INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

INTESTINAL SIGNALS constitute an important subset of the postingestive cues that inhibit feeding behavior. Chronically implanted duodenal catheters allow infusions of nutrients directly into the intestine, thereby bypassing oral and gastric sensory receptors. Duodenal infusions of nutritionally complete liquid diet (8) or specific macronutrients (16, 31) suppress feeding in animals with open gastric fistulas (sham feeding) and in animals that are feeding normally (26).

Recent work has focused on the role of specific macronutrients in intestinal satiety, the location and nature of the receptors that detect intestinal nutrients, and contributions of the enteric nervous system. On the basis of results from several studies, individual fatty acids, mixtures of lipids, and some amino acids suppress feeding more than equicaloric infusions of monosaccharides or disaccharides (26, 31); therefore, caloric monitoring does not appear to be the primary mechanism for macronutrient detection. Receptors for lipid and glucose appear to be associated with the intestinal mucosa, inasmuch as duodenal infusions of various simple sugars (24) or fatty acids (5) suppress sham feeding but comparable infusions into the hepatic portal vein do not (6, 24). Also, different peripheral pathways appear to signal the presence of lipids or carbohydrates vs. amino acids, as total subdiaphragmatic vagotomy (31) completely blocks the intake-suppressing effects of lipid or carbohydrate but only partially attenuates the effect of amino acids.

We have begun to ask the question, How are these signals generated in the gastrointestinal tract used by the brain to initiate the satiation process? Our working hypothesis is that at least some signals from different gastrointestinal sites and/or sensors initially utilize different pathways but eventually converge at higher levels of the neuraxis to terminate an ongoing meal. To this end, we have shown that gastric balloon distension with two paradigms designed to preferentially stimulate different types of gastric mechanosensors produced similar topographical patterns of neuronal activation in the rat caudal brain stem, with a similar proportion of catecholaminergic neurons recruited (28). This finding suggested that different distension signals from within the stomach converge already at the caudal brain stem level. In the present study we wanted to test the hypothesis with respect to different small intestinal nutrient sensors and compare it with gastric distension. Although the induction of nuclear Fos as a measure of neuronal activation has important limitations (20), it is nevertheless the method of choice to obtain an initial in toto overview of the neural substrate underlying this particular sensory system, and its combination with transmitter and peptide immunocytochemistry will allow the identification of the neurochemical phenotype of activated neurons.

The effect of duodenal nutrient infusion on the expression of Fos in the rat brain has been investigated by Zittel et al. (33). Expression of Fos-like immunoreactivity (FLI) was seen in the nucleus of the solitary tract (NTS) after infusions of lipid emulsion (Intralipid, 2.7 kcal) or D-glucose (2.9 kcal), but not after infusions of glucose at 1.1 kcal or mannitol, hydrochloric acid, or casein hydrolysate (1.2 kcal). However, the study did not address the question of possible different topographical distribution patterns of activated neurons, and the effects on sham feeding were not assessed in the same animals. McCaffery et al. (9), in a preliminary study, found that the relative order of FLI induction by duodenal infusions of various lipids (linoleic acid, Intralipid, oleic acid, linolenic acid) was similar to the relative potency of the infusates for inducing satiety in separate studies. However, here again, the level of analysis did not allow detection of differences in FLI distribution patterns, and effects on behavior and neural activation were not directly compared. In addition, the Fos method was used in conjunction with a variety of other stimuli and treatments that increase or decrease food intake, including cholecystokinin (CCK) (13, 14, 18), dehydration (13), food ingestion or deprivation (13), hypoglycemia (13), and gastric distension (28). The fact that such treatments resulted in greatly different patterns of FLI in the caudal brain stem indicates that this method is a useful tool in probing the neural substrate underlying various aspects of ingestive behavior, including the central pathways of satiety signals.

One aim of the present study was, therefore, to compare the effects of duodenal infusions of different macronutrients on sham feeding and brain stem FLI expression by use of identical experimental parameters and, to some extent, the same animals for the behavioral and anatomic analyses. The second aim was to perform a fine-grained, subnuclear analysis of the neural FLI activation patterns achieved with the various infusions, allowing detection of possible quantitative and qualitative differences.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Animals. Male Sprague-Dawley rats were obtained from Harlan; they weighed 250-300 g at the first surgery. They were housed individually and maintained on a 12:12-h light-dark cycle with lights on at 0700. Food (Purina Chow 5001) and water were available ad libitum, except during 18-h deprivation periods before infusion experiments.

