INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

PERIPHERAL ADMINISTRATION of cholecystokinin (CCK) evokes a number of behavioral effects, including satiety and decrease in exploratory behavior, as well as various alterations of gastrointestinal (GI) function, for example inhibition of gastric motility and emptying (1, 8, 28). Endogenous CCK is released from mucosal endocrine cells in response to nutrients, such as fat and protein, entering the duodenum (28). Autoradiographic studies revealed two types of CCK receptors. CCK-A receptors are primarily localized peripherally in the GI tract and centrally in the nucleus of the solitary tract (NTS) as well as in the area postrema (AP). CCK-B receptors are predominantly distributed throughout the brain, but also have been detected in the periphery on the vagus nerve (7, 12, 25, 32). Studies using selective and specific CCK antagonists showed that administration of lipid into the upper small intestine mimics some of the effects of exogenous CCK on GI function and satiety via CCK receptor-mediated mechanisms (1, 28).

Evidence has accumulated that the effects of exogenously administered and endogenously released CCK in the periphery on behavioral parameters as well as on GI function are, at least partially, mediated by mechanisms involving central nervous system (CNS) structures (1, 28, 35). It has been shown, for example, that the CCK-induced inhibition of gastric emptying and stimulation of pancreatic secretion are dependent on intact vagal reflex pathways (28) and that the satiating effect as well as behavioral consequences of exogenous and endogenous CCK in the periphery are mediated by action of the peptide on vagal afferent fibers projecting to the NTS in the brain stem (1, 8). Furthermore, selective bilateral lesions at the level of the vagus, NTS, midbrain, and paraventricular nucleus of the hypothalamus (PVN) abolish satiety and exploratory behavior induced by peripherally administered CCK (1, 8).

CCK administered peripherally induces alterations in neuronal activity in the PVN, NTS, AP, and locus ceruleus complex (LCC) [i.e., locus ceruleus (LC) plus subceruleus nuclei (SC), the subjacent area], brain regions known to play a pivotal role in the CNS control of GI function, as well as in the integration of behavioral, autonomic, and neuroendocrine response mechanisms of the organisms (9, 22, 24, 28, 34, 35). Some of this information has been obtained by proving an activation of c-fos expression in these brain areas induced by peripheral injection of CCK (6, 9, 20, 22, 30, 31, 37). This can be determined by detection of changes in c-Fos-like immunoreactivity (c-FLI) within brain nuclei, which identifies polysynaptically activated neurons on a cellular level. This method also allows further characterization of the neurochemical phenotype of activated neurons (13).

It has been demonstrated before that ingestion or direct intraduodenal administration of nutrients in awake rodents induces activation of c-fos expression in the NTS and AP (10, 11, 31, 37). Also, there is some evidence that capsaicin-sensitive vagal afferents and predominantly CCK-A receptors are involved in the pathways mediating the activation of c-fos expression in the NTS, AP, LC/SC, and PVN, induced by exogenous sulfated CCK octapeptide (CCK-8S) administered peripherally (6, 9, 20, 22). However, it is controversial whether release of peripheral CCK due to physiological stimuli plays a significant role in the induction of c-fos expression in the NTS and AP (10, 31, 37). In addition, it still needs to be established whether entry of nutrients into the intestine induces alterations of c-fos expression in the PVN or LC/SC and whether endogenous CCK and vagal pathways play a significant role in afferent transmission of information about such conditions from the gut to these brain nuclei.

Therefore, the initial purpose of the present study was to determine whether lipid entering the small intestine in physiological amounts influences neuronal activity in the LC/SC, dorsal vagal complex (DVC), and PVN, as assessed by c-FLI. It then became of further interest whether endogenous CCK and vagal afferents play a significant role in the intraduodenal lipid-induced activation of c-fos expression in these brain nuclei. Thus we also investigated whether intraduodenal lipid-induced changes in c-FLI in the LC/SC, NTS, AP, and PVN would be affected by pretreatment with specific CCK-A and CCK-B receptor antagonists or by sensory vagal denervation due to bilateral perivagal treatment with the sensory neurotoxin capsaicin (22, 28, 29). In additional experiments, we assessed the contribution of CCK receptors not localized on capsaicin-sensitive vagal afferents to changes of neuronal activity in the LC/SC, DVC, or PVN at entry of lipid into the upper small intestine, by combining pretreatment with perivagal capsaicin and selective CCK-A or CCK-B receptor antagonists.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Animals. Male Sprague-Dawley rats (strain of Deutsche Tierversuchsanstalt, Hannover, Germany; distributed by Winkelmann, Borchen, Germany), weighing 300-400 g, were housed in colony cages under conditions of controlled illumination (14:10-h light-dark cycle), humidity, and temperature (22 ± 2°C) for at least 7 days before the surgical procedure. They were fed a standard rat diet (Altromin, Lage, Germany) and tap water ad libitum. The animals were deprived of food but not water 18 h before each experiment.

Chemicals. The lipid emulsion (Intralipid 20%, containing 200 g soya bean oil, 12 g phosphatidylcholine, and 22 g glycerol in sterile water; Pharmacia, Erlangen, Germany) was warmed up immediately before the experiment to 38°C. Mannitol was made up in sterile water to a solution (0.33 M) isosmolar to the lipid emulsion (270 mosmol) and warmed up immediately before the experiment to 38°C. The CCK-A receptor antagonist MK-329 (Merck, Rahway, NJ) and the CCK-B receptor antagonist L-365,260 (Merck) were dissolved in dimethyl sulfoxide (Sigma, St. Louis, MO), Tween 80 (Sigma), and sterile 0.15 M NaCl (8:1:1 vol/vol/vol). Capsaicin (10 mg) was sonicated with 0.1 ml Tween 80 for 10 min, made up to 1 ml with olive oil (Sigma), and mixed thoroughly.

