Hindbrain administration of MK-801, a noncompetitive N-methyl-d-aspartate (NMDA) channel blocker, increases meal size, suggesting NMDA receptors in this location participate in control of food intake. However, dizocilpine (MK-801) reportedly antagonizes some non-NMDA ion channels. Therefore, to further assess hindbrain NMDA receptor participation in food intake control, we measured deprivation-induced intakes of 15% sucrose solution or rat chow after intraperitoneal injection of either saline vehicle or d(-)-2-amino-5-phosphonopentanoic acid (AP5), a competitive NMDA receptor antagonist, to the fourth ventricular, or nucleus of the solitary tract (NTS). Intraperitoneal injection of AP5 (0.05, 0.1, 1.0, 3.0, and 5.0 mg/kg) did not alter 30-min sucrose intake at any dose (10.7 ± 0.4 ml, saline control) (11.0 ± 0.8, 11.2 ± 1.0, 11.2 ± 1.0, 13.1 ± 2.2, and 11.0 ± 1.9 ml, AP5 doses, respectively). Fourth ventricular administration of both 0.2 μg (16.7 ± 0.6 ml) and 0.4 μg (14.9 ± 0.5 ml) but not 0.1 and 0.6 μg of AP5 significantly increased 60-min sucrose intake compared with saline (11.2 ± 0.4 ml). Twenty-four hour chow intake also was increased compared with saline (AP5: 31.5 ± 0.1 g vs. saline: 27.1 ± 0.6 g). Furthermore, rats did not increase intake of 0.2% saccharin after fourth ventricular AP5 administration (AP5: 9.8 ± 0.7ml, vs. saline: 10.5 ± 0.5ml). Finally, NTS AP5 (20 ng/30 nl) significantly increased 30- (AP5: 17.2 ± 0.7 ml vs. saline: 14.6 ± 1.7 ml), and 60-min (AP5: 19.4 ± 0.6 ml vs. saline: 15.5 ± 1.4 ml) sucrose intake, as well as 24-h chow intake (AP5: 31.6 ± 0.3 g vs. saline: 26.1 ± 1.2 g). These results support the hypothesis that hindbrain NMDA receptors participate in control of food intake and suggest that this participation also may contribute to control of body weight over a 24-h period.
- nucleus of the solitary tract
- N-methyl-d-aspartate channel blocker
n-methyl-d-aspartate (NMDA)-type glutamate receptors are expressed in many neurons in the central nervous system (4, 24, 33, 52), as well as by some peripheral neurons, including primary vagal afferents (1) and the intrinsic neurons of the gastrointestinal tract (7, 19, 52). Over the past decade, evidence of NMDA receptor participation in meal termination has steadily accumulated. Studies from our laboratory (6), as well as others (17), have demonstrated that intraperitoneal administration of dizocilpine (MK-801), a noncompetitive antagonist of NMDA receptors, increases intake of both solid chow and palatable food, but not water, in rats after overnight food deprivation. Administration of MK-801 into the fourth ventricle or directly into the caudal medial nucleus of the solitary tract (NTS) increases the size of a 15% sucrose meal (49). Furthermore, lesions of the dorsal vagal complex abolish increased meal size evoked by systemic administration of MK-801 (50). Finally, systemic MK-801 administration increases only nutritive sucrose but not nonnutritive saccharin solution (5), indicating that blockade of NMDA receptors by MK-801 delays meal termination by interfering with nutrient-related negative feedback signals generated in a meal. Collectively, these findings indicate that systemic administration of MK-801 increases food intake by acting within the dorsal hindbrain, an area that receives vagal afferent input from the gut (2, 45). Although these results are consistent with NMDA receptor participation in control of meal termination, MK-801 has been reported to antagonize not only NMDA receptors, but also to interfere with activation of some other ionotropic neurotransmitter receptors, including nicotinic and 5-HT3 receptor channels (35, 51, 54). For example, electrophysiological evidence shows that MK-801 blocks activation of nicotinic ACh ion channels (12, 35) Therefore, it remains conceivable that MK-801 participation in control of food intake involves interactions with non-NMDA receptors.
