| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
-Expressing Cells in the Nucleus Tractus Solitarius of Ovariectomized RatsPhysiology and Behaviour Group (L.A., N.G.), Institute of Animal Science, ETH (Swiss Federal Institute of Technology) Zurich, 8603 Schwerzenbach, Switzerland; and Department of Psychiatry (N.G.), Weill Medical College of Cornell University, New York, New York 10027
Address all correspondence and requests for reprints to: Lori Asarian, Ph.D., Institute of Animal Science, ETH (Swiss Institute of Technology) Zürich, Schorenstrasse 16, 8603 Schwerzenbach, Switzerland. E-mail: lasarian{at}ethz.ch.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
) signaling, inhibits feeding in rats and mice is increasing the satiating potency of cholecystokinin (CCK) acting on peripheral CCK-1 receptors. Ingested lipid is a principal secretagogue of intestinal CCK, and intraduodenal lipid infusions elicit CCK-mediated satiation in animals and humans. Here we tested whether estradiol affects the satiating potency of intraduodenal lipid infusions in ovariectomized rats and, using c-Fos immunocytochemistry, searched for potential brain sites of ER involved. Food-deprived ovariectomized rats with open gastric cannulas sham fed 0.8 M sucrose 2 d after estradiol (estradiol benzoate, 10 µg, sc) or vehicle injection. Estradiol markedly increased the satiating potency of intraduodenal infusions of Intralipid but not the satiating potency of L-phenylalanine (10 min infusions, 0.44 ml/min, 0.13 kcal/ml), which in male rats satiates via a CCK-independent mechanism. Estradiol had no significant effect in rats pretreated with the CCK-1 receptor antagonist Devazepide (1 mg/kg, ip). The effect of estradiol on intraduodenal Intralipid-induced satiation was mirrored by selective increases in the number of cells expressing c-Fos immunoreactivity in a circumscribed region of the nucleus tractus solitarius (NTS), just caudal to the area postrema (cNTS) but not elsewhere in the NTS or the hypothalamic paraventricular or arcuate nuclei. In addition, a significant proportion of cNTS c-Fos-positive cells also expressed ER. These data provide behavioral and cellular evidence that estradiol-ER signaling in cNTS neurons increases the satiating potency of endogenous CCK released in response to ingested lipid. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In female rats, the satiating potency of CCK is modulated by an activational effect of estradiol (14, 15, 16). In ovariectomized (OVX) rats, the actions on feeding of CCK and of CCK antagonism were increased by estradiol treatment (17, 18, 19, 20), and in intact, cycling females, CCK-1 receptor antagonism increased feeding more during estrus (i.e. the periovulatory period of the ovarian cycle) than during diestrus (17, 21). These data are likely to be directly relevant to human eating because food intake also decreases during the periovulatory phase of the ovarian cycle in women (reviewed in Refs. 14, 15, 16).
The mechanisms mediating the effect of estradiol on CCK satiation have not been fully determined (14, 15, 16). Differences in CCK secretion do not appear to account for the increase in CCK satiation in estradiol-treated OVX rats because, as described above, the satiating effect of exogenous CCK is also increased. Estrogen receptor (ER)- seems to be necessary because CCK-1 receptor antagonism did not increase food intake in transgenic mice lacking ER (22). Clues as to the site of the crucial ER come from immunocytochemical studies of CCK- and food-induced expression of c-Fos, a marker of increased neuronal electrophysiological activity. In OVX rats, estradiol increased the number of cells expressing c-Fos after either food intake or ip CCK injection in regions of the nucleus tractus solitarius (NTS) that receive gut vagal afferents, the paraventricular nucleus of the hypothalamus (PVN), and the central nucleus of the amygdala (CeA) (23, 24). Four aspects of these data were interesting. First, feeding and ip CCK did not increase c-Fos expression in several brain areas in female rats that did respond in similar tests of male rats, including the area postrema (AP) and ventrolateral medulla (25, 26, 27, 28, 29). This may reflect a sex difference, the smaller amounts of food and smaller CCK doses used by us, or some other difference in method. Second, the number of cells expressing c-Fos in the NTS, PVN, and CeA also increased as a function of the amount of food ingested during test meals. This result is similar to a previous report that small ip CCK doses produced graded c-Fos responses in the NTS (30) and indicates the utility of c-Fos as a correlate of the neural representation of the graded, postingestional intrameal negative feedback that leads to satiation (31). Third, the NTS, PVN, and CeA all contain ER (32, 33, 34). Fourth, in transgenic mice lacking ER, CCK did not increase c-Fos expression in the NTS (22). These data, taken together with the prior result that estradiol did not increase the expression of CCK-1 receptors in the NTS (35), suggest that estradiol may act in the NTS to influence the neural processing of the incoming vagal CCK satiation signal.