Perivagal capsaicin treatment. This method was performed as described elsewhere (29). Atropine (0.5 mg/kg ip; Sigma) was administered before surgery to reduce the acute actions of capsaicin on cardiovascular and respiratory systems. Rats were anesthetized with xylazine (10 mg/kg ip; Bayer, Leverkusen, Germany) and ketamine (100 mg/kg ip; Parke-Davis, Freiburg, Germany). Under aseptic conditions, both cervical vagi were exposed by a midline neck incision and the vagal trunk was carefully freed from the right and left carotid arteries and exposed for a distance of 3-4 mm. To avoid diffusion of the capsaicin solution into the surrounding tissue, strips of parafilm were placed around the exposed nerve. Then a small pledge of cottonwool soaked in capsaicin solution was applied to the nerves for 30 min. Further drops of capsaicin solution were applied every 10 min to a maximum of 0.05 ml (0.5 mg/ml). Finally, the area was thoroughly rinsed with saline and dried with swabs, and the neck incision was closed. Experiments were performed 2-3 wk after perivagal capsaicin treatment. Implantation of duodenal catheters in capsaicin-treated rats was performed 7 days after this preparation.

Duodenal catheters. This method was performed as described elsewhere (18). Rats were anesthetized with a mixture of ketamine (75 mg/kg ip; Parke-Davis) and xylazine (5 mg/kg ip; Bayer) and chronically implanted with a polyethylene catheter (1.2 mm ID, 1.7 mm OD) into the duodenum, which was inserted through an enterotomy localized ~2 cm distally from the pylorus. The catheter was fixed at the GI wall by a purse-string suture and routed subcutaneously to the interscapular region, where it was exteriorized through the skin and secured. Experimental procedures were performed 7-10 days after surgery.

Experimental procedure. The freely moving rats were infused with warmed-up lipid emulsion, mannitol, or vehicle (sterile water) into the duodenum for 30 min (0.05 ml/min, total volume 1.5 ml, infusion pump; Braun, Melsungen, Germany) in sawdust-devoid home cages. At 90 min after the end of the intraduodenal infusion, animals were anesthetized with ketamine (100 mg/kg ip, Parke-Davis) and xylazine (10 mg/kg ip, Bayer) and transcardially perfused with phosphate-buffered saline (PBS) buffer (0.1 M, pH 7.4) followed by Zamboni's fixative (2% formaldehyde and 2% picric acid in 0.1 M PBS buffer, pH 7.4). The brains were removed, postfixed in Zamboni's fixative, cryoprotected in 25% sucrose, and stained according to the protocol outlined below.

c-Fos immunohistology. After cryoprotection, free-floating brain sections (30 µm) were stained for c-FLI using the avidin-biotin-peroxidase method. Primary antibody (rabbit anti c-Fos oncoprotein polyclonal antiserum, a rabbit polyclonal antibody directed against the NH₂-terminal peptide 417 of the Fos protein; Dianova, Hamburg, Germany) was applied at a dilution of 1:1,000 with 0.5% Triton X-100 (Serva, Heidelberg, Germany) and 1% goat normal serum (Serva) in PBS (0.1 M) for 48 h at 4°C with gentle agitation. The biotinylated secondary antibody (goat anti-rabbit immunoglobulin G; Vector Laboratories, Burlingame, CA) was applied for 90 min. Diaminobencidine (Vector Laboratories) was used as the chromogen. Sections were mounted on chrome alum gelatin-coated slides and air dried before being dehydrated via a graded ethanol series, cleared in xylol, and placed under a coverslip with DePeX (Serva). Sections were examined with the use of bright-field microscopy.

Experimental design. In the first part of the study, our aim was to investigate whether intraduodenal infusion of lipid or mannitol (isosmotic control) would alter neuronal activity in the LCC, DVC, and PVN. Therefore, induction of c-fos expression at infusion of lipid or mannitol into the duodenum was determined by assessing changes in c-Fos immunoreactivity in these brain areas.

In the second part of this work, we studied whether CCK-A or CCK-B receptors are involved in intraduodenal lipid-induced modulation of neuronal activity in these brain nuclei. Therefore, we determined the effectiveness of selective and potent CCK-A and CCK-B receptor antagonists to block the activation of c-fos expression in LCC, DVC, or PVN induced by intraduodenal infusion of lipid (14). Accordingly, rats were pretreated with doses of 0.1 mg/kg or 1 mg/kg of the CCK-A receptor antagonist MK-329 or the CCK-B receptor antagonist L-365,260, or with vehicle, injected intraperitoneally at 45 min before intraduodenal infusion of lipid or vehicle (sterile water).

In the third part of the study, we wanted to determine whether capsaicin-sensitive vagal pathways are involved in the afferent mechanism mediating the activation of c-fos expression in the LCC, PVN, and DVC induced by intraduodenal infusion of lipid. In this experiment, rats were perivagally pretreated with capsaicin as described above to interrupt afferent neurotransmission by vagal C fibers at intraduodenal infusion of lipid. Sham-operated animals (which were treated with the vehicle, 10% Tween 80 in olive oil, bilaterally applied to the vagus nerve for 30 min as described above) intraduodenally infused with lipid and capsaicin-treated animals infused with vehicle into the duodenum served as controls.

In the fourth part of this study, our aim was to investigate whether the activation of c-fos expression in the LCC, DVC, or PVN induced by intraduodenal infusion of lipid is partially mediated by CCK receptors not localized on capsaicin-sensitive vagal afferents. Therefore, we performed a combined pretreatment with perivagal capsaicin plus intraperitoneal injection of the CCK-A or CCK-B antagonist, given at a dose of 1 mg/kg ip at 45 min before intraduodenal infusion of lipid or vehicle. Perivagal capsaicin-pretreated animals, as well as sham-operated rats treated with vehicle injected intraperitoneally and lipid or vehicle infused intraduodenally, served as controls.

