- Email: [email protected]
The Effects on Feeding of Galanin and M40 When Injected Into the Nucleus of the Solitary Tract, the Lateral Parabrachial Nucleus, and the Third Ventricle
Physiology & Behavior, Vol. 67, No. 2, pp. 259–267, 1999 © 1999 Elsevier Science Inc. Printed in the USA. All rights reserved 0031-9384/99/$–see front matter
PII S0031-9384(99)00075-X
The Effects on Feeding of Galanin and M40 When Injected Into the Nucleus of the Solitary Tract, the Lateral Parabrachial Nucleus, and the Third Ventricle FRANK H. KOEGLER, DAVID A. YORK1 AND GEORGE A. BRAY Pennington Biomedical Research Center, Louisiana State University, 6400 Perkins Road, Baton Rouge, LA 70808 Received 17 December 1998; Accepted 23 March 1999 KOEGLER, F. H., D. A. YORK AND G. A. BRAY. The effects on feeding of galanin and M40 when injected into the nucleus of the solitary tract, the lateral parabrachial nucleus, and the third ventricle. PHYSIOL BEHAV 67(2) 259–267, 1999.—Several reports indicate that central injection of galanin stimulates feeding, and that there is macronutrient specificity in this response. In addition, the galanin receptor antagonist, M40, reduces food intake when injected centrally. The nucleus of the solitary tract (NTS) and the lateral parabrachial nucleus (PBN) contain galanin receptors, and are involved in the control of food intake. Hence, we sought to compare the feeding response to galanin injection into these areas with that of third ventricle (3V) galanin injection. The feeding response to injection of galanin was greatest for the 3V. Hindbrain injection of galanin stimulated food intake only at the beginning of the dark period. NTS injection of M40 inhibited intake of a macronutrient diet in food-deprived rats, but was ineffective at reducing dark-onset feeding or deprivation-induced chow intake. 3V injection of M40 did not reduce deprivation-induced intake. PBN injection of galanin at dark onset had no effect in a group of fat-preferring rats. These results suggest that hindbrain galanin may contribute to feeding by inhibiting satiety, and that hypothalamic galanin receptors are involved with stimulation of intake. Furthermore, the absence of a consistent pattern of the stimulation of macronutrient intake suggests that galanin may not be a significant effector of macronutrient selection during individual meals. © 1999 Elsevier Science Inc. Galanin
Macronutrients
NTS
1PBN
M40
Feeding
THE peptide galanin is synthesized in many brain and gut sites, and several galanin receptor subtypes from both human and rat have been cloned (6,21,22). Galanin is implicated in regulation and control of memory, nociception, food intake, and both insulin and gastric acid secretion (2). Several brain sties specifically related to feeding, such as the hypothalamus, the amygdala, the lateral parabrachial nucleus (PBN), and the caudal hindbrain contain high levels of galaninergic cells and processes, as well as galanin receptors (13,14,18). Injection of galanin into some of these areas, namely the paraventricular nucleus of the hypothalamus (PVN), the amygdala, and the nucleus of the solitary tract (NTS), stimulates feeding in a variety of paradigms (7,8,10,19,20). In addition, injection of the galanin receptor antagonist, M40 (1), into several of these sites can reduce feeding initiated by various stimuli, including 1To
palatable foods, blockade of fatty acid metabolism, and galanin itself (4,5,7). There are conflicting reports regarding the macronutrient specificity of galanin-stimulated intake in rodents. some reports suggest that galanin preferentially increases fat intake (11,20), while other suggest a stimulation of intake that is proportional to baseline intake (19). The demonstration that central injection of galanin antagonists decreases the overconsumption of fat intake by Brattleboro rats, which are deficient in AVP and overexpress galanin (15), supports a link between galanin and fat intake. We hypothesized that the activity of galanin in different brain regions would have different effects on food intake. The PBN has high levels of galanin and receives dense galaninergic innervation from the NTS as well as descending input from the hypothalamus. This interconnec-
whom requests for reprints should be addressed. E-mail: [email protected]
259
260
KOEGLER, YORK AND BRAY
tion between hypothalamic and hindbrain nuclei that are involved in feeding, and the lateral PBN’s proximity to the medial PBN, which receives gustatory information, suggested that galanin in the PBN may have a modulatory role on macronutrient selection. Thus, we hypothesized that the stimulation of galanin receptors in the hypothalamus would initiate food intake, whereas stimulation of hindbrain galanin receptors would alter macronutrient selection and suppress satiety. We further hypothesized that the activity of galanin in the PBN would not affect total caloric intake. The experiments described in this article were designed to evaluate the effects of galanin and M40 on food intake and macronutrient selection after injection into the nucleus of the solitary tract, the lateral parabrachial nucleus, and the third ventricle. MATERIALS AND METHODS
Animals, Housing, Diet All experiments used adult male Sprague–Dawley rats (Harlan–Sprague–Dawley Inc., Indianapolis, IN) weighing between 300 and 360 g at the beginning of experimentation. Upon arrival, rats underwent a 1-week quarantine period before being moved into the experimental rooms where they remained and underwent feeding tests. Rats were individually housed in hanging wire mesh cages in temperature controlled rooms (22–238C). Lights were automatically controlled and set to turn off at 1900 h and turn on 12 h later. Automatic red lighting was used for dark-cycle experimentation. Tap water was available ad lib via an automated watering system. Except during the overnight deprivation experiments, food was available at all times. Chow diet consisted of standard laboratory rodent chow (#5001; PMI Nutrition International, Inc., St Louis, MO). The diet used in the macronutrient experiments provided pure carbohydrate, protein, and fat sources, each supplemented with vitamins and minerals. These diets were made in our laboratory from commercially available ingredients (Table 1). The high-fat diet provided 55% of energy from fat at an energy density of 4.84 kcal/g. Three-choice and high-fat diets were fed to animals in individual glass food jars, whereas chow was fed from hoppers except during experiments when it was placed on the cage floors. When macronutrient or high-fat diets were used in an experiment, or when diets were switched, there was a 10-day minimum exposure time to the new diet prior to testing. Surgeries All cannulation surgeries were performed in an approved surgery suite according to institutional and federal regulations. Animals were anesthetized with Nembutal (sodium pentobarbital; 50 mg/mL/kg, IP) and treated with atropine sulfate (0.14 mg/0.25 mL, SC) to reduce anesthesia-induced respiratory distress. Routine stereotaxic implantation of 15mm long, 25-gauge, stainless steel guide cannulas was performed, and animals were allowed at least 7 days of recovery after surgery before testing commenced. Coordinates for third-ventricle cannulations were: 2.5 mm caudal to bregma, 7.7 mm subdural, in the midline. For 3V cannula implantation, the superior sagittal sinus was gently retracted away from the midline during cannula placement to prevent vessel rupture. Coordinates for the NTS were 1.8 mm rostral to the posterior occipital crest, 5.9 mm subdural, and 0.5 mm to the right side of midline. For NTS cannulations, the incisor bar was positioned at 1.5 mm, and the head was positioned in the
TABLE 1 COMPOSITON OF DIETS Three-Choice Diet
Corn starch Powdered Sugar Casein DL-Methionine Vegetable shortening Corn oil AIN-76A vitamin mix AIN-76A mineral mix Choine chloride Cellulose (Alphacel) Energy density (kcal/g)
Carbohydrate
Fat
Protein
High Fat
58.11 29.06 — 0.11 — — 0.77 3.07 0.18 8.72 3.53
— — — 0.20 75.12 — 1.49 5.95 0.34 16.91 6.85
— — 87.17 0.11 — — 0.77 3.07 0.18 8.72 3.53
23.91 — 28.69 0.14 26.60 3.20 1.06 4.21 0.24 11.95 4.84
*Ingredients expressed as percent by weight.