Here we present a series of experiments further investigating the role of estradiol in intestinal satiation. We tested OVX rats that received intraduodenal nutrient infusions to determine: 1) whether estradiol differentially affects the potency of nutrients that induce satiation via CCK-dependent (e.g. Intralipid, Kabi Pharmacia, Stockholm, Sweden) or CCK-independent [e.g. L-phenylalanine; L-Phen)] mechanisms (3, 4, 5); 2) whether Intralipid or L-Phen-induced c-Fos expression in the brain is differentially affected by estradiol; and 3) whether Intralipid or L-Phen induce c-Fos expression in neurons that express ER.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Surgery and estradiol treatment
In a single procedure, rats were OVX and implanted with chronic gastric sham-feeding cannulas and intraduodenal infusion catheters, as described previously (19, 36). Briefly, rats were deprived of food overnight and anesthetized with ip injections (1 ml/kg) of a combination of 70 mg/kg ketamine (Ketaset; Fort Dodge, Fort Dodge, IA) and 4.5 mg/kg xylazine (Rompun; Henry Schein, Melville, NY). A midline 3-cm laparotomy was done. Rats were first bilaterally OVX. Then a small stab wound was made in the limiting ridge of the stomach and the duodenal catheter, a 30-cm length of SILASTIC brand silicon tubing (Dow Corning Corp., Midland, MI; 0.075 mm inner diameter, 0.17 mm outer diameter; Technical Products, Decatur, GA) was led 7 cm into the duodenum and sutured to the intestinal wall, approximately 2 cm distal to the pylorus, using 7–0 Prolene suture (Ethicon, VWR, Piscataway, NJ), and through a 5 x 5 mm square of surgical mesh (Marlex; Bard Implants, Billerica, MA) that was placed on the serosal surface of the intestine for reinforcement. The distal end of the catheter was passed through holes in the flange of a custom-made stainless steel gastric cannula and was held in place by friction. The flanged end of the cannula was then inserted into the stomach and held in place by a 3–0 Vicryl (Ethicon, VWR) purse string suture. An annular piece of surgical mesh, approximately 2 cm outer diameter, was placed around the shaft of the cannula to promote adhesion to the abdominal muscle wall. The gastric cannula was then exteriorized through a stab wound about 2 cm to the left of the midline, and a washer was screwed onto the externalized cannula shaft. The end of the intraduodenal catheter was jam-fit onto the flanged nub of the screw cap that closed the gastric cannula. When the screw cap was opened, several centimeters of the catheter were accessible. When the cannula was closed, the catheter was coiled in the stomach. The presence of the duodenal catheter in the stomach did not appear to interfere with food intake, weight gain, or gut function. The washer was removed after 3–5 d and preoperative body weight was recovered within 5–6 d.
After 7–10 d of recovery, rats were divided into two groups of approximately equal body weights. One group received intrascapular sc injections of 10 µg 17ß-estradiol-3-benzoate (Sigma, St. Louis, MO) in 100 µl sesame oil (Sigma) and the other 100 µl sesame oil alone. Injections were done between 0900 and 0930 h every Tuesday and Wednesday. Rats were tested on Fridays. This cyclic estradiol treatment produces plasma estradiol concentrations that are sufficient to maintain normal body weight and, when combined with progesterone treatments on Fridays, normal sexual receptivity (17, 19, 37, 38). Furthermore, it does not produce progressive changes in behavioral and neurochemical measures that are produced by continuous estradiol treatment, e.g. with estradiol-containing silicone capsule implants (37, 38). We tested the effects of estradiol on intestinal satiation on Fridays because previously it increased the satiating potency of CCK and increased food- and ip CCK-induced c-Fos expression in several brain areas (23, 24).
Sham feeding
Sham feeding adaptation was begun approximately 10 d after surgery. On 4–5 d/wk food, but not water, was removed at 1630 h. Between 1000 and 1020 h the next day, the gastric cannula was opened, the outer end of the duodenal catheter was externalized, and the stomach was lavaged with 10-ml aliquots of warm 0.15 M NaCl until the drainage was free of food particles. The duodenal catheter was attached to a longer length of the same-size SILASTIC brand tubing, which was passed through a larger (0.27 mm inner diameter, 0.32 mm outer diameter; Technical Products) SILASTIC brand drainage tube sheathed in a stainless steel spring. The proximal end of the drainage tube was attached to the gastric cannula, the distal end and the duodenal catheter were passed through a slot in the cage floor, and finally the duodenal catheter was attached to an infusion pump (Harvard Bioscience, South Natick, MA). Rats could move around freely during sham-feeding tests. Rats were offered 0.8 M sucrose solution in graduated burettes (±1.0 ml) from 1030 to 1115 h. Intraduodenal infusions of saline (0.15 M NaCl; Sigma; 0.44 ml/min for 10 min) began at 1036 h. We began the infusion at this time point because this is when intraduodenal nutrient infusions produced the most potent inhibitory effect on sham feeding (39). Sham intakes were recorded every 3 min. When 45-min sucrose sham intakes did not vary significantly on 3 consecutive days, nutrient infusion tests began.
Adaptation trials were done Monday through Thursday, and nutrient infusion tests were done only on Fridays. During tests the gastric drainage was collected in pans under the cages. Data from tests in which the volume of drainage collected during the test was not at least as much as the volume of the sucrose and water ingested during the test were not used. In addition, saline and L-Phen solutions were colored with green food dye to detect reflux of intestinal infusate into the stomach, and data from tests in which green- (saline or L-Phen) or white (Intralipid)-colored drainage appeared were not used.
