- Email: [email protected]
Behavioral deficits in monosodium glutamate rats: Specific changes in the structure of feeding behavior
Life seicneq vd. ag No. 23, pp 21n-2m 1990 cTqyfi@~1998EQNicrsdenceInc. RintedintheIJsk AurighrsIc!sc~ 0024-m/9a $19.00 t .a0
PII SOOZA-3205(!W)oo187-8
ELSEVIER
BEHAVIORAL DEFICITS IN MONOSODIUM GLUTAMATE RATS: SPECIFIC CHANGES IN THE STRUCTURE OF FEEDING BEHAVIOR A. Stricker-Krongrad,
C. Burlet and B. Beck1
INSERM U-308. MRCA. Equipe de Neurobiologie et Physiologie Experimentales 38, rue Lionnois F-54000 NANCY. (Received in final form March 17,1998) Summary We studied the feeding rhythms and feeding patterns of adult Long-Evans rats treated with monosodium glutamate (MSG) in their early post-natal period. This treatment is known to induce neuronal degeneration in the arcuate nucleus (ARC), a major hypothalamic site implicated in the regulation of feeding. Neonatal rats were treated intraperitoneally with MSG or saline (controls) alone on the first days of life. At age of 6 months, male control and male MSG rats were placed in our automatic feeding system, and the structure of feeding behavior and diurnal feeding rhythms were analysed. On a 24 hours basis, MSG rats ate less than control rats (- 24 %). This hypophagia resulted from a mild diurnal hyperphagia (+ 6 %) and a pronounced nocturnal hypophagia (- 34 %). This hypophagia was the main consequence of a decrease of meal size in MSG rats (- 37 %) and was associated with an increase in meal duration (+ 52 %). It was also associated with a total disappearence of the two feeding peaks that normally occur at light and dark onset in the rat (- 90 % 2 h after dark onset and - 49 % 2 h before light onset). These results indicate that neonatal treatment with MSG induces important changes in feeding patterns and feeding rhythms in the adulthood. These changes might be related to the disappearance of neurotransmitters located in the arcuate nucleus. Z&YW0rd.r: mormsodium glutamate feeding Peaks
toxicity,
feeding behavior, hypophagia,arcuate nucleus lesion, meal size,
Monosodium glutamate (MSG) treatment in early post-natal period is known to induce profound neuroanatomical disturbances in rodents. Neuronal loss following MSG are localized in the circumventricular organs (1, 2, 3, 4) such as the arcuate nucleus (3, 4). A general loss of neuromodulators is observed in the medio-lasal hypothalamus (1, 2, 3, 5). These disturbances are associated with feeding behaviour perturbations (1, 5, 6, 7, 8, 9, 10, 1 l), such as diminished food intake (5,7, 9, 10, 1 l), persistent preference for carbohydrates (7, 10) and an alterated response to some orexigenic agents (5, 9). Actually, little is known about the feeding pattern and feeding rhythms that underly the hypophagia of the MSG treated rats. Their investigation might help us to understand the eating behavior disturbances induced by arcuate nucleus lesions. Methods Experimental animals
Neonatal Long-Evans rats were injected subcutaneously with 4 g/kg body weight monosodium glutamate (Sigma, LaVerpilliere, France) or with saline at days 3, 5 and 7. Injections were counterbalanced among litters and between sex. At weaning, female rats were discarded from the 1Dr B. Beck. INSERM U308 M.R.C.A. 38, me Lionnois. 54000 Nancy. France. Phone: (3) 83 36 41 45. Fax: (3) 83 37 62 44. e-mail: [email protected]
Feediog Behavior cd MSG Treakd Rats
Vol. 62, No. 23, I998
study and male rats were fed on standard rat chow diet. Six months later, control and MSG rats were randomly taken among saline and MSG-treated animals, respectively. Feeding pattern experiment
Sixteen rats (eight controls and eight MSG-treated) were housed in single wire cages with food and water ad lib&urn in an air-conditioned room with a 12 hr light : 12 hr dark cycle. They were fed on a well-balanced diet supplying 54 % of energy from carbohydrate and 30 % from fat (12). The diet was mixed with water (40/50 (w/w)) in order to obtain a paste usable in our automatic feeding system. Behavioral record
Each cage included a complete automatic feeding system with food delivery under the animal’s control (FeedBAC (13, 14)). Briefly, a plastic syringe containing the pasty diet is connected to infrared light diodes which control its delivery with a step-to-step motor. Data from the feeding system are registered in a computer for treatment and analysis in a second step. Behavioral recording was performed during multiple periods of 24 hours. Before this period, an intra-animal reliability calculation was performed during a minimum of three times 24 hours in order to ensure that the animals were well adapted to the feeding system. Design and analysis