Biochemical analysis of plasma CCK. Male Sprague-Dawley rats, weighing 300-350 g, were intraduodenally infused with lipid as described above. Starting 30 min before the intraduodenal lipid infusion, blood samples were taken retroorbitally every 30 min (4 animals per time point, at 2 time points during the experiment: 1 ml per rat and per time point, i.e., a total of 2 ml/animal in each of 12 rats) under general anesthesia with xylazine and ketamine as described above during an observation period of 150 min. Immediately after completion of each test, the blood samples were centrifuged at 4°C, and the plasma was stored at 25°C until assayed. Determination of plasma CCK concentrations included extraction and subsequent radioimmunoassay as described earlier (17). The CCK antibody (G-160, which was generously provided by Dr. I. Koop, Berlin, Germany) binds to the sulfated tyrosyl group of the CCK molecule and therefore identifies the biologically active form of CCK. It does not cross-react with nonsulfated gastrin and cross-reacts only to 1.5% with sulfated gastrin. Results are presented in picograms CCK per milliliter.

Data and statistical analysis. Semiquantitative assessment of c-fos expression was achieved by counting the number of c-FLI-positive cells in brain areas where intraduodenal lipid infusion induced c-fos expression, i.e., LCC, AP, NTS, and PVN. Cells with dark brown or black nuclear c-FLI staining above the generally low background at bright-field microscopy were identified as c-FLI-positive cells. Quantification of c-FLI-positive cells in the LCC, AP, NTS, and PVN was performed in every third of all coronal sections throughout the rostrocaudal extent of these brain areas. Anatomic correlations of the LCC, AP, NTS, and PVN were made according to landmarks given in a stereotaxic atlas (27). c-FLI-positive cells were counted in identical numbers of sections per rat for the various brain nuclei investigated, and in every immunohistological reaction identical numbers of brains were processed for all groups investigated to accomplish maximal consistency of the results. This proceeding was strictly performed in a single-blind method. Finally, the average number of c-FLI-positive cells per section was calculated in each rat for the LCC, NTS, AP, and PVN. Data are expressed as means ± SE of the average number of cells/section of all rats. The data were analyzed by analysis of variance, and differences between groups were evaluated by the Student-Newman-Keuls test; P <0.05 was considered significant.

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

Effects of intraduodenal infusion of lipid on c-fos expression. c-FLI was densely distributed in neuronal nuclei in the LC and also in the nucleus subceruleus alpha of the SC after intraduodenal infusion of the lipid emulsion (Fig. 1). In the same rats, c-FLI-positive cells were localized in high density in the lateral, medial, and ventral hypothalamic PVN (Fig. 2), as well as in the dorsolateral and dorsomedial region of the NTS and the caudal AP (Fig. 3). Minimal staining was also evident in the commissural NTS. In contrast, rats showed only minimal nuclear c-Fos staining in the LCC, AP, NTS, and PVN under control conditions and after intraduodenal infusion of mannitol (Figs. 1-3).

View larger version (149K): [in this window] [in a new window]   Fig. 1.   Effects of intraduodenal infusion of lipid on c-fos expression in locus ceruleus/subceruleus nuclei (LC/SC). A: expression of c-fos is only minimal under control conditions, as shown by the sparse number of c-Fos-like immunoreactivity (c-FLI)-positive neurons. B: lipid infused intraduodenally induces changes of neuronal activity in the LC/SC, as revealed by immunohistochemical detection of various c-FLI-positive neurons.

View larger version (131K): [in this window] [in a new window]   Fig. 2.   Effects of intraduodenal infusion of lipid on c-fos expression in the paraventricular nucleus of the hypothalamus (PVN). A: expression of c-fos is only minimal under control conditions, as shown by the low number of c-FLI-positive neurons. B: lipid infused intraduodenally induces changes of neuronal activity in the PVN, as shown by densely distributed c-FLI-positive neurons.

View larger version (143K): [in this window] [in a new window]   Fig. 3.   Effects of intraduodenal infusion of lipid on c-fos expression in the nucleus of the solitary tract (NTS) and the area postrema (AP). A: expression of c-fos is only minimal under control conditions, as shown by a low number of c-FLI-positive cells in these brain stem areas. B: lipid infused intraduodenally induces profound changes of neuronal activity in the NTS and AP, as proved by the abundant c-FLI-positive neurons.

The average number of c-FLI-positive cells per section in the LC/SC significantly increased 13-fold from 0.4 ± 0.1 cells/section under control conditions to 5.3 ± 0.1 cells/section at infusion of lipid into the duodenum. However, mannitol administered intraduodenally had no effect on c-fos expression in the LCC, as proved by an average number of 0.5 ± 0.4 c-FLI-positive cells/section (Fig. 4). The density of c-Fos immunoreactivity in the AP amounted to an average number of 5.8 ± 3.2 c-FLI-positive cells/section under basal conditions. Treatment with lipid infused intraduodenally increased the c-Fos immunoreactivity in the AP nearly sevenfold to 39.9 ± 5.4 c-FLI-positive cells/section. In contrast, mannitol infused intraduodenally had no effect on c-fos expression in the AP, as shown by an unchanged average number of 8.1 ± 2.4 c-FLI-positive cells/section (Fig. 4). Infusion of the lipid emulsion into the duodenum activated c-fos expression also in the NTS, increasing the average number of c-FLI-positive cells per section more than 40-fold, from 2.0 ± 0.3 cells/section under control conditions to 81.4 ± 28.3 cells/section. Intraduodenal mannitol did not alter neuronal activity in the NTS, as shown by an unchanging c-FLI density of 2.2 ± 0.6 positive cells/section (Fig. 4). Intraduodenal infusion of lipid induced a considerable activation of c-fos expression in the PVN, as shown by the 12-fold increase in the average number of c-FLI-positive cells/section from 6.4 ± 1.3 to 78.7 ± 20.0 in this hypothalamic nucleus. The average number of 11.5 ± 2.9 c-FLI positive cells/section at intraduodenal infusion of mannitol revealed that this treatment had no significant effect on neuronal activity in the PVN (Fig. 4).