After sham feeding tests, the drainage tubes were removed, the duodenal catheter reattached to the gastric cannula screw cap and coiled into the stomach, the screw cap replaced, and chow returned.
Experiment 1. Estradiol and intestinal satiation
As described above, endogenous CCK mediates intestinal satiation elicited by intraduodenal infusions of Intralipid, but not L-Phen, in male rats (3, 4, 5, 40), and estradiol potently affects CCK-induced satiation in female rats (15, 16, 17). Intestinal satiation has not been tested in female rats. Therefore, we sought to determine whether estradiol differentially affects Intralipid and L-Phen induced intestinal satiation in OVX rats and whether any such difference is CCK mediated.
Experiment 1a: effects of estradiol on intestinal satiation.
Six estradiol-treated and seven oil-treated OVX rats were tested. Two independent cross-over designs were done, with order randomized for individual rats. One cross-over test compared intraduodenal infusions of 10% Intralipid (prepared by mixing saline with 20% Intralipid, Kabi Pharmacia) vs. saline, and the other compared intraduodenal infusions of 50 mg/ml L-Phen (Sigma) dissolved in saline vs. saline. Infusate pH was adjusted to 7.4 with HCl or NaOH (Sigma) as required and tonicity to near 300 mOsM with NaCl (Sigma), as checked using a vapor pressure osmometer (Wescor 5130A, Logan, UT). Infusates were brought to room temperature before use. The Intralipid infusion approximates the intrameal rate of delivery of lipid to the intestine when male rats ingest similar lipid emulsions (3, 6) and, also in males, produced a marked, but submaximal, inhibition of sham feeding under similar conditions that was reversed by pretreatment with a CCK-1 receptor antagonist (36). Intestinal infusion of this dose of L-Phen also significantly inhibited sham feeding in male rats, but this inhibition was not reduced by CCK-1 receptor antagonist pretreatment (40).
Experiment 1b: effects of CCK-1 receptor antagonism on Intralipid-induced intestinal satiation.
Six new estradiol-treated and six new oil-treated OVX rats were tested using the procedure above, with two differences. First, intraduodenal infusions were done from min 5–15 of the sham feeding test, rather than min 6–16, and sham intakes were measured every 5 min, rather than every 3 min. Second, either Devazepide (1 mg/kg; Merck, Sharpe & Dome Research Laboratories, West Point, PA) or its vehicle alone (1 ml/kg 0.5% carboxymethyl cellulose in distilled water) was ip injected at 1000 h. Again there were two cross-over tests. One compared intraduodenal infusions of 10% Intralipid vs. saline without Devazepide pretreatment and the other compared Intralipid vs. saline with Devazepide.
Data analysis
Two analyses were done. First, the overall effects of estradiol or Devazepide treatment on intestinal satiation were analyzed using total sham intakes from the start of intraduodenal infusions until the end of the test (i.e. min 6–45 in experiment 1a and min 5–45 in experiment 1b). In experiment 1a, a two-way ANOVA, with hormone treatment as between-subjects factor and infusate (saline, L-Phen, Intralipid) as within-subjects factor, was done. Sham intakes during and after intraduodenal saline infusions did not differ in either the oil- or estradiol-treated rats, so these data were collapsed. ANOVA were followed up with Bonferroni-Holm tests contrasting each saline vs. nutrient difference within hormone treatment, each oil vs. estradiol difference within nutrient, and the relative effect of estradiol in L-Phen vs. Intralipid tests (i.e. [oil/Intralipid – oil/L-Phen] vs. [estradiol/Intralipid – estradiol/L-Phen]). Experiment 1b, was similarly analyzed with a two-way ANOVA, with hormone treatment as between-subjects factor and nutrient/drug treatment (i.e. the four combinations of saline or Devazepide and saline or Intralipid) as within-subjects factor, followed up with the analogous Bonferroni-Holm tests. Second, to characterize the time course of intestinal satiation, the effects of estradiol or Devazepide treatment on intestinal satiation from the start of intraduodenal infusions until the end of the test were analyzed for each nutrient separately. That is, in experiment 1a, for each 3-min interval and each nutrient, difference scores (saline-nutrient infusion) were subjected to two-way ANOVA, with hormone treatment as between-subjects factor and time as within-subjects factor, and in experiment 1b, 5-min difference scores (saline-Intralipid infusion) were subjected to two-way ANOVA, with drug treatment (saline, Devazepide) as within-subjects factor and hormone treatment as between-subjects factor. Both these ANOVAs were followed up with t tests contrasting oil vs. estradiol and saline vs. Devazepide, respectively. ANOVAs were done with the SAS GLM procedure (SAS Institute, Cary, NC). Data are reported as means ± SEM, with the SE of the difference (SED) given to indicate experiment-wide residual variability. Because our hypotheses are directional, i.e. that both estradiol and intraduodenal nutrient infusions inhibit sham feeding, one-tailed tests were used. Significance levels of P < 0.025 (identical with P < 0.05 two-tailed) and 0.05 are reported.
