SEARCH

SEARCH BY CITATION

Keywords:

  • cholecystokinin-A receptor;
  • enteric nervous system;
  • intraganglionic laminar endings;
  • intramuscular arrays;
  • nodose ganglia;
  • exocrine pancreas;
  • immunohistochemistry

Abstract

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. CONCLUSIONS
  7. Acknowledgements
  8. REFERENCES

A large body of evidence derived from electrophysiological recording and pharmacological/behavioral experiments suggests the presence of CCKA-receptors on vagal primary afferent fibers innervating the gastrointestinal tract. With the availability of antibodies specific for the CCKA-receptor, we wanted to demonstrate its presence and distribution on identified vagal afferent fibers and different types of terminals in the mucosa, myenteric plexus, and external muscle layers of the stomach and duodenum. In the duodenal mucosa, neither a C-terminal (Ab-1) nor an N-terminal (Ab-2) specific antibody produced any specific staining; in the myenteric plexus, non-vagal enteric neurons and their processes, but not vagal intraganglionic laminar endings (IGLEs), exhibited CCKAR-immunoreactivity. Similarly, in the gastric myenteric plexus, a population of enteric neurons and their processes, but not identified vagal IGLEs, were labeled by both antibodies. In both external muscle layers of the stomach, CCKAR-immunoreactive axons were in close register with labeled vagal afferent intramuscular arrays, but the two labels were not contained in the same varicosities. Ab-1 immunoreactivity was found in the cell membrane of vagal afferent perikarya in the nodose ganglia and in pancreatic acinar cells. The failure to detect CCKAR-immunoreactivity in peripheral vagal afferent terminals cannot be due to methodological problems because it was present in enteric neurons in the same sections, and because it did not stain structures resembling IGLEs in material without the potentially masking vagal afferent label. We conclude that CCKA-receptors on vagal afferent terminals: 1) are below the immunohistochemical detection threshold, 2) exhibit a conformation or affinity state inaccessible to the two antibodies, or 3) are not transported to the peripheral terminals. Anat Rec 266:10–20, 2002. © 2002 Wiley-Liss, Inc.

An impressive number of studies implicate cholecystokinin released from enteroendocrine cells in the upper small intestine through a vagal afferent mechanism in the control of food satiation. At the behavioral level, it was demonstrated that CCK administered ip loses its capacity to suppress food intake after subdiaphragmatic vagotomy (Smith et al., 1981; Joyner et al., 1993), systemic or local capsaicin treatment (South and Ritter, 1988; Chavez et al., 1997), or selective surgical vagal deafferentation (Smith et al., 1985). At the electrophysiological level, CCK induced increased firing of primary vagal afferent fibers (Blackshaw and Grundy, 1990; Schwartz and Moran, 1994) and second-order neurons in the nucleus tractus solitarius (NTS) (Ritter et al., 1989), the latter depending again on the integrity of the vagus nerve. At the anatomic/morphological level, a close anatomical relationship was demonstrated between CCK-producing enteroendocrine cells and at least some of the vagal afferent fibers in the duodenal mucosa (Berthoud and Patterson, 1996). Furthermore, local infusion of CCK into the pancreatico-duodenal artery required the lowest doses of the hormone to suppress food intake, suggesting that the relevant receptors are located in the proximal duodenum (Cox, 1998).

In addition, the CCKA (but not the CCKB) receptor is implicated in this peripheral mechanism, since the selective CCKA-receptor antagonists devazepide and lorglumide (but not selective CCKB-receptor antagonists) were able to block CCK-induced suppression of food intake (Corwin et al., 1991; Miesner et al., 1992). CCKA-receptor mRNA has been demonstrated by polymerase chain reaction (PCR) in nodose ganglia (Moriarty et al., 1997), and autoradiography has shown that CCKA-binding sites are concentrated at the ligation site of a ligated vagus nerve, suggesting transport of CCKA-receptors from the nodose ganglion to the periphery (Corp et al., 1993).

The aim of the present study was to identify CCKAR-immunoreactivity to specific terminal sites of the vagal afferent fibers innervating gastrointestinal targets. Using anterograde tracing of vagal afferents from the nodose ganglia, we (Berthoud and Powley, 1992; Berthoud et al., 1995; Williams et al., 1997; Kressel et al., 1994) and others (Phillips et al., 1997; Wang and Powley, 2000) have identified the specifics of vagal afferent innervation in the various layers of the gastrointestinal tract. Three major different terminal structures were identified. In the mucosa of both the stomach and small intestine a moderate number of terminal axons were found reaching all the way up to the tip of the duodenal villi and gastric glands (Berthoud and Neuhuber, 1994; Berthoud et al., 1995; Williams et al., 1997). In both the longitudinal and circular external smooth-muscle layers of primarily the stomach, arrays of branching axonal terminals, so-called intramuscular arrays (IMAs) have been described (Berthoud and Powley, 1992; Wang and Powley, 2000). Finally, in the myenteric plexus throughout the esophagus and gastrointestinal tract, a large number of profusely arborizing leafy terminal structures, so-called intraganglionic laminar endings (IGLEs), have been identified (Nonidez, 1946; Kolossow and Milochin, 1963; Rodrigo et al., 1975; Neuhuber, 1987; Berthoud and Powley, 1992; Berthoud et al., 1997; Wang and Powley, 2000).