View larger version (27K): [in this window] [in a new window]   Fig. 4.   Effects of intraduodenal infusion of lipid on c-fos expression in LC/SC (A), AP (B), NTS (C), and PVN (D). Expression of c-fos in these brain nuclei, as determined by c-Fos immunoreactivity, is minimal under control conditions (n = 6). Intraduodenal infusion of mannitol (n = 4) has no effect on c-fos expression. In contrast, lipid infused into the duodenum (n = 9) markedly induces activation of c-fos expression in the LC/SC, AP, NTS, and PVN, as shown by significantly increased numbers of c-FLI-positive cells. * P < 0.05 vs. controls; x P < 0.05 vs. mannitol; analysis of variance (ANOVA) and Student-Newman-Keuls test. Bars represent means ± SE.

Effects of intraduodenal infusion of lipid on CCK plasma levels. The average basal CCK plasma level in the rats (n = 4) was 3.3 ± 1.4 pg/ml. Intraduodenal infusion of lipid increased the CCK plasma concentration 3.9-fold to a maximum of 13.0 ± 2.5 pg/ml at the end of the 30-min treatment period. The CCK plasma levels remained significantly elevated for at least 30 min after the treatment period and were back to basal levels within 60 min after the end of intraduodenal lipid infusion (Table 1).

Effects of CCK-A and CCK-B antagonists on intraduodenal lipid-induced c-fos expression. In vehicle-pretreated rats, intraduodenal infusion of lipid induced neuronal activation in the LCC (LC and nucleus subceruleus alpha), PVN, NTS, and AP, as shown by a significant increase in the average number of c-FLI-positive cells per section (Fig. 5). Pretreatment with the CCK-A receptor antagonist MK-329 at a dose of 0.1 mg/kg ip significantly inhibited the intraduodenal lipid-induced activation of c-fos expression in the LC/SC, as revealed by a reduction of the increase in the number of c-FLI-positive cells per section of 39% (Fig. 5). The CCK-A antagonist also markedly reduced the intraduodenal lipid-induced activation of c-fos expression in the other brain nuclei investigated: pretreatment with this dose of MK-329 reduced the increase in the average number of c-FLI-positive cells per section in the AP by 100%, in the NTS by 73%, and in the PVN by 71% (Fig. 5). In contrast, pretreatment with 0.1 mg/kg ip of the CCK-B receptor antagonist had no effect on the intraduodenal lipid-induced activation of c-fos expression in the LC/SC, NTS, and PVN. The reduction of the intraduodenal lipid-induced increase of c-FLI-positive cells per section in the AP of 54% after pretreatment with L-365,260 was statistically not significant but nevertheless seems discernible (Fig. 5).

View larger version (37K): [in this window] [in a new window]   Fig. 5.   Effects of cholecystokinin (CCK-A) and CCK-B receptor antagonists (0.1 mg/kg ip) on intraduodenal lipid-induced activation of c-fos expression in LC/SC, AP, NTS, and PVN. A: CCK-A receptor antagonist MK-329 significantly inhibited intraduodenal lipid-induced activation of c-fos expression in the LC/SC, as determined by c-FLI. In contrast, CCK-B receptor antagonist L-365,260 had no effect. B: MK-329 markedly inhibited intraduodenal lipid-induced activation of c-fos expression in the AP, as determined by the density of c-FLI-positive cells. In contrast, CCK-B receptor antagonist L-365,260 had no significant effect. C: CCK-A antagonist MK-329 significantly inhibited intraduodenal lipid-induced activation of c-fos expression in the NTS, as determined by c-FLI. In contrast, CCK-B antagonist L-365,260 had no effect on c-fos expression in the NTS. D: MK-329 significantly inhibited intraduodenal lipid-induced activation of c-fos expression in the PVN. In contrast, CCK-B receptor antagonist L-365,260 had no effect. * P < 0.05 vs. vehicle intraperitoneally + vehicle intraduodenally; x P < 0.05 vs. vehicle intraperitoneally + lipid intraduodenally; + P < 0.05 vs. MK-329 intraperitoneally + lipid intraduodenally; ANOVA and Student-Newman-Keuls test. Bars represent means ± SE; n = 4 animals/group.

Increasing the dose of the CCK-A receptor antagonist to 1 mg/kg ip additionally inhibited the intraduodenal lipid-induced activation of c-fos expression in the LC/SC, causing a more pronounced reduction of the increase in the number of c-FLI-positive cells per section in the LC/SC by 59%. However, this higher dose of MK-329 did not further reduce the lipid-induced increase in the density of c-FLI-positive cells in the AP (98%), NTS (55%), or PVN (70%) (Table 2). In contrast to the 10-fold lower dosage of the CCK-B receptor antagonist, pretreatment with 1 mg L-365,260/kg ip significantly inhibited the intraduodenal lipid-induced activation of c-fos expression in the LC/SC and AP, reducing the increase in the average number of c-FLI-positive cells per section in the LC/SC by 70% and in the AP by 77%. However, this dose of the receptor antagonist still had no effect on the lipid-induced activation of c-fos expression in the NTS, and the reduction of the increase in the number of c-FLI-positive cells per section in the PVN by 36% after pretreatment with L-365,260 was statistically not significant (Table 2).