stereotaxic holder such that there was a nasal inclination of 58. NTS cannulas were 14 mm in length. Cannulas targeted at the lateral parabrachial nucleus were 12 mm in length. Lateral parabrachial nucleus coordinates were: 9.7 mm caudal to bregma, 5.1 mm subdural, and 6 1.9 mm from the midline. When not in use, all cannulas were obstructed with 32-gauge stainless steel stylettes made to extend 0.25 mm beyond the cannula tip. Feeding Tests All feeding tests took place in the home cages of the animals using the maintenance diets of either chow, three-choice macronutrient, or high-fat diet. At 12–20 h prior to each feeding test, food containers were replenished with fresh food. Before and during the experiments, food jar weights and spilled food were semiautomatically recorded to the nearest 0.01 g. Jars were manually removed from cages and weighed, and data from a balance was automatically captured into a spreadsheet. For experiments measuring chow intake, pellets were placed onto the cage floor and collected manually and weighed at each time point. Rats were trained to eat chow from the floor prior to the start of the experiments. Feeding tests involving overnight food deprivation consisted of removing food jars within 1 h prior to the dark period, and replacing the food immediately after treatment the following morning approximately 14 h later. The experiments were designed to minimize effects of test day on drug treatment by employing a full Latin square design whenever possible, and always including a control group on any given test day. Rats served as their own control subjects, and were not tested more than once in any consecutive 3-day period. Galanin was obtained from American Peptide Company (Sunnyvale, CA) and Bachem (Torrance, CA), and M40 was purchased from Peninsula Laboratories (Belmont, CA). Doses of galanin and M40 ranged from 0.1 to 2.5 nmol, based on previous reports (1,4,5,7), and the specific doses used for each experiment are described below and in the Results section text and figures. Central Injection Procedures Behavioral tests. Injectors for central drug administration were constructed from 32-gauge stainless steel tubing fastened to PE-10 flexible tubing (20–30 cm). The tubing was filled with distilled water and connected to a glass microliter
GALANIN, M40: 3V, NTS, PBN syringe to allow manual microinfusion of volumes of 300 nL up to 1 mL. The meniscus between the injection solution and water was visible in the clear tubing, and allowed confirmation of solution delivery. Infusions were performed manually while the rat was in its home cage, and lasted between 30 and 120 s, depending on the solution volume and brain area. Injections into the neuropil were 300 nL in volume, whereas ventricular injections were 1 mL in volume (see individual experiments). After the infusions, injectors were left in place for a minimum of 30 additional seconds to reduce back pressure and possible reflux when the injector was removed and replaced by the stylette. The procedure consisted of removing an animal from its home cage, removing the stylette, inserting the prefilled injector, and replacing the animal in its cage for the infusion. After the postinfusion period, the animal was immediately removed from the cage, the injector was replaced by a stylette, and the animal was replaced into its cage for the remainder of the test. Excitotoxin lesions. An injection procedure similar to that described above was used to deliver ibotenic acid (3 mg/300 nL/side; Sigma Chemical Company, St. Louis, MO) to the lateral parabrachial nuclei. Pilot experiments revealed that simultaneous bilateral ibotenic acid injections were often lethal; therefore, lesions were made unilaterally and a 1-week recovery period was allowed before the contralateral lesion was performed. Specific Experiments Experiment 1: Effect of 3V and NTS galanin and M40 injection on macronutrient diet intake during daytime. Cannulas were implanted into the NTS of 12 animals, and another 12 animals received cannulas targeted to the third ventricle. Doses of galanin injected into the 3V and NTS were 0 (saline), 0.1, 0.5, and 2.5 nmol. These experiments began at approximately 0900 h. After the galanin experiments, the same group of rats was tested with the galanin antagonist, M40. M40 (0.5 nmol) or saline was injected after overnight food deprivation and food intake was measured after 30, 60, and 120 min. Experiment 2: Effect of NTS galanin and M40 on early dark-period intake of macronutrient diet. The 24 rats with NTS cannulas were tested with galanin and M40 (0.5 or 1.5 nmol for each) during the first hour of the dark phase. To make timed central injections possible, the group of 24 was split into two and tested on successive days, with an identical treatment regimen for both groups. Intake was measured for 30 and 60 min after the lights went off, and overnight. Prior to galanin and M40 drug tests, two 48-h periods of baseline intake were measured. Experiment 3: Effect of NTS injection of M40 or galanin on daytime chow intake and the effect of galanin on high-fat diet intake. A group of 12 rats maintained on chow diets with cannulas targeted at the NTS was tested during the daytime (0900 h) with both saline vehicle (0.3 mL) and M40 (0.5 nmol) after overnight food deprivation. Food intake was measured after 30 and 60 min. Because galanin was ineffective in stimulating daytime feeding when rats were eating a chow diet that is relatively low in fat content, the diet was switched to a highfat composite diet (Table 1). Experiment 4: Effect of PBN galanin on daytime chow intake and dark-period macronutrient intake. A separate group of 12 rats received bilateral PBN injection of galanin (0.5 nmol) or saline vehicle (0.3 nmol) at 0900 h, and intake of chow was measured at 1 and 3 h. The same group of rats was adapted to the three-choice macronutrient diet for 10 days
261 and retested with bilateral galanin (0.5 or 1.5 nmol) or vehicle during the early part of the dark period when rats normally increase food intake. Food intake was measured 30 min, 60 min, and 14 h after injection of galanin into the PBN. Experiment 5: Effect of PBN galanin on dark-onset macronutrient intake. Another group of 14 rats was implanted with bilateral PBN cannulas and was maintained and tested with the three-choice macronutrient diet. Galanin (1.5 nmol) or vehicle was injected bilaterally into the PBN immediately after the onset of the dark period, and intake was measured after 1 and 12 h. After completion of the galanin injection experiments, these animals were lesioned with ibotenic acid to determine if there would be a change in macronutrient intake after chemical ablation of the PBN. Histology After completion of the feeding tests, all animals were prepared for brain histology by deep anesthetization and transcardial perfusion with buffer followed by perfusion with formaldehyde. Brains were removed, sectioned, and stained with cresyl violet. With the aid of a light microscope, cannula tracks were mapped onto brain atlas map drawings to confirm proper cannula location. Histological criteria for exclusion from data analysis were nay of the following: 1) cannulas more than 0.3 mm away from the target nucleus (for neuropil injections); 2) destruction of the target nucleus (except for the lesion experiments); 3) extraventricular cannula (for ventricular injection experiments). Ibotenic acid lesions were evaluated by inspecting the injection site for gliosis and loss of neuronal cell bodies. If data from an animal was eliminated because of meeting the exclusion criteria, it is noted in the results section for that particular experiment. Data Analysis Data are presented as the mean intake in kilocalories (or grams) 6 SEM. For experiments with multiple time-point observations, intake data for subsequent time points is presented as cumulative. Because all animals received both control and drug treatments, all statistical tests used either paired (t-test) or repeated measures (analysis of variance) design. Most tests used two-factor, mutilevel, repeated-measures analysis of variance (ANOVA), computed with either SigmaStat 2.03 (SPSS Inc.), or the SAS system (SAS Institute). Unless otherwise noted, when multiple post hoc comparisons were performed on a series of data, p-value corrections were made to reduce Type I significance errors (Bonferroni’s correction). As a consequence, some statements of significance tend to be conservative. RESULTS
Experiment 1a: Effect of 3V and NTS Galanin Injection on Spontaneous Three-Choice Macronutrient Diet Intake During Daytime One animal from each group was eliminated due to improper cannula placement. The 24-h baseline measurements revealed a relatively stable composition of macronutrient intake throughout the duration of the experiments, with animals selecting calories primarily from carbohydrates. Third ventricle cannulated animals selected approximately 57% of daily calories from carbohydrate, 27% from fat, and 16% protein. Animals with NTS cannulas selected 53%, 33%, and 15% of daily calories from carbohydrate, fat, and protein, respectively.
262
KOEGLER, YORK AND BRAY
At 30 min there was a main effect of 3V injection of galanin on total food intake, F(3, 11) 5 6.76, p 5 0.00112 (Fig. 1, left side). Bonferroni-adjusted multiple comparison revealed that total intake after 2.5 nmol was greater than after saline (p , 0.05). There also were significant main effects of 3V drug treatment and macronutrient intake, F(3, 33) 5 6.75, p 5 0.0011; F(2, 22) 5 9.62, p 5 0.0010. There was an interaction between macronutrient and galanin, F(6, 66) 5 4.67, p 5 0.0005. Comparison of dose effects collapsed across macronutrients revealed a statistically significant, dose-dependent intake for 0.5 and 2.5 nmol galanin. At 30 min, multiple comparisons to saline revealed that carbohydrate intake after 2.5 nmol galanin was greater than after saline (p , 0.05). The 2.5nmol galanin dose marginally stimulated protein intake (p 5 0.08). No dose of galanin injected into the 3V reliably increased fat intake relative to saline. At 60 min the results of statistical analysis of the 3V galanin response similar to results obtained at 30 min [main effect on total intake, F(3, 11) 5 5.76, p 5 0.00277]. Bonferroni’s comparison revealed that intake after 2.5 nmol was greater than after saline (p , 0.05). A twofactor repeated-measures ANOVA revealed a significant effect of galanin and macronutrient on intake, F(3, 33) 5 5.76, p 5 0.0028, F(2, 22) 5 11.8, p 5 0.0003. There was an interaction between macronutrient and treatment, F(6, 66) 5 2.26, p 5 0.0059. At 60 min multiple comparisons of galanin-induced intake to intake after saline revealed that carbohydrate intake after 3V injection of 2.5 nmol galanin was greater than after
saline (p , 0.05). There was no statistically reliable increase in fat or protein intake at 60 min. Galanin (0, 0.1, 0.5, 2.5 nmol) failed to affect energy intake or macronutrient selection when injected into the NTS (Fig. 1, right side). Total intake at 30 and 60 min was not affected by galanin treatment [one-way, repeated-measures ANOVA, F(3, 11) 5 0.477, p 5 0.701; F(3, 11) 5 0.0974, p 5 0.961, respectively]. There was no effect of galanin on intake of any particular macronutrient at either 30 or 60 min, F(3, 10) 5 0.789, p 5 0.5096; F(3, 10) 5 0.124, p 5 0.9454, respectively. Experiment 1b: Effect 3V and NTS M40 on DeprivationInduced Intake of Three Choice Macronutrient Diet Two-way, repeated-measures ANOVA with factors of cannulation site and drug treatment revealed a marginally significant effect of M40 on the suppression of food intake (Fig. 2) at 30 min, F(1, 18) 6 3.7587, p 5 0.0684, but no interaction, F(1, 18) 5 0.04681, p 5 0.8311. Total food intake after 60 min was suppressed by M40, F(1, 18) 5 6.88, p 5 0.0173. There was no difference in food intake with cannulation site, F(1, 18) 5 2.45, p 5 0.1351, nor any interaction effect, F(1, 18) 5 1.33, p 5 0.2642. Effects of M40 after 120 min paralleled those at 60 min, with a main effect of M40 but no effect of cannulation site, nor was there an interaction between M40 treatment and cannulation site. A two-way, repeated-measures ANOVA (treatment 3 macronutrient) of the effect of M40 on individ-
FIG. 1. Effect of galanin injected into the 3V or NTS. Galanin reliably stimulated 30 and 60 min intake when injected into the 3V but not when injected into the NTS. The greatest effect on the stimulation of intake was for carbohydrate. There was a trend towards the stimulation of fat intake; however, after adjusting for multiple comparisons, this result was not statistically relevant. “a” denotes difference from saline, p , 0.05.