Experiment 2. Effects of estradiol on c-Fos expression after intraduodenal nutrient infusions
We sought to determine what brain regions might mediate the feeding responses established in experiment 1, first, by measuring neuronal activation, as revealed by c-Fos expression, after intraduodenal Intralipid or L-Phen infusions in estradiol- or oil-treated rats (experiment 2a) and, second, by establishing whether estradiol-mediated changes in intraduodenal nutrient-induced c-Fos expression occur in neurons expressing ER (experiment 2b).
Procedure.
Rats (n = 28) were ovariectomized, maintained on estradiol or oil treatment, and adapted to intraduodenal saline infusions and sham feeding for 2 wk, as above. Then, saline, Intralipid, or L-Phen was infused as before, except that rats were not allowed to sham feed. Ninety minutes later the rats were anesthetized with pentobarbital (1 ml/kg, 50 mg/ml) and transcardially perfused with 100 ml of isotonic heparinized saline containing 0.5% NaNO2 and then with 400 ml of 4% paraformaldehyde in 0.1 M sodium phosphate buffer (PB). Their brains were removed, postfixed for 2 h in paraformaldehyde at room temperature, and cryoprotected over 3–4 d in 10, 20, and 30% sucrose in PB at 4 C, each until saturated. Forty-micrometer sections were cut coronally on a freezing stage microtome from the pyramids through the anterior commissure [i.e. from about 14.1–14.6 mm posterior to bregma according to the atlas of Paxinos and Watson (41)]. Sections were collected in six serially ordered sets, which were stored at –20 C in cryoprotectant.
Experiment 2a: c-Fos immunocytochemistry.
One of the six sets of sections from each rat was: 1) incubated for 10 min in 0.5% hydrogen peroxide solution in PB; 2) incubated for 1 h in PB containing 0.3% Triton X-100 (0.3% PBTx) and 1% normal goat serum (NGS); 3) incubated 20 h at room temperature with rabbit polyclonal anti-c-Fos peptide (1:10,000, Ab-5; Oncogene, Cambridge, MA) diluted in 0.3% PBTx and 1% NGS; 4) incubated for 1 h at room temperature with a biotinylated goat antirabbit secondary antibody (Vector Laboratories, Burlingame, CA) diluted 1:400 in 0.3% PBTx; 5) incubated for 1 h in avidin-biotin complex (Vector) diluted 1:500 in 0.3% PBTx.; and 6) immersed for 5 min in 2% nickel-intensified diaminobenzidine (DAB; 25 mg/ml solution in PB; KPL, Gaithersburg, MD). Tissue was mounted on gelatin-coated slides, air dried overnight, defatted in ascending ethanol and xylene rinses, coverslipped with Permount (Fisher Scientific, Pittsburgh, PA), and then digitally imaged (KP-D50, Hitachi Denshi, Japan).
The number of cells expressing c-Fos was quantified bilaterally using Image-Pro Plus software (version 3.0; Media Cybernetics, Gaithersburg, MD) in the following regions [locations are caudal to bregma as per (41)]: NTS subregions (nomenclature is our own): NTS just caudal to the AP (cNTS; 14.4–14.1 mm); subpostremal NTS (spNTS; 14.0–13.7 mm); intermediate NTS (iNTS; 13.3–13.1 mm, i.e. the area rostral to the AP in which the NTS abuts the fourth ventricle); PVN (1.8–2.1 mm), and arcuate hypothalamic nucleus (Arc; 2.8–3.1 mm). Cells were considered c-Fos positive if their nuclei contained dark, punctate blue-black immunolabeling and were counted using constant minimum and maximum OD and object size criteria, which were validated with visual counts. Two sections spaced 100–150 µm apart were counted per rat for each anatomical region. Mean count/section were analyzed.
Experiment 2b: colocalization of c-Fos and ER.
Another set of brain sections was processed for double-label immunofluorescence to identify c-Fos and ER. To improve the quality of the ER fluorescent label and allow the use of two primary antibodies raised in rabbits, we amplified the ER immunofluorescence signaling using biotinylated tyramide, according to a published technique (42, 43). Free-floating tissue sections were: 1) incubated for 10 min in 0.5% H2O2 in PB solution; 2) incubated for 1 h in 0.1% PBTx and 1% NGS; 3) incubated for 20 h at room temperature with polyclonal anti-ER peptide (1:100,000, c1355; Upstate Biotechnology, Lake Placid, NY) diluted in 0.3% PBTx and 1% NGS; 4) incubated for 1 h with biotinylated goat antirabbit secondary antibody (Vector) diluted 1:400 in PB containing 0.3% PBTx; 5) washed in 0.1% PBTx and incubated for 30 min in biotinylated tyramide solution (60 µl/10 ml PBS + 1:1000 H2O2) for 30 min; 6) washed in 0.1% PBTx and incubated in avidin-biotin complex (Vector) diluted 1:500 in 0.3% PBTx; 7) washed in 0.1% PBTx and then incubated with Cy2 conjugated egg-white avidin (Jackson ImmunoResearch, West Grove, PA) to visualize the ER-fluorescent label; 8) washed and incubated for 1 h with 0.3 PBTx and 2% normal donkey serum; 9) washed and incubated for 20 h at room temperature with anti-c-Fos antiserum (1:10,000, Ab-5; Oncogene) diluted in 0.3% PBTx and 1% normal donkey serum; 10) rinsed in 0.1% PBTx and incubated for 1 h with Cy3-conjugated affinipure donkey antirabbit IgG (Jackson ImmunoResearch) to visualize the c-Fos; and 11) washed in 0.1% PBTx, mounted on gelatinized slides, dried at room temperature for 24 h, dehydrated, and coverslipped using Krystalon medium (EMD Chemicals, Gibbstown, NJ). Steps 7–11 were done in the dark. As a control for label specificity, in additional sections the c-Fos primary antibody was omitted from the run. When this was done, the Cy2 signal remained strong, but there was no Cy3 signal (data not shown).
