- Email: [email protected]
Tissue distribution and innervation pattern of peptide immunoreactivities in the rat pancreas
Peptides,Vol. 13, pp. 91-98, 1992
0196-9781/92 $5.00 + .00 Copyright © 1992 Pergamon Press Ltd.
Printed in the USA.
Tissue Distribution and Innervation Pattern of Peptide Immunoreactivities in the Rat Pancreas ROBERTO
DE GIORGIO,*
CATIA
STERNINI,*~ ~ KARL ANDERSON,* AND VAY LIANG W. GO*~
NICHOLAS
C. B R E C H A * t ~
Center for Ulcer Research and Education/Digestive Disease Center, Departments of*Medicine and i'Anatomy & Cell Biology, ~Brain Research Institute, UCLA School of Medicine, and VAMC at Wadsworth, Los Angeles, CA 90073 R e c e i v e d 11 J u l y 1991 DE GIORGIO, R., C. STERNINI, K. ANDERSON, N. C. BRECHA AND V. L. W. GO. Tissuedistribution and innervation pattern ofpeptide immunoreactivities in the rat pancreas. PEPTIDES 13(1) 91-98, 1992.--The distribution of calcitonin generelated peptide (CGRP), substance P/tachykinin (SP/TK), vasoactive intestinal polypeptide (VIP), neuropeptide Y (NPY) and gastrin-releasing peptide (GRP) immunoreactivities (IR) in the rat pancreas was investigated using radioimmunoassay and immunohistochemistry. CGRP, NPY and VIP tissue contents are much higher than GRP and SP/TK concentrations. Peptidecontaining nerves are distributed to both the exocrine and endocrine pancreas. However, differences exist in terms of density and targets of innervation for each peptidergic system. In the acini and through the stroma, fibers IR for CGRP, NPY and VIP are > GRP- and SP/TK-containing processes. The vasculature is supplied by a prominent NPY, CGRP and, to a lesser extent, SP/ TK innervation. VIP-IR is found occasionally, and GRP-IR is never detected, in fibers associated with blood vessels. Around ducts, CGRP- and NPY-positive neurites are > SP/TK- >/VIP-1R fibers, whereas GRP-containing nerves are not visualized. In the islets, the density of peptidergic nerves is: VIP-, GRP- >/CGRP-IR > NPY or SP/TK. In intrapancreatic ganglia, VIP- and, to a lesser extent, NPY-IRs are found in numerous neuronal cell bodies and in nerve fibers; GRP-IR is present in numerous nerve processes and in few cell bodies; CGRP- and SP/TK-1Rs are detected only in fibers wrapping around unlabeled ganglion cells. The majority of CGRP-IR fibers contain SP/TK-IR. The existence of differential patterns of peptidergic nerves suggests that peptides exert their effects on pancreatic functions via different pathways. Pancreatic innervation Peptidergic nerves Calcitonin gene-related peptide Vasoactive intestinal polypeptide Neuropeptide Y Gastrin-releasing peptide
Substance P Tachykinins Intrapancreatic ganglia Rat
(duodenal versus splenic portion) of the rat pancreas, and we compared the patterns of distribution of nerve fibers containing these peptide IRs. In addition, since C G R P - I R s colocalize with SP/TK-IRs in many structures, including sensory ganglia, cutaneous, visceral and vascular sensory nerves (13,14,26), we evaluated to what extent C G R P - and SP/TK-IRs coexist in neuronal processes of the rat pancreas.
C O N S I D E R A B L E evidence indicates that a variety of biologically active neuropeptides contribute to the control of digestive functions, i.e., motor, absorptive and secretory activities (6). Peptide-containing nerves represent a prominent c o m p o n e n t of the innervation of the digestive system, including the pancreas. In fact, pancreatic nerves immunoreactive (IR) for vasoactive intestinal polypeptide (VIP) (3,5,22-25), substance P (SP), enkephalin, cholecystokinin (22,25,34), bombesin and its related peptide gastrin-releasