3
4
5
6
7
I
I
I
I
| •
| ,
i ,
i °
8 9 10 11 12 13 14 15 DAYS OF TREATMENT
FIG. 4. Mean body weight change of female and male rats during 14 day regimens of either no drug (Control) or DFEN (3 mg/kg/day via minipump). These rats had chow ad lib in addition to daily 30 rain dessert tests. Control vs. DFEN groups differ (p<0.05) on days 1, 2, 5 and 14 (females) and on days 1, 3, 5 and 8 (males).
The dessert intake (mean baseline = 16.0 ml) shown in Fig. 3 remained stable in the controls, and was suppressed by DFEN, F(1,8) = 21.9, p<0.001. There was also an interaction between drug and days, F(5,40)=2.9, p < 0 . 0 5 , because the intake in the DFEN-treated rats increased during the treatment and, by day 11 was not different from controls. The mean body weight of the rats also was decreased by DFEN, F(1,8)=9.99, p = 0 . 0 1 , with a net loss of 20 g on day 1, and a slow regain to 5 g below starting weight by day 11. This sustained, modest weight loss is comparable to previous findings with DFEN (3,12).
148
SOUQUET AND ROWLAND
Study B In contrast to the above result, the effects of DFEN were small in study B. In females, the intake of milk was reduced to about 70% of controls on days 1-3, and thereafter was at control values. The overall drug effect, F(1,12)=3.5, p = 0 . 0 8 , did not reach customary levels of significance. In the males, no drug effect on milk intake was observed, F(1,12)=0.0. The DFEN evidently had some effects in these animals. The 24 hr intakes of chow were significantly decreased by 60 and 30% on days 1 and 2, respectively, in females and by 30% on day 1 in males, although intakes were at control levels for the rest of the 14 days. Thus, overall ANOVAs for food intake were not significant. Body weight, expressed as weight gain from day 0, was decreased by DFEN [females: F(1,14)=7.12, p<0.02; males: F(1,14)= 5.63, p<0.04], as shown in Fig. 4. In the acute tolerance test, a combining data from males and females, the milk intake was comparable in all vehicle-treated groups, regardless of prior chronic treatment. As expected, the original controls receiving acute DFEN were profoundly anorexic (intake 20% of vehicle group), while the chronic DFEN rats showed marked tolerance to the anorectic action of the acute dose (intake about 80% of vehicle group), see Table 1. DISCUSSION In apparent contrast to the tail pinch results in Experiment 1, the effects of DFEN on dessert intake are less marked and/or show considerable tolerance. The effects are clearly greater in the older and heavier females of Study A, than in the more rapidly growing, younger rats of Study B. We have previously noted that the weight reducing effect of DFEN is greatest in mature rats that either are overweight and/or are not gaining weight rapidly (3, 12, 15, t7). The present result suggests this is true with respect to dessert intake, although we cannot rule out two other possibilities. First, the time of day differed between studies A and B, and the rats tested in the middle of the day (A) showed a larger effect than at the end of the day (B). However, Leibowitz (9) has argued that the anorectic action of DFEN on carbohydrate intake (our milk dessert is 80% carbohydrate and 20% protein in calories) is maximal at the
end of day/early night period. Based on this, we anticipated larger effects of DFEN in study B, which is clearly at variance with our findings. A second possibility is that the dose and/or duration of the minipump regimen in Study B was inadvertently lower than in Study A. We have no direct plasma measures of DFEN to address this issue, but the fact that body weight reduction relative to controls was evident in Study B, and that there was substantial tolerance to the acute test dose [as reported before (12)] argues that drug was delivered by the pumps. It is noteworthy that body weight was consistently lower during chronic DFEN, despite little overall anorexia. This presumably reflects decreased feed efficiency in the DFEN-treated rats, as has been suggested in acute studies in rats (4,11). CONCLUSIONS We have previously shown partial tolerance to the effect of d,l-fenfluramine (daily injections) on both dessert intake and tail pinch-induced eating (13,16). We did not observe full tolerance in the present tail pinch studies, using either d,1- or d-fenfluramine. The effects of DFEN on the milk intake were less marked than on stress-related eating, and some tolerance was evident. This result is consistent with findings that the 50% inhibitory dose of acute DFEN is lower in tail pinch than dessert paradigms (17). Even when behavioral tolerance did occur, sustained effects of DFEN on body weight were evident. The chronic mode of administration may be less conducive to development of tolerance than daily injections. We did not measure brain levels of 5HT in the present work, although we generally find little or no depletion of 5HT with either chronic or acute DFEN at these doses (3, 6, 12, 17). Thus, while at higher and/or intermittent doses, DFEN may reduce brain 5HT and so contribute to anorectic tolerance (8), this is unlikely to be an important mechanism at low clinical doses. The DFEN regimens used here produce plasma levels of DFEN (and norfenfluramine) that are 5 - 1 0 × those reported as therapeutic doses in humans (17,18). ACKNOWLEDGEMENT We thank L'Institut des Recherches Servier for support of this paper.