The most plausible site for CCK to excite vagal afferents seems to be in the lamina propria of the duodenal mucosa. Some behavioral evidence supports this view of a paracrine mechanism as opposed to a true hormonal mechanism. It has been demonstrated that food intake suppression induced by intraduodenal infusion of various nutrients is not correlated with the obtained plasma CCK increases (Brenner et al., 1993). However, the participation of gastric vagal afferents, directly or indirectly, is also indicated. Single vagal afferents responding to gastric distension in vivo have been clearly shown to be sensitive to CCK (Schwartz et al., 1994), and vagal afferent recording from an in vitro stomach preparation in the presence of smooth-muscle stabilizing agents found sensitivity of some, but not all, tested vagal afferents to CCK. (Berthoud and Blackshaw, unpublished observations). Furthermore, a recent immunohistochemical study (Sternini et al., 1999) demonstrated CCKA-receptor immunoreactivity on enteric neurons in the stomach. The study also suggested the presence of CCKA-receptors on vagal fibers in the gastric mucosa on the basis of a reduced number of CCKAR-immunoreactive profiles in rats with subdiaphragmatic vagotomy. Surprisingly, no CCKAR-immunoreactive fibers were found in the duodenal mucosa.

To clarify the distribution and exact location of CCKA-receptors on the various terminal structures of gastrointestinal vagal afferent fibers, we used two different anti-CCKA-receptor antibodies, either alone or in combination with a new method for labeling vagal afferent fibers (Kressel, 1998). A newly available antibody directed against the C-terminal region (intracellular domain) of the CCKAR molecule was compared to the N-terminal specific antibody (extracellular domain) used by Sternini et al. (1999). Vagal afferents were anterogradely traced from the nodose ganglion with horseradish peroxidase (HRP), and visualized by directly attaching fluorochrome-conjugated streptavidin to the HRP via biotin (Kressel, 1998). Pancreatic acinar tissue and nodose ganglia served as positive control tissues, wherein CCKA-receptor activity has been shown previously (Moriarty et al., 1997).

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. CONCLUSIONS
  7. Acknowledgements
  8. REFERENCES

Animals

Fifty-five young adult, male Sprague-Dawley rats (100–250 g) were housed under normal laboratory conditions (12:12 hr light/dark cycle, lights on at 0700 hr; 22 ± 2°C), with lab chow (Purina 5001, Richmond, IN) and water available ad libitum. All animal procedures were approved by the Institutional Animal Care and Use Committee and conformed to the guidelines of the National Institutes of Health.

Anterograde Tracing of Vagal Afferents

Under ketamine/xylazine/acepromazine (80/4/1.6 mg/kg, sc) anesthesia and after atropine administration (2 mg/kg, ip), the left nodose ganglion was exposed by a ventral approach in 39 rats. Using a glass micropipette with a tip diameter of ∼10 μm, wheat germ agglutinin-conjugated horseradish peroxidase (WGA-HRP, 2 μl of 4% aqueous solution, Sigma (St. Louis, MO) grade IV) was then pressure-injected. The wound was closed with wound clips, and 24–48 hr were allowed for anterograde transport. In some rats (n = 35), Fluorogold (1–2 mg in 1.2 ml sterile saline, ip; Fluorochrome, Denver, CO) was injected 2–5 days before perfusion in order to label all enteric and autonomic neurons (Powley and Berthoud, 1991).

Tissue Preparation

Rats were euthanized with an overdose of pentobarbital sodium (200 mg/kg, ip) and transcardially perfused with heparinized (20 U/ml) saline, followed by freshly prepared 4% phosphate-buffered (pH 7.4) paraformaldehyde solution. The stomach, duodenum, pancreas, brainstem, and nodose ganglia were extracted, postfixed for 0.5–2 hr, and then placed in 15% sucrose in 0.1 M phosphate buffer overnight at 4°C, after which they were frozen in cold isopentane over dry ice and stored at –20°C until sectioned. Tissue not labeled with WGA-HRP was blocked and stored for up to 12 months at −20°C in cryoprotectant solution (50% PBS, 30% ethylene glycol, 20% glycerol) or at 4°C overnight in 25% sucrose in 4% paraformaldehyde before sectioning.