Surgery. Gastric fistulas and duodenal infusion catheters were implanted at least 10 days before rats began adaptation to the testing conditions. After anesthesia with a mixture of ketamine (80 mg/kg), xylazine (4 mg/kg), and acepromazine (1.6 mg/kg) and treatment with atropine (0.5 mg/kg), rats received laparotomies along the linea alba. The stomach and duodenum were externalized, and a stab wound was made in the center of a purse string suture (6-0 monofilament suture) in the greater curvature of the nonglandular stomach. A silastic duodenal catheter (0.63 mm ID, 1.19 mm OD, 14 cm long) with an anchoring sleeve of slightly larger silastic (1.02 mm ID, 2.16 mm OD, 3 mm long; secured 1 cm from the internal end with cyanoacrylic adhesive) was installed through the stomach wound and secured in the duodenum (two 8-0 monofilament sutures and Marlex mesh) with the internal end 4 cm beyond the pylorus.

Adaptation to testing conditions and environment. Ten of the animals were tested for the effects of duodenal nutrient infusions on sham feeding, with each rat receiving all nutrient and control infusions in randomized order. Subsequently, each of these animals received a single nutrient infusion just before it was killed and its brain was removed for c-Fos processing. The remaining animals were not used in sham-feeding experiments; rather, each animal was given only one nutrient or control infusion just before brain removal.

Sham-feeding tests with duodenal infusions. After reaching the sham-feeding criterion described above, 10 rats were given sham-feeding tests while receiving duodenal infusions of specific macronutrients or control infusions (see below), with each rat receiving all treatments in random order (within-subjects design). Before sham-feeding tests the animals' fistulas were opened, and stomach contents were flushed with 6 ml of warm water. Then their duodenal catheters were attached to infusion leads running to 10-ml syringes on infusion pumps (model 210, Kd Scientific). Infusion leads were constructed from an 80-cm length of polyethylene tubing (PE-100, 0.82 mm ID, 1.52 mm OD) attached to a 20-gauge needle for connection to an infusion syringe and joined to a 25-cm length of silicone tubing (1.02 mm ID, 2.16 mm OD) with 19-gauge tubing. A 2-cm length of 19-gauge tubing also connected the silicone tubing to the duodenal catheter. After a 60-min adaptation period without sham feeding, animals were presented with 30% sucrose in graduated feeding tubes. After 10 min of feeding without duodenal infusions, rats were given 30-min infusions followed by 20 min of sham feeding without infusion (total sham-feeding bout equaled 60 min). The volume consumed was determined at 5-min intervals through the sham-feeding test. Fistula drainage was ensured by excluding animals when collected gastric drainage did not equal or exceed intake.

Final infusion test for Fos immunocytochemistry. In the second part of this study, 51 rats, including the 10 used in the sham-feeding study, received the duodenal infusions described above but were not allowed to sham feed. Thus eight or nine animals were assigned to each infusion group. Again, animals that showed duodenogastric reflux >20% of the infusion volume were excluded, and only 3 animals showed reflux between 10 and 20%.

Fos immunocytochemistry. After postfixation, brains were cut into blocks (medulla, midbrain, posterior forebrain, anterior forebrain) for ease of sectioning. The medulla was then placed in 25% sucrose in 4% paraformaldehyde overnight before it was cut into 30-µm coronal sections. Sections were collected in PBS for subsequent free-float processing. Some brains were stored at 20°C in cryoprotectant (50% PBS, 20% glycerol, 30% ethylene glycol) for several weeks with no apparent effect on c-Fos immunocytochemistry (27). Tissue sections from three to eight brains, spread across several different treatments, were processed at the same time. Sections were first removed from PBS and rinsed in 1% sodium borohydride in PBS for 30 min to reduce aldehyde cross-linking of proteins, thereby enhancing antibody penetration. After they were rinsed in PBS, sections were pretreated for 20 min with 1.5% hydrogen peroxide, 20% methanol, and 0.1% Triton X-100 (TX) in PBS to reduce endogenous peroxidase. Further rinsing in PBS alone continued until bubbling activity stopped. Sections were immersed in 5% normal goat serum, 1% BSA, and 0.5% TX for 30 min before incubation with an antibody to the protein product Fos of the c-fos gene (1:50,000 dilution in PBS with 0.5% TX, 0.1% gelatin, 0.05% sodium azide; Ab-5, Oncogene Sciences, Cambridge, MA; 48-60 h at 8°C). Sections were rinsed four times in PBS containing 0.1% gelatin and then incubated for 90 min in biotinylated goat anti-rabbit secondary antibody (1:1,000 dilution in PBS with TX; Biotin-SP IgG H + L, Jackson ImmunoResearch Labs, West Grove, PA). After three rinses in PBS alone, sections were immersed in an avidin-biotin complex (Vectastain ABC Elite Kit PK-6100, Vector Laboratories, Burlingame, CA) for 60 min. The Fos-antibody complex was then visualized by immersing sections for 4 min in a metal-enhanced diaminobenzidine substrate (1:9 dilution of diaminobenzidine in stable peroxide; Pierce Chemical, Rockford, IL). Cells labeled for FLI exhibited a dark blue-black nuclear stain. After three rinses in PBS, sections were immersed in 70% glycerol before they were mounted on slides with 100% glycerol.