A meal was defined as the ingestion of a minimum of 1.2 g of the pasty diet followed by at least 10 minutes during which no feeding occured. The following dependent variables and derived measures were evaluated : meal number, meal size (g), total amount of food eaten (g), meal duration (min), eating rate (g/min), meal intervals (min) and time spent eating (min) were calculated as the mean of each variable for each animal. The following analyses were performed on the values registered during the 24 h period : light on, light off and entire periods feeding pattern analyses. Feeding rhythms experiment
Data from the feeding systems were registered into a computer for analysis. Behavioral recording was performed during ten consecutive days with the same animals. Each two hours, food intake (in g) was compiled for each animal and means were calculated for the control and MSG-treated groups. Data from the ten days were matched together in a single 24 hours period. Statistical analyses All the analyses were conducted
following dependent or repeated measures experimental design. Gaussian-distributed variables were analyzed with Fischer two-way analysis of variance for repeated or independent measures and compared one to the other with Student’s paired-t test. In case of nonGaussian variables (i.e. discrete values) or of unequal variances, they were analyzed with Friedman’s two-way analysis of variance for dependent variables and compared one to the other with paired Wiicoxon T test. Only probability values less than 0.05 (two-tailed) were taken into account, Non-discrete variables are presented as mean corrected by standard error of the mean. Discrete variables are presented as median. Results Experimental animals
At the beginning of the behavioral experiment. MSG-treated rats weighed less than control rats (346 + 7 vs 376 + 9 g, p
Vol. 62, No. ?3,1998
Feeding Behavior of MSG Treated Rats
2129
Feeding pattern experiment
Table I: Feeding patterns of MSG and control rats during the light/dark cycle. * p
LIGHT
127k 11 28.5 + 1.3
213 + 28 9.1 + 0.5
2.:: 2.4 * 25.0 + 0.72 +
3.7: 2.3 f 6.5 + 0.61 +
0 5 0:2 0.5 0.03
05 0.i 0.1 0.04
DARK
83kO4 19.5 f 1.1 7 2.5 f 0.2 2.4 f 0.2 18.4 + 0.2 0.70 * 0.07
122 If:12 36.1 k 1.4**
146 f 14** 10.5 * 0.4 *
86f06 25.6 k 1.2**
3.8’: 1.5 f 19.0 + 0.45 f
2.4: 03 * 1.5 + 0.; * 6.9 k O.l* 0.53 Z!c0.04
3.4: 1.5 * 12.1 f 0.45 +
0 3* o.i** 0.6** 0.05**
0.3* 0.1** 0.3** 0.05*>
Feeding behavior during the 24 hours period MSG-treated animals ate less than control (- 24 % ; t = 7.68 ; p < 0.001). This diminution in food intake was associated with an increase in time spent eating (+ 29 % ; t = 4.29 ; p < 0.001) and in meal duration (+ 52 % ; t = 2.23 ; p < 0.05). Meal size decreased (- 37 % ; t = 3.18 ; p< 0.01) and consequently, eating rate decreased (- 37 % ; t = 4.63 ; p < 0.001). Feeding behavior during the light period MSG-treated animals ate slightly more than control (+ 5 % ; t = 2.82 ; p < 0.05). During this period, MSG treated animals spent more time to eat ( t = 2.18 ; p c 0.05), ate less during meals (t = 2.52 ; p < 0.05) and their meals were shorter (t = 2.22 ; p < 0.05) than control animals. Feeding behavior during the dark period MSG-treated animals ate less than control (- 34 % ; t = 17.4 ; p c 0.0001). During this period, MSG-treated animals spent much more time to eat (t = 3.74 ; p < O.Ol), ate less during meals (t = 4.02 ; p < O.Ol), but their meals were longer (t = 2.49 ; p c 0.05) than control animals. Consequently, eating rate decreased (t = 2.90 ; p c 0.05). Feeding rhythms
Control animals Results are shown in figure 1. Average food intake ranged from 0 to 6 grams per 2 hours. During the light period, two intervals without meals occured at its middle and its end. At the beginning of dark period, food intake reached a peak (5.1 f 0.8 g/2 h (12 h - 14 h) vs 0.6 + 0.4 g/2 h (14 h -16 h) ; t = 5.03 ; p < 0.001). At its end, another peak occured (4.1 ?I 0.3 gl2h (22 h - 24 h) vs 0.9 f 0.4 gI2h (20 h - 22 h) ; t = 6.40 ; p < 0.001). These two peaks were not statistically different (t = 1.17 ; NS).
Fading Behaviorof MN3 Treated Rats
2l30
:. :::.
FOOD INTAKE
(g)
6
Vol. 62, No. 23,1!998
.._....._..~.~.~,~,~.~,~.~ ::_~_~_~.~.~_~,~.~.~,‘_~.‘.~ iiilililiiliiia~~~~~~~~~~~~~~~~~~~~~~ :. .. .‘_ .‘.‘.~.‘.‘.-.‘.~.‘.‘.’ . . . . .. . .::;,: ~.~_~_~.~.~.~.~,~,~,~,‘.~.‘.~.‘.’.~.’.~.’,’, .._,........................ ..:.. ..~_~_~.‘_‘_‘.‘_‘_‘_‘_~.~.~.‘.‘.~.~.~.’.’.-.~.’.‘.‘. . . . . .“““‘... . . . . . . . . ....‘...“.. ..... ..... _~_~,~,~.~.~_‘_‘_~.~_~_‘_‘.~.~,~.~.~.~.~.~ ,~,~,‘,‘,~.~_‘_‘_‘.~.‘.‘.‘,‘,‘,‘.’.’.’.’.’. .~.~_~.‘.~_~_~_~_~.~.~.‘.‘.~.‘.‘.’.’.’.’.’ .~.~_~.~,~,~.~,~.~.~.~.‘.‘.~.‘.~.’.’.’,’,’ .~.~.~.~.~.~.~.~.~_..~.~.~.~.~.~.~.’.’.’.’.~
:.:.:.....