Samples (∼5 × 5 mm) of the ventral fundus and corpus, proximal duodenum (D1-D2), and pancreas head, as well as the right nodose ganglion and the medulla were prepared for cryostat sectioning. In the proximal duodenum, 20-μm-thick flat sections (in the plane of the myenteric plexus) and cross sections both parallel to the longitudinal and the circular muscle (for analysis of the mucosa) were taken from areas close to the mesentery. Similar 20 μm thick flat- and/or cross sections were obtained from the mid-ventral regions of the corpus and fundus. Frontal sections of medulla were cut at 25 μm and nodose ganglia were sectioned at 14 μm. Sections from the pancreas, nodose ganglion, and some of the mucosal cross sections were directly thawed, dried, and processed on slides. Other mucosal cross sections of the corpus and duodenum, as well as the flat gut and brainstem sections, were collected in 0.1 M phosphate buffer or phosphate-buffered saline (PBS) for free-floating immunohistochemical processing.

Characterization and Specificity of Antibodies

The polyclonal antibody (Ab-1) was raised in rabbits against a peptide with the amino acid sequence SHMSTSAPPP, corresponding to the intracellular domain near the C-terminal of the rat CCKA receptor (antiserum B 380-1; Euro-Diagnostica, Sweden, distributed by Accurate Chemicals, New York, NY). Specificity of the antibody was recently tested by preabsorption with the synthetic peptide fragment used for immunization, and by comparing it with in situ hybridization (ISH) (45-mer riboprobe, nucleotides 982–1,026 of cDNA of rat CCKA-receptor) in rat pancreas (Ohlsson et al., 2000). Staining was completely absent in preabsorption controls, and results with ISH confirmed location of the receptor in the pancreas (Ohlsson et al., 2000). We tested the specificity of staining only by omitting the primary antibody.

The AB-2 antibody (94159) used in this study was obtained from the CURE Digestive Diseases Research Center at UCLA, Los Angeles, CA. The specificity of this antibody to an amino acid sequence (CCKA 45-57) near the extracellular N-terminal domain of the rat CCKA-receptor was described by Sternini et al. (1999). Specificity was further confirmed by Western blots, cell-surface staining of CCKAR-transfected cells, and internalization of the CCKA-receptor in cells treated with CCK (Sternini et al., 1999). In our immunohistochemical procedures, preabsorption with the peptide revealed no staining in rat tissue.

Simple CCKAR-Immunohistochemistry Protocols

For tissue processed with AB-1, sections were initially rinsed in fresh 0.1% sodium borohydride in PBS. Following PBS washes, the tissue was blocked with a solution of 5% normal goat serum (NGS) and 1% bovine serum albumin (BSA) in PBS with 0.5% Triton X-100 (PBS/T) to minimize background staining. Primary antibody was then applied at 1:1,000–1:4,000 concentrations with 2% BSA in PBS/T as the diluent. The dilutions differed depending on the tissue: the pancreas stained best at 1:4,000, whereas the gut stained at 1:1,000. Incubation was 18–20 hr at room temperature or 40 hr at 7°C. Secondary antibody incubation followed thorough PBS washing. Alexa 594 goat anti-rabbit IgG (Molecular Probes, Eugene, OR) was diluted 1:2,000 in 2% BSA-PBS/T. Sections were incubated for 2 hr at room temperature in the dark; then they were washed, placed in 70% glycerol for 1 hr, and mounted in 100% glycerol with 5% n-propyl gallate added as an anti-fade agent. Pancreas tissue was used as a positive control for AB-1. Nonspecific staining was absent when the primary antibody was omitted.

The protocol for AB-2 differed from the above in that the primary antibody concentration used was 1:100, using simple immunofluorescence in the gut tissue. With biotinylated tyramine amplification, primary antibody was diluted 1:3,000, and Texas Red streptavidin (1:500, 2 hr; Vector, Burlingame, CA) or Cy-2 streptavidin (1:800, 2 hr; Jackson, West Grove, PA) was then used to visualize the CCKAR immunoreactivity. Because of variability in staining, we tried other buffers in our protocols, including the Tris buffer used in the Kressel protocol (1998) and 0.1-M phosphate buffer in the Sternini et al. (1999) report. All the buffers gave similar results. As with AB-1, there was no staining of tissue when the primary antibody (AB-2) was omitted.

Combined Immunohistochemistry and Vagal Afferent Labeling

Vagal afferents traced from the nodose ganglion were detected with a protocol developed by Kressel (1998), in which biotinylated tyramine amplified the HRP and fluorescent-labeled streptavidin was used for visualization. Briefly, the tissue was treated with 0.05% sodium borohydride in 0.1 M Tris-HCl, 0.15 M NaCl, pH 7.5 (TN), and then washed in TN with 0.05% Tween 20 (TNT). Sections were then permeabilized with 0.1% Triton X-100 in TN and placed in 2% BSA-TNT. After an hour, the tissue was washed in TNT and biotinylated tyramine (1:50, Dupont-NEN, Boston, MA) was applied for 15 min. For sequentially labeled sections, the tissue was again washed in TNT and the WGA-HRP was visualized with Cy-2 conjugated streptavidin (1:800 in 2% BSA-TNT for 2 hr). PBS followed TNT washes before the tissue was treated for single CCKAR labeling, as described above. Simultaneous labeling for vagal afferents and AB-1 or AB-2 called for post-tyramine TNT washes followed by PBS. Blocking and primary antibody incubations were the same as for single labeling, but detection combined Cy-2 streptavidin for WGA-HRP with Alexa 594 for CCKAR, both diluted in 2% BSA-PBS/T for 2 hr at 7°C. Washing, mounting, and coverslipping were the same as in the simple immunofluorescence protocol.