GALANIN, M40: 3V, NTS, PBN ual macronutrients for 3V injection site at 30, 60, and 120 min failed to reveal any significant effects. Analysis of individual macronutrient intake following M40 injection into the NTS failed to reveal significant effects at 30 min (p 5 0.1240). After 60 min, however, M40 injection into the NTS significantly suppressed total energy intake, F(1, 9) 5 6.818, p 5 0.0282. There was no statistically significant interaction between macronutrient and treatment, F(2, 18) 5 0.684, p 5 0.5173. Experiment 2: Effect of Galanin and M40 Injected Into the NTS on Early Dark-Period Intake of Three-Choice Macronutrient Diet Analysis of baseline intake revealed a macronutrient composition of 45 6 6, 36 6 7, and 18 6 6% from carbohydrate, fat, and protein, respectively. From the 23 animals included after histological analysis of brain tissue, six had greater than two-thirds of total daily intake from carbohydrate; one animal had greater than two-thirds of its total intake from fat.
263 There was a main effect of galanin on total energy intake (Fig. 3) measured from the start of the dark phase, F(2, 21) 5 4.06, p 5 0.0244. Repeated-measures ANOVA of carbohydrate intake revealed a significant effect of drug treatment, F(2, 21) 5 6.33, p 5 0.0031. There was no effect of galanin on fat intake, F(2, 21) 5 0.607, p 5 0.5496. There was no main effect of M40 on intake, F(2, 21) 5 0.9564, p 5 0.4004. Multiple comparisons revealed that there was no effect of M40 on intake of any macronutrient during the first 30 min of the dark period. Experiment 3: Effect of NTS Injection of M40 or Galanin on Daytime Chow Intake Ten days after the M40 injection experiments, the response to galanin (1 nmol) during daytime feeding of rodent chow was measured in the same group of nonfood-deprived rats (Fig. 4). Galanin had no effect on chow intake at any time point, F(1, 11) 5 0.575, p 5 0.464.
FIG. 2. Effect of the galanin receptor antagonist, M40, on overnight deprivationinduced refeeding. M40 (0.5 nmol) injected into the nucleus of the solitary tract, but not 3V, suppressed 30-, 60-, and 120-min intake induced by overnight food deprivation. “a” denotes difference from saline, p , 0.05.
264
KOEGLER, YORK AND BRAY tively. Bilateral injection of 0.5 or 1.5 nmol of galanin into the PBN was ineffective in altering total macronutrient intake, or changing the percentage of macronutrient intake during the dark phase (Fig. 5). Experiment 5: Effect of PBN Galanin on Dark Onset Intake of a Three-Choice Macronutrient Diet (Carbohydrate-Preferring Animals)
FIG. 3. Effect of galanin and M40 injection into the NTS during dark-onset feeding of a three-choice macronutrient diet. Galanin dose dependently increased carbohydrate intake during the initial 30 min after treatment. M40 had no effect. “a” denotes difference from saline; “b” denotes difference from other dose of same peptide, p , 0.05.
Subsequently, rats were adapted to a high-fat diet and retested. There was no effect of galanin injection (1 nmol) into the NTS when rats were tested during the light period (0900 h), F(1, 11) , 0.001, p 5 0.987; data not shown. Experiment 4: Effect of PBN Galanin on Daytime Chow Intake and Dark-Period Macronutrient Intake (Fat-Preferring Animals) Bilateral PBN injection of 0.5 nmol galanin during daytime did not stimulate 1 or 3 h intake of rodent chow when rats were maintained with a chow diet (data not shown). After adaptation to the three-choice macronutrient diet analysis of basal diet intake revealed that this group of animals was particularly and unusually fat preferring, with a mean percentage of daily energy intake from fat being 60 6 5%. Protein and carbohydrate intake were 10 6 1% and 30 6 4%, respec-
In contrast to the animals in the previous experiment, baseline 24-h intake preference of this group averaged 60% for carbohydrate, 26% for rat, and 14% for protein. Injection of 1.5 nmol galanin bilaterally into the PBN immediately after the onset of the dark period failed to increase total intake when measured after 1 or 12 h (Fig. 6). There was no stimulation of carbohydrate (p 5 0.211, not significant) but an inhibition of fat intake (p 5 0.0428) during the first hour, suggesting a shift in intake. Comparisons of prelesion diet intake with the amount eaten 6 weeks after ibotenic acid lesions showed that both groups decreased fat intake and increased carbohydrate intake. However, the lesioned group decreased its fat intake to a lesser degree, whereas the sham-lesioned group increased carbohydrate intake to a greater degree. This effect was not statistically reliable (lesion effect: p 5 0.07). It is noteworthy that this group of animals was predominantly carbohydrate preferring. There was no effect of the lesions on body weight. DISCUSSION
The major observations reported in this study were that the 3V injection site is the most effective site at which galanin stimulates intake, that M40 was ineffective at decreasing deprivation-induced intake when injected into the 3V, and that galanin was effective at stimulating intake only during the dark-onset feeding period when injected into the NTS. At this time galanin weakly stimulated carbohydrate intake, which was the preferred macronutrient of the group of animals tested. This effect was not present when rats were tested during the daytime with a three-choice macronutrient diet, high-
FIG. 4. Effect of galanin and M40 (NTS) on chow intake. Both galanin and M40 failed to affect intake of a chow diet in 1-h tests. Galanin was injected during the daytime in presumed sated rats; M40 was injected in the daytime in rats that had been overnight food deprived. (Note: ordinate scale is grams.)
GALANIN, M40: 3V, NTS, PBN
265
FIG. 6. Effect of galanin injection into the PBN on dark-onset feeding of a three-choice macronutrient diet. In a group of carbohydrate preferring rats, galanin did not affect total intake during the first hour following treatment nor during the following 11 h. Fat intake was suppressed during the 1-h period. “a” denotes difference from saline, p , 0.05.
FIG. 5. Effect of galanin injection into the PBN during the darkonset and overnight feeding period. Galanin injected bilaterally into the PBN failed to affect intake of a three-choice macronutrient diet. There was a minimal, nonsignificant stimulation of overnight fat intake in this group of fat-preferring animals.