Analysis. The volume of 30% sucrose sham fed during the 10 min before infusions began, the 30-min infusion period, and the 20-min period after the infusions was subjected to a randomized-block one-way ANOVA with infusion type as the main factor. When appropriate, differences between intakes after particular treatments were analyzed with Duncan's post hoc tests.

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

Effects of macronutrient infusions on sham feeding. During the 10-min period of sham feeding before the infusions began, no significant differences in intake were seen between the different infusion conditions (Fig. 1). Rats sham fed rapidly (~2 ml/min) and with few interruptions. However, during the 30-min infusion period, a significant treatment effect was seen [Fig. 1; F(5,45) = 12.12, P < 0.001]. Infusions of 4.5 kcal of glucose, 1.5 kcal of amino acid mixture, and 1.5 kcal of linoleic acid suppressed sham feeding relative to infusions of saline, water, and 1.5 kcal of glucose (P < 0.01). No difference was seen between intake after 1.5 kcal of glucose and intake after saline or water.

View larger version (30K): [in this window] [in a new window]   Fig. 1.   Effect of various nutrient and control infusions into duodenum on sham feeding of 30% sucrose. Suppression of sham feeding by infusion of water, 1.5 and 4.5 kcal of glucose, 1.5 kcal of an amino acid mixture, and 1.5 kcal of linoleic acid relative to saline infusion was calculated for each individual animal. Values are means ± SE of 10 rats. Amount sham fed was measured for periods of 10 min before infusion, 30 min during infusion, and 20 min after infusion. Water and 1.5 kcal of glucose did not significantly suppress sham feeding in any period. Higher dose of glucose, as well as amino acid mixture and linoleic acid, significantly suppressed sham feeding during and after infusion. * P < 0.05.

Fos expression in the dorsal vagal complex. As the NTS appears to be viscerotopically organized along its rostrocaudal axis (1), for purposes of analysis the NTS was subdivided into three regions relative to the AP. Across all three regions, animals that received saline, water, or 1.5-kcal glucose infusions showed very few Fos-positive neurons (Fig. 2), and no significant differences in Fos labeling were detected between animals receiving these infusion conditions in any portion of the NTS or AP. Animals that received 1.5-kcal infusions of amino acids or linoleic acid or 4.5-kcal infusions of glucose showed extensive expression of Fos-positive neurons (Fig. 2), so detailed comparisons were made only between animals that received these infusions and animals that received saline infusions.

View larger version (151K): [in this window] [in a new window]   Fig. 2.   Distribution of Fos-positive neurons in dorsal vagal complex at rostrocaudal level of area postrema of rats intraduodenally infused with saline, water, 1.5 or 4.5 kcal of glucose, 1.5 kcal of amino acids, or 1.5 kcal of linoleic acid. Images were obtained using transmitted light mode of a laser scanning microscope. For subnuclear organization see Fig. 4.

View larger version (29K): [in this window] [in a new window]   Fig. 3.   Fos expression at different rostrocaudal levels of nucleus of solitary tract (NTS) and in area postrema (AP) induced by control and various nutrient infusions into duodenum of unanesthetized rats. Individual scores for each rat were obtained by calculating average number of Fos-positive neurons from two to four 30-µm-thick sections, taken at 150-µm intervals, for each area (the most rostral and caudal NTS were not taken into account). Values are means ± SE of 8 or 9 rats. For each area, columns that do not share same letter are significantly different (P < 0.05, Duncan's multiple-range tests following separate ANOVAs).