Effects of perivagal capsaicin treatment on intraduodenal lipid-induced c-fos expression. After pretreatment with perivagal capsaicin, the expression of c-fos at intraduodenal infusion of vehicle was similar to animals not pretreated with the sensory neurotoxin, as determined by the density of c-FLI-positive cells in the LC/SC of 0.7 ± 0.2 cells/section, in the AP of 2.2 ± 0.8 cells/section, in the NTS of 0.6 ± 0.3 cells/section, and in the PVN of 2.8 ± 1.0 cells/section (Figs. 4-6). In sham-operated rats, intraduodenal infusion of lipid induced a marked activation of c-fos expression in these brain nuclei, shown by the pronounced increase in the average number of c-FLI-positive cells in the LCC to 4.3 ± 0.5 cells/section, in the PVN to 82.3 ± 10.2 cells/section, in the NTS to 88.0 ± 11.7 cells/section, and in the AP to 34.0 ± 1.7 cells/section (Fig. 6). Perivagal capsaicin pretreatment significantly diminished the effect of intraduodenal lipid on c-fos expression in the LC/SC, reducing the increase in the average number of c-FLI-positive cells per section by 86% to 88.0 ± 11.7 cells/section (Fig. 6). In addition, pretreatment with capsaicin reduced the increase in the average number of c-FLI-positive cells/section in the NTS by 66% to 30.4 ± 7.5 cells/section, in the AP by 46% to 19.5 ± 2.1 cells/section, and in the PVN by 76% to 21.9 ± 11.3 cells/section (Fig. 6).

View larger version (32K): [in this window] [in a new window]   Fig. 6.   Effects of perivagal capsaicin treatment on intraduodenal lipid-induced activation of c-fos expression in LC/SC, AP, NTS, and PVN. A: perivagal capsaicin treatment blocked intraduodenal lipid-induced activation of c-fos expression in the LC/SC, as determined by c-FLI. B: intraduodenal lipid-induced activation of c-fos expression in the AP was significantly reduced by perivagal capsaicin treatment, as determined by c-FLI. C: perivagal capsaicin treatment significantly inhibited intraduodenal lipid-induced activation of c-fos expression in the NTS. D: intraduodenal lipid-induced activation of c-fos expression in the PVN was significantly reduced by perivagal capsaicin treatment, as determined by c-FLI. * P < 0.05 vs. capsaicin + vehicle intraduodenally; x P < 0.05 vs. sham operation + lipid intraduodenally; ANOVA and Student-Newman-Keuls. Bars represent means ± SE; n = 4 animals/group.

Effects of combined pretreatment with perivagal capsaicin and CCK-A or CCK-B receptor antagonists on intraduodenal lipid-induced activation of c-fos expression. In sham-operated rats treated intraperitoneally with vehicle, the intraduodenal infusion of lipid induced a pronounced activation of c-fos expression in the LC/SC, AP, NTS, and PVN, as shown by the marked increase in the number of c-FLI-positive cells per section in these brain nuclei (Fig. 7).

View larger version (36K): [in this window] [in a new window]   Fig. 7.   Effects of treatment with perivagal capsaicin and CCK-A or CCK-B receptor antagonists on intraduodenal lipid-induced activation of c-fos expression in LC/SC, AP, NTS, and PVN. A: pretreatment with perivagal capsaicin inhibited intraduodenal lipid-induced activation of c-fos expression in the LC/SC. Treatment with the CCK-A receptor antagonist MK-320 or the CCK-B receptor antagonist L-365,260 (1 mg/kg ip) in perivagal capsaicin-pretreated animals had no significant additional inhibitory effect on the lipid-induced activation of c-fos expression in the LC/SC. Combined pretreatment with perivagal capsaicin and the CCK-A or CCK-B receptor antagonist completely abolished significant differences from the sham-operated control group. B: pretreatment with perivagal capsaicin inhibited intraduodenal lipid-induced activation of c-fos expression in the AP. Treatment with the CCK-A receptor antagonist MK-320 or the CCK-B receptor antagonist L-365,260 (1 mg/kg ip) in perivagal capsaicin-pretreated animals additionally inhibited the lipid-induced activation of c-fos expression in the AP, as shown by a significant further reduction of the number of c-FLI-positive cells. Combined pretreatment with perivagal capsaicin and the CCK-A or CCK-B receptor antagonist completely abolished significant differences from the sham-operated control group. C: pretreatment with perivagal capsaicin inhibited intraduodenal lipid-induced activation of c-fos expression in the NTS. Treatment with the CCK-A receptor antagonist MK-320 or the CCK-B receptor antagonist L-365,260 (1 mg/kg ip) in perivagal capsaicin-pretreated animals had no significant additional inhibitory effect on intraduodenal lipid-induced activation of c-fos expression in the NTS. Combined pretreatment with perivagal capsaicin and the CCK-A receptor antagonist, but not with the CCK-B receptor antagonist, completely abolished significant differences from the sham-operated control group. D: pretreatment with perivagal capsaicin inhibited intraduodenal lipid-induced activation of c-fos expression in the PVN. Treatment with the CCK-A receptor antagonist MK-320 or the CCK-B receptor antagonist L-365,260 (1 mg/kg ip) in perivagal capsaicin-pretreated animals had no significant additional inhibitory effect on intraduodenal lipid-induced activation of c-fos expression in the PVN. Combined pretreatment with perivagal capsaicin and the CCK-A receptor antagonist, but not with the CCK-B receptor antagonist, completely abolished significant differences from the sham-operated control group. * P < 0.05 vs. sham operated + vehicle intraperitoneally + lipid intraduodenally; + P < 0.05 vs. sham operated + vehicle intraperitoneally + vehicle intraduodenally; x P < 0.05 vs. capsaicin + vehicle intraperitoneally + lipid intraduodenally; ANOVA and Student-Newman-Keuls. Bars represent means ± SE; n = 4 animals/group.

Perivagal capsaicin pretreatment in animals treated intraperitoneally with vehicle inhibited the lipid-induced activation of c-fos expression in the LC/SC, DVC, and PVN, reducing the increase of the average number of c-FLI-positive cells per section in the LC/SC by 84%, in the NTS by 84%, and in the PVN by 78%. However, the 43% reduction of the lipid intraduodenally induced increase in c-FLI-positive cells in the AP observed in this part of the studies was statistically not significant (Fig. 7).

