Physiology&Behavior, Vol. 49, pp. 1013-1018. ©Pergamon Press plc, 1991. Printed in the U.S.A.
0031-9384/91 $3.00 + .00
Effects on Metabolic and Hormonal Parameters of Monosodium Glutamate (Umami Taste) Ingestion in the Rat C. V I A R O U G E , P. E V E N , C. R O U G E O T * A N D S. N I C O L A I D I S 1
Laboratoire de Neurobiologie des Rdgulations, C.N.R.S. UA 637 Collige de france, 11 Place Marcelin Berthelot 75231 Paris Cedex 05, France *Laboratoire de Pharmacologie des Mddiateurs Chimiques INSERM U 207 Institut Pasteur, 28 rue du Docteur Roux 75015 Paris, France
VIAROUGE, C., P. EVEN, C. ROUGEOT AND S. NICOLAJ~IS. Effects on metabolicand hormonalparameters of monosodium glutamate (umami taste) ingestion in the rat. PHYSIOL BEHAV 49(5) 1013-1018, 1991.--Umami taste appears to signal, at the gustatory level, the intake of proteins, therefore the working hypothesis was: does umami taste of a monosodiurn glutamate (MSG) solution elicit changes in both glucagon and insulin release, similar to those elicited by amino acids, and consequently, changes in plasma glucose and in overall cellular metabolism? In a first experiment, rats were equipped with indwelling jugular and oral catheter and serial samplings were made in the free moving, undisturbed rat before and after an oral or IV infusion of MSG (0.05 M). None of the plasma parameters showed any significant response. In a second experiment, energy expenditure was monitored by means of an original computer-based calorimeter capable of calculating, besides the classical parameters, resting metabolism in a moving animal (designated by background metabolism). The addition of MSG to a low calorie, low-protein meal did not modify background metabolism or respiratory quotient. Therefore MSG ingestion does not by itself affect plasma levels of hormones of glucose and protein metabolism, total metabolism rate, or nutrient utilization. However, examination of individual data and those from a pilot experiment for future work suggests that MSG becomes an efficient metabolic effector if added to a caloric diet, and so enhances proper thermogenesis of macronutrients. Monosodium glutamate Respiratory quotient
Umami Metabolism
Insulin Glucagon Glycemia Indirect calorimetry Thermic effect of feeding Anticipatory oro-vegetative reflexes
ACCORDING to our traditional thinking, nutritional substrates in the internal milieu act upon their specific effectors in order to produce the appropriate endocrine responses and, as a result, their proper cellular utilization. This action obeys mainly the law of "mass action." For example, the more plasma glucose the more insulin release. However, before reaching the internal milieu, nutrients have been in contact with the oral and gastrointestinal walls. This obligatory passage is not neutral as far as the endocrine and metabolic processing is concerned. The role of sensory afferents in endocrine and metabolic processing of ingestion was first recognized in 1963 and in 1969 in relation to water, NaC1 (11) and glucose (10) ingestion. It was shown that specific neuronal messages from the gustatory and gastrointestinal system reach hypothalamic structures (10,14) before triggering endocrine responses able to change the metabolic state in an anticipatory way (anticipatory in the sense that they will precede the postahsorptive responses that the same nutrients will evoke). The anticipatory responses were shown to be sensory specific, i.e., glucose can be replaced by saccharin and water elicits a re-
action opposite to salt. More recently, specific gastrointestinal receptors were shown to sense various amino acids (17). These anticipatory reflexes are often referred to as cephalic responses although they can be triggered not only from cephalic but also from gastrointestinal sensors. Monosodium glutamate (MSG) associated with other nutrients enhances their hedonic impact and modifies their acceptability (8). On the other hand, these compounds have been suggested to be the sensory message alerting the organism to the arrival of amino acids (16). Amino acids require appropriate pancreatic secretions in order to be properly processed. Unlike glucose, amino acids elicit a double ct and 13 islet activation, i.e., a concomitant glucagon and insulin secretion (8 cell, somatostatin activation was not examined in this work). Given the above notions, we asked, in the first experiment, the following questions: Is oral sensing of MSG followed by changes in circulating glucose, insulin, and glucagon concentrations similar to those seen after protein ingestion? Are the possible changes in circulating glucose, insulin, and glucagon specifically
IRe,quests for reprints shouM be addressed to Dr. S. Nicolai'dis, Laboratoire de Neurobiologie des Rdgulations, Coll6ge de fiance, 11 Place Marcelin Berthelot, 75231 Paris Cedex 05, France.