View larger version (129K): [in this window] [in a new window]   Fig. 4.   High-magnification photomicrograph of Fos expression in NTS subnuclei induced by an intraduodenal infusion of D-glucose (4.5 kcal). Note almost complete absence of Fos expression in dorsal motor nucleus. ap, Area postrema; c, central canal; cen, central subnucleus; com, commissural subnucleus; dm, dorsomedial subnucleus; dmnX, dorsal motor nucleus of vagus; is, interstitial subnucleus; med, medial subnucleus; nXII, hypoglossal nucleus; ts, solitary tract.

View larger version (25K): [in this window] [in a new window]   Fig. 5.   Fos expression in different subnuclei of NTS of rats induced by intraduodenal control and various nutrient infusions. Individual scores were obtained from 1 section of each rat located at rostrocaudal level of mid-AP. Values are means ± SE of 8 rats. For each subnucleus, columns that do not share same letter are significantly different (P < 0.05, Duncan's multiple-range tests following separate ANOVAs). For abbreviations of subnuclei see Fig. 4 legend.

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

Several conclusions can be drawn from the results of this study. 1) The intake-inhibiting effects of macronutrients involve more than simple caloric monitoring, because equicaloric duodenal infusions of different macronutrients produce different effects. 2) Different macronutrients trigger different levels of neural activation, as indicated by Fos expression. 3) The relative effects of different macronutrients on sham feeding and neural activation are not well correlated. 4) Although different macronutrients trigger different levels of neural activation, no differences in pattern of activation were detected in this study. 5) Qualitative differences in pattern do appear to exist between neural activity induced by duodenal infusions of macronutrients and gastric distension, two distinct but complementary inhibitors of ingestion.

Effects on sham feeding. Duodenal infusions of different macronutrients had dramatically different effects on intake during sham feeding. Linoleic acid and the amino acid mixture suppressed sham feeding more than equicaloric infusions of glucose and more than control infusions. These results agree with previous studies in which duodenal infusions of fatty acids or amino acids suppressed sham feeding (31) or normal feeding (26) similarly and both of these macronutrients suppressed feeding more than equicaloric infusions of glucose or maltose. In this study, linoleic acid suppressed feeding even more than glucose infusions containing three times as many calories. Although equicaloric infusions of linoleic acid tended to suppress feeding more than amino acids, this tendency did not reach significance.

Effects on Fos expression. In the analyses of FLI, macronutrient infusions as a whole had their greatest effect on the region of the NTS immediately subjacent to the AP, with fewer cells expressing Fos in the NTS rostral to the AP and in the AP itself and even fewer cells showing Fos in the NTS caudal to the AP. This result was not surprising, because the NTS is larger and contains more neurons in the region of the AP than in its more rostral or caudal regions. Duodenal nutrient infusions had no detectable effect on Fos in the dmnX. These results agree with work by Zittel et al. (33), where lipid and glucose infusions induced greater Fos labeling in the more rostral NTS (subjacent to the AP and the 4th ventricle) than in the caudal NTS (at obex). In an earlier study by Olson et al. (13), Fos expression was determined after treatments that inhibited food intake (e.g., CCK, food ingestion) or potentiated food intake (deprivation or insulin). Although the rostrocaudal distribution of Fos expression was not described, intake-inhibiting treatments were associated with substantial expression in the NTS and, to a lesser extent, in the AP, but no significant Fos expression appeared in the dmnX. Conversely, intake-potentiating treatments triggered substantial Fos expression in the dmnX but had little effect in the NTS or the AP. In the study of Olson et al. and in another study (28), gastric distension triggered substantial Fos expression in the NTS and the dmnX.

Convergence vs. separate pathways of potential satiety signals. Although several NTS subnuclei exhibited significantly greater Fos expression after nutrient infusions than after control treatments, the rank ordering of effects from different nutrients was similar across all affected subnuclei. The greatest degree of labeling was always triggered by the 4.5-kcal glucose and 1.5-kcal linoleic acid infusions, followed by amino acid infusions. Only in the medial subnucleus did 4.5 kcal of glucose trigger significantly more Fos expression than linoleic acid. Although these similar distribution patterns of activated neurons across subnuclei strongly indicate a high degree of convergence of sensory signals, additional experiments are necessary to substantiate this explanation. It is possible that the different sensory stimuli activate different populations of neurons in the same general areas. Such populations may differ in their neurochemical phenotype and or in their projection patterns (see Perspectives). Because the Fos method does not allow testing for more than one stimulus for a given animal, individual neurons cannot be compared across subjects. To detect true convergence of two stimuli on a given second-order neuron, electrophysiological recording will be necessary.

Perspectives