5 4
0
4
8
12
16
20
24
28
32
36
HOURS Fig I: Distribution of feeding episodes at two hours cumulated data of ten consecutive days)
intervals
in the control
rats (n=8
;
MSG-treated animals Results are shown in figure 2. Average food intake ranged from 0 to 3 grams per 2 hours. During the light period, no intervals without meals occured and during the dark period, no peaks occured. Food intake of MSG rats at the beginning and the end of this period was significantly smaller than control rats (0.5 f 0.3 g/2 h (MSG) vs 5.1 f 0.8 g/2 h (C) (12 h -14 h) : t = 5.38, p < 0.001 and 2.1 + 0.2 g/2 h (MSG) vs 4. I f 0.3 g/2 h (C) (22 h - 24 h) ; t = 5.54 ; p < 0.001).
0
8
12
16
20
24
28
32
36
HOURS Fig 2: Distribution of feeding episodes at two hours intervals in the MSG-treated cumulated data of ten consecutive days).
rats (n=8 ;
Vol. 62, No. 23, 19!98
Feeding Behavior of MSG Treated Rats
2131
Discussion The present experiment showed that MSG-treated rats are characterized by disturbances in feeding pattern and feeding rhythms, and extends the analyses of its behavioral syndrome. MSG rats ate less than normal rats mainly by decreasing meal size and eating rate. The feeding pattern disturbances of MSG rats mainly originates in its profound perturbations in the dark phase. This agrees with previous experiments recording gross food intake only (1, 5,6, 8, 10). Meal size was the most affected parameter of the microstructure of the feeding behavior in MSG rats. There are evidence that the triggering of this parameter is not under circadian control, but only depends on the energy regulation level (15, 16, 17, 18, 19). This hypothesis is supported by experiments showing that in case of lower ambient temperature (16), diabetes (20), and refeeding (21), rats eat more food by increasing meal size and not meal frequency. The same phenomenon occurs in the obese Zucker rats with a meal size around 3 g/meal whereas in the lean rats it is less than 2 g/meal (14, 22). In these two cases, meal sizes are not affected by circadian variations (15). In the present experiment, circadian variations in meal sizes were not found either in control or in MSG rats according to this hypothesis. On the other hand, the other parameters of the microstructure were affected differently during the circadian cycle. Disappearence of the two feeding peaks, which normally occur in rats (16, 23, 24), indicates a disruption of the circadian control of feeding in MSG rats, as already described (11). This might be a component of a more general disruption of circadian rhythmicity, as demonstrated by changes in the sleep-wake cycle and locomotor activity (25,26). Among brain transmitters systems that play a role in food intake and circadian control altogether, and might support the behavioral syndrome of the MSG rats, the intrahypothalamic arcuateparaventricular neuropeptide Y (NPY) network is a good candidate. There are major argument in its favor. The first one comes from central injection studies. When continuously infused in cerebral ventricles, NPY induces a disruption of dark/light rhythms of food intake independently of its orexigenic properties (12). Its stimulation properties in the paraventricular nucleus (PVN) vary along the daily cycle, being more effective in the first portion of the dark phase (27). When injected into the suprachiasmatic nucleus, the circadian clock, it phase-shifts the activity of the hamsters (28). The second one comes from biochemical studies which show the existence of daily rhythms of endogenous NPY levels within hypothalamic areas (29). In the parvocellular part of the PVN. it increases at the onset of the dark period, while in the ARC levels reach peaks at the two phase shifts (29). These endogenous peaks are in phase with those observed in food intake in the control animals of this experiment. The third one comes from studies showing that the activity of the hypothalamic NPY-containing pathway is higher in obese Zucker rats (14, 30, 3 1, 32), fasted rats (33, 34, 35,36) and diabetic rats (34, 37); all of which are characterized by having increased meal sizes (15, 20, 2 1). The fourth one comes from studies showing that hypothalamic NPY is depleted in MSG rats (5, 9, 10, 38). Recent evidence also suggest that leptin, the ob gene product, might contribute in association with NPY to the day-night feeding alterations observed in MSG-treated rats (39). Therefore the absence in amplitude variations in food intake during the light/dark cycle are associated with reduced meal size and increased meal duration in the MSG rats. These results describe the profound eating behavior disturbances that are induced by monosodium glutamate-induced lesion of the arcuate nucleus. References 1. R. DAWSON and J. F. LORDEN. J. Comp. Physiol. Psycho]. 95 71-84 (1981). 2. N. LEMKEY-JOHNSTON and W.A. REYNOLDS. J. Neuropathol. Exp. Neurol. 33 74-97 (1974). 3. B. MEISTER, S. CECCATELLI, T. HOKFELT, N.E. ANDEN, M. ANDEN and E. THEODORSSON. Exp. Brain Res. 76 343-368 (1989). 4. J. W. OLNEY Science. 164 7 19-721 ( 1969). 5. R. DAWSON, D.R. WALLACE and S.M. GABRIEL. Pharmacol. Biochem. Behav. 32 391398 ( 1989).