MK-801 is a noncompetitive NMDA receptor antagonist that inhibits NMDA receptor activation by binding to the phencyclidine recognition site within the open NMDA channel pore. Binding of MK-801 to the NMDA receptor does not directly affect the binding ability of glutamate. In fact, the dissociation rate of bound MK-801 is increased in the presence of glutamate binding (38). Unlike MK-801, competitive NMDA antagonists compete directly with glutamate for binding on the ion channel. Therefore, competitive NMDA receptor antagonists not only are more selective for NMDA receptors than their noncompetitive cogeners, but interpretation of their pharmacological effects is more straightforward.
d-(-)-2-Amino-5-phosphonopentanoic acid (AP5) is a potent and selective competitive NMDA-type glutamate receptor antagonist (20, 53). In this study, we tested the hypothesis that blockade of NMDA receptors by AP5 would increase food intake. To do this, we first measured feeding response of a palatable sucrose solution after systemic intraperitoneal administration of various doses of AP5. Secondly, we examined intakes of both solid (rat chow) and liquid (sucrose solution) foods in response to fourth ventricular administration of AP5. To test the hypothesis that AP5 interferes with feedback signals from nutritive component in a meal, we measured sucrose, as well as saccharin intake, in response to fourth ventricular AP5 administration. Finally, we delivered AP5 directly into the dorsal vagal complex and measured the intakes of chow and sucrose.
MATERIALS AND METHODS
Adult male Sprague-Dawley rats (Harlan, Indianapolis, IN), weighing 225–249 g were housed individually in a temperature-controlled vivarium and adapted to a 12:12-h light-dark cycle (lights on at 0600). Rats had ad libitum access to water and standard rodent chow (Purina 5001), except during experiments or overnight deprivation, as described below. Animal protocols were approved by The Pennsylvania State University Institutional Animal Care and Use Committee.
d-(-)-2-amino-5-phosphonopentanoic acid (AP5; AG Scientific, San Diego, CA) was dissolved in 0.9% sterile saline. For intraperitoneal administration, AP5 was injected in a volume of 1 ml/kg body wt. For central administration, all rats received a total volume of 3 μl into the fourth ventricle or 30 nl into the NTS. Sterile saline (0.9%) served as the vehicle control in all experiments.
Intracerebroventricular Cannula Implantation
After overnight food deprivation, rats anesthetized with a mixture of ketamine (50 mg/kg), xylazine (5 mg/kg), and acepromazine (1 mg/kg) were secured in a stereotaxic frame (David Kopf Instruments, Tujunga, CA) and implanted with an intracerebroventricular cannula. For fourth-ventricle injections, a 23-gauge guide cannula was implanted toward the fourth ventricle, using the following coordinates: 2.0 mm from the occipital crest, 0.0 mm lateral from the midline, and 6.7 mm ventral from the dura level according to Paxinos and Watson (31). For NTS injections, a 26-gauge guide cannula was implanted toward the caudomedial NTS, using the following coordinates: 0.0 mm from the occipital crest, 0.8 mm lateral from the midline, and 7.9 mm ventral from the dura level. Cannulae were secured to the skull by stainless-steel screws and methacrylate cement. All rats were allowed a minimum of 7 days postsurgical recovery period and weighed daily until they returned to presurgical weight (270–280 g) before testing began. The protocol employed here was similar to earlier reports by this laboratory (6).
Feeding responses Following Intraperitoneal Administration of AP5
After overnight (16 h) food deprivation, rats (n = 8) were removed from their home cages and received an intraperitoneal injection of AP5 (0.05, 0.1, 1.0, 3.0, and 5.0 mg/kg) or saline. Fifteen minutes after the injection, calibrated drinking burettes containing 15% sucrose solution (wt/vol) were presented, and intake was measured to the nearest 0.1 ml every 5 min for 30 min. Chow and water were returned after completion of the experiment. Tests were conducted at 48-h intervals. All rats received the same drug treatment on a given testing day bracketed by control, saline injections, with each drug dose tested a minimum of two times.