In perivagal capsaicin-pretreated rats also treated with the CCK-A receptor antagonist MK-329 (1 mg/kg ip), no significant additional inhibition of the lipid-induced activation of c-fos expression was observed in the LC/SC, NTS, or PVN. Nevertheless, the combined pretreatment with perivagal capsaicin and MK-329 at intraduodenal infusion of lipid completely abolished significant differences in the number of c-FLI-positive cells per section in the LC/SC, AP, NTS, and PVN, in comparison with sham-operated rats treated with vehicle intraperitoneally as well as intraduodenally. Also, treatment of capsaicin-pretreated animals with the CCK-A receptor antagonist further inhibited the intraduodenal lipid-induced activation of c-fos expression in the AP, as shown by the significant reduction of c-FLI-positive cells in the AP under these experimental conditions (Fig. 7).

Treatment of capsaicin-pretreated rats with the CCK-B receptor antagonist L-365,260 (1 mg/kg ip) also additionally inhibited the intraduodenal lipid-induced activation of c-fos expression in the AP, as assessed by a significant reduction of the c-FLI-positive cells per section, in comparison with pretreatment with capsaicin alone. However, the CCK-B receptor antagonist had no additional inhibitory effects on the intraduodenal lipid-induced activation of c-fos expression in the LC/SC, NTS, and PVN of capsaicin-pretreated animals (Fig. 7). In contrast to treatment with the CCK-A receptor antagonist, combined pretreatment with perivagal capsaicin and L-365,260 at intraduodenal infusion of lipid did not abolish significant differences in the number of c-FLI-positive cells in the NTS and PVN in comparison with sham-operated rats treated intraperitoneally and intraduodenally with vehicle (Fig. 7).

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

The c-Fos protein, product of the c-fos immediate-early gene, is an intranuclear phosphoprotein, which after translation is rapidly transported into the cell nucleus; here it binds to DNA to regulate gene transcription. It has been shown in a variety of models that c-fos expression in CNS-neurons can be induced by various physiological stimuli. Thus changes in intranuclear c-FLI indicate alterations of neuronal activity. Labeling of neuronal tissue for the c-Fos protein provides an activity map with single-cell resolution and allows tracing of the neuronal pathways activated by physiological or pharmacological stimuli (13).

After intraduodenal infusion of the lipid emulsion, c-FLI was densely distributed in neuronal nuclei in the LC and nucleus subceruleus alpha. This observation provides the first experimental evidence that nutrients in the small intestine alter neuronal activity in the LCC, i.e., LC and SC, the subjacent region. There is convincing evidence that osmoreceptors are located in the small intestine, and it has been shown that a hyperosmotic solution in the stomach can induce c-fos expression in the brain (26). However, it seems unlikely that the effect of the lipid emulsion is due to activation of osmoreceptors or luminal distension, because intraduodenal infusion of an isosmolar mannitol solution under the same conditions had no significant effect on c-fos expression in the LCC or in the other brain nuclei investigated.

The LCC receives substantial afferent projections, among them from medullary areas like the DVC, which is well established to mediate viscerosensory and endocrine afferent information to the brain, and also from the hypothalamic PVN (5, 35). Physiological experiments have shown that the DVC and PVN play a significant role in the CNS regulation of feeding, satiety, and GI function (1, 28, 35). Furthermore, the dose of lipid infused into the duodenum in the present study is in the range of the amount of fat naturally encountered by the small intestine after feeding in rats (16). Therefore, the present data provide indirect evidence that the LCC might be involved in CNS-mediated GI and behavioral responses of the organism associated with feeding.

We have recently shown that CRF in the LC/SC mimics stress-induced alterations of GI secretion and motility, proposing a role for the LCC in the CNS regulation of the GI tract, and of CRF at this site in stress-related alterations of GI function (23, 24). Thus the data of the present study suggest that intestinal nutrient-related afferent influences on neuronal activity in the LC/SC could modulate efferent effects of this brain area on GI secretion and motility.

The results reveal that a CCK receptor-related mechanism and afferent vagal pathways play an important role in the intraduodenal lipid-induced activation of neurons in the LC and nucleus subceruleus alpha. Pretreatment with the CCK-A receptor antagonist MK-329 administered peripherally (0.1 mg/kg and 1 mg/kg ip) dose dependently reduced the intraduodenal lipid-induced activation of c-fos expression in the LC/SC by 39 and 59%, respectively. In contrast, pretreatment with the CCK-B receptor antagonist L-365,260 at a dose of 0.1 mg/kg ip had no effect, but L-365,260 significantly inhibited the intraduodenal lipid-induced activ ation of c-fos expression in the LC/SC at a dose of 1 mg/kg ip, reducing the number of c-FLI-positive cells in the LC/SC by 70%. There are two receptor subtypes for CCK: the CCK-A receptor and the CCK-B receptor. The CCK-A receptor has a high affinity for sulfated CCK and is the predominant CCK receptor type in the periphery. In contrast, the CCK-B receptor, which has a high affinity also to nonsulfated CCK, predominates in the brain (7, 12, 28, 32). The CCK-A antagonist MK-329 (devazepide) is well known to have a potent and specific affinity to peripheral CCK-A receptors and shows >1,000-fold selectivity over the CCK-B receptor (14). Therefore, the observation that intraduodenal lipid-induced activation of c-fos expression in the LCC can be suppressed by the low dose of MK-329 but not of L-365,260 suggests that the CCK-related effects of intraduodenal lipid on neuronal activity in the LCC are predominantly mediated via peripheral CCK-A receptors. The effect of the higher dose of the CCK-B receptor antagonist, which is of similar magnitude as at pretreatment with 1 mg MK-329/kg ip, could be due to nonspecific binding of L-365,260 to peripheral CCK-A receptors. However, CCK-B receptors have also been identified on peripheral endings of vagal nerve fibers as well as on sensory circumventricular organs like the AP (7, 32). Therefore, these data do not entirely exclude that CCK-related effects of intraduodenal lipid on c-fos expression in the LC/SC are partially mediated via CCK-B receptors.