2132
6.
Feeding Behavior of MSG Treated Rats
Vol. 62, No. 23.1998
R. DAWSON and Z. ANNAU. Neurobehav. Toxicol. Teratol. 5 399-406 (1983). R. B. KANAREK, J. MEYERS. R. MADE and J. MAYER. Pharmacol. Biochem. Behav. 10 717-721 (1979). 8. J. F. LORDEN and A. CAUDLE. Neurobehav. Toxicol. Teratol.. 8 509-519 (1986). 9. A. STRICKER-KRONGRAD, C. BURLET and B. BECK. Eur J Pharmacol. 295 27-34 (1996). IO. B. BECK, A. STRICKER-KRONGRAD. A. BURLET, N. MUSSE. J.P. NICOLAS and C. BURLET. Neurosci. Letters 225 153- 156 ( 1997). I I W. J. RIETVELD. J.H. MEIJER. J. RUIS and P. BUYS. J. Interdiscipl. Cycle Res. 17 327334 ( 1986). 12. B. BECK. A. STRICKER-KRONGRAD. J.P. NICOLAS and C. BURLET. Ann. N.Y. Acad. Sci.’ 611 49 l-494 ( 1990). 13. A. STRICKER-KRONGRAD, B. BECK. J.P. NICOLAS and C. BURLET. Pharmacol. Biochem. Behav. 43 881-886 (1992). 14 A. STRICKER-KRONGRAD, J.P. MAX, N. MUSSE, J.P. NICOLAS, C. BURLET and B. BECK. Brain Res. 660 162-I 66 (1994). A. DE JONG-NAGELSMIT, J. KEIJSER and J. H. STRUBBE. I5 A. B. ALINGH-PRINS, Physiol. Behav. 38 423-426 (1986). 16 S. ARMSTRONG Neurosci. Biobehav. 4 27-53 (1980). 17 R.F. DAVIES J. Comp. Physiol. Psychol. 91 91 I-916 (1982). J. KEIJSER and J.H. STRUBBE. Physiol. Behav. 18 N.J. SPITERI, A.J. ALINGH-PRINZ, 25 775-777 (1980. Physiol Behav. 19 J.H. STRUBBE. J. KEIJSER. T. DIJKSTRA and A.J. ALINGH-PRINZ. 36 489-493 ( 1986). 20 D.W. THOMAS, E. SCHARRER and J. MAYER. Physiol. Behav. 17 345-349 (1976). 21 D.A. LEVITSKY Physiol. Behav. 5 291-300 ( 1970). 22. E.E. BECKER and J.A. GRINKER. Physiol. Behav. 18 685-692 (1977). 23. J. LE MAGNEN and S. TALLON. J. Physiol. Paris. 58 323-349 (1966). 24. P.S. SIEGAL J. Comp. Physiol. Psychol. 54 294-301 (1961). 25. M. OLIVO, K. KITAHAMA, J.L. VALATX and M. JOUVET. Neurosci. Lett. 67 186-190 (1986). 26. S. MIYABO, I. YAMAMURA, E. OOYA. N. AOYAGI, Y. HORIKAWA and S. HAYASHI. Brain Res. 339 20 I-208. 27. B.J. STANLEY and W.J. THOMAS. Peptides. 14 475-481 (1986). 28. H.E. ALBERS and C.F. FERRIS. Neurosci. Lett. 50 163-168 (1984). B. BECK. C. BURLET and S.F. LEIBOWITZ. Brain. Res. 536 29. M. JHANWAR-UNIYAL, 33 l-334 (1990). 30. B. BECK, A. BURLET, R. BAZIN. J.P. NICOLAS and C. BURLET. J. Nutr. 123 11681172 (1993). 31. H.D. MC CARTHY. P.E. MC KIBBIN. B. HOLLOWAY, R. MAYERS and G. WILLIAMS. Life Sci. 49 1491-1497 (1991). 32. G. SANACORA, M. KERSHAW, J.A. FINKELSTEIN and J.D. WHITE. Endocrinology 127 730-737 (I 990). 33. B. BECK, M. JHANWAR-UNIYAL, A. BIJRLET, J.P. NICOLAS and C. BURLET. Brain Res. 528 245-249 (1990). 34. H.M. FRANKISH, D. MCCARTHY. S. DRYDEN, A. KILPATRICK and G. WILLIAMS. Peptides. 14 94 I-948 ( 1993). 35. A. SAHU, P.S. KALRA and S.P. KALRA. Peptides 9 83-86 (1988). 36. J.D. WHITE and M. KERSHAW. Mol. Cell Neurosci. 141-48 (1990). 37. G. WILLIAMS. Y. LEE, H.M. CARDOSO. B.E. OKPERE and S.R. BLOOM. Diabetes. 3 8 321-327 (1989). 38. M. ABE, M. SAITO and T. SHIMUZU. Brain Rea. Bull. 24 289-291 (1990). 39. R. DAWSON, M.A. PELLEYMOUNTER, W.J. MILLARD, S. LIU and B. EPPLER. Am. J. Physiol. 36 E202-E206. 7.