fat diet, or a chow diet. M40 was effective at reducing threechoice diet intake stimulated by an overnight fast, but had no effect on initial dark-onset feeding, or deprivation-induced intake of rodent chow. In addition, galanin injection into the PBN did not stimulate dark-onset feeding in a group of fat-preferring rats. However, in a separate group of carbohydrate-preferring rats, the
same dose of galanin increased carbohydrate intake and inhibited fat intake. PBN galanin injection had no effect on spontaneous daytime intake when a chow diet was used as the maintenance diet. Finally, bilateral lesion of the PBN with ibotenic acid had no effect on body weight, but had a tendency to shift intake towards fat and away from carbohydrate relative to nonlesioned, intact animals. The results of these experiments suggest that the role of extrahypothalamic galanin in the regulation of food intake is a secondary, modulatory one, rather than a primary effector of meal initiation or macronutrient selection. In comparison to the powerful stimulatory effect of NPY or inhibitory effect of CRH, the relatively weak effects of galanin and M40 on food intake and macronutrient selection, regardless of time of day, diet, or injection site, establish a lesser position for this peptide in the control of food intake. The effects of hindbrain injection of galanin on food intake are weak in comparison with that of third ventricle injection. There were no effects of hindbrain galanin on spontaneous daytime intake, regardless of maintenance diet. This is in contrast to previous reports in which the stimulatory effect of hindbrain injection of galanin was more pronounced during the daytime. Because the previous experiments utilized a semisynthetic high-fat diet, it is possible the differences are
266
KOEGLER, YORK AND BRAY
diet related. Only in carbohydrate-preferring rats did hindbrain galanin stimulate dark-onset food intake. These results are incongruent with some previous reports that suggest a relationship between fat intake and galanin administration. Previous reports suggest a link between galanin and both fat ingestion and body weight (20). However, the macronutrient diet fat source in the current experiments was primarily of vegetable origin (shortening) rather than fat of animal origin (lard). This difference may be responsible for the disparate results. In another study that used the same diet, intraventricular galanin injection increased the intake a preferred macronutrient, i.e., that which was most often selected during daily intake, rather than the intake of fat intake per se (19). The hypothesis that hindbrain injections of galanin would stimulate fat intake was not supported; however, it was shown that the galanin receptor antagonist, M40, was most effective at reducing deprivation-induced food intake when injected into the hindbrain rather than in the third ventricle. This suggests that hindbrain galanin receptors may not be involved in the outright stimulation or initiation of food intake, but rather, may contribute to intake by reducing tonic or meal-induced satiety signals. Specifically, galanin receptors in the NTS, being primarily inhibitory to neuronal membrane potential, may serve to reduce the effect of ascending visceral afferent information that is classically thought to reduce food intake. This would permit an organism in a food-deprived state to override viscerally derived meal termination signals such as gastric distention, thus allowing greater food intake. It is interesting that M40 injections into the NTS had no effect on dark-onset feeding in contrast to its inhibitory effect on deprivation-induced feeding. It is possible that the satiety signals, for example, gastric distension, and other postprandial signals, may be greater in the deprivation-refeeding state, rather than during the normal dark-onset feeding state. Food intake after an overnight deprivation was greater than that during spontaneous daytime feeding or dark onset feeding. Therefore, interference with a galaninergic inhibition of a satiety signal may not be as easily detectable in the dark-onset feeding daytime feeding paradigm.
The lack of a robust effect of galanin on dark-onset feeding, or daytime feeding when injected into the PBN suggests that galanin receptors in this nucleus are not intrinsically involved in the initiation of feeding. The PBN is classically recognized as a relay nucleus for ascending visceral afferent information, and has direct and reciprocal connections with the nucleus of the solitary tract, as well as the hypothalamus. Indeed, the galaninergic projection from the NTS is particularly well defined, and the terminal fields of these fibers are very dense in the central, dorsal, and external parts of the PBN. In particular, these areas of the PBN are thought to be involved in the control of food intake: many treatments that either stimulate (b-mercaptoacetate, 2-deoxyglucose) or inhibit (dexfenfluramine, a-melanocyte stimulating hormone) food intake activate neurons in this area, as shown by induction of the c-fos gene (9,12,16). Furthermore, lesions in these areas can prevent the development of conditioned taste aversions and the ability to respond with feeding to lipoprivic stimuli (3,17). Despite this information, galanin was ineffective at affecting food intake when injected into this area. It is also possible that the doses of galanin used in these experiments were not high enough to affect feeding behavior; however, using larger doses of the peptide at volumes suitable for neuropil injections becomes problematic due to solubility and tonicity. In conclusion, these experiments demonstrate that 1) hypothalamic galanin can stimulate feeding directly, but caudal hindbrain galanin systems are more likely to be involved in the stimulation of food intake via indirect mechanisms, namely the inhibition of satiety; 2) the control of macronutrient selection by galanin and its receptors in these experimental paradigms failed to reveal a consistent pattern of activity; and 3) galanin in the lateral parabrachial nucleus is not critical to acute paradigm feeding behaviors. ACKNOWLEDGEMENTS
Special thanks to Drs. Lori Singer and Noel Calingasan for advice and help. This work was supported by grant DK 31988-09.
REFERENCES 1. Bartfai, T.; Langel, U.; Bedecs, K.; Andell, S.; Land, T.; Gregersen, S.; Ahren, B.; Girotti, P.; Consolo, S.; Corwin, R.; et al.: Galanin-receptor ligand M40 peptide distinguishes between putative galanin-receptor subtypes. Proc. Natl. Acad. Sci. USA 90:11287–11291; 1993. 2. Bedecs, K.; Berthold, M.; Bartfai, T.: Galanin—10 years with a neuroendocrine peptide. Int. J. Biochem. Cell Biol. 27:337–349; 1995. 3. Calingasan, N. Y.; Ritter, S.: Lateral parabrachial subnucleus lesions abolish feeding induced by mercaptoacetate but not by 2-deoxy-D-glucose. Am. J. Physiol. 265:R1168–1178; 1993. 4. Corwin, R. L.; Robinson, J. K.; Crawley, J. N.: Galanin antagonists block galanin-induced feeding in the hypothalamus and amygdala of the rat. Eur. J. Neurosci. 5:1528–1533; 1993. 5. Crawley, J. N.; Robinson, J. K.; Langel, U.; Bartfai, T.: Galanin receptor antagonists M40 and C7 block galanin-induced feeding. Brain. Res. 600:268–272; 1993. 6. Fathi, Z.; Cunningham, A. M.; Iben, L. G.; Battaglino, P. B.; Ward, S. A.; Nichol, K. A.; Pine, K. A.; Wang, J.; Goldstein, M. E.; Iismaa, T. P.; Zimanyi, I. A.: Cloning, pharmacological characterization and distribution of a novel galanin receptor. Brain. Res. Mol. Brain. Res. 51:49–59; 1997. 7. Koegler, F. H.; Ritter, S.: Feeding induced by pharmacological
8.
9.
10. 11. 12. 13.
blockade of fatty acid metabolism is selectively attenuated by hindbrain injections of the galanin receptor antagonist, M40. Obes. Res. 4:329–336; 1996. Koegler, F. H.; Ritter, S.: Galanin injection into the nucleus of the solitary tract stimulates feeding in rats with lesions of the paraventricular nucleus of the hypothalamus. Physiol. Behav. 63:521–527; 1998. Koegler, F. H.; York, D. A.; Berthoud, H. R.; Bray, G. A.: Central injection of the anorectic peptide a-MSH in rat activates the arcuate and lateral parabrachial nuclei. FASEB Summer Research Conference: Behavioral and metabolic sub-phenotypes in obesities, genetic and molecular aspects, pathophysiology and therapeutic implications. Snowmass, CO., June 20–25, 1998. Kyrkouli, S. E.; Stanley, B. G.; Leibowitz, S. F.: Galanin: stimulation of feeding induced by medial hypothalamic injection of this novel peptide. Eur. J. Pharmacol. 122:159–160; 1986. Leibowitz, S. F.; Kim, T.: Impact of a galanin antagonist on exogenous galanin and natural patterns of fat ingestion. Brain. Res. 599:148–152; 1992. Li, B. H.; Fowland, N. E.: Dexfenfluramine induces Fos-like immunoreactivity in discrete brain regions in rats. Brain. Res. Bull. 31:43–48; 1993. Melander, T.; Hokfelt, T.; Rokaeus, A.: Distribution of galanin-
GALANIN, M40: 3V, NTS, PBN
14.
15. 16. 17. 18.
like immunoreactivity in the rat central nervous system. J. Comp. Neurol. 248:475–517; 1986. Melander, T.; Kohler, C.; Nilsson, S.; Hokfelt, T.; Brodin, E.; Theodorsson, E.; Bartfai, T.: Autoradiographic quantitation and anatomical mapping of 125I galanin binding sites in the rat central nervous system. J. Chem. Neuroanat. 1:213–233; 1988. Odorizzi, M.; Fernette, B.; Burlet, C.; Max, J. P.; Burlet, A.: Galanin antagonists reverse the overconsumption of fat in the Brattleboro Rat. Int. J. Obes. 22(Suppl. 3):P112; S127; 1988. Ritter, S.; Dinh, T. T.: 2-Mercaptoacetate and 2-deoxy-D-glucose induce Fos-like immunoreactivity in rat brain. Brain. Res. 641:111– 120; 1994. Sakai, N.; Yamamoto, T.: Role of the medial and lateral parabrachial nucleus in acquisition and retention of conditioned taste aversion in rats. Behav. Brain. Res. 93:63–70; 1998. Skofitsch, G.; Sills, M. A.; Jacobowitz, D. M.: Autoradiographic
267
19.
20. 21. 22.
distribution of 125I-galanin binding sites in the rat central nervous system. Peptides 7:1029–1042; 1986. Smith, B. K.; Berthoud, H. R.; York, D. A.; Bray, G. A.: Differential effects of baseline macronutrient preferences on macronutrient selection after galanin, NPY, and an overnight fast. Peptides 18:207–211; 1997. Tempel, D. L.; Leibowitz, K. J.; Leibowitz, S. F.: Effects of PVN galanin on macronutrient selection. Peptides 9:309–314; 1988. Wang, S.; Hashemi, T.; Fried, S.; Clemmons, A. L.; Hawes, B. E.: Differential intracellular signaling of the GalR1 and GalR2 galanin receptor subtypes. Biochemistry 37:6711–6717; 1998. Wang, S.; He, C.; Hashemi, T.; Bayne, T.: Cloning and expressional characterization of a novel galanin receptor. Identification of different pharmacophores within galanin for the three galanin receptor subtypes. J. Biol. Chem. 272:31949–31952; 1997.