Labeled cells were counted by eye in images made with the Cy2 and Cy3 excitation wavelengths individually, and together. The criteria for double labeling of individual cells was: 1) presence of red labeling in the Cy3 image, 2) green labeling in the Cy2 image, and 3) yellow labeling in the Cy3-Cy2 image. Apparent double labeling of nuclear c-Fos and ER immunoreactivities appeared as a yellow spot. To preclude the possibility that apparent double labeling resulted from the overlay of different cells, double labeling was verified by examination of 1-µm confocal images produced by an LSM 410 confocal laser-scanning microscope (Carl Zeiss Microimaging, Thornwood, NY).
Data analysis
c-Fos, ER, and double-labeled counts/section were analyzed with two-way ANOVA, with infusate (saline, Intralipid, L-Phen) and hormone treatment (estradiol, oil) each as between-subjects factors. Each anatomical region described above was analyzed in experiment 2a, and the cNTS and Arc were analyzed in experiment 2b. Tissue for some rats was lost in processing in experiment 2b, so there were three to five rats/group. ANOVA were followed up with two-tailed Bonferroni-Holm tests contrasting each saline vs. nutrient difference within hormone treatment, each oil vs. estradiol difference within nutrient, and the relative effect of estradiol in L-Phen vs. Intralipid tests (i.e. [oil/Intralipid – oil/L-Phen] vs. [estradiol/Intralipid – estradiol/L-Phen]).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
). As previously shown in male rats (36, 40), rats began sham feeding sucrose avidly (
6–9 ml/3 min) and gradually slowed but after 45 min were still ingesting approximately 0.5–1 ml/min so that total 45-min sham intakes during intraduodenal saline tests were greater than 60 ml in each condition. Intraduodenal infusions of L-Phen and Intralipid each significantly reduced total amount sham fed during and after infusions in both oil- and estradiol-treated rats (Fig. 1C). Estradiol significantly increased the inhibitory effect of intraduodenal Intralipid but not that of intraduodenal L-Phen, and the difference in estradiol’s relative effects on the satiating effects of the two nutrients was significant (Fig. 1C). Post hoc tests were based on main effects of nutrient treatment (F2,33 = 13.99, P < 0.001) and intraduodenal infusions (F2,33 = 4.71, P < 0.03, SED = 7.0 ml); the interaction effect was not significant.
|
). Estradiol increased the inhibitory effect of intraduodenal Intralipid during six of the 11 3-min intervals beginning 12 min after infusion onset. The main effects of time, F13,154 = 2.77, P < 0.001, and estradiol treatment, F1,154 = 20.37, P < 0.002, SED = 1.3 ml, were both significant for Intralipid infusions. In contrast, only the main effect of time, F13,154 = 1.86, P < 0.04, SED = 1.6 ml, and not that of estradiol, F1,154 = 0.51, was significant for L-Phen infusions. Neither interaction effect was significant. Informal observations indicated that during saline infusions rats rarely moved away from the drinking tubes, whereas during the nutrient infusions they often moved away to groom and rest, i.e. to exhibit the behaviors typical of behavioral satiety. This is similar to what was reported in a test of intraduodenal Intralipid infusions in male rats in which behavioral sequence of postprandial satiety was explicitly measured (36).
Experiment 1b. Role of CCK in Intralipid-induced intestinal satiation
Although the rats used in experiment 1b did not sham feed quite as vigorously as those used in experiment 1a, the outcomes were similar: 1) estradiol had no effect on sham feeding in rats receiving intraduodenal saline infusions, 2) intraduodenal Intralipid infusion reduced sham feeding in oil-treated rats, and 3) estradiol pretreatment significantly increased the feeding-inhibitory effect of intraduodenal Intralipid infusion (Fig. 2, A and C). In addition, Devazepide pretreatment did not significantly affect the feeding-inhibitory effect of Intralipid in oil-treated rats but markedly reduced it in estradiol-treated rats (Fig. 2C). Post hoc tests were based on main effects of hormone treatment, F1,28 = 7.02, P < 0.02, and drug/nutrient, F3,28 = 9.05, P < 0.001, SED = 6.2 ml; the interaction effect was not significant. In estradiol-treated rats, the effect of Devazepide to reduce the feeding-inhibitory effect of intraduodenal Intralipid became apparent 10 min after infusions started (Fig. 2B), i.e. just when experiment 1a revealed that estradiol increased Intralipid’s effect. In estradiol-treated rats, the main effect of Devazepide treatment, F1,96 = 8.20, P < 0.005, SED = 1.3 ml, was significant, but the effect of time and the interaction were not. In oil treated-rats, ANOVA revealed no significant differences.