1013
1014
VIAROUGE ET AL
due to the sensory effect of MSG or are they also produced by systemic action of MSG? In the second experiment, we investigated the action of MSG at the oral level on overall metabolism. To achieve these measurements, we used an original calorimeter (3) which allows instantaneous measurement of the usual metabolic parameters and, in addition, the resting metabolism in a freely moving animal: this is what we call the background metabolism which is the part of total metabolism from which the cost of locomotor activity has been subtracted. Respiratory quotient (CO2 release/O2 consumption of the rat) was also monitored in order to assess possible changes in relational carbohydrates versus protein and lipid utilization.
1"1-1
METHOD
Experiment 1: Effect of MSG on Circulating Metabolic Factors Animals and surgical procedures. To assess the changes in plasma levels of glucose, glucagon, and insulin induced when MSG is applied directly (intravenously: IV) in the inner milieu, 6 male Wistar rats weighing approximately 300 g were prepared with indwelling intrajugular catheters according to a previously described technique (12). The tubing (silastic o.d. 0.037 inch) was inserted into the right external jugular vein down to the atrium. The same tubing allowed for blood sampling. To assess the changes of plasma level of glucose, glucagon, and insulin induced when MSG was applied in the oral cavity (intraoral: IO), 6 other rats were prepared the same way as above, and intraoral fistula were also implanted (15) in order to produce the oral stimulation with MSG. Blood-sampling procedure. After recovery of the animals (body weight back to control levels) the catheters were attached to a pump capable of injecting or withdrawing blood samples via a two-channel watertight swivel. The swivel was held on a counterweighted boom to ensure complete freedom of movement and feeding. Chronically implanted rats were deprived of food from 10:00 a.m. to 2:00 p.m. At 2:00 p.m., a blood sample of 0.4 ml was drawn before the MSG injections. Then, either MSG (Ajinomoto, Japan) (0.5 ml, 0.05 M) or vehicle (distilled water) was injected via either the IV or oral catheter. Then 3 other samples of 0.4 ml were withdrawn 5, 15 and 30 min posttreatment. Samples were immediately centrifuged at 4°C and the plasma stored in a freezer. The concentration of 0.05 M was chosen for MSG as it had been previously reported to be the preference peak for this substance in the rat (9). Pure MSG was dissolved in distilled water. The control stimulation used the water vehicle. Responses to water and MSG were compared at both the oral and the IV level. Chemical determinations. Plasma glucose levels were measured using a sample of 0.03 ml of plasma with an automatic glucose analyzer (Glucose oxydase method, Yellow Springs Instruments, model S3A). The remaining plasma was stored at -20°C for hormonal assays. Plasma glucagon levels were measured by an immunoassay Amersham kit (N1601). Plasma insulin level evaluation used antiserum to insulin prepared from guinea pigs by the Institut Pasteur, Paris. The radioiodination of insulin was performed according to a previously described method (7). Blood cells resuspended in saline were given back to the rat at the subsequent sampling. Statistical analysis. All results are presented as mean -*-STDV. Student's paired t-test was used to compare the results between vehicle and MSG injection for each time after treatment. Experiment 2: Effect of MSG on Energy Expenditure Experimental device. Continuous measurement of energy ex-
$IIt^I~ / /