PII S0031-9384(99)00075-X
The Effects on Feeding of Galanin and M40 When Injected Into the Nucleus of the Solitary Tract, the Lateral Parabrachial Nucleus, and the Third Ventricle FRANK H. KOEGLER, DAVID A. YORK1 AND GEORGE A. BRAY Pennington Biomedical Research Center, Louisiana State University, 6400 Perkins Road, Baton Rouge, LA 70808 Received 17 December 1998; Accepted 23 March 1999 KOEGLER, F. H., D. A. YORK AND G. A. BRAY. The effects on feeding of galanin and M40 when injected into the nucleus of the solitary tract, the lateral parabrachial nucleus, and the third ventricle. PHYSIOL BEHAV 67(2) 259–267, 1999.—Several reports indicate that central injection of galanin stimulates feeding, and that there is macronutrient specificity in this response. In addition, the galanin receptor antagonist, M40, reduces food intake when injected centrally. The nucleus of the solitary tract (NTS) and the lateral parabrachial nucleus (PBN) contain galanin receptors, and are involved in the control of food intake. Hence, we sought to compare the feeding response to galanin injection into these areas with that of third ventricle (3V) galanin injection. The feeding response to injection of galanin was greatest for the 3V. Hindbrain injection of galanin stimulated food intake only at the beginning of the dark period. NTS injection of M40 inhibited intake of a macronutrient diet in food-deprived rats, but was ineffective at reducing dark-onset feeding or deprivation-induced chow intake. 3V injection of M40 did not reduce deprivation-induced intake. PBN injection of galanin at dark onset had no effect in a group of fat-preferring rats. These results suggest that hindbrain galanin may contribute to feeding by inhibiting satiety, and that hypothalamic galanin receptors are involved with stimulation of intake. Furthermore, the absence of a consistent pattern of the stimulation of macronutrient intake suggests that galanin may not be a significant effector of macronutrient selection during individual meals. © 1999 Elsevier Science Inc. Galanin
Macronutrients
NTS
1PBN
M40
Feeding
THE peptide galanin is synthesized in many brain and gut sites, and several galanin receptor subtypes from both human and rat have been cloned (6,21,22). Galanin is implicated in regulation and control of memory, nociception, food intake, and both insulin and gastric acid secretion (2). Several brain sties specifically related to feeding, such as the hypothalamus, the amygdala, the lateral parabrachial nucleus (PBN), and the caudal hindbrain contain high levels of galaninergic cells and processes, as well as galanin receptors (13,14,18). Injection of galanin into some of these areas, namely the paraventricular nucleus of the hypothalamus (PVN), the amygdala, and the nucleus of the solitary tract (NTS), stimulates feeding in a variety of paradigms (7,8,10,19,20). In addition, injection of the galanin receptor antagonist, M40 (1), into several of these sites can reduce feeding initiated by various stimuli, including 1To
palatable foods, blockade of fatty acid metabolism, and galanin itself (4,5,7). There are conflicting reports regarding the macronutrient specificity of galanin-stimulated intake in rodents. some reports suggest that galanin preferentially increases fat intake (11,20), while other suggest a stimulation of intake that is proportional to baseline intake (19). The demonstration that central injection of galanin antagonists decreases the overconsumption of fat intake by Brattleboro rats, which are deficient in AVP and overexpress galanin (15), supports a link between galanin and fat intake. We hypothesized that the activity of galanin in different brain regions would have different effects on food intake. The PBN has high levels of galanin and receives dense galaninergic innervation from the NTS as well as descending input from the hypothalamus. This interconnec-
whom requests for reprints should be addressed. E-mail: [email protected]
259
260
KOEGLER, YORK AND BRAY
tion between hypothalamic and hindbrain nuclei that are involved in feeding, and the lateral PBN’s proximity to the medial PBN, which receives gustatory information, suggested that galanin in the PBN may have a modulatory role on macronutrient selection. Thus, we hypothesized that the stimulation of galanin receptors in the hypothalamus would initiate food intake, whereas stimulation of hindbrain galanin receptors would alter macronutrient selection and suppress satiety. We further hypothesized that the activity of galanin in the PBN would not affect total caloric intake. The experiments described in this article were designed to evaluate the effects of galanin and M40 on food intake and macronutrient selection after injection into the nucleus of the solitary tract, the lateral parabrachial nucleus, and the third ventricle. MATERIALS AND METHODS
Animals, Housing, Diet All experiments used adult male Sprague–Dawley rats (Harlan–Sprague–Dawley Inc., Indianapolis, IN) weighing between 300 and 360 g at the beginning of experimentation. Upon arrival, rats underwent a 1-week quarantine period before being moved into the experimental rooms where they remained and underwent feeding tests. Rats were individually housed in hanging wire mesh cages in temperature controlled rooms (22–238C). Lights were automatically controlled and set to turn off at 1900 h and turn on 12 h later. Automatic red lighting was used for dark-cycle experimentation. Tap water was available ad lib via an automated watering system. Except during the overnight deprivation experiments, food was available at all times. Chow diet consisted of standard laboratory rodent chow (#5001; PMI Nutrition International, Inc., St Louis, MO). The diet used in the macronutrient experiments provided pure carbohydrate, protein, and fat sources, each supplemented with vitamins and minerals. These diets were made in our laboratory from commercially available ingredients (Table 1). The high-fat diet provided 55% of energy from fat at an energy density of 4.84 kcal/g. Three-choice and high-fat diets were fed to animals in individual glass food jars, whereas chow was fed from hoppers except during experiments when it was placed on the cage floors. When macronutrient or high-fat diets were used in an experiment, or when diets were switched, there was a 10-day minimum exposure time to the new diet prior to testing. Surgeries All cannulation surgeries were performed in an approved surgery suite according to institutional and federal regulations. Animals were anesthetized with Nembutal (sodium pentobarbital; 50 mg/mL/kg, IP) and treated with atropine sulfate (0.14 mg/0.25 mL, SC) to reduce anesthesia-induced respiratory distress. Routine stereotaxic implantation of 15mm long, 25-gauge, stainless steel guide cannulas was performed, and animals were allowed at least 7 days of recovery after surgery before testing commenced. Coordinates for third-ventricle cannulations were: 2.5 mm caudal to bregma, 7.7 mm subdural, in the midline. For 3V cannula implantation, the superior sagittal sinus was gently retracted away from the midline during cannula placement to prevent vessel rupture. Coordinates for the NTS were 1.8 mm rostral to the posterior occipital crest, 5.9 mm subdural, and 0.5 mm to the right side of midline. For NTS cannulations, the incisor bar was positioned at 1.5 mm, and the head was positioned in the
TABLE 1 COMPOSITON OF DIETS Three-Choice Diet
Corn starch Powdered Sugar Casein DL-Methionine Vegetable shortening Corn oil AIN-76A vitamin mix AIN-76A mineral mix Choine chloride Cellulose (Alphacel) Energy density (kcal/g)
Carbohydrate
Fat
Protein
High Fat
58.11 29.06 — 0.11 — — 0.77 3.07 0.18 8.72 3.53
— — — 0.20 75.12 — 1.49 5.95 0.34 16.91 6.85
— — 87.17 0.11 — — 0.77 3.07 0.18 8.72 3.53
23.91 — 28.69 0.14 26.60 3.20 1.06 4.21 0.24 11.95 4.84
*Ingredients expressed as percent by weight.