PII SOOZA-3205(!W)oo187-8
ELSEVIER
BEHAVIORAL DEFICITS IN MONOSODIUM GLUTAMATE RATS: SPECIFIC CHANGES IN THE STRUCTURE OF FEEDING BEHAVIOR A. Stricker-Krongrad,
C. Burlet and B. Beck1
INSERM U-308. MRCA. Equipe de Neurobiologie et Physiologie Experimentales 38, rue Lionnois F-54000 NANCY. (Received in final form March 17,1998) Summary We studied the feeding rhythms and feeding patterns of adult Long-Evans rats treated with monosodium glutamate (MSG) in their early post-natal period. This treatment is known to induce neuronal degeneration in the arcuate nucleus (ARC), a major hypothalamic site implicated in the regulation of feeding. Neonatal rats were treated intraperitoneally with MSG or saline (controls) alone on the first days of life. At age of 6 months, male control and male MSG rats were placed in our automatic feeding system, and the structure of feeding behavior and diurnal feeding rhythms were analysed. On a 24 hours basis, MSG rats ate less than control rats (- 24 %). This hypophagia resulted from a mild diurnal hyperphagia (+ 6 %) and a pronounced nocturnal hypophagia (- 34 %). This hypophagia was the main consequence of a decrease of meal size in MSG rats (- 37 %) and was associated with an increase in meal duration (+ 52 %). It was also associated with a total disappearence of the two feeding peaks that normally occur at light and dark onset in the rat (- 90 % 2 h after dark onset and - 49 % 2 h before light onset). These results indicate that neonatal treatment with MSG induces important changes in feeding patterns and feeding rhythms in the adulthood. These changes might be related to the disappearance of neurotransmitters located in the arcuate nucleus. Z&YW0rd.r: mormsodium glutamate feeding Peaks
toxicity,
feeding behavior, hypophagia,arcuate nucleus lesion, meal size,
Monosodium glutamate (MSG) treatment in early post-natal period is known to induce profound neuroanatomical disturbances in rodents. Neuronal loss following MSG are localized in the circumventricular organs (1, 2, 3, 4) such as the arcuate nucleus (3, 4). A general loss of neuromodulators is observed in the medio-lasal hypothalamus (1, 2, 3, 5). These disturbances are associated with feeding behaviour perturbations (1, 5, 6, 7, 8, 9, 10, 1 l), such as diminished food intake (5,7, 9, 10, 1 l), persistent preference for carbohydrates (7, 10) and an alterated response to some orexigenic agents (5, 9). Actually, little is known about the feeding pattern and feeding rhythms that underly the hypophagia of the MSG treated rats. Their investigation might help us to understand the eating behavior disturbances induced by arcuate nucleus lesions. Methods Experimental animals
Neonatal Long-Evans rats were injected subcutaneously with 4 g/kg body weight monosodium glutamate (Sigma, LaVerpilliere, France) or with saline at days 3, 5 and 7. Injections were counterbalanced among litters and between sex. At weaning, female rats were discarded from the 1Dr B. Beck. INSERM U308 M.R.C.A. 38, me Lionnois. 54000 Nancy. France. Phone: (3) 83 36 41 45. Fax: (3) 83 37 62 44. e-mail: [email protected]
Feediog Behavior cd MSG Treakd Rats
Vol. 62, No. 23, I998
study and male rats were fed on standard rat chow diet. Six months later, control and MSG rats were randomly taken among saline and MSG-treated animals, respectively. Feeding pattern experiment
Sixteen rats (eight controls and eight MSG-treated) were housed in single wire cages with food and water ad lib&urn in an air-conditioned room with a 12 hr light : 12 hr dark cycle. They were fed on a well-balanced diet supplying 54 % of energy from carbohydrate and 30 % from fat (12). The diet was mixed with water (40/50 (w/w)) in order to obtain a paste usable in our automatic feeding system. Behavioral record
Each cage included a complete automatic feeding system with food delivery under the animal’s control (FeedBAC (13, 14)). Briefly, a plastic syringe containing the pasty diet is connected to infrared light diodes which control its delivery with a step-to-step motor. Data from the feeding system are registered in a computer for treatment and analysis in a second step. Behavioral recording was performed during multiple periods of 24 hours. Before this period, an intra-animal reliability calculation was performed during a minimum of three times 24 hours in order to ensure that the animals were well adapted to the feeding system. Design and analysis