Confocal Microscopy

Glycerol-mounted tissue sections were viewed with conventional epifluorescence microscopy and images of select motifs were generated in a Zeiss LSM-310 confocal microscope using the 568 nm and 488 nm laser lines of an argon/krypton laser for the Alexa 594 and Cy-2 fluorophores, respectively. Generally, extended focus images were generated by collapsing three to 25 optical sections, 0.6–2 μm apart. Sharpening algorithms were used for some images.

RESULTS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. CONCLUSIONS
  7. Acknowledgements
  8. REFERENCES

Nodose Ganglia

A portion of vagal afferent perikarya in the nodose ganglia expressed CCKAR-immunoreactivity with Ab-1 (n = 4 rats), but not with Ab-2 (n = 6 rats, Fig. 1A–E). Omission of the primary antibody resulted in complete absence of staining (Fig. 1B). Immunoreactivity was localized to the neuronal cell membrane (Figs. 1C and D, and 4M). No staining was obtained with Ab-2, even at high antibody concentrations up to 1:50 (Fig. 1E).

thumbnail image

Figure 1. CCKA-receptor immunoreactivity in (A–E) rat nodose ganglion and (F–J) pancreas, using either (A–D and F–H) a C-terminal specific antibody (Ab-1) or (E and J) an N-terminal specific antibody (Ab-2). Extended focus images represent three to six optical sections, 0.6–1.5 μm apart, obtained with confocal microscope. A–D: About half of neurons in the nodose ganglia show CCKAR-immunoreactivity restricted to cell membrane with Ab-1. B: Omission of primary antibody results in a lack of any staining. E: Ab-2 does not result in any specific staining. (F) Acinar cell membranes and (G) some islet cells show CCKAR-immunoreactivity using Ab-1. H: No translocation of receptor from membrane to cytoplasm was detected in animal stimulated with CCK (100 μg/kg, ip, 1 hr). J: No specific staining was obtained in acinar tissue with Ab-2. Scale bar in J: (A–C and E) 50 μm; (D) 20 μm; (F–J) 30 μm.

Download figure to PowerPoint

Exocrine Pancreas

Similarly, pancreatic acinar cells strongly expressed CCKAR-immunoreactivity when using Ab-1 (n = 2 rats), but not when using Ab-2 (n = 2 rats, Fig. 1F, H, and J). Omission of the primary antibody resulted in complete absence of staining (not shown). Immunoreactivity was limited to acinar cell membranes (Fig. 1F). Stimulation with CCK (100 μg/kg, ip) did not result in detectable receptor translocation into the cytoplasm (Fig. 1H). Ab-2 did not result in any staining, even at high antibody concentrations of 1:50 (Fig. 1J). In addition to acinar cells, certain cell populations within islets of Langerhans also expressed membrane-bound CCKAR-immunoreactivity (Fig. 1G).

Duodenum

Neither antibody, used at any concentration, with or without amplification protocols, and alone or in combination with vagal afferent labeling, was able to specifically label any nerve fibers in the duodenal mucosa (Fig. 2A and B). We have extensively inspected more than 100 cross sections sampled from several different locations within the proximal duodenum of five rats. We previously reported (Berthoud et al., 1995) the presence of DiI-labeled vagal afferent fibers in the lamina propria of both the crypt and villous areas (Fig. 2G and H). Although many cross sections did not contain any DiI-labeled fibers, they were quite abundant in certain areas. In contrast, HRP-labeled vagal afferents were very rarely seen in the duodenal mucosa (Fig. 2F), and no attempt for double labeling was made.

thumbnail image

Figure 2. CCKA-receptor immunoreactivity in relation to vagal afferent innervation in rat duodenum. A–C: Absence of CCKAR-immunoreactive fibers in lamina propria of (A) crypts and (B) villi of duodenal mucosa. Note that CCKAR-immunoreactive fiber profiles are located in cross-sectioned myenteric ganglion only (arrow in A), shown at higher magnification in C (shown for Ab-2, but same negative results with Ab-1). D and E: In whole mounts or flat sections through the myenteric plexus, CCKAR-immunoreactivity is contained in varicose fibers in (D) interconnective strands (Ab-2) and (E) myenteric ganglia (Ab-1). F–H: Vagal afferent fibers traced with either (F) HRP or (G and H) DiI are shown for comparison. Vagal afferent fibers distribute both (F and G) in villous lamina propria and (H) around crypts. cl = crypts of Lieberkühn, cm = circular muscle, e = epithelium, lm = longitudinal muscle, lp = lamina propria, sm = submucosa. Scale bar in B: (A, B, and H) 75 μm; (C and E) 50 μm; (F and G) 60 μm; (D) 250 μm.