@ k/ FIG. 1. The metabolic device. penditure from which the portion due to locomotor activity can be separately computed (background metabolism) was achieved using an open circuit calorimeter (3) as shown in Fig. 1. The device consisted of a small cage (9 liters), to minimize dead space, and the animal platform rested on three dynamic force transducers so that the work produced on the platform by the animal could be quantitatively measured. The food cup was independently weighed and the temperature in the cage was closely controlled by a powerful feedback-controlled thermoregulatory system. The data from the analyzers and monitors, i.e., oxygen consumption, carbon dioxide release, air flow, food intake, locomotor activity and temperature, were recorded at 10 second bins, then treated on line by a computer. Subsequently, respiratory quotient and background metabolism could be further computed from the original data. Background metabolism computation, particularly within small bins (10 s), needed precise and quantitative measurement of the rate of locomotor effort, the estimation of the corresponding cost, and the subtraction of this cost from the total metabolism. The difficulty came from the fact that the dynamic force transducers monitored activity instantly whereas respiratory gas changes followed a complex and necessarily delayed path, through the cage and the tubing, before being monitored. Therefore, these two desynchronized magnitudes had to be matched. To match delayed and nondelayed parameters, a particular procedure was used. This was recently published in detail (6). Briefly, it consisted in modeling the deformation induced by the
UMAMI TASTE AND METABOLISM
1015
I Io Injection I
100 90
30
80
20
A
(5 T--
70
A
I:
60 "-
WATER
] IO3
10 0
50
E
MSG
z --I
I IV Injection I
m 100 0 o ~--1
I10 Injection I
40
z_. 90
35.
WATER
25
1
15 80
5 70
-10
i
!
i
0
10
20
-10
FIG. 2. Changes in blood glucose levels before and after IO (upper) or IV (lower) injection (time: 0 min) of 0.5 ml MSG (0.05 M) (filled squares) or vehicle (white squares). Data are mean---STDV.
aerodynamics of the system measuring the animal's gas exchanges and in matching the outcome of the model with the instantaneously monitored locomotor profile. In this way, the equation of diffusion of 02 and CO2 in the calorimeter was adjusted and the correction for this diffusion was made by using a processing of digital filtering of the data developed by Kalman. Finally, both parameters, i.e., locomotion and gas exchanges, were modeled as if they were analyzed at the level of the nostrils of the rat. Animals and housing. Experiments were carried out on adult male Wistar rats (230-300 g). Before the experimental procedure, the animals were housed in individual wire cages of the same shape as the experimental cage. These cages were located in a temperature-controlled enclosure in which the air was confined and maintained at 26°(2 in order to obtain the same conditions as in the experimental cage. Lights were on from 7:00 a.m. to 7:00 p.m. Standard chow (Extralabo M25) and water were available ad lib except during the test periods. Experimental meals. Control meals consisted of a gel obtained by dissolving 40 g of alimentary gelatin powder (Proscience, France) in one liter of tap water. To obtain test meals we just added MSG 0.05 M (i.e., 2 g/l) to the control meals. These meals provided 0.16 kcal/g. In order to eliminate neophobia, samples of these meals were presented twice to the rats, a few days before the experiment. Experimental protocol. The food was removed at 9:30 a.m. At least two hours before the test meal presentation (between 4:00 and 5:00 p.m.), the rat was transferred from its mock cage to the experimental one; water was provided ad lib but no food was offered. Between 4:00 and 5:00 p.m., 6 g of gel were introduced in the food cup through a thin tube without opening the cage. The experiment lasted until the next day at 9:00 a.m. Twelve experiments were performed in 3 different subjects and a total of 12 control meals and 9 test meals were recorded.