Feeding Responses After Central Administration of AP5
There are currently no reports examining the feeding effects of AP5 administered into the hindbrain. Therefore, to establish an efficacious dose of AP5, we first conducted a dose-response experiment in which various doses of AP5 were administered into the fourth ventricle. During testing, overnight food-deprived, rats (n = 10) were removed from their home cages, the obturators were removed from the cannulas and a 30-gauge injector connected to a 25-μl Hamilton syringe was inserted into the cannulas. AP5 (0.1, 0.2, 0.4, and 0.6 μg) or saline were infused for 30 s, and the injectors were allowed to remain seated in the cannulas for an additional 1 min. Upon completion of the injection, the injector was removed and the obturator was replaced. Immediately after injections, rats were returned to their home cages and 15% sucrose intake was presented. Intake was measured for 60 min. On the basis of the dose-response data, our next experiment tested the ability of fourth-ventricular AP5 to increase chow intake. Briefly, overnight food-deprived rats (n = 12) received an injection of AP5 (0.2 μg) into the fourth ventricle as described above. A preweighed amount of rat chow was presented immediately after the injection and intake was measured (accounting for spillage) at 0.5, 1, 2, 4, and 24 h. All injections were separated by a minimum of 48 h. A separate group of rats (n = 16) implanted with fourth-ventricular cannula was used in the experiment, in which intake of saccharin after AP5 administration was measured. Rats were randomly divided into two groups and were adapted to ingest either 15% sucrose (n = 8) or 0.2% saccharin solution (n = 8). On the test days, rats received a fourth ventricular administration of either saline or AP5 (0.2 μg) as described above. Immediately after the injection, calibrated burettes filled with 15% sucrose solution were given to sucrose group, while the saccharin group was given a 0.2% saccharin solution. This concentration of saccharin has been shown to be close to “isopreferred” to a 15% sucrose solution (26, 43). Intake was measured for 60 min. To account for changes in baselines, each AP5 treatment was bracketed by several saline injections (∼2–5) until a stabilized baseline was achieved. In all experiments, at least two AP5 tests were conducted.
The final experiment tested the effect of AP5 on sucrose and chow intake when administered directly into the NTS. Food-deprived rats (n = 8) received an injection of AP5 (20 ng in 30 nl) into the caudal medial NTS via a 26-gauge injector connected to a motorized 10-μl microsyringe (World Precision Instruments, Sarasota, FL). Both 60-min sucrose intake, as well as 24-h chow intake, was measured after completion of the injections using the protocol described above. A minimum of three AP5 tests were conducted. All injections were separated by 48 h, during which no experimental manipulations occurred. Each AP5 test was bracketed by a saline control test.
Verification of Cannula Patency and Placement
Cannula placement and patency were verified both functionally and histologically. Previous studies have demonstrated that rats increase their 30-min sucrose intake after fourth ventricular, as well as NTS administration, of MK-801 (49). For functional verification, 15% sucrose intake was measured for 30 min after MK-801 was administered into the fourth ventricle (2.0 μg/3.0 μl) or the NTS (198 ng/30 nl). The functional tests were conducted both before and after the experiment. After completion of the final functional feeding test, rats were anesthetized and killed by transcardial perfusion with 0.1 M PBS, followed by 4% formalin. Brains were removed and fixed in formalin for 4 h and were overnight cryoprotected in 20% sucrose solution. The location of the guide cannulas was verified by histological examination of brain slices. Brains were sectioned at a thickness of 40 μm with a cryostat and mounted on the slides. The sections were stained with cresyl violet and examined under the microscope. Location of each cannula tip was mapped according to the rat brain atlas of Paxinos and Watson (31). Only data from rats whose cannula placements were in the fourth ventricle or NTS, respectively, and that responded to functional feeding test were included in the statistical analysis.
Unless otherwise noted, for all experiments, control, saline injections were run the day before and the day after each drug injection, with the intake from these days averaged and compared with average intakes after drug administration. Thus each experimental datum represents the mean sucrose intake ± SE from at least two tests for each drug dose compared with the average of the bracketing saline control tests. Cumulative sucrose and chow intakes were analyzed using two-way repeated-measures ANOVA, with drug injection and time as independent variables, using PC-SAS (version 8.02, SAS Institute, Carey, NC) mixed procedure, followed by Tukey's pair-wise multiple-comparison tests. In all cases, a P value of <0.05 was considered significant. All data are presented as means ± se.