Download figure to PowerPoint

In the same cross sections that did not exhibit any CCKAR-immunoreactivity in the mucosa, immunorectivity was present in the myenteric plexus in the form of varicose fibers surrounding myenteric neurons (Fig. 2A and C), and this was seen with both antibodies. In flat sections through the duodenal myenteric plexus, an abundance of CCKAR-immunoreactive fibers could be seen (Fig. 2D and E), and a few neuronal perikarya with membrane-bound CCKAR-immunoreactivity were also present (not shown). In material processed for CCKAR only, no IGLE-like structures were found.

In double-labeled material, both HRP-labeled vagal afferent fibers and CCKAR-immunoreactive fibers were present in the same duodenal myenteric ganglia (Fig. 4K), but no colocalization was seen.

In addition to neuronal elements, Ab-1 CCKAR-immunoreactivity was detected in Brunner's glands (not shown) and in interstitial cells of Cajal (Fig. 4L) near the pylorus (reported in Patterson et al., 2001).

Stomach

In the gastric mucosa, Ab-2 yielded staining of some nerve fibers coursing through the lamina propria surrounding the gastric glands (Fig. 3A and B). Only very few and weakly labeled fibers were found with Ab-1 (Fig. 3C–E). In contrast, gland cells near the base of gastric glands exhibited strong immunoreactivity in the cell membrane when Ab-1 was used (Fig. 3F and G), but not Ab-2. Only a few HRP-labeled vagal afferent fibers were present in the gastric mucosa, and no colocalization with CCKAR-immunoreactivity was detected.

thumbnail image

Figure 3. CCKA-receptor immunoreactivity in rat gastric mucosa. A–C: Varicose nerve fibers running in the lamina propria parallel to the gastric glands expressing CCKAR-immunoreactivity with (A and B) antibody Ab-2, but not with (C) Ab-1. D and E: Vagal afferent fibers traced with HRP are relatively scarce, and did not colocalize CCKAR-immunoreactivity. F and G: Antibody Ab-1 (but not Ab-2) labeled membranes of gastric gland cells near the base of mucosa. Scale bar in F: (A) 150 μm; (B–E) 75 μm; (F and G) 60 μm.

Download figure to PowerPoint

In the gastric myenteric plexus, CCKAR-immunoreactive varicose nerve fibers and cell bodies were present. Using Ab-1, characteristically labeled neuronal cell bodies were present in most of the myenteric ganglia (Figs. 4F and 5B–F). Ab-1-generated CCKAR-immunoreactivity typically delineated the cell membrane and axon, as well as either short laminar dendrites (Figs. 4F and 5C–E), or longer, sometimes leafy dendrites (Fig. 5F). In contrast, Ab-2-generated CCKAR-immunoreactivity was typically limited to the cytoplasm and axon, without labeling any dendrites (Figs. 4B and C, and 5A). In material processed for the receptor only, no CCKAR-immunoreactive structures resembled that of vagal IGLEs. HRP-labeled vagal afferent fibers and IGLEs were abundant (Fig. 4E and F), but in no instance was colocalization of CCKAR-immunoreactivity observed. Some IGLEs were in close contact with CCKAR-immunoreactive myenteric neurons (Fig. 4F).

thumbnail image

Figure 4. Dual-channel confocal images showing the relationship between CCKA-receptor immunoreactive structures (red) with traced vagal afferent fibers (green) in (A–J) rat stomach, (K) duodenum, (L) pylorus, and (M) nodose ganglion. Extended focus images obtained from six to 20 optical sections, 0.8–2.0 μm apart. A–F: Absence of colocalization of CCKAR-immunoreactivity in labeled vagal afferent fibers and IGLEs in gastric myenteric plexus. CCKAR-immunoreactivity was expressed in varicose fibers within myenteric ganglia and interconnecting strands with both antibodies. Some enteric neurons also expressed CCKAR-immunoreactivity, primarily in the (B and C) cytoplasm and proximal axon with antibody Ab-2, and (F) membrane, filamentous dendrites, and axon with antibody Ab-1 (see also Fig. 5). G–J: Although some CCKAR-immunoreactive fibers ran completely parallel to vagal afferent fibers in the external smooth-muscle layers (arrows), individual fluorophore-containing vesicles were not colocalized, as indicated by the absence of yellow. K: In this cross section of duodenal myenteric plexus there was no colocalization of CCKAR-immunoreactive and vagal afferent fibers. L: Vagal afferent fibers in the pyloric sphincter muscle, although occasionally in close proximity, were not in a particular anatomical relationship with CCKAR-immunoreactive interstitial cells of Cajal (red). M: CCKAR-immunoreactivity in cell membrane of nodose ganglion neurons (red) retrogradely labeled (green cytoplasm) by injecting Fluorogold into the gastric wall. For abbreviations, see legend of Figure 3. Scale bar in M: (A) 75 μm; (B–D and K) 30 μm; (E) 60 μm; (F, H, and J) 50 μm; (G) 100 μm; (L and M) 40 μm.