A meal was defined as the ingestion of a minimum of 1.2 g of the pasty diet followed by at least 10 minutes during which no feeding occured. The following dependent variables and derived measures were evaluated : meal number, meal size (g), total amount of food eaten (g), meal duration (min), eating rate (g/min), meal intervals (min) and time spent eating (min) were calculated as the mean of each variable for each animal. The following analyses were performed on the values registered during the 24 h period : light on, light off and entire periods feeding pattern analyses. Feeding rhythms experiment
Data from the feeding systems were registered into a computer for analysis. Behavioral recording was performed during ten consecutive days with the same animals. Each two hours, food intake (in g) was compiled for each animal and means were calculated for the control and MSG-treated groups. Data from the ten days were matched together in a single 24 hours period. Statistical analyses All the analyses were conducted
following dependent or repeated measures experimental design. Gaussian-distributed variables were analyzed with Fischer two-way analysis of variance for repeated or independent measures and compared one to the other with Student’s paired-t test. In case of nonGaussian variables (i.e. discrete values) or of unequal variances, they were analyzed with Friedman’s two-way analysis of variance for dependent variables and compared one to the other with paired Wiicoxon T test. Only probability values less than 0.05 (two-tailed) were taken into account, Non-discrete variables are presented as mean corrected by standard error of the mean. Discrete variables are presented as median. Results Experimental animals
At the beginning of the behavioral experiment. MSG-treated rats weighed less than control rats (346 + 7 vs 376 + 9 g, p
Vol. 62, No. ?3,1998
Feeding Behavior of MSG Treated Rats
2129
Feeding pattern experiment
Table I: Feeding patterns of MSG and control rats during the light/dark cycle. * p
LIGHT
127k 11 28.5 + 1.3
213 + 28 9.1 + 0.5
2.:: 2.4 * 25.0 + 0.72 +
3.7: 2.3 f 6.5 + 0.61 +
0 5 0:2 0.5 0.03
05 0.i 0.1 0.04
DARK
83kO4 19.5 f 1.1 7 2.5 f 0.2 2.4 f 0.2 18.4 + 0.2 0.70 * 0.07
122 If:12 36.1 k 1.4**
146 f 14** 10.5 * 0.4 *
86f06 25.6 k 1.2**
3.8’: 1.5 f 19.0 + 0.45 f
2.4: 03 * 1.5 + 0.; * 6.9 k O.l* 0.53 Z!c0.04
3.4: 1.5 * 12.1 f 0.45 +
0 3* o.i** 0.6** 0.05**
0.3* 0.1** 0.3** 0.05*>
Feeding behavior during the 24 hours period MSG-treated animals ate less than control (- 24 % ; t = 7.68 ; p < 0.001). This diminution in food intake was associated with an increase in time spent eating (+ 29 % ; t = 4.29 ; p < 0.001) and in meal duration (+ 52 % ; t = 2.23 ; p < 0.05). Meal size decreased (- 37 % ; t = 3.18 ; p< 0.01) and consequently, eating rate decreased (- 37 % ; t = 4.63 ; p < 0.001). Feeding behavior during the light period MSG-treated animals ate slightly more than control (+ 5 % ; t = 2.82 ; p < 0.05). During this period, MSG treated animals spent more time to eat ( t = 2.18 ; p c 0.05), ate less during meals (t = 2.52 ; p < 0.05) and their meals were shorter (t = 2.22 ; p < 0.05) than control animals. Feeding behavior during the dark period MSG-treated animals ate less than control (- 34 % ; t = 17.4 ; p c 0.0001). During this period, MSG-treated animals spent much more time to eat (t = 3.74 ; p < O.Ol), ate less during meals (t = 4.02 ; p < O.Ol), but their meals were longer (t = 2.49 ; p c 0.05) than control animals. Consequently, eating rate decreased (t = 2.90 ; p c 0.05). Feeding rhythms
Control animals Results are shown in figure 1. Average food intake ranged from 0 to 6 grams per 2 hours. During the light period, two intervals without meals occured at its middle and its end. At the beginning of dark period, food intake reached a peak (5.1 f 0.8 g/2 h (12 h - 14 h) vs 0.6 + 0.4 g/2 h (14 h -16 h) ; t = 5.03 ; p < 0.001). At its end, another peak occured (4.1 ?I 0.3 gl2h (22 h - 24 h) vs 0.9 f 0.4 gI2h (20 h - 22 h) ; t = 6.40 ; p < 0.001). These two peaks were not statistically different (t = 1.17 ; NS).
Fading Behaviorof MN3 Treated Rats
2l30
:. :::.
FOOD INTAKE
(g)
6
Vol. 62, No. 23,1!998
.._....._..~.~.~,~,~.~,~.~ ::_~_~_~.~.~_~,~.~.~,‘_~.‘.~ iiilililiiliiia~~~~~~~~~~~~~~~~~~~~~~ :. .. .‘_ .‘.‘.~.‘.‘.-.‘.~.‘.‘.’ . . . . .. . .::;,: ~.~_~_~.~.~.~.~,~,~,~,‘.~.‘.~.‘.’.~.’.~.’,’, .._,........................ ..:.. ..~_~_~.‘_‘_‘.‘_‘_‘_‘_~.~.~.‘.‘.~.~.~.’.’.-.~.’.‘.‘. . . . . .“““‘... . . . . . . . . ....‘...“.. ..... ..... _~_~,~,~.~.~_‘_‘_~.~_~_‘_‘.~.~,~.~.~.~.~.~ ,~,~,‘,‘,~.~_‘_‘_‘.~.‘.‘.‘,‘,‘,‘.’.’.’.’.’. .~.~_~.‘.~_~_~_~_~.~.~.‘.‘.~.‘.‘.’.’.’.’.’ .~.~_~.~,~,~.~,~.~.~.~.‘.‘.~.‘.~.’.’.’,’,’ .~.~.~.~.~.~.~.~.~_..~.~.~.~.~.~.~.’.’.’.’.~
:.:.:.....