Sucrose Intake After Systemic Administration of AP5
As shown in Fig. 1, intraperitoneal injection of AP5 (0.05, 0.1, 1.0, 3.0, and 5.0 mg/kg) did not significantly increase 30-min sucrose intake at any dose tested (11.0 ± 0.8, 11.2 ± 1.0, 11.2 ± 1.0, 13.1 ± 2.2, and 11.0 ± 1.9 ml, for 0.05, 0.1, 1.0, 3.0, and 5.0 mg/kg AP5, respectively) compared with control (10.7 ± 0.4 ml; P > 0.05) for all AP5 doses.
Sucrose Intake After Fourth-Ventricle Administration of AP5
Two rats did not meet cannula placement criteria; therefore, data from 8 of the 10 implanted rats were included in the statistical analysis. There was a significant effect of injection [F(4,118) = 52.26; P < 0.001], and a significant effect of time [F(11,118) = 22.76; P < 0.001] but not a significant interaction between injection and time [F(44,118) = 0.49; P = 1.0]. Figure 2 shows the effects of various doses of fourth-ventricular AP5 injection on 60-min sucrose intake. Administration of 0.1 μg AP5 did not significantly increase sucrose intake compared with saline (12.7 ± 0.9 ml, AP5 vs. 11.2 ± 0.4 ml, saline; P = 0.20). However, rats increased their 60-min sucrose intake after 0.2 μg (16.7 ± 0.6 ml; P < 0.001) and 0.4 μg (14.9 ± 0.5 ml; P = 0.002) AP5 injections compared with saline (11.2 ± 0.4 ml). Both 0.2 μg (11.5 ± 0.7 ml; P = 0.007) and 0.4 μg (11.6 ± 0.5 ml; P = 0.005) dose of AP5 significantly increased sucrose intake starting from 10-min post injection compared with saline injection (8.2 ± 0.9 ml). We did not observe a dose-dependent increasing effect of fourth-ventricular AP5 on sucrose intake given that administration of 0.6 μg AP5 did not significantly increase 60-min sucrose intake compared with saline (11.7 ± 0.9 ml, AP5 vs. 11.2 ± 0.4 ml, saline; P = 0.65). Rather, 60-min sucrose intake after 0.6 μg AP5 (11.7 ± 0.9 ml) administration was significantly decreased compared with intakes following 0.2 μg (16.7 ± 0.6 ml; P < 0.001) and 0.4 μg (14.9 ± 0.5 ml; P = 0.008) of AP5.
Chow Intake After Intracerebroventricular Administration of AP5
A significant effect of injection [F(1,34) = 102.28; P < 0.001], time [F(8,34) = 1,704.44; P < 0.001], as well as a significant interaction between injection and time [F(8,34) = 8.81; P < 0.001] on chow intake was demonstrated by repeated-measures ANOVA. As illustrated in Fig. 3A, AP5 (0.2 μg/3 μl) produced an increasing effect on solid chow intake. After AP5 administration, rats significantly increased chow intake from 0 to 0.5 h (AP5: 5.9 ± 0.2 g vs. saline: 4.9 ± 0.4 g; P = 0.02), from 2 to 4 h (AP5: 2.3 ± 0.2 g vs. saline: 1.4 ± 0.4 g; P = 0.04), and from 4 to 24 h time periods (AP5: 19.7 ± 0.2 g vs. saline: 17.4 ± 0.3 g; P < 0.001; see Fig. 3B). Figure 3A shows that chow intake after AP5 administration was not significantly increased at 0.5–1 h (AP5: 2.5 ± 0.1 g vs. saline: 2.2 ± 0.4 g; P = 0.55) and 1–2 h (AP5: 1.1 ± 0.2g vs. saline: 1.3 ± 0.4 g; P = 0.76) time periods. Figure 3, C and D, illustrate the cumulative chow intake after fourth-ventricular administration of AP5 starting from 30-min postinjection (AP5: 5.9 ± 0.2 g vs. saline: 4.9 ± 0.4 g; P = 0.02) until the end of cumulative 24-h measurement (AP5: 31.5 ± 0.1 g vs. saline: 27.1 ± 0.6 g; P < 0.001). AP5-treated rats consumed significantly more chow compared with saline control throughout the experimental period.