Download figure to PowerPoint

thumbnail image

Figure 5. CCKA-receptor immunoreactivity in myenteric plexus neurons of (A–F) rat stomach and (G) duodenum obtained with antibody (A) Ab-2 or (B–G) Ab-1. A: Primarily cytoplasmic labeling with Ab-2. B: Low-magnification image showing scattered CCKAR-immunoreactive enteric neurons (AB-1). C–G: Labeling of neuronal membrane, lamellar dendrites, and proximal axon obtained with antibody Ab-1. Note cluster of three positive neurons in D. Scale bar in G: (A, C, D, F, and G) 40 μm; (B) 80 μm; (E) 60 μm.

Download figure to PowerPoint

In the external muscle layers, both CCKAR-immunoreactive nerve fibers and vagal afferent fibers were present (Fig. 4G–J). Some of the CCKAR-immunoreactive fibers looked exactly like vagal afferent intramuscular arrays (IMAs). Although some of these fibers were in intimate anatomical contact with HRP-labeled vagal afferents (Fig. 4G–J), confocal microscopic analysis showed that individual varicosities containing CCKAR-immunoreactivity (red) were not identical with varicosities containing the vagal afferent marker (green); in other words, there were no yellow varicosities (Fig. 4G–J). It is thus possible that some vagal afferent fibers in the external muscle layers did contain CCKAR-immunoreactivity, but that it was localized in different varicosities (vesicles) than the HRP-label.

Strong CCKAR-immunoreactivity was also found in subserosal ganglia that are typically associated with larger branches of the vagus nerve as they penetrate the external muscle layer of the stomach.

The relative abundance of CCKAR-immunoreactivity on vagal and nonvagal neuronal elements for the various tissues and compartments is provided in Table 1.

Table 1. Abundance of CCKAR-immunoreactivity in vagal and non-vagal fibers, terminals and neuronal cell bodies in different compartments of the gastrointestinal tract
 Ab-1(C-terminal)Ab-2(N-terminal)
  • a

    CCKAR-IR and vagal afferent marker not in same varicosities. A minimum of 10 sections from at least 3 different rats, and as many as 100 sections from 10 different rats were inspected for each site.

  • +++, Very abundant; ++, abundant; +, consistently present but few; +/–, few and inconsistent;–, not found; n.d., not determined.

Duodenum
 Mucosa
  Non-vagal fibers
  Vagal fibers
 Myenteric plexus
  Fibers++++
  Cell bodies++
  Vagal IGLEs
 External muscle
  Non-vagal++
  Vagal
Stomach
 Mucosa (corpus, antrum)
  Non-vagal fibers++
  Vagal fibers
  Gastric glands++
 Myenteric plexus (fundus, corpus, antrum)
  Fibers++++
  Cell bodies++
  Vagal IGLEs
 Circular & longit. muscle
  Non-vagal fibers++
  Vagal IMAs+/–a
 Subserosal ganglia
  Unidentified fibers++n.d.
Pylorus
 Sphincter muscle
  ICC+++
Pancreas
 Acinar cells+++
 Islet cells++
 Interlobular ganglia
Nodose ganglion
  Cell bodies++
  Fibers+

DISCUSSION

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. CONCLUSIONS
  7. Acknowledgements
  8. REFERENCES

CCKAR-Immunoreactivity in the Duodenum

Earlier studies strongly suggested the presence of CCKA-receptors, and we expected to find CCKAR-immunoreactivity, on vagal afferent nerve terminals in the duodenal mucosa. However, neither antibody used was able to detect CCKAR-immunoreactivity in this compartment, no matter what antibody concentration and amplification protocol was used, and whether we processed only for CCKAR or in combination with vagal afferent labeling. Using Ab-2, Sternini et al. (1999) also found no immunoreactive nerve fibers in the duodenal mucosa. Because we (and Sternini et al. (1999)) found specific CCKAR-immunoreactivity in the myenteric plexus of the same sections, the failure to find mucosal staining cannot be due to problems with the protocols used. Furthermore, because there was no staining in tissue processed for CCKAR only, the failure cannot be due to masking or quenching of the fluorescent secondary antibody by the HRP-detection protocol.