5 4
0
4
8
12
16
20
24
28
32
36
HOURS Fig I: Distribution of feeding episodes at two hours cumulated data of ten consecutive days)
intervals
in the control
rats (n=8
;
MSG-treated animals Results are shown in figure 2. Average food intake ranged from 0 to 3 grams per 2 hours. During the light period, no intervals without meals occured and during the dark period, no peaks occured. Food intake of MSG rats at the beginning and the end of this period was significantly smaller than control rats (0.5 f 0.3 g/2 h (MSG) vs 5.1 f 0.8 g/2 h (C) (12 h -14 h) : t = 5.38, p < 0.001 and 2.1 + 0.2 g/2 h (MSG) vs 4. I f 0.3 g/2 h (C) (22 h - 24 h) ; t = 5.54 ; p < 0.001).
0
8
12
16
20
24
28
32
36
HOURS Fig 2: Distribution of feeding episodes at two hours intervals in the MSG-treated cumulated data of ten consecutive days).
rats (n=8 ;
Vol. 62, No. 23, 19!98
Feeding Behavior of MSG Treated Rats
2131
Discussion The present experiment showed that MSG-treated rats are characterized by disturbances in feeding pattern and feeding rhythms, and extends the analyses of its behavioral syndrome. MSG rats ate less than normal rats mainly by decreasing meal size and eating rate. The feeding pattern disturbances of MSG rats mainly originates in its profound perturbations in the dark phase. This agrees with previous experiments recording gross food intake only (1, 5,6, 8, 10). Meal size was the most affected parameter of the microstructure of the feeding behavior in MSG rats. There are evidence that the triggering of this parameter is not under circadian control, but only depends on the energy regulation level (15, 16, 17, 18, 19). This hypothesis is supported by experiments showing that in case of lower ambient temperature (16), diabetes (20), and refeeding (21), rats eat more food by increasing meal size and not meal frequency. The same phenomenon occurs in the obese Zucker rats with a meal size around 3 g/meal whereas in the lean rats it is less than 2 g/meal (14, 22). In these two cases, meal sizes are not affected by circadian variations (15). In the present experiment, circadian variations in meal sizes were not found either in control or in MSG rats according to this hypothesis. On the other hand, the other parameters of the microstructure were affected differently during the circadian cycle. Disappearence of the two feeding peaks, which normally occur in rats (16, 23, 24), indicates a disruption of the circadian control of feeding in MSG rats, as already described (11). This might be a component of a more general disruption of circadian rhythmicity, as demonstrated by changes in the sleep-wake cycle and locomotor activity (25,26). Among brain transmitters systems that play a role in food intake and circadian control altogether, and might support the behavioral syndrome of the MSG rats, the intrahypothalamic arcuateparaventricular neuropeptide Y (NPY) network is a good candidate. There are major argument in its favor. The first one comes from central injection studies. When continuously infused in cerebral ventricles, NPY induces a disruption of dark/light rhythms of food intake independently of its orexigenic properties (12). Its stimulation properties in the paraventricular nucleus (PVN) vary along the daily cycle, being more effective in the first portion of the dark phase (27). When injected into the suprachiasmatic nucleus, the circadian clock, it phase-shifts the activity of the hamsters (28). The second one comes from biochemical studies which show the existence of daily rhythms of endogenous NPY levels within hypothalamic areas (29). In the parvocellular part of the PVN. it increases at the onset of the dark period, while in the ARC levels reach peaks at the two phase shifts (29). These endogenous peaks are in phase with those observed in food intake in the control animals of this experiment. The third one comes from studies showing that the activity of the hypothalamic NPY-containing pathway is higher in obese Zucker rats (14, 30, 3 1, 32), fasted rats (33, 34, 35,36) and diabetic rats (34, 37); all of which are characterized by having increased meal sizes (15, 20, 2 1). The fourth one comes from studies showing that hypothalamic NPY is depleted in MSG rats (5, 9, 10, 38). Recent evidence also suggest that leptin, the ob gene product, might contribute in association with NPY to the day-night feeding alterations observed in MSG-treated rats (39). Therefore the absence in amplitude variations in food intake during the light/dark cycle are associated with reduced meal size and increased meal duration in the MSG rats. These results describe the profound eating behavior disturbances that are induced by monosodium glutamate-induced lesion of the arcuate nucleus. References 1. R. DAWSON and J. F. LORDEN. J. Comp. Physiol. Psycho]. 95 71-84 (1981). 2. N. LEMKEY-JOHNSTON and W.A. REYNOLDS. J. Neuropathol. Exp. Neurol. 33 74-97 (1974). 3. B. MEISTER, S. CECCATELLI, T. HOKFELT, N.E. ANDEN, M. ANDEN and E. THEODORSSON. Exp. Brain Res. 76 343-368 (1989). 4. J. W. OLNEY Science. 164 7 19-721 ( 1969). 5. R. DAWSON, D.R. WALLACE and S.M. GABRIEL. Pharmacol. Biochem. Behav. 32 391398 ( 1989).