REFERENCES 1. Antelman, S. M.; Caggiula, A. R.; Eichler, A. J.; Lucik, R. R. The importance of stress in assessing the effects of anorectic drugs. Curr. Med. Res. Opin. 6(Suppl. 1):73-82; 1979. 2. Carlton, J.; Rowland, N. E. Effects of initial body weight on anorexia and tolerance to fenfluramine in rats. Pharmacol. Biochem. Behav. 23:551-554; 1985. 3. Carlton, J.; Rowland, N. E. Long term actions of d-fenfluramine in two rat models of obesity: sustained reductions in body weight and adiposity without depletion of brain serotonin. Int. J. Obes.; in press. 4. Even, P.; Coulaud, H.; Nicolaidis, S. Lipostatic and ischymetric mechanisms originate dexfenfluramine-induced anorexia. Pharmacol. Biochem. Behav. 30:89-99; 1988. 5. Fantino, M.; Faion, F.; Rolland, Y. Effect of dexfenfluramine on body weight set point: study in the rat with hoarding behaviour. In: Nicolaidis, S., ed. Serotoninergic system, feeding and body weight regulation. London: Academic Press; 1986:115-126. 6. Garattini, S. Effects of d-fenfluramine on eating disorders. In: Ferrari, E., ed. Disorders of eating behaviour: a psychoneuroendocrine approach. Oxford: Pergamon; 1986:327-341. 7. Garattini, S.; Bizzi, A.; Caccia, S.; Mennini, T.; Samanin, R. Progress in assessing the role of serotonin in the control of food intake. Clin. Neuropharmacol. 1l(Suppl. 1):$8-$32; 1988. 8. Kleven, M. S.; Schuster, C, R.; Seiden, L. S. Effect of depletion of brain serotonin by repeated fenfluramine on neurochemical and
9.
10. 11. 12. 13. 14. 15. 16.
anorectic effects of acute fenfluramine. J. Pharmacol. Exp. Ther. 246:822-828; 1988. Leibowitz, S. F.; Weiss, G. F.; Shor-Posner, G. Hypothalamic serotonin: pharmacological, biochemical, and behavioral analyses of its feeding-suppressive action. Clin. Neuropharmacol. l l(Suppl. 1): $51-$71; 1988. Levitsky, D. A.; Strupp, B. J.; Lupoli, J. Tolerance to anorectic drugs: Pharmacological or artifactual. Phannacol. Biochem. Behav. 14:661-667; 1981. Levitsky, D. A.; Schuster, J. A.; Stallone, D.; Strupp, B. J. Modulation of the thermic effect of food by fenfluramine. Int. J. Obes. 10:169-173; 1986. Rowland, N. E. Effect of continuous infusions of dexfenfluramine on food intake, body weight and brain amines in rats. Life Sci. 39: 2581-2586; 1986. Rowland, N. E.; Antelman, S. M.; Bartness, T. J. Comparison of the effects of fenfluramine and other anorectic agents in different feeding and drinking paradigms in rats. Life Sci. 36:2295-2300; 1985. Rowland, N. E.; Carlton, J. Different behavioral mechanisms underlie tolerance to the anorectic effects of fenfluramine and quipazine. Psychopharmacology (Berlin) 81:155-157; 1983. Rowland, N. E.; Carlton, J. Neurobiology of an anorectic drug: fenfluramine. Prog. Neurobiol. 27:13-62; 1986. Rowland, N. E.; Carlton, J. Tolerance to fenfluramine anorexia: fact
CHRONIC DEXFENFLURAMINE AND FEEDING
or fiction? In: Nicolaidis, S., ed. Serotoninergic system, feeding and body weight regulation. London: Academic Press; 1987:71-83. 17. Rowland, N. E.; Carlton, J. Dexfenfiuramine: effects on food intake in various animal models. Clin. Neuropharmacol. ll(Suppl. 1): $33-$50; 1988. 18. Rowland, N. E.; Souquet, A.-M.; Edwards, D. J. Long term actions
149
of dexfenfluramine on food intake, body weight and brain serotonin in rodents. In: Paoletti, R.; Vanhoutte, P. M., eds. Serotonin: From cell biology to pharmacology and therapeutics. Dordrecht: Kluwer Academic Publishers; in press. 19. Stunkard, A. J. Anorectic agents lower a body weight set point. Life Sci. 30:2043-2055; 1982.