I
i
i
10
20
30
TIME (min)
30
TIME (rain)
i
0
FIG. 3. Changes in blood insulin levels before and after IO (upper) or IV (lower) injection (time: 0 rain) of 0.5 rrd MSG (0.05 M) (filled squares) or vehicle (white squares). Data are mean-+ STDV. *p
0031-9384/91 $3.00 + .00
Effects on Metabolic and Hormonal Parameters of Monosodium Glutamate (Umami Taste) Ingestion in the Rat C. V I A R O U G E , P. E V E N , C. R O U G E O T * A N D S. N I C O L A I D I S 1
Laboratoire de Neurobiologie des Rdgulations, C.N.R.S. UA 637 Collige de france, 11 Place Marcelin Berthelot 75231 Paris Cedex 05, France *Laboratoire de Pharmacologie des Mddiateurs Chimiques INSERM U 207 Institut Pasteur, 28 rue du Docteur Roux 75015 Paris, France
VIAROUGE, C., P. EVEN, C. ROUGEOT AND S. NICOLAJ~IS. Effects on metabolicand hormonalparameters of monosodium glutamate (umami taste) ingestion in the rat. PHYSIOL BEHAV 49(5) 1013-1018, 1991.--Umami taste appears to signal, at the gustatory level, the intake of proteins, therefore the working hypothesis was: does umami taste of a monosodiurn glutamate (MSG) solution elicit changes in both glucagon and insulin release, similar to those elicited by amino acids, and consequently, changes in plasma glucose and in overall cellular metabolism? In a first experiment, rats were equipped with indwelling jugular and oral catheter and serial samplings were made in the free moving, undisturbed rat before and after an oral or IV infusion of MSG (0.05 M). None of the plasma parameters showed any significant response. In a second experiment, energy expenditure was monitored by means of an original computer-based calorimeter capable of calculating, besides the classical parameters, resting metabolism in a moving animal (designated by background metabolism). The addition of MSG to a low calorie, low-protein meal did not modify background metabolism or respiratory quotient. Therefore MSG ingestion does not by itself affect plasma levels of hormones of glucose and protein metabolism, total metabolism rate, or nutrient utilization. However, examination of individual data and those from a pilot experiment for future work suggests that MSG becomes an efficient metabolic effector if added to a caloric diet, and so enhances proper thermogenesis of macronutrients. Monosodium glutamate Respiratory quotient
Umami Metabolism
Insulin Glucagon Glycemia Indirect calorimetry Thermic effect of feeding Anticipatory oro-vegetative reflexes
ACCORDING to our traditional thinking, nutritional substrates in the internal milieu act upon their specific effectors in order to produce the appropriate endocrine responses and, as a result, their proper cellular utilization. This action obeys mainly the law of "mass action." For example, the more plasma glucose the more insulin release. However, before reaching the internal milieu, nutrients have been in contact with the oral and gastrointestinal walls. This obligatory passage is not neutral as far as the endocrine and metabolic processing is concerned. The role of sensory afferents in endocrine and metabolic processing of ingestion was first recognized in 1963 and in 1969 in relation to water, NaC1 (11) and glucose (10) ingestion. It was shown that specific neuronal messages from the gustatory and gastrointestinal system reach hypothalamic structures (10,14) before triggering endocrine responses able to change the metabolic state in an anticipatory way (anticipatory in the sense that they will precede the postahsorptive responses that the same nutrients will evoke). The anticipatory responses were shown to be sensory specific, i.e., glucose can be replaced by saccharin and water elicits a re-
action opposite to salt. More recently, specific gastrointestinal receptors were shown to sense various amino acids (17). These anticipatory reflexes are often referred to as cephalic responses although they can be triggered not only from cephalic but also from gastrointestinal sensors. Monosodium glutamate (MSG) associated with other nutrients enhances their hedonic impact and modifies their acceptability (8). On the other hand, these compounds have been suggested to be the sensory message alerting the organism to the arrival of amino acids (16). Amino acids require appropriate pancreatic secretions in order to be properly processed. Unlike glucose, amino acids elicit a double ct and 13 islet activation, i.e., a concomitant glucagon and insulin secretion (8 cell, somatostatin activation was not examined in this work). Given the above notions, we asked, in the first experiment, the following questions: Is oral sensing of MSG followed by changes in circulating glucose, insulin, and glucagon concentrations similar to those seen after protein ingestion? Are the possible changes in circulating glucose, insulin, and glucagon specifically
IRe,quests for reprints shouM be addressed to Dr. S. Nicolai'dis, Laboratoire de Neurobiologie des Rdgulations, Coll6ge de fiance, 11 Place Marcelin Berthelot, 75231 Paris Cedex 05, France.