Saccharin Intake After Fourth-Ventricular Administration of AP5
There was a significant effect of injection [F(3,35) = 122.68; P < 0.001], time [F(11,659) = 137.23; P < 0.001], and a significant interaction between injection and time [F(33,659) =2.14; P < 0.001]. As shown in Fig. 4, when an injection of AP5 (0.2 μg) was given, rats in sucrose group significantly consumed more sucrose during the 60-min tests (16.0 ± 0.9 ml) compared with control (12.8 ± 0.8 ml; P < 0.001). However, rats in saccharin group did not drink more saccharin after receiving AP5 injection (AP5: 9.8 ± 0.7 ml vs. saline: 10.5 ± 0.5 ml; P = 0.23).
Feeding Responses After NTS Administration of AP5
Repeated-measures ANOVA revealed a significant effect of dose [F(1,69) = 77.52; P < 0.001] and a significant effect of time [F(11,69) = 24.11; P < 0.001] but not a significant interaction between dose and time [F(11,69) = 0.53; P = 0.88]. Figure 5 shows 60-min sucrose intake after 20 ng of AP5 administration. Starting from 5-min postinjection (8.5 ± 0.4 ml), rats consumed a significantly larger amount of sucrose compared with saline (6.8 ± 1.0 ml; P = 0.05). This increased intake continued throughout the 60-min testing period. Injection of AP5 into the NTS produced a significant increase of cumulative 60-min sucrose intake (19.4 ± 0.6 ml) compared with saline control (15.5 ± 1.4 ml; P < 0.001).
There was a significant effect of injection [F(1,51) = 42.36; P < 0.001] and time [F(8,51) = 183.14; P < 0.001], as well as a significant interaction between injection and time [F(8,51) = 4.27; P < 0.001], as demonstrated by repeated-measures ANOVA. Figure 6A illustrates that rats consumed significantly more chow 30 min after 20 ng of AP5 was administered into the NTS (6.7 ± 0.3 g) compared with saline control (5.5 ± 0.3 g; P = 0.05). As shown in Fig. 6B, chow intake from 4–24 h (AP5: 19.1 ± 0.7g vs. saline: 15.7 ± 0.8 g; P < 0.01) was also significantly enhanced after AP5 administration. In contrast, intakes after AP5 and saline administrations were not significantly different from each other from 0.5 to 1 h (AP5: 1.6 ± 0.2 g vs. saline: 1.6 ± 0.5 g; P = 0.96), 1–2 h (AP5: 1.6 ± 0.4 g vs. saline: 1.2 ± 0.2 g; P = 0.77), and 2–4 h (AP5: 2.7 ± 0.3 g vs. saline: 2.1 ± 0.5 g, saline; P = 0.65) after the injections were given (see Fig. 6A). Figure 6, C and D depict cumulative chow intake after AP5 administration. AP5 significantly increased 24-h chow intake (31.6 ± 0.3 g) compared with saline control (26.1 ± 1.2 g; P < 0.001) starting from 30-min postinjection (AP5: 6.7 ± 0.3 g, vs. saline: 5.5 ± 0.3 g; P = 0.05). The increase in food intake continued to be significant throughout the experimental period.