Bilateral perivagal capsaicin treatment significantly diminished the increase in the number of c-FLI-positive cells per section in the LCC at intraduodenal infusion of lipid by 85% (study 3: 86%, study 4: 84%). The sensory neurotoxin capsaicin applied locally to the vagus nerve produces an impairment of the afferent innervation by vagal C-fibers (29). Thus the present data suggest that afferent neurotransmission of information about entry of lipid into the duodenum altering neuronal activity in the LC/SC is predominantly facilitated by vagal C-fibers.

In perivagal capsaicin-pretreated animals, simultaneous pretreatment with the CCK-A receptor antagonist MK-329 or the CCK-B receptor antagonist L-365,260 at a dose of 1 mg/kg ip, shown to be effective in study 2, had no significant additional inhibitory effect on intraduodenal lipid-induced activation of c-fos expression in the LC/SC. Electrophysiological experiments have demonstrated the sensitivity of vagal afferents to CCK (28, 33). Furthermore, CCK-A and CCK-B receptors have been found to be localized on the vagus nerve and to be transported within afferent vagal fibers (7, 25). Therefore, the present findings propose that predominantly CCK receptors on capsaicin-sensitive vagal fibers mediate the induction of neuronal activity in the LC/SC, induced by peripheral release of CCK at entry of lipid into the intestine.

Our data showing Fos protein expression in the NTS in response to intestinal lipid confirm previous results in this experimental model (37). However, the significance of CCK and vagal afferent pathways in intraduodenal lipid-induced activation of c-fos expression in the NTS has not been investigated before. In addition, the present data demonstrate that entry of lipid into the upper small intestine induces a pronounced increase in neuronal activity, as assessed by c-FLI, in the PVN and in the AP.

Pretreatment with the CCK-A receptor antagonist MK-329 at a dose of 0.1 mg/kg ip reduced the lipid-induced increase in c-FLI-positive neurons in the NTS by 73% and in the PVN by 71%. In both brain nuclei, a 10-fold higher dose of MK-329 had no additional inhibitory effect. Pretreatment with perivagal capsaicin reduced the increase in the number of activated neurons in the NTS by ~70% (study 3: 66%, study 4: 84%) and in the PVN by ~77% (study 3: 76%, study 4: 78%); these effects are in the same range as with treatment with MK-329. In contrast, pretreatment with these doses of the CCK-B receptor antagonist L-365,260 had no effect on c-fos expression in both brain nuclei. It is well established that entry of fat into the upper small intestine causes release of CCK from mucosal endocrine cells (28), and we proved that in this experimental model infusion of lipid induces release of CCK in the periphery. Thus the present results demonstrate that CCK-A receptors and capsaicin-sensitive vagal afferents play a central role in the intraduodenal lipid-induced activation of c-fos expression in the NTS and PVN.

In perivagal capsaicin-pretreated animals, simultaneous pretreatment with the CCK-A receptor antagonist, at a dose shown to be effective in study 2, or with the CCK-B receptor antagonist had no significant additional inhibitory effect on intraduodenal lipid-induced activation of c-fos expression in the NTS and PVN. These observations and the comparable effectiveness of pretreatment with perivagal capsaicin and MK-329 in our study suggest that predominantly CCK-A receptors on capsaicin-sensitive vagal fibers facilitate the CCK-mediated afferent neurotransmission of information about entry of lipid into the duodenum from the periphery to the NTS and PVN, resulting in increased Fos protein expression in these brain nuclei. These results are in agreement with reports that activation of c-fos expression in the NTS and PVN induced by exogenous sulfated CCK-8 administered peripherally is dependent on vagal nerve integrity and predominantly mediated by CCK-A receptors (6, 9, 22). The recent report that activation of c-fos expression in the DVC induced by ingestion of food is not dependent on CCK is not in conflict with our results, because oral, pharyngeal, esophageal, and gastric cues contribute to meal-induced c-fos expression (10, 11).

Patterns of c-fos expression in the brain differ between treatments associated with decreased or increased food intake and gastric motility; the distribution of c-FLI observed in our study closely resembles the pattern viewed at situations associated with decreased gastric motility and food intake (26). These results, as well as the effects of perivagal capsaicin and CCK receptor antagonists observed in our experiments, are in accordance with physiological and behavioral studies showing that intraduodenal infusion of lipid inhibits food intake as well as gastric emptying and secretion via vagal afferent pathways and CCK receptor-mediated mechanisms (28, 36). Therefore, activation of c-fos expression in brain nuclei under physiological circumstances is not just useful for mapping neuroanatomical circuits of activated neurons but most likely is also of physiological relevance. This might explain why in a previous study intraduodenal infusion of casein, which is thought to increase circulating levels of CCK in the rat but is only a weak inhibitor of food intake in this species, had no effect on c-fos expression in the NTS in this experimental model (37).

The NTS is the primary relay station of vagally mediated afferent neurotransmission of information from the viscera to midbrain and forebrain areas (28, 35). It has been shown that exogenous CCK activates NTS neurons projecting to the PVN, and functional neuronal connections between the DVC and the PVN have been described (30, 35). Therefore, direct projections from the NTS to the PVN might provide the neuroanatomical basis for the intraduodenal lipid-induced activation of c-fos expression in the PVN. However, the LCC and catecholaminergic neurons in the ventrolateral medulla also receive input from the DVC, project to the PVN, and are activated by peripheral exogenous CCK (15, 19, 22, 30). Consequently, these brain pathways could also be involved in the afferent transmission of information about entry of nutrients into the intestine to the PVN.