1013
1014
VIAROUGE ET AL
due to the sensory effect of MSG or are they also produced by systemic action of MSG? In the second experiment, we investigated the action of MSG at the oral level on overall metabolism. To achieve these measurements, we used an original calorimeter (3) which allows instantaneous measurement of the usual metabolic parameters and, in addition, the resting metabolism in a freely moving animal: this is what we call the background metabolism which is the part of total metabolism from which the cost of locomotor activity has been subtracted. Respiratory quotient (CO2 release/O2 consumption of the rat) was also monitored in order to assess possible changes in relational carbohydrates versus protein and lipid utilization.
1"1-1
METHOD
Experiment 1: Effect of MSG on Circulating Metabolic Factors Animals and surgical procedures. To assess the changes in plasma levels of glucose, glucagon, and insulin induced when MSG is applied directly (intravenously: IV) in the inner milieu, 6 male Wistar rats weighing approximately 300 g were prepared with indwelling intrajugular catheters according to a previously described technique (12). The tubing (silastic o.d. 0.037 inch) was inserted into the right external jugular vein down to the atrium. The same tubing allowed for blood sampling. To assess the changes of plasma level of glucose, glucagon, and insulin induced when MSG was applied in the oral cavity (intraoral: IO), 6 other rats were prepared the same way as above, and intraoral fistula were also implanted (15) in order to produce the oral stimulation with MSG. Blood-sampling procedure. After recovery of the animals (body weight back to control levels) the catheters were attached to a pump capable of injecting or withdrawing blood samples via a two-channel watertight swivel. The swivel was held on a counterweighted boom to ensure complete freedom of movement and feeding. Chronically implanted rats were deprived of food from 10:00 a.m. to 2:00 p.m. At 2:00 p.m., a blood sample of 0.4 ml was drawn before the MSG injections. Then, either MSG (Ajinomoto, Japan) (0.5 ml, 0.05 M) or vehicle (distilled water) was injected via either the IV or oral catheter. Then 3 other samples of 0.4 ml were withdrawn 5, 15 and 30 min posttreatment. Samples were immediately centrifuged at 4°C and the plasma stored in a freezer. The concentration of 0.05 M was chosen for MSG as it had been previously reported to be the preference peak for this substance in the rat (9). Pure MSG was dissolved in distilled water. The control stimulation used the water vehicle. Responses to water and MSG were compared at both the oral and the IV level. Chemical determinations. Plasma glucose levels were measured using a sample of 0.03 ml of plasma with an automatic glucose analyzer (Glucose oxydase method, Yellow Springs Instruments, model S3A). The remaining plasma was stored at -20°C for hormonal assays. Plasma glucagon levels were measured by an immunoassay Amersham kit (N1601). Plasma insulin level evaluation used antiserum to insulin prepared from guinea pigs by the Institut Pasteur, Paris. The radioiodination of insulin was performed according to a previously described method (7). Blood cells resuspended in saline were given back to the rat at the subsequent sampling. Statistical analysis. All results are presented as mean -*-STDV. Student's paired t-test was used to compare the results between vehicle and MSG injection for each time after treatment. Experiment 2: Effect of MSG on Energy Expenditure Experimental device. Continuous measurement of energy ex-
$IIt^I~ / /