|
). In addition, the increment in Intralipid-induced c-Fos was significantly greater in estradiol- vs. oil-treated rats, and the estradiol-induced increment in c-Fos was larger in rats infused with Intralipid vs. L-Phen. Each of these three comparisons was highly significant (experiment wide P < 0.01). L-Phen did not reliably affect c-Fos expression in the cNTS. ANOVA detected no significant effects in the spNTS or iNTS (Fig. 3, B and C); the effects of infusate, hormone, and their interaction for spNTS were F2,167 = 2.47, P = 0.12, F1,167 = 3.34, P = 0.086, and F1,167 = 0.09, P = 0.91, and for iNTS were F2,17 = 0.05, P = 0.96, F1,17 = 0.06, P = 0.81, and F1,17 = 0.02, P = 0.98.
|
). In the mPVN, the ANOVA effect of hormone was highly significant, F1, 18 = 47.50, P < 0.001, SED = 10 cells/section, but again the effects of infusate and its interaction with hormone were not, F2, 18 = 2.29, P = 0.13, and F1, 18 = 1.90, P = 0.18. None of the post hoc tests was significant. No significant ANOVA effects were detected in the parvocellular PVN (data not shown).
|
).
Experiment 2b. Colocalization of c-Fos and ER
ER immunoreactivity.
As previously reported (32, 33, 34), ER was expressed in what we designated the cNTS, i.e. just posterior to the AP (14.4–14.1 mm posterior to bregma), a region of the NTS to which most abdominal vagal afferents project. Less ER was expressed slightly more caudally. Little ER was expressed more rostrally (i.e. in the spNTS). Virtually no ER immunoreactivity was detected in the iNTS or further rostrally. Our assay detected ER in the MPA and Arc in similar densities as previously reported. Finally, estradiol treatment alone failed to affect the numbers of cells expressing ER in either the cNTS or Arc, F1,15 = 0.18 and 0.65, Ps = 0.68 and 0.43, respectively.
c-Fos and ER colocalization in the cNTS and Arc.
Fluorescence labeling, like DAB labeling, indicated that Intralipid infusions, but not L-Phen infusions, significantly increased cNTS c-Fos expression in estradiol- but not oil-treated rats (Fig. 5A). Neither estradiol treatment nor intraduodenal nutrient infusions significantly altered the number of cNTS cells that expressed ER (Fig. 5B). The number of cells activated by intraduodenal Intralipid that also expressed ER was significantly greater in estradiol- than oil-treated rats (Fig. 5C). In terms of percentages, in the cNTS of estradiol-treated rats after Intralipid infusions, about 30% of the ER-positive cells expressed c-Fos and about 40% of the c-Fos-expressing cells were ER positive, whereas in estradiol-treated rats after saline or L-Phen infusions and in oil-treated rats after any infusion, the corresponding percentages were 5 and 13%. Figure 5, D–F, shows representative confocal images of c-Fos/ER colocalization in the cNTS of an estradiol-treated rat after intraduodenal Intralipid infusion. In contrast, no effect of L-Phen on c-Fos and ER colocalization was detected. Post hoc tests for c-Fos were based on a significant effect of estradiol treatment, F1,15 = 4.70, P < 0.05, and infusate, F2,15 = 3.99, P < 0.05, SED = 3.2 cells/section, and post hoc tests for c-Fos/ER colocalization were based on significant infusate and hormone-infusate interaction effects, F1,15 = 4.39, P < 0.05, and F2,15 = 12.20, P < 0.001, SED = 1.2 cells/section.
|
colocalization in the Arc (Fig. 6). The c-Fos ANOVA results for hormone, infusate and their interaction were F1,15 = 3.09, P = 0.10, F2,15 = 1.00, P = 0.39, and F2,15 = 3.05, P = 0.08, respectively, and the c-Fos/ER colocalization ANOVA results were F1,15 = 0.64, P = 0.43, F2,15 = 0.21, P = 0.81, and F2,15 = 1.63, P = 0.23, respectively.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
on neurons in the cNTS to increase the satiating potency of endogenous CCK. The behavioral evidence is: 1) in OVX rats, estradiol treatment increased the inhibition of sham feeding elicited by intraduodenal infusions of Intralipid, which is CCK dependent, but not L-Phen, which is not CCK dependent; and 2) Devazepide, a potent and selective CCK-1 receptor antagonist, reduced the inhibitory effect of Intralipid on sham feeding more in estradiol- than oil-treated rats. The cellular evidence is that ER-positive neurons in the cNTS were selectively activated by intraduodenal infusions of Intralipid, but not L-Phen, in estradiol-treated rats but not oil-treated rats. Ingested food acts on preabsorptive receptors during the meal to elicit changes in peripheral neural activity and local or systemic levels of chemical signals, including several gut peptides. Information so encoded functions as feedback controls of feeding (1, 2, 3, 4, 5). Information arising from oropharyngeal sites provides primarily positive feedbacks, whereas information arising from postingestive sites, especially gastric mechanoreceptors and intestinal chemoreceptors, provides mainly negative feedbacks. Estradiol appears not to affect oropharyngeal feedbacks (14, 15, 16) but rather to influence the actions of gastrointestinal feedbacks, including those encoded as changes in secretion of CCK (14, 15, 16), glucagon (44), and ghrelin (45). Here we further examined estradiol’s effects on intestinal feedbacks by investigating for the first time estradiol’s effects on the satiating actions of intraduodenal nutrient infusions in sham-feeding OVX rats. The effects of gastrointestinal, especially intraduodenal, infusions on feeding have been extensively analyzed in male rats but, as far as we know, not in females. To maximize the physiological relevance of our data, the rate of intraduodenal infusions was closely matched to intrameal rates of gastric emptying of similar lipids previously reported (4, 6). Although our infusions were continuous rather than pulsatile, whereas the transpyloric flow of nutrients is predominantly pulsatile, continuous and pulsatile lipid infusions in humans similarly stimulated CCK release and inhibited eating (46).