stereotaxic holder such that there was a nasal inclination of 58. NTS cannulas were 14 mm in length. Cannulas targeted at the lateral parabrachial nucleus were 12 mm in length. Lateral parabrachial nucleus coordinates were: 9.7 mm caudal to bregma, 5.1 mm subdural, and 6 1.9 mm from the midline. When not in use, all cannulas were obstructed with 32-gauge stainless steel stylettes made to extend 0.25 mm beyond the cannula tip. Feeding Tests All feeding tests took place in the home cages of the animals using the maintenance diets of either chow, three-choice macronutrient, or high-fat diet. At 12–20 h prior to each feeding test, food containers were replenished with fresh food. Before and during the experiments, food jar weights and spilled food were semiautomatically recorded to the nearest 0.01 g. Jars were manually removed from cages and weighed, and data from a balance was automatically captured into a spreadsheet. For experiments measuring chow intake, pellets were placed onto the cage floor and collected manually and weighed at each time point. Rats were trained to eat chow from the floor prior to the start of the experiments. Feeding tests involving overnight food deprivation consisted of removing food jars within 1 h prior to the dark period, and replacing the food immediately after treatment the following morning approximately 14 h later. The experiments were designed to minimize effects of test day on drug treatment by employing a full Latin square design whenever possible, and always including a control group on any given test day. Rats served as their own control subjects, and were not tested more than once in any consecutive 3-day period. Galanin was obtained from American Peptide Company (Sunnyvale, CA) and Bachem (Torrance, CA), and M40 was purchased from Peninsula Laboratories (Belmont, CA). Doses of galanin and M40 ranged from 0.1 to 2.5 nmol, based on previous reports (1,4,5,7), and the specific doses used for each experiment are described below and in the Results section text and figures. Central Injection Procedures Behavioral tests. Injectors for central drug administration were constructed from 32-gauge stainless steel tubing fastened to PE-10 flexible tubing (20–30 cm). The tubing was filled with distilled water and connected to a glass microliter
GALANIN, M40: 3V, NTS, PBN syringe to allow manual microinfusion of volumes of 300 nL up to 1 mL. The meniscus between the injection solution and water was visible in the clear tubing, and allowed confirmation of solution delivery. Infusions were performed manually while the rat was in its home cage, and lasted between 30 and 120 s, depending on the solution volume and brain area. Injections into the neuropil were 300 nL in volume, whereas ventricular injections were 1 mL in volume (see individual experiments). After the infusions, injectors were left in place for a minimum of 30 additional seconds to reduce back pressure and possible reflux when the injector was removed and replaced by the stylette. The procedure consisted of removing an animal from its home cage, removing the stylette, inserting the prefilled injector, and replacing the animal in its cage for the infusion. After the postinfusion period, the animal was immediately removed from the cage, the injector was replaced by a stylette, and the animal was replaced into its cage for the remainder of the test. Excitotoxin lesions. An injection procedure similar to that described above was used to deliver ibotenic acid (3 mg/300 nL/side; Sigma Chemical Company, St. Louis, MO) to the lateral parabrachial nuclei. Pilot experiments revealed that simultaneous bilateral ibotenic acid injections were often lethal; therefore, lesions were made unilaterally and a 1-week recovery period was allowed before the contralateral lesion was performed. Specific Experiments Experiment 1: Effect of 3V and NTS galanin and M40 injection on macronutrient diet intake during daytime. Cannulas were implanted into the NTS of 12 animals, and another 12 animals received cannulas targeted to the third ventricle. Doses of galanin injected into the 3V and NTS were 0 (saline), 0.1, 0.5, and 2.5 nmol. These experiments began at approximately 0900 h. After the galanin experiments, the same group of rats was tested with the galanin antagonist, M40. M40 (0.5 nmol) or saline was injected after overnight food deprivation and food intake was measured after 30, 60, and 120 min. Experiment 2: Effect of NTS galanin and M40 on early dark-period intake of macronutrient diet. The 24 rats with NTS cannulas were tested with galanin and M40 (0.5 or 1.5 nmol for each) during the first hour of the dark phase. To make timed central injections possible, the group of 24 was split into two and tested on successive days, with an identical treatment regimen for both groups. Intake was measured for 30 and 60 min after the lights went off, and overnight. Prior to galanin and M40 drug tests, two 48-h periods of baseline intake were measured. Experiment 3: Effect of NTS injection of M40 or galanin on daytime chow intake and the effect of galanin on high-fat diet intake. A group of 12 rats maintained on chow diets with cannulas targeted at the NTS was tested during the daytime (0900 h) with both saline vehicle (0.3 mL) and M40 (0.5 nmol) after overnight food deprivation. Food intake was measured after 30 and 60 min. Because galanin was ineffective in stimulating daytime feeding when rats were eating a chow diet that is relatively low in fat content, the diet was switched to a highfat composite diet (Table 1). Experiment 4: Effect of PBN galanin on daytime chow intake and dark-period macronutrient intake. A separate group of 12 rats received bilateral PBN injection of galanin (0.5 nmol) or saline vehicle (0.3 nmol) at 0900 h, and intake of chow was measured at 1 and 3 h. The same group of rats was adapted to the three-choice macronutrient diet for 10 days
261 and retested with bilateral galanin (0.5 or 1.5 nmol) or vehicle during the early part of the dark period when rats normally increase food intake. Food intake was measured 30 min, 60 min, and 14 h after injection of galanin into the PBN. Experiment 5: Effect of PBN galanin on dark-onset macronutrient intake. Another group of 14 rats was implanted with bilateral PBN cannulas and was maintained and tested with the three-choice macronutrient diet. Galanin (1.5 nmol) or vehicle was injected bilaterally into the PBN immediately after the onset of the dark period, and intake was measured after 1 and 12 h. After completion of the galanin injection experiments, these animals were lesioned with ibotenic acid to determine if there would be a change in macronutrient intake after chemical ablation of the PBN. Histology After completion of the feeding tests, all animals were prepared for brain histology by deep anesthetization and transcardial perfusion with buffer followed by perfusion with formaldehyde. Brains were removed, sectioned, and stained with cresyl violet. With the aid of a light microscope, cannula tracks were mapped onto brain atlas map drawings to confirm proper cannula location. Histological criteria for exclusion from data analysis were nay of the following: 1) cannulas more than 0.3 mm away from the target nucleus (for neuropil injections); 2) destruction of the target nucleus (except for the lesion experiments); 3) extraventricular cannula (for ventricular injection experiments). Ibotenic acid lesions were evaluated by inspecting the injection site for gliosis and loss of neuronal cell bodies. If data from an animal was eliminated because of meeting the exclusion criteria, it is noted in the results section for that particular experiment. Data Analysis Data are presented as the mean intake in kilocalories (or grams) 6 SEM. For experiments with multiple time-point observations, intake data for subsequent time points is presented as cumulative. Because all animals received both control and drug treatments, all statistical tests used either paired (t-test) or repeated measures (analysis of variance) design. Most tests used two-factor, mutilevel, repeated-measures analysis of variance (ANOVA), computed with either SigmaStat 2.03 (SPSS Inc.), or the SAS system (SAS Institute). Unless otherwise noted, when multiple post hoc comparisons were performed on a series of data, p-value corrections were made to reduce Type I significance errors (Bonferroni’s correction). As a consequence, some statements of significance tend to be conservative. RESULTS
Experiment 1a: Effect of 3V and NTS Galanin Injection on Spontaneous Three-Choice Macronutrient Diet Intake During Daytime One animal from each group was eliminated due to improper cannula placement. The 24-h baseline measurements revealed a relatively stable composition of macronutrient intake throughout the duration of the experiments, with animals selecting calories primarily from carbohydrates. Third ventricle cannulated animals selected approximately 57% of daily calories from carbohydrate, 27% from fat, and 16% protein. Animals with NTS cannulas selected 53%, 33%, and 15% of daily calories from carbohydrate, fat, and protein, respectively.
262
KOEGLER, YORK AND BRAY
At 30 min there was a main effect of 3V injection of galanin on total food intake, F(3, 11) 5 6.76, p 5 0.00112 (Fig. 1, left side). Bonferroni-adjusted multiple comparison revealed that total intake after 2.5 nmol was greater than after saline (p , 0.05). There also were significant main effects of 3V drug treatment and macronutrient intake, F(3, 33) 5 6.75, p 5 0.0011; F(2, 22) 5 9.62, p 5 0.0010. There was an interaction between macronutrient and galanin, F(6, 66) 5 4.67, p 5 0.0005. Comparison of dose effects collapsed across macronutrients revealed a statistically significant, dose-dependent intake for 0.5 and 2.5 nmol galanin. At 30 min, multiple comparisons to saline revealed that carbohydrate intake after 2.5 nmol galanin was greater than after saline (p , 0.05). The 2.5nmol galanin dose marginally stimulated protein intake (p 5 0.08). No dose of galanin injected into the 3V reliably increased fat intake relative to saline. At 60 min the results of statistical analysis of the 3V galanin response similar to results obtained at 30 min [main effect on total intake, F(3, 11) 5 5.76, p 5 0.00277]. Bonferroni’s comparison revealed that intake after 2.5 nmol was greater than after saline (p , 0.05). A twofactor repeated-measures ANOVA revealed a significant effect of galanin and macronutrient on intake, F(3, 33) 5 5.76, p 5 0.0028, F(2, 22) 5 11.8, p 5 0.0003. There was an interaction between macronutrient and treatment, F(6, 66) 5 2.26, p 5 0.0059. At 60 min multiple comparisons of galanin-induced intake to intake after saline revealed that carbohydrate intake after 3V injection of 2.5 nmol galanin was greater than after
saline (p , 0.05). There was no statistically reliable increase in fat or protein intake at 60 min. Galanin (0, 0.1, 0.5, 2.5 nmol) failed to affect energy intake or macronutrient selection when injected into the NTS (Fig. 1, right side). Total intake at 30 and 60 min was not affected by galanin treatment [one-way, repeated-measures ANOVA, F(3, 11) 5 0.477, p 5 0.701; F(3, 11) 5 0.0974, p 5 0.961, respectively]. There was no effect of galanin on intake of any particular macronutrient at either 30 or 60 min, F(3, 10) 5 0.789, p 5 0.5096; F(3, 10) 5 0.124, p 5 0.9454, respectively. Experiment 1b: Effect 3V and NTS M40 on DeprivationInduced Intake of Three Choice Macronutrient Diet Two-way, repeated-measures ANOVA with factors of cannulation site and drug treatment revealed a marginally significant effect of M40 on the suppression of food intake (Fig. 2) at 30 min, F(1, 18) 6 3.7587, p 5 0.0684, but no interaction, F(1, 18) 5 0.04681, p 5 0.8311. Total food intake after 60 min was suppressed by M40, F(1, 18) 5 6.88, p 5 0.0173. There was no difference in food intake with cannulation site, F(1, 18) 5 2.45, p 5 0.1351, nor any interaction effect, F(1, 18) 5 1.33, p 5 0.2642. Effects of M40 after 120 min paralleled those at 60 min, with a main effect of M40 but no effect of cannulation site, nor was there an interaction between M40 treatment and cannulation site. A two-way, repeated-measures ANOVA (treatment 3 macronutrient) of the effect of M40 on individ-
FIG. 1. Effect of galanin injected into the 3V or NTS. Galanin reliably stimulated 30 and 60 min intake when injected into the 3V but not when injected into the NTS. The greatest effect on the stimulation of intake was for carbohydrate. There was a trend towards the stimulation of fat intake; however, after adjusting for multiple comparisons, this result was not statistically relevant. “a” denotes difference from saline, p , 0.05.