@ k/ FIG. 1. The metabolic device. penditure from which the portion due to locomotor activity can be separately computed (background metabolism) was achieved using an open circuit calorimeter (3) as shown in Fig. 1. The device consisted of a small cage (9 liters), to minimize dead space, and the animal platform rested on three dynamic force transducers so that the work produced on the platform by the animal could be quantitatively measured. The food cup was independently weighed and the temperature in the cage was closely controlled by a powerful feedback-controlled thermoregulatory system. The data from the analyzers and monitors, i.e., oxygen consumption, carbon dioxide release, air flow, food intake, locomotor activity and temperature, were recorded at 10 second bins, then treated on line by a computer. Subsequently, respiratory quotient and background metabolism could be further computed from the original data. Background metabolism computation, particularly within small bins (10 s), needed precise and quantitative measurement of the rate of locomotor effort, the estimation of the corresponding cost, and the subtraction of this cost from the total metabolism. The difficulty came from the fact that the dynamic force transducers monitored activity instantly whereas respiratory gas changes followed a complex and necessarily delayed path, through the cage and the tubing, before being monitored. Therefore, these two desynchronized magnitudes had to be matched. To match delayed and nondelayed parameters, a particular procedure was used. This was recently published in detail (6). Briefly, it consisted in modeling the deformation induced by the
UMAMI TASTE AND METABOLISM
1015
I Io Injection I
100 90
30
80
20
A
(5 T--
70
A
I:
60 "-
WATER
] IO3
10 0
50
E
MSG
z --I
I IV Injection I
m 100 0 o ~--1
I10 Injection I
40
z_. 90
35.
WATER
25
1
15 80
5 70
-10
i
!
i
0
10
20
-10
FIG. 2. Changes in blood glucose levels before and after IO (upper) or IV (lower) injection (time: 0 min) of 0.5 ml MSG (0.05 M) (filled squares) or vehicle (white squares). Data are mean---STDV.
aerodynamics of the system measuring the animal's gas exchanges and in matching the outcome of the model with the instantaneously monitored locomotor profile. In this way, the equation of diffusion of 02 and CO2 in the calorimeter was adjusted and the correction for this diffusion was made by using a processing of digital filtering of the data developed by Kalman. Finally, both parameters, i.e., locomotion and gas exchanges, were modeled as if they were analyzed at the level of the nostrils of the rat. Animals and housing. Experiments were carried out on adult male Wistar rats (230-300 g). Before the experimental procedure, the animals were housed in individual wire cages of the same shape as the experimental cage. These cages were located in a temperature-controlled enclosure in which the air was confined and maintained at 26°(2 in order to obtain the same conditions as in the experimental cage. Lights were on from 7:00 a.m. to 7:00 p.m. Standard chow (Extralabo M25) and water were available ad lib except during the test periods. Experimental meals. Control meals consisted of a gel obtained by dissolving 40 g of alimentary gelatin powder (Proscience, France) in one liter of tap water. To obtain test meals we just added MSG 0.05 M (i.e., 2 g/l) to the control meals. These meals provided 0.16 kcal/g. In order to eliminate neophobia, samples of these meals were presented twice to the rats, a few days before the experiment. Experimental protocol. The food was removed at 9:30 a.m. At least two hours before the test meal presentation (between 4:00 and 5:00 p.m.), the rat was transferred from its mock cage to the experimental one; water was provided ad lib but no food was offered. Between 4:00 and 5:00 p.m., 6 g of gel were introduced in the food cup through a thin tube without opening the cage. The experiment lasted until the next day at 9:00 a.m. Twelve experiments were performed in 3 different subjects and a total of 12 control meals and 9 test meals were recorded.
I
i
i
10
20
30
TIME (min)
30
TIME (rain)
i
0
FIG. 3. Changes in blood insulin levels before and after IO (upper) or IV (lower) injection (time: 0 rain) of 0.5 rrd MSG (0.05 M) (filled squares) or vehicle (white squares). Data are mean-+ STDV. *p