In the present study, we observed that fourth-ventricular administration of AP5 significantly increased 60 min intake of 15% sucrose but was unable to affect intake of saccharin solution. In addition, intracerebroventricular administration of AP5 increased 24-h intake of solid rodent chow. Likewise, when AP5 was injected into the caudomedial NTS, a significant increase of sucrose, as well as chow intake, was observed. However, increased feeding was not evoked when AP5 was injected systemically. Although previous reports indicate that AP5 is a highly specific NMDA receptor antagonist (20, 53), peripheral AP5 activity appears to be limited. Our previous work shows that the central site is well established for the feeding effects of MK-801, and the current findings also indicate that AP5 only works centrally, since peripheral administration of AP5 at doses as high as 5 mg/kg did not increase food intake. Therefore, the question is whether peripherally injected AP5 can penetrate the brain in sufficient amounts and, if so, whether the doses used in this study were high enough to compete with glutamate effectively at the central site. Although the epilepsy literature suggests that AP5 does get into the brain (32, 42), it seems that the doses that we used systemically may not provide sufficient levels to effectively compete with endogenous glutamate in the dorsal vagal complex (DVC). Clearly, local injection must put a higher concentration of the competitor at the active site than peripheral injection. If we assume a 300-g rat with ∼60 ml of extracellular fluid (ECF) and no sequestration of AP5 or clearance, then the average brain ECF concentration would be far below what would be achieved with local injection. Our data clearly show that AP5 acts centrally, as the dose used in the brain is greatly lower than the dose that was ineffective when injected peripherally. This finding is consistent with previous results with MK-801. That is, nanogram doses are effective in the NTS but not the periphery. Also, dorsal hindbrain lesion abolishes the response to high doses of peripheral MK-801 (50). This is in line with electrophysiology data showing that when AP5 was injected into the dorsal hindbrain, gastric tone was increased in rabbits (3). However, intravenous administration of AP5 does not produce this effect nor does peripheral AP5 attenuate firing rate of gastric vagal sensory afferent fibers, innervating the distal stomach in response to gastric distention (44). This further suggests that AP5's effects are centrally mediated. Together, these results support the conclusion that AP5 is more likely to act through central, rather than peripheral, mechanisms. This is supported by our current experiments, in which injections of AP5 into the fourth ventricle or the NTS increased both liquid and solid food intake, and intraperitoneal injections of the antagonist did not. In other words, the concentration of AP5 that reaches DVC via permeable circumventricular routes after intraperitoneal injection may not be sufficient to compete with endogenous glutamate.
Data from our dose-response experiment revealed that the lowest effective dose of AP5 is between 0.1 μg and 0.2 μg. Therefore, to minimize any disruptive side effects from high doses of AP5, we chose the 0.2 μg AP5 dose for all subsequent experiments. AP5 did not increase sucrose intake at the dose of 0.1 μg but did increase sucrose consumption at doses of 0.2 μg and 0.4 μg. However, there was no significant difference in intake between these two doses. This fact might indicate a lack of dose dependency for AP5-induced increase in food intake. However, a lack of dose dependency would not be predicted on the basis of AP5 binding characteristics, which reveal a typical sigmoidal-shaped curve, which approaches a plateau stage as the binding reaches saturation (28, 29). An alternative explanation for the apparent lack of dose dependency is the possible existence of disruptive side effects at higher AP5 doses. In this regard, it is interesting that we observed obvious impaired locomotor function in 3 out of 10 rats after they received the 0.6 μg AP5 dose. Given the close proximity to our injection site to areas responsible for controlling posture and locomotion, it seems likely that the dosage range over which fourth-ventricle and NTS administration of AP5 can increase food intake may be limited by competing side effects that occur when the higher doses diffuse to sites involved in motor function. Such competing side effects may explain the lack of response at the highest AP5 dose in the current experiments.
Administration of AP5 into the fourth ventricle produced a significant increase in sucrose intake, which was similar in magnitude to that reported after injection of the noncompetitive NMDA receptor antagonist, MK-801(5, 57). MK-801 produces a dramatic enhancement of sucrose ingestion (6, 57) but does not increase 24-h food intake. In contrast to the MK-801 results, our current findings show that increased sucrose consumption by fourth-ventricle AP5 (0.2 μg and 0.4 μg) was not limited to a 30-min feeding period, but enhanced sucrose intake persisted over an entire 60-min recording period, suggesting that the antagonist might increase the size of a meal by extending its duration. Therefore, it is possible that AP5 exerts effects on 24-h food intake.