There is no direct evidence for CCKA-receptor localization on vagal or other nerve fibers in the small intestinal mucosa. The strongest indirect evidence comes from the generally accepted view that CCK's effect on vagal afferents is paracrine rather than endocrine. This view is mainly based on the observations that 1) vagal mucosal afferents show an extraordinary sensitivity to CCK that is not mediated by smooth-muscle motor effects (Richards et al., 1996); 2) food intake suppression induced by intraduodenal infusion of various nutrients is not correlated with the obtained plasma CCK increases (Brenner et al., 1993); and 3) vagal afferent terminals are in close proximity to the site of CCK release from enteroendocrine cells (Berthoud and Patterson, 1996). This paracrine hypothesis assumes that CCKA receptors are located on mucosal terminals of primary vagal afferents. CCKA receptor mRNA (Moriarty et al., 1997) and protein (present results) expression in cell bodies of the nodose ganglia support this assumption.

Why then were we unable to identify CCKAR-immunoreactivity in the duodenal mucosa? The simplest explanation is low receptor concentration, too low for immunohistochemical detection. Although this is a reasonable conclusion, it is intriguing that the receptor concentration was apparently high enough in neuronal processes in the myenteric plexus and external muscle layers to be detected. An alternative explanation could be the presence of a different receptor subtype or affinity-state, not recognized by either antibody (Wank et al., 1994). Recording CCK-induced spike activity of gastric vagal afferents in Otsuka-Long-Evans Tokushima Fatty (OLETF) rats that was not affected by either CCKA or CCKB receptor antagonists, Kurosawa et al. (1999) suggested the involvement of a novel (non-A, non-B) CCK receptor.

There is evidence for a truncated receptor form lacking its N-terminal extracellular domain in pancreatic acinar tissue (Poirot et al., 1994), and this may be why Ab-2, which is directed against the extracellular domain, does not stain acinar cell membranes. There is also evidence for low- and high-affinity states of the CCKA-receptor. It was suggested that low-affinity CCKA-receptors mediate vagal satiety signals and high affinity receptors mediate pancreatic protein secretion through a vagal afferent mechanism (Li et al., 1997). It is thus possible that vagal afferents in the duodenal mucosa picking up satiety signals carry only low-affinity CCKA-receptors, and that other vagal afferent fibers in the myenteric plexus carrying high-affinity receptors pick up signals relevant for pancreatic secretion. The paucity of immunoreactive fibers in the duodenal mucosa could thus be explained if neither one of the antibodies recognizes the low-affinity receptor.

In our previous studies using the carbocyanine dye DiI as an anterograde tracer, we found many more labeled fibers in the small intestinal mucosa (Berthoud et al., 1995; Williams et al., 1997) than in the present study using HRP. Because the labeling intensity in the myenteric plexus (IGLEs) and external muscle layers (IMAs) is quite similar for the two tracers, we conclude that HRP is more rapidly degraded in the mucosa than in the other compartments. However, tracing with DiI was not indicated in the present study because the use of detergents such as Triton X-100, which guarantee optimal visualization of CCKA receptors, abolishes the DiI label.

In the myenteric plexus of the duodenum, CCKAR-immunoreactivity was present in many varicose fibers surrounding enteric neurons, and in a few cell bodies, but was not colocalized in identified vagal fibers and terminals. The most abundant and widely distributed vagal afferent terminal structures are the IGLEs. We found no identified IGLEs containing CCKAR-immunoreactivity. Again, it is unlikely that this failure is due to masking of the CCKAR signal by HRP processing, because in the absence of the latter, no IGLE-like structures were ever delineated by CCKAR-immunoreactivity. Because the IGLE structure is so idiosyncratic, it is difficult to miss it or confuse it with other structures in the myenteric plexus. Until recently, it was not known what type of sensory information the IGLEs, which are situated around myenteric plexus ganglia, might be picking up, although Neuhuber (1987) speculated that they serve both a mechanosensory and a local effector function. The recent study by Zagorodnyuk and Brookes (2000) clearly identified IGLEs as mechanosensors, although an additional chemosensory function cannot be excluded.

Recording studies in rats from teased vagal afferent fibers previously showed that fibers sensitive to duodenal (Schwartz et al., 1995) or gastric loads (Schwartz et al., 1994; Schwartz and Moran, 1994) were also sensitive to CCK, and classified them as polymodal afferents. However, it was disputed whether the effects of CCK on distension-sensitive vagal afferents were direct, or mediated via smooth-muscle motor responses (Blackshaw and Grundy, 1990; Schwartz et al., 1994; Schwartz et al., 1997). Persistence of the response to CCK after atropine was taken as evidence for a direct action (Schwartz et al., 1994), but noncholinergic motor effects cannot be excluded. Based on the recent observations by Zagorodnyuk and Brookes (2000), and the relative paucity of IMAs in the duodenum, it is very likely that IGLEs represent the duodenal load-sensitive vagal units recorded by Schwartz et al. (1995).

Why then did we not find CCKA-receptors on IGLEs in the duodenum? As for the mucosa, it could be that CCKA-receptors on IGLEs are of the low affinity-state and are not recognized by either one of the antibodies. It is interesting that the low-affinity CCKA receptor antagonist CCK-JMC-180 blocked the response to CCK in gastric load-sensitive vagal afferents (Schwartz et al., 1994). Alternatively, the electrophysiological results do not prove that the receptor is actually located on vagal afferents. CCKAR-bearing enteric neurons could stimulate vagal afferents by acting on IGLEs.