Our principal behavioral result was that, although intraduodenal infusions of Intralipid and L-Phen inhibited sham feeding similarly in oil-treated rats, estradiol treatment increased the satiating potency of Intralipid only and not L-Phen infusions. Furthermore, the effect of estradiol to increase intraduodenal Intralipid-induced satiation was CCK dependent because it was reversed by CCK-1 receptor antagonism. This finding complements our previous reports that CCK’s satiating potency during spontaneous feeding is increased by estradiol treatment in OVX rats and increased during estrus in intact rats, estrus being the phase of the ovarian cycle when the inhibitory effect of endogenous estradiol on food intake is maximal (14, 15, 16, 17, 21).
In contrast to its effect on Intralipid-induced satiation, neither estradiol treatment nor CCK-1 receptor antagonism affected L-Phen-induced satiation. This is novel evidence that estradiol selectively affects only certain gastrointestinal feedbacks, CCK-mediated feedbacks among them. It would be interesting therefore to determine whether estradiol also increases the satiating potency of intraduodenal carbohydrate, protein, or mixed nutrient infusions. We assume it is likely to do so because the satiating effect of intraduodenal infusions of at least some carbohydrates and proteins is CCK dependent (4, 5) and because CCK-1 receptor antagonism increased food intake during estrus in rats maintained on low-fat diets (i.e. chow) (17, 21). It is also possible that estradiol might affect intraduodenal L-Phen-induced satiation under other conditions if L-Phen synergizes with signals related to gastric volume or gastric emptying that are not present during sham feeding. We know of no evidence for that possibility.
Estradiol appears to act centrally rather than locally to affect feeding (14, 15, 16). The brain site or sites in which estradiol does act to inhibit feeding, however, have not been identified (14, 15, 16, 47, 48). Therefore, one goal of our work was to identify potential sites for the action of estradiol on CCK-induced satiation by determining whether lipid-induced intestinal satiation is associated with increases in neuronal activation, as measured with c-Fos immunocytochemistry, in neurons expressing ER. This occurred in one site, the cNTS. In estradiol-treated, Intralipid-infused rats, 38% of the c-Fos-positive cells in the cNTS also expressed ER, whereas less than 10% of c-Fos-positive cells did so in the other groups. This result suggests that estradiol increases Intralipid-induced intestinal satiation by acting on cNTS neurons in a way that increases their response to negative feedback controls of satiation. Our finding that small amounts of estradiol placed on the surface of the hindbrain over the cNTS are sufficient to reduce food intake (49) is consistent with this hypothesis. Whether the ER-positive NTS neurons mediating these effects receive direct inputs from abdominal vagal afferents, which are known to relay satiating signals arising from intraduodenal lipid infusions and endogenous and exogenous CCK in male rats (4, 5), and whether they also mediate estradiol’s other effects on feeding will require further research.
Intraduodenal Intralipid infusions failed to increase c-Fos expression in the spNTS, iNTS, or PVN. In contrast, in previous work both food ingestion (23) and ip CCK (24) increased c-Fos expression in these areas in OVX rats, and these responses were increased by the same estradiol treatment we used here. These differences suggest that Intralipid produces less widespread neural activation than either feeding or ip CCK. Because the amount of c-Fos expression in the saline-infused rats was much higher in the spNTS and iNTS than the cNTS, however, we cannot exclude the possibility that a ceiling effect may have obscured effects of intraduodenal nutrient infusion on c-Fos expression. In any case, it is tempting to speculate that the activation of c-Fos in the cNTS after feeding (23), ip CCK (24), or intraduodenal Intralipid infusions (present results) represents the initial effect of estradiol on feeding-related circuits, which is then propagated to other brain areas.