Previously our laboratory (6), as well as others (17), have shown that MK-801 administration increases chow intake up to 4-h postinjection, but not on 24 h intake. These observations suggested either that MK-801's efficacy was limited to a few hours duration or that other mechanisms allowed rats to compensate for increased meal size early in the test period with reduction of subsequent intake. In our present experiments, rats significantly increased cumulative 24-h chow intake after receiving AP5. The largest increase occurred during the first meal, 30-min after chow was presented. However, significant increases in intake were also detected between 2–4 h and 4–24 h postinjection. It appears from these results that NMDA receptors may participate in control of either meal size, intermeal duration, or both and that AP5 has a prolonged effect on food intake compared with MK-801 (21). This is in agreement with binding data and the fact that AP5 possesses an extended rate of dissociation from its receptor. In addition, the results of this study also suggest that any reduction in the size or frequency of meals over the 24-h measurement period did not compensate for the increased intake during the first 4-h after AP5. It is worth noting, however, that in the present study, food intake was measured after overnight food deprivation, which is different from the typical experimental design, examining hyperphagia by testing food intake in nondeprived animals at the onset of the dark cycle, when the spontaneous drive of feeding is triggered. Although the results from the present study demonstrate unequivocally that AP5 increases food intake in food-deprived rats by acting in the hindbrain, it is also possible that rats may increase food intake in response to administration of AP5 when they are not in a food-deprived condition during which deprivation- or scheduled-induced appetitive signals are minimized. Previous data from our laboratory have shown that MK-801 increases deprivation-induced intake of sucrose and chow but not water (6). Furthermore, MK-801 increases intake of a 15% nutritive sucrose solution but does not increase intake of a 0.2% nonnutritive saccharin solution (5), indicating the mechanism by which NMDA receptor blockade delays satiation is rather nutrient-specific. In the present study, the previous hypothesis was further proven by blocking hindbrain NMDA receptors using AP5. Fourth ventricular administration of AP5 increases sucrose intake but not saccharin, which is consistent with previous data using MK-801. These results suggest that the nutrient content of a meal is critical to NMDA-dependent negative feedback mechanism and blockade of the receptor by either MK-801 or AP5 enlarges meal size by interfering with the nutrient-related satiation signals.
Consistent with our fourth-ventricular administration results, NTS administration of AP5 at a dose as small as 20 ng significantly increased the size of a 15% sucrose meal, as well as 24-h cumulative chow consumption. In fact, increased sucrose intake reached significance during the first 5 min after the injection was given. There is a growing body of evidence supporting NMDA receptor participation in the caudomedial NTS in control of food intake. First, vagal afferent fibers that innervate the stomach and intestine terminate in this NTS region (27, 45), and NMDA subtype glutamate receptors are widely distributed in this area (41). Moreover, subpopulations of these neurons are activated by anorectic signals, contributing to meal termination such as CCK (39), intestinal nutrients (34, 58), and gastric distention (11). Taken together with previous findings, these results support the evidence that NMDA receptor plays a role in the control of food intake at the level of NTS.
The mechanisms by which NMDA receptors participate in control of food intake are not completely elucidated. NMDA receptor blockade produces changes in gastric tone and motility (46). Our previous work indicates that MK-801 can produce changes in gastric emptying (10). However, the feeding effect is independent of the emptying effect for MK-801 (9), and although not tested yet, this would be expected to be true for AP5 as well. Because MK-801 has been reported to interact with other receptors that might participate in the process of satiation, it can be argued that the effect of MK-801 administration on food intake may be due to these non-NMDA actions. For example, intracerebroventricular administration of dopamine receptor antagonists such as SCH 23390, pimozide, and haloperidol have been shown to dose-dependently attenuate food intake elicited by MK-801 (25). Although no data exist on MK-801 action at dopamine receptors, it is conceivable that such an action exists. MK-801 has been shown to directly act on some nicotinic ACh ion channels and dose-dependently reduces the activation period of these channels (12, 35). Therefore, it is possible that the MK-801-induced increase of food intake may partially result of the inactivation of nicotinic ACh ion channel. Activation of the receptor by nicotine has been shown to suppress appetite (13, 14). However, studies examining the effect of nicotinic receptor antagonists on feeding do not support this explanation. In one study, showing the effect of the nicotinic receptor antagonist mecamylamine (MEC) on attenuating nicotine-induced hypophagia in food-restricted rats, MEC, by itself, did not have any effect and was only effective when administered along with nicotine (55). Results from another study in freely fed rats also showed that food intake is attenuated only when MEC was injected in the presence of nicotine (15). Therefore, as far as food intake, it seems that the relationship between MK-801 and nicotinic ACh is not clear, and to our knowledge, there are currently no reports examining the feeding effects of MK-801 on nicotine-induced hypophagia. In any event, we are aware of no data implicating AP5 actions at non-NMDA receptors. Therefore, although it remains conceivable that increased food intake by MK-801 might involve some effects at nonglutamatergic receptors, our finding of increased food intake produced by AP5 strengthens the claim for specific participation of hindbrain NMDA receptor in the control of food intake.