GALANIN, M40: 3V, NTS, PBN ual macronutrients for 3V injection site at 30, 60, and 120 min failed to reveal any significant effects. Analysis of individual macronutrient intake following M40 injection into the NTS failed to reveal significant effects at 30 min (p 5 0.1240). After 60 min, however, M40 injection into the NTS significantly suppressed total energy intake, F(1, 9) 5 6.818, p 5 0.0282. There was no statistically significant interaction between macronutrient and treatment, F(2, 18) 5 0.684, p 5 0.5173. Experiment 2: Effect of Galanin and M40 Injected Into the NTS on Early Dark-Period Intake of Three-Choice Macronutrient Diet Analysis of baseline intake revealed a macronutrient composition of 45 6 6, 36 6 7, and 18 6 6% from carbohydrate, fat, and protein, respectively. From the 23 animals included after histological analysis of brain tissue, six had greater than two-thirds of total daily intake from carbohydrate; one animal had greater than two-thirds of its total intake from fat.
263 There was a main effect of galanin on total energy intake (Fig. 3) measured from the start of the dark phase, F(2, 21) 5 4.06, p 5 0.0244. Repeated-measures ANOVA of carbohydrate intake revealed a significant effect of drug treatment, F(2, 21) 5 6.33, p 5 0.0031. There was no effect of galanin on fat intake, F(2, 21) 5 0.607, p 5 0.5496. There was no main effect of M40 on intake, F(2, 21) 5 0.9564, p 5 0.4004. Multiple comparisons revealed that there was no effect of M40 on intake of any macronutrient during the first 30 min of the dark period. Experiment 3: Effect of NTS Injection of M40 or Galanin on Daytime Chow Intake Ten days after the M40 injection experiments, the response to galanin (1 nmol) during daytime feeding of rodent chow was measured in the same group of nonfood-deprived rats (Fig. 4). Galanin had no effect on chow intake at any time point, F(1, 11) 5 0.575, p 5 0.464.
FIG. 2. Effect of the galanin receptor antagonist, M40, on overnight deprivationinduced refeeding. M40 (0.5 nmol) injected into the nucleus of the solitary tract, but not 3V, suppressed 30-, 60-, and 120-min intake induced by overnight food deprivation. “a” denotes difference from saline, p , 0.05.
264
KOEGLER, YORK AND BRAY tively. Bilateral injection of 0.5 or 1.5 nmol of galanin into the PBN was ineffective in altering total macronutrient intake, or changing the percentage of macronutrient intake during the dark phase (Fig. 5). Experiment 5: Effect of PBN Galanin on Dark Onset Intake of a Three-Choice Macronutrient Diet (Carbohydrate-Preferring Animals)
FIG. 3. Effect of galanin and M40 injection into the NTS during dark-onset feeding of a three-choice macronutrient diet. Galanin dose dependently increased carbohydrate intake during the initial 30 min after treatment. M40 had no effect. “a” denotes difference from saline; “b” denotes difference from other dose of same peptide, p , 0.05.
Subsequently, rats were adapted to a high-fat diet and retested. There was no effect of galanin injection (1 nmol) into the NTS when rats were tested during the light period (0900 h), F(1, 11) , 0.001, p 5 0.987; data not shown. Experiment 4: Effect of PBN Galanin on Daytime Chow Intake and Dark-Period Macronutrient Intake (Fat-Preferring Animals) Bilateral PBN injection of 0.5 nmol galanin during daytime did not stimulate 1 or 3 h intake of rodent chow when rats were maintained with a chow diet (data not shown). After adaptation to the three-choice macronutrient diet analysis of basal diet intake revealed that this group of animals was particularly and unusually fat preferring, with a mean percentage of daily energy intake from fat being 60 6 5%. Protein and carbohydrate intake were 10 6 1% and 30 6 4%, respec-
In contrast to the animals in the previous experiment, baseline 24-h intake preference of this group averaged 60% for carbohydrate, 26% for rat, and 14% for protein. Injection of 1.5 nmol galanin bilaterally into the PBN immediately after the onset of the dark period failed to increase total intake when measured after 1 or 12 h (Fig. 6). There was no stimulation of carbohydrate (p 5 0.211, not significant) but an inhibition of fat intake (p 5 0.0428) during the first hour, suggesting a shift in intake. Comparisons of prelesion diet intake with the amount eaten 6 weeks after ibotenic acid lesions showed that both groups decreased fat intake and increased carbohydrate intake. However, the lesioned group decreased its fat intake to a lesser degree, whereas the sham-lesioned group increased carbohydrate intake to a greater degree. This effect was not statistically reliable (lesion effect: p 5 0.07). It is noteworthy that this group of animals was predominantly carbohydrate preferring. There was no effect of the lesions on body weight. DISCUSSION
The major observations reported in this study were that the 3V injection site is the most effective site at which galanin stimulates intake, that M40 was ineffective at decreasing deprivation-induced intake when injected into the 3V, and that galanin was effective at stimulating intake only during the dark-onset feeding period when injected into the NTS. At this time galanin weakly stimulated carbohydrate intake, which was the preferred macronutrient of the group of animals tested. This effect was not present when rats were tested during the daytime with a three-choice macronutrient diet, high-
FIG. 4. Effect of galanin and M40 (NTS) on chow intake. Both galanin and M40 failed to affect intake of a chow diet in 1-h tests. Galanin was injected during the daytime in presumed sated rats; M40 was injected in the daytime in rats that had been overnight food deprived. (Note: ordinate scale is grams.)
GALANIN, M40: 3V, NTS, PBN
265
FIG. 6. Effect of galanin injection into the PBN on dark-onset feeding of a three-choice macronutrient diet. In a group of carbohydrate preferring rats, galanin did not affect total intake during the first hour following treatment nor during the following 11 h. Fat intake was suppressed during the 1-h period. “a” denotes difference from saline, p , 0.05.
FIG. 5. Effect of galanin injection into the PBN during the darkonset and overnight feeding period. Galanin injected bilaterally into the PBN failed to affect intake of a three-choice macronutrient diet. There was a minimal, nonsignificant stimulation of overnight fat intake in this group of fat-preferring animals.
fat diet, or a chow diet. M40 was effective at reducing threechoice diet intake stimulated by an overnight fast, but had no effect on initial dark-onset feeding, or deprivation-induced intake of rodent chow. In addition, galanin injection into the PBN did not stimulate dark-onset feeding in a group of fat-preferring rats. However, in a separate group of carbohydrate-preferring rats, the
same dose of galanin increased carbohydrate intake and inhibited fat intake. PBN galanin injection had no effect on spontaneous daytime intake when a chow diet was used as the maintenance diet. Finally, bilateral lesion of the PBN with ibotenic acid had no effect on body weight, but had a tendency to shift intake towards fat and away from carbohydrate relative to nonlesioned, intact animals. The results of these experiments suggest that the role of extrahypothalamic galanin in the regulation of food intake is a secondary, modulatory one, rather than a primary effector of meal initiation or macronutrient selection. In comparison to the powerful stimulatory effect of NPY or inhibitory effect of CRH, the relatively weak effects of galanin and M40 on food intake and macronutrient selection, regardless of time of day, diet, or injection site, establish a lesser position for this peptide in the control of food intake. The effects of hindbrain injection of galanin on food intake are weak in comparison with that of third ventricle injection. There were no effects of hindbrain galanin on spontaneous daytime intake, regardless of maintenance diet. This is in contrast to previous reports in which the stimulatory effect of hindbrain injection of galanin was more pronounced during the daytime. Because the previous experiments utilized a semisynthetic high-fat diet, it is possible the differences are
266
KOEGLER, YORK AND BRAY
diet related. Only in carbohydrate-preferring rats did hindbrain galanin stimulate dark-onset food intake. These results are incongruent with some previous reports that suggest a relationship between fat intake and galanin administration. Previous reports suggest a link between galanin and both fat ingestion and body weight (20). However, the macronutrient diet fat source in the current experiments was primarily of vegetable origin (shortening) rather than fat of animal origin (lard). This difference may be responsible for the disparate results. In another study that used the same diet, intraventricular galanin injection increased the intake a preferred macronutrient, i.e., that which was most often selected during daily intake, rather than the intake of fat intake per se (19). The hypothesis that hindbrain injections of galanin would stimulate fat intake was not supported; however, it was shown that the galanin receptor antagonist, M40, was most effective at reducing deprivation-induced food intake when injected into the hindbrain rather than in the third ventricle. This suggests that hindbrain galanin receptors may not be involved in the outright stimulation or initiation of food intake, but rather, may contribute to intake by reducing tonic or meal-induced satiety signals. Specifically, galanin receptors in the NTS, being primarily inhibitory to neuronal membrane potential, may serve to reduce the effect of ascending visceral afferent information that is classically thought to reduce food intake. This would permit an organism in a food-deprived state to override viscerally derived meal termination signals such as gastric distention, thus allowing greater food intake. It is interesting that M40 injections into the NTS had no effect on dark-onset feeding in contrast to its inhibitory effect on deprivation-induced feeding. It is possible that the satiety signals, for example, gastric distension, and other postprandial signals, may be greater in the deprivation-refeeding state, rather than during the normal dark-onset feeding state. Food intake after an overnight deprivation was greater than that during spontaneous daytime feeding or dark onset feeding. Therefore, interference with a galaninergic inhibition of a satiety signal may not be as easily detectable in the dark-onset feeding daytime feeding paradigm.