The highly restricted rostral-caudal extent of c-Fos activation in the NTS here is especially interesting in comparison with, first, the much larger rostral-caudal extent of the NTS that receives vagal projections from the intestines (50) and, second, the lack of rostral-caudal selectivity in NTS c-Fos responses in female rats (as described above) and in male rats after food ingestion, intragastric or intraduodenal nutrient infusions, or ip CCK (25, 29, 30, 51, 52, 53). Whether this contrast with our results reflects a selective sex difference in brain responsivity to intraduodenal infusions or simply methodological differences is not clear. Finally, in contrast to the circumscribed anterior-posterior distribution of activated NTS neurons, we observed little selectivity of c-Fos responses in the medial, dorsomedial, and commissural subnuclei of the NTS. This is similar to most, but not all (e.g. Ref. 30) previous reports. In summary, the specific activation of ER-positive cells in the cNTS in estradiol-treated OVX rats after intraduodenal Intralipid infusion seems to present a unique opportunity for delineation of an important functional neural network of satiation.
The lack of effect of estradiol to increase intraduodenal Intralipid-induced c-Fos in the PVN suggests either that these areas are not involved in the estrogenic inhibition of feeding or that its involvement is limited to other contexts. Butera and colleagues (54, 55) suggested that the PVN was involved in the estrogenic inhibition of eating based on observation of decreased feeding after microinjections of estradiol into the PVN, but others have failed to confirm this (47, 48). The effect of estradiol here to increase mPVN c-Fos after intraduodenal infusions of saline, L-Phen, and Intralipid that were all isotonic with the plasma suggests that estradiol may mediate PVN processing of a neural signal related to intraduodenal volume.
In the DAB-labeled samples but not in the fluorescence-labeled samples, both Intralipid and L-Phen decreased c-Fos in the Arc. We do not know why the reason for this apparent discrepancy. Lo et al. (51) recently reported a similar decrease in Arc c-Fos 1 h after Intralipid gavage, although Arc c-Fos measured 6 h after gavage was markedly increased. In two other relevant studies, ip CCK alone had no effect on Arc c-Fos in rats but antagonized the increase in Arc c-Fos after ip ghrelin in rats (56), and 24-h food deprivation increased Arc c-Fos in mice, which was then rapidly reduced by eating or ip peptide YY (57). Taken together, these data suggest that the Arc has a multifaceted response to feeding-related stimuli. Nevertheless, although Clegg et al. (58) and Gao et al. (59) have presented data implicating the Arc in the estrogenic control of eating, here we saw no significant effect of estradiol on c-Fos expression in the Arc.
In conclusion, estradiol has been shown previously to increase CCK-induced satiation during the periovulatory phase of the rat ovarian cycle. Here we showed that estradiol does this in part by increasing CCK-dependent lipid-induced intestinal satiation in OVX rats and that this effect may be produced by an action of estradiol on a ER neurons in the cNTS. Determining whether this interactive satiating effect of estradiol and CCK also occurs in women warrants investigation, especially in view of the facts that CCK also mediates lipid-induced intestinal satiation in young male human volunteers (1), that normal reproductive function is associated with clinically relevant and apparently estradiol-mediated changes in eating in women, and that women have increased vulnerability to disordered eating (14, 15, 16).
|
Acknowledgments |
|---|
|
Footnotes |
|---|
The authors have nothing to disclose.
First Published Online September 6, 2007
Abbreviations: AP, Area postrema; Arc, arcuate hypothalamic nucleus; CCK, cholecystokinin; CeA, central nucleus of the amygdala; cNTS, NTS just caudal to the AP; DAB, diaminobenzidine; ER, estrogen receptor; iNTS, NTS just caudal to the intermediate NTS; L-Phen, L-phenylalanine; mPVN, magnocellular PVN; NGS, normal goat serum; NTS, nucleus tractus solitarius; OVX, ovariectomized; PB, phosphate buffer; PBTx, PB containing Triton X-100; PVN, paraventricular nucleus of the hypothalamus; spNTS, NTS just caudal to the subpostremal NTS.
Received March 12, 2007.
Accepted for publication August 23, 2007.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
null mice. Endocrinology 142:4751–4757 and -ß in the rat central nervous system. J Comp Neurol 388:507–525[CrossRef][Medline] and ß immunoreactivity by oxytocin neurons of rat paraventricular nucleus. J Neuroendocrinol 9:803–806[CrossRef][Medline]This article has been cited by other articles:
 |
|
 |
|
  L. Asarian Loss of cholecystokinin and glucagon-like peptide-1-induced satiation in mice lacking serotonin 2C receptors Am J Physiol Regulatory Integrative Comp Physiol, January 1, 2009; 296(1): R51 - R56. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. A. Roepke, C. Xue, M. A. Bosch, T. S. Scanlan, M. J. Kelly, and O. K. Ronnekleiv Genes Associated with Membrane-Initiated Signaling of Estrogen and Energy Homeostasis Endocrinology, December 1, 2008; 149(12): 6113 - 6124. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Thammacharoen, T. A. Lutz, N. Geary, and L. Asarian Hindbrain Administration of Estradiol Inhibits Feeding and Activates Estrogen Receptor-{alpha}-Expressing Cells in the Nucleus Tractus Solitarius of Ovariectomized Rats Endocrinology, April 1, 2008; 149(4): 1609 - 1617. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