Exactly, how AP5 interferes with the satiation process to increase food intake remains uncertain. However, results from previous studies may provide a possible explanation. First, hindbrain microinjection of L-glutamate into DVC, has been known to evoke both inhibitory (47, 48) as well as excitatory (18) gastric responses. Second, NTS administration of glutamate tends to produce gastric inhibition (47, 48), whereas activation of dorsal motor nucleus of the vagus (DMV) evokes gastric excitation (18, 30). It has been shown that AP5 attenuates inhibition of firing rate of the DMV produced by gastric and intestinal distention (56). However, microinjection of glutamate into the NTS reduces firing rate of neurons in the DMV, suggesting that glutamate might activate inhibitory neurons in the NTS, which project to the DMV (40). Moreover, AP5 has been shown to increase gastric tone and decrease gastric volume in rabbits when injected into the area postrema (3). Therefore, it is possible that centrally administered AP5 increases food intake by reducing the negative feedback signals generated during ingestion of a meal. We observed that rats increased 24-h chow consumption after receiving an AP5 administration into the fourth ventricle or NTS, suggesting a long-term effect of AP5 in control of food intake. However, rats did not increase 24-h chow intake after receiving MK-801 administration (6). This may be due to the different functional nature of competitive NMDA receptor antagonist such as AP5 compared with the noncompetitive ones such as MK-801. For example, noncompetitive NMDA receptor antagonist MK-801 can only bind to the receptor when it is activated by depolarization, which causes removal of magnesium (23). However, the ability of MK-801 but not AP5 to block the NMDA-activated ion channel is voltage-dependent and reduced by depolarization (16, 22). On the other hand, the competitive NMDA receptor antagonist AP5 shares the same binding site as glutamate and has a higher binding affinity compared with MK-801 (21). It has been shown that there are two identical cooperative binding sites per ion channel, which have similar affinities and that the unbinding rate of two binding sites is slower than a single binding site (8). Therefore, even though MK-801 has a longer dissociation rate (21), it is possible that its binding duration is shorter compared with AP5 because of the voltage-dependent binding ability so that the effect of the competitive NMDA receptor antagonist lasts longer. Moreover, MK-801 has been shown to increase turnover of dopamine and serotonin, a potentially complicating effect, which has not been observed after intracerebroventricular administration of AP5 (36, 37). Hence, these differences in the neuropharmacology between AP5 and MK-801 may lead to explanations of subtle differences in the effects of the two drugs on feeding. For example, our results reveal a possibility of 24-h intake effects of AP5. However, meal pattern will be required to confirm and understand potential differences in the feeding effects of AP5 and MK-801.
In summary, our results show that fourth-ventricular, as well as NTS administration, of AP5, a potent and specific competitive antagonist of NMDA-type glutamate receptors, increases both palatable sucrose and solid chow intake. These findings support the conclusion that the site of AP5 action is in the hindbrain, and, more specifically, within the NTS. Furthermore, our current results help to allay concerns that previously reported effects of MK-801 might be attributable to action at non-NMDA ion channels. Finally, our data raise the possibility that manipulation of hindbrain NMDA receptors may impact control of food intake over a 24-h period and hence could influence the control of body weight.
This work was supported by National Institute of Diabetes and Digestive and Kidney Diseases Grant DK-52849 and National Institute of Neurological Disorders and Stroke Grant NS-20561.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2006 the American Physiological Society