2132
6.
Feeding Behavior of MSG Treated Rats
Vol. 62, No. 23.1998
R. DAWSON and Z. ANNAU. Neurobehav. Toxicol. Teratol. 5 399-406 (1983). R. B. KANAREK, J. MEYERS. R. MADE and J. MAYER. Pharmacol. Biochem. Behav. 10 717-721 (1979). 8. J. F. LORDEN and A. CAUDLE. Neurobehav. Toxicol. Teratol.. 8 509-519 (1986). 9. A. STRICKER-KRONGRAD, C. BURLET and B. BECK. Eur J Pharmacol. 295 27-34 (1996). IO. B. BECK, A. STRICKER-KRONGRAD. A. BURLET, N. MUSSE. J.P. NICOLAS and C. BURLET. Neurosci. Letters 225 153- 156 ( 1997). I I W. J. RIETVELD. J.H. MEIJER. J. RUIS and P. BUYS. J. Interdiscipl. Cycle Res. 17 327334 ( 1986). 12. B. BECK. A. STRICKER-KRONGRAD. J.P. NICOLAS and C. BURLET. Ann. N.Y. Acad. Sci.’ 611 49 l-494 ( 1990). 13. A. STRICKER-KRONGRAD, B. BECK. J.P. NICOLAS and C. BURLET. Pharmacol. Biochem. Behav. 43 881-886 (1992). 14 A. STRICKER-KRONGRAD, J.P. MAX, N. MUSSE, J.P. NICOLAS, C. BURLET and B. BECK. Brain Res. 660 162-I 66 (1994). A. DE JONG-NAGELSMIT, J. KEIJSER and J. H. STRUBBE. I5 A. B. ALINGH-PRINS, Physiol. Behav. 38 423-426 (1986). 16 S. ARMSTRONG Neurosci. Biobehav. 4 27-53 (1980). 17 R.F. DAVIES J. Comp. Physiol. Psychol. 91 91 I-916 (1982). J. KEIJSER and J.H. STRUBBE. Physiol. Behav. 18 N.J. SPITERI, A.J. ALINGH-PRINZ, 25 775-777 (1980. Physiol Behav. 19 J.H. STRUBBE. J. KEIJSER. T. DIJKSTRA and A.J. ALINGH-PRINZ. 36 489-493 ( 1986). 20 D.W. THOMAS, E. SCHARRER and J. MAYER. Physiol. Behav. 17 345-349 (1976). 21 D.A. LEVITSKY Physiol. Behav. 5 291-300 ( 1970). 22. E.E. BECKER and J.A. GRINKER. Physiol. Behav. 18 685-692 (1977). 23. J. LE MAGNEN and S. TALLON. J. Physiol. Paris. 58 323-349 (1966). 24. P.S. SIEGAL J. Comp. Physiol. Psychol. 54 294-301 (1961). 25. M. OLIVO, K. KITAHAMA, J.L. VALATX and M. JOUVET. Neurosci. Lett. 67 186-190 (1986). 26. S. MIYABO, I. YAMAMURA, E. OOYA. N. AOYAGI, Y. HORIKAWA and S. HAYASHI. Brain Res. 339 20 I-208. 27. B.J. STANLEY and W.J. THOMAS. Peptides. 14 475-481 (1986). 28. H.E. ALBERS and C.F. FERRIS. Neurosci. Lett. 50 163-168 (1984). B. BECK. C. BURLET and S.F. LEIBOWITZ. Brain. Res. 536 29. M. JHANWAR-UNIYAL, 33 l-334 (1990). 30. B. BECK, A. BURLET, R. BAZIN. J.P. NICOLAS and C. BURLET. J. Nutr. 123 11681172 (1993). 31. H.D. MC CARTHY. P.E. MC KIBBIN. B. HOLLOWAY, R. MAYERS and G. WILLIAMS. Life Sci. 49 1491-1497 (1991). 32. G. SANACORA, M. KERSHAW, J.A. FINKELSTEIN and J.D. WHITE. Endocrinology 127 730-737 (I 990). 33. B. BECK, M. JHANWAR-UNIYAL, A. BIJRLET, J.P. NICOLAS and C. BURLET. Brain Res. 528 245-249 (1990). 34. H.M. FRANKISH, D. MCCARTHY. S. DRYDEN, A. KILPATRICK and G. WILLIAMS. Peptides. 14 94 I-948 ( 1993). 35. A. SAHU, P.S. KALRA and S.P. KALRA. Peptides 9 83-86 (1988). 36. J.D. WHITE and M. KERSHAW. Mol. Cell Neurosci. 141-48 (1990). 37. G. WILLIAMS. Y. LEE, H.M. CARDOSO. B.E. OKPERE and S.R. BLOOM. Diabetes. 3 8 321-327 (1989). 38. M. ABE, M. SAITO and T. SHIMUZU. Brain Rea. Bull. 24 289-291 (1990). 39. R. DAWSON, M.A. PELLEYMOUNTER, W.J. MILLARD, S. LIU and B. EPPLER. Am. J. Physiol. 36 E202-E206. 7.