The observation that lipid in the upper small intestine induces activation of c-fos expression in the AP is in accordance with previous studies reporting increased c-FLI in the AP after ingestion of nutrients (10, 11, 31). Pretreatment with the CCK-A receptor antagonist MK-329 at a dose of 0.1 mg/kg ip already completely blocked the induction of c-fos expression in the AP. These data suggest that the intraduodenal lipid-induced activation of neuronal activity in the AP, as assessed by changes in c-FLI, is completely mediated by CCK receptors.

Perivagal capsaicin pretreatment reduced the increase in the average number of c-FLI-positive cells in the AP in study 3 by 43% and in study 4 by 46%. Thus the reduction of c-fos expression at sensory vagal denervation is markedly lower in the AP than in the other brain nuclei investigated (LCC: 85%, NTS: 70%, PVN: 77%). Infusion of lipid into the duodenum significantly increases the level of circulating CCK in this experimental model. Furthermore, CCK-B and also CCK-A receptors have been detected in the AP, and the AP, as one of the sensory circumventricular organs, is directly accessible to circulating CCK (12, 32). Therefore, the present observations indicate that intraduodenal lipid-induced neuronal activation in the AP is partially mediated by vagal afferent pathways and partially results from direct effects of circulating CCK at the AP itself. It has been shown that NTS neurons receive afferent input from the AP (2). The data of the present study provide indirect evidence that the AP also receives input from the NTS. The idea that the lipid-induced activation of c-fos expression in the AP is mediated by vagal and nonvagal pathways is supported by the outcome of combined pretreatment with perivagal capsaicin and CCK-receptor antagonists. These experiments in capsaicin-pretreated animals revealed additional inhibitory effects of the CCK-receptor antagonists on the intraduodenal lipid-induced activation of c-fos expression in the AP, in contrast to the other brain nuclei investigated.

Intraduodenal infusion of lipid elevates the levels of circulating CCK in this experimental model. Therefore, the induction of c-fos expression in the AP by intraduodenal lipid could result from paracrine or endocrine effects of CCK on vagal afferents or from direct actions of circulating CCK on the AP (28, 33). Pretreatment with the CCK-B receptor antagonist at a dose of 0.1 mg/kg ip, which had no effect on c-fos expression in the other brain nuclei investigated, reduced the lipid-induced increase in c-FLI in the AP by 54%. The higher doses of L-365,260 (1 mg/kg ip), which still had no significant consequence on expression of the c-Fos protein in the NTS and PVN, reduced the increase of c-FLI in the AP by 77%. It is well established that CCK-B receptors are located in the AP but have also been detected on the vagus nerve (7, 32). Thus the inhibitory effect of the CCK-B receptor antagonist on the lipid-induced increase of c-FLI in the AP could be due to action on vagal afferents or on the AP itself. However, CCK-B receptors are predominantly located in the brain, and L-365,260 was still effective in capsaicin-pretreated animals. Therefore, the present results suggest that CCK-B receptors in the AP are involved in the intraduodenal lipid-induced activation of c-fos expression in this brain area. Nevertheless, the CCK-A receptor antagonist MK-329 was more effective than L-365,260. Thus the effect of the CCK-B receptor antagonist also could result from nonspecific binding to the CCK-A receptor. However, the marked differences in the efficacy of L-365,260 to inhibit neuronal activation in the AP versus the other brain nuclei investigated and the high density of CCK-B receptors in the AP suggest that CCK-B binding sites in this circumventricular organ are involved in the activation of AP neurons by lipid entering the small intestine.

It has been shown that AP neurons and vagal afferents converge to excite cells in the NTS and that NTS neurons receive afferent input from the AP (2). Therefore, the incomplete (70-85%) blockade of lipid-induced activation of c-fos expression in the LCC, NTS, and PVN after pretreatment with perivagal capsaicin could be due to input from the AP to these brain areas. However, the observation of complete blockade of c-fos induction in the AP after pretreatment with MK-329, but not in the other brain nuclei investigated, makes this very unlikely. After neonatal capsaicin treatment, the density of CCK receptors on the vagus nerve is reduced only by ~50% (21). Nevertheless, it is unclear whether the incomplete blockade of lipid-induced activation of c-fos expression in the LCC, NTS, and PVN after pretreatment with perivagal capsaicin is due to incomplete functional impairment of capsaicin-sensitive afferent C-fibers or whether afferent pathways different from capsaicin-sensitive vagal fibers and CCK receptors are involved in these effects.

In summary, lipid entering the duodenum induces alterations of neuronal activity in the LCC, the DVC, and the PVN. CCK-A receptors on capsaicin-sensitive, afferent vagal fibers predominantly facilitate the transmission of information about the presence of lipid in the upper small intestine from the gut to the LCC, NTS, and PVN, leading here to activation of c-fos expression. Capsaicin-sensitive vagal afferents and nonvagal pathways mediate the activation of c-fos expression by intraduodenal lipid in the AP, presumably via CCK-A and CCK-B receptors.

Perspectives

Intraduodenal lipids induce c-fos expression in the LCC by similar pathways as in DVC and PVN, suggesting that the LCC is part of this circuitry. The LCC plays an important role in the stress response of the organism (34), most likely also in stress-related alterations of GI function (23, 24), and has been shown to be involved in CNS mechanisms of anxiety and depression (4). Thus these present observations may have implications for the understanding of the pathophysiology of functional GI disorders and other diseases where emotional distress and alterations of GI function interact (3).

The afferent information about entry of lipids into the intestine, which induces c-fos expression in DVC, LCC, and PVN predominantly by peripheral release of CCK, most likely also affects efferent activity of these brain nuclei with consecutive alterations of GI function and behavior, e.g., feeding (35). The neurochemical phenotypes of these activated neurons and the brain neurotransmitters mediating the physiological and behavioral responses still need to be established.