CCKA Receptors in the Stomach

Much that has been discussed above could also be said for the stomach. However, there are at least two fundamental differences between the small intestine and stomach with respect to CCK and its potential receptor distribution. First, there is no CCK production in the gastric mucosa—all CCK enteroendocrine cells are located in the intestinal mucosa (Berthoud and Patterson, 1996). Therefore, CCK cannot act in a paracrine fashion in the stomach, except for CCK released from a few enteric neurons. Second, there are many IMAs in the gastric external muscle layers, but very few in the intestines (Wang and Powley, 2000).

The strongest evidence for CCK-sensitivity exists for receptors located in the gastric mucosa, as they also respond to light mucosal stroking, acidity, and hypertonicity, and the response to CCK was not decreased after cholinergic blockade (as shown in the ferret (Blackshaw and Grundy, 1990)). Furthermore, the number of mucosal fibers exhibiting CCKAR-immunoreactivity with Ab-2 was reduced by 50% in vagotomized rats (Sternini et al., 1999). However, the paracrine argument for CCK's satiety effects cannot be made for gastric mucosal sensors, as the ligand would have to come through the general circulation. In the present study, few vagal afferent fibers could be identified with the HRP-tracing method, and in these few fibers we could not demonstrate colocalization of (Ab-2) CCKAR-immunoreactivity. With the Ab-1 antibody we were unable to label many gastric mucosal fibers, while cell membranes of basal gland cells stained strongly. Again, one explanation for the apparent discrepancy is the possibility that gastric mucosal CCKA-receptors are of the low-affinity type and may thus not be recognized by Ab-1. The fact that Ab-2 recognizes them would suggest still another subtype of CCKA-receptor in this compartment.

As discussed above, IGLEs are the likely transducer sites for vagal units responding to gastric loads and distension. If the assumption is correct—that these mechanosensitive units are also directly excited by CCK—then one would have expected to find CCKA-receptors on gastric IGLEs. However, as for the duodenum, the antibodies we used in the present study did not result in CCKAR-immunoreactivity in IGLEs.

The only indication of perhaps some degree of double labeling of the same neuronal processes was found in the external muscle layers of the stomach. There, some CCKAR-immunoreactive varicose fibers were in register with vagal afferent fibers (IMAs), although the two fluorescent markers were not contained in the same varicosities. Furthermore, the CCKAR-immunoreactive structures very much resembled vagal afferent IMAs. We can thus not completely exclude the presence of CCKAR-immunoreactivity on some vagal IMAs, but for conclusive proof, ultrastructural analysis will be necessary.

We have recently found strong CCKAR-immunoreactivity in interstitial cells of Cajal (ICC) located within the pyloric sphincter muscle and proximal duodenal circular muscle (Patterson et al., 2001). Although there were some close anatomical appositions between CCKAR-bearing ICC and the numerous vagal IMAs in the sphincter muscle, none of the many IMAs were double-labeled.

Before the discovery of the mechanosensitivity of IGLEs (Zagorodnyuk and Brookes, 2000, Zagorodnyuk et al., 2001), the so-called vagal intramuscular arrays (Berthoud and Powley, 1992; Wang and Powley, 2000) were considered the ideal structures to detect gastric distension. Although located in parallel to the respective smooth-muscle bundles, they were thought to act as in-series tension receptors, activated by both distension and muscular contraction (Grundy, 1988; Blackshaw and Grundy, 1990). Since in the Zagorodnyuk et al. (2001) study, no activity could be elicited by mechanically probing near IMAs, it is no longer clear whether IMAs really are tension receptors, or alternatively may serve local effector functions. If some of these IMAs indeed possess CCKA-receptors, CCK that potentially activates them can only come through the blood stream.

CONCLUSIONS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. CONCLUSIONS
  7. Acknowledgements
  8. REFERENCES

Using immunohistochemistry, we were unable to demonstrate CCKA-receptors on identified vagal afferent fibers and terminals in the rat duodenum and stomach. Considering its presence in perikarya of vagal afferent neurons in the nodose ganglia, as well as in a population of enteric neurons and their axonal and dendritic processes, we conclude that either the receptor concentration is below the immunohistochemical detection threshold, or that the receptor located on peripheral vagal terminals has a different conformation or affinity state, making it inaccessible for the two antibodies tested.

Acknowledgements

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. CONCLUSIONS
  7. Acknowledgements
  8. REFERENCES

We thank Helen Wong and John Walsh for supplying us with the N-terminal CCKA-receptor antibody. 94159 provided by CURE/Gastroenteric Biology Center, Antibody/RIA Core, NIH grant #DK41301.

REFERENCES

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. CONCLUSIONS
  7. Acknowledgements
  8. REFERENCES