The lack of a robust effect of galanin on dark-onset feeding, or daytime feeding when injected into the PBN suggests that galanin receptors in this nucleus are not intrinsically involved in the initiation of feeding. The PBN is classically recognized as a relay nucleus for ascending visceral afferent information, and has direct and reciprocal connections with the nucleus of the solitary tract, as well as the hypothalamus. Indeed, the galaninergic projection from the NTS is particularly well defined, and the terminal fields of these fibers are very dense in the central, dorsal, and external parts of the PBN. In particular, these areas of the PBN are thought to be involved in the control of food intake: many treatments that either stimulate (b-mercaptoacetate, 2-deoxyglucose) or inhibit (dexfenfluramine, a-melanocyte stimulating hormone) food intake activate neurons in this area, as shown by induction of the c-fos gene (9,12,16). Furthermore, lesions in these areas can prevent the development of conditioned taste aversions and the ability to respond with feeding to lipoprivic stimuli (3,17). Despite this information, galanin was ineffective at affecting food intake when injected into this area. It is also possible that the doses of galanin used in these experiments were not high enough to affect feeding behavior; however, using larger doses of the peptide at volumes suitable for neuropil injections becomes problematic due to solubility and tonicity. In conclusion, these experiments demonstrate that 1) hypothalamic galanin can stimulate feeding directly, but caudal hindbrain galanin systems are more likely to be involved in the stimulation of food intake via indirect mechanisms, namely the inhibition of satiety; 2) the control of macronutrient selection by galanin and its receptors in these experimental paradigms failed to reveal a consistent pattern of activity; and 3) galanin in the lateral parabrachial nucleus is not critical to acute paradigm feeding behaviors. ACKNOWLEDGEMENTS
Special thanks to Drs. Lori Singer and Noel Calingasan for advice and help. This work was supported by grant DK 31988-09.
REFERENCES 1. Bartfai, T.; Langel, U.; Bedecs, K.; Andell, S.; Land, T.; Gregersen, S.; Ahren, B.; Girotti, P.; Consolo, S.; Corwin, R.; et al.: Galanin-receptor ligand M40 peptide distinguishes between putative galanin-receptor subtypes. Proc. Natl. Acad. Sci. USA 90:11287–11291; 1993. 2. Bedecs, K.; Berthold, M.; Bartfai, T.: Galanin—10 years with a neuroendocrine peptide. Int. J. Biochem. Cell Biol. 27:337–349; 1995. 3. Calingasan, N. Y.; Ritter, S.: Lateral parabrachial subnucleus lesions abolish feeding induced by mercaptoacetate but not by 2-deoxy-D-glucose. Am. J. Physiol. 265:R1168–1178; 1993. 4. Corwin, R. L.; Robinson, J. K.; Crawley, J. N.: Galanin antagonists block galanin-induced feeding in the hypothalamus and amygdala of the rat. Eur. J. Neurosci. 5:1528–1533; 1993. 5. Crawley, J. N.; Robinson, J. K.; Langel, U.; Bartfai, T.: Galanin receptor antagonists M40 and C7 block galanin-induced feeding. Brain. Res. 600:268–272; 1993. 6. Fathi, Z.; Cunningham, A. M.; Iben, L. G.; Battaglino, P. B.; Ward, S. A.; Nichol, K. A.; Pine, K. A.; Wang, J.; Goldstein, M. E.; Iismaa, T. P.; Zimanyi, I. A.: Cloning, pharmacological characterization and distribution of a novel galanin receptor. Brain. Res. Mol. Brain. Res. 51:49–59; 1997. 7. Koegler, F. H.; Ritter, S.: Feeding induced by pharmacological
8.
9.
10. 11. 12. 13.
blockade of fatty acid metabolism is selectively attenuated by hindbrain injections of the galanin receptor antagonist, M40. Obes. Res. 4:329–336; 1996. Koegler, F. H.; Ritter, S.: Galanin injection into the nucleus of the solitary tract stimulates feeding in rats with lesions of the paraventricular nucleus of the hypothalamus. Physiol. Behav. 63:521–527; 1998. Koegler, F. H.; York, D. A.; Berthoud, H. R.; Bray, G. A.: Central injection of the anorectic peptide a-MSH in rat activates the arcuate and lateral parabrachial nuclei. FASEB Summer Research Conference: Behavioral and metabolic sub-phenotypes in obesities, genetic and molecular aspects, pathophysiology and therapeutic implications. Snowmass, CO., June 20–25, 1998. Kyrkouli, S. E.; Stanley, B. G.; Leibowitz, S. F.: Galanin: stimulation of feeding induced by medial hypothalamic injection of this novel peptide. Eur. J. Pharmacol. 122:159–160; 1986. Leibowitz, S. F.; Kim, T.: Impact of a galanin antagonist on exogenous galanin and natural patterns of fat ingestion. Brain. Res. 599:148–152; 1992. Li, B. H.; Fowland, N. E.: Dexfenfluramine induces Fos-like immunoreactivity in discrete brain regions in rats. Brain. Res. Bull. 31:43–48; 1993. Melander, T.; Hokfelt, T.; Rokaeus, A.: Distribution of galanin-
GALANIN, M40: 3V, NTS, PBN
14.
15. 16. 17. 18.
like immunoreactivity in the rat central nervous system. J. Comp. Neurol. 248:475–517; 1986. Melander, T.; Kohler, C.; Nilsson, S.; Hokfelt, T.; Brodin, E.; Theodorsson, E.; Bartfai, T.: Autoradiographic quantitation and anatomical mapping of 125I galanin binding sites in the rat central nervous system. J. Chem. Neuroanat. 1:213–233; 1988. Odorizzi, M.; Fernette, B.; Burlet, C.; Max, J. P.; Burlet, A.: Galanin antagonists reverse the overconsumption of fat in the Brattleboro Rat. Int. J. Obes. 22(Suppl. 3):P112; S127; 1988. Ritter, S.; Dinh, T. T.: 2-Mercaptoacetate and 2-deoxy-D-glucose induce Fos-like immunoreactivity in rat brain. Brain. Res. 641:111– 120; 1994. Sakai, N.; Yamamoto, T.: Role of the medial and lateral parabrachial nucleus in acquisition and retention of conditioned taste aversion in rats. Behav. Brain. Res. 93:63–70; 1998. Skofitsch, G.; Sills, M. A.; Jacobowitz, D. M.: Autoradiographic
267
19.
20. 21. 22.
distribution of 125I-galanin binding sites in the rat central nervous system. Peptides 7:1029–1042; 1986. Smith, B. K.; Berthoud, H. R.; York, D. A.; Bray, G. A.: Differential effects of baseline macronutrient preferences on macronutrient selection after galanin, NPY, and an overnight fast. Peptides 18:207–211; 1997. Tempel, D. L.; Leibowitz, K. J.; Leibowitz, S. F.: Effects of PVN galanin on macronutrient selection. Peptides 9:309–314; 1988. Wang, S.; Hashemi, T.; Fried, S.; Clemmons, A. L.; Hawes, B. E.: Differential intracellular signaling of the GalR1 and GalR2 galanin receptor subtypes. Biochemistry 37:6711–6717; 1998. Wang, S.; He, C.; Hashemi, T.; Bayne, T.: Cloning and expressional characterization of a novel galanin receptor. Identification of different pharmacophores within galanin for the three galanin receptor subtypes. J. Biol. Chem. 272:31949–31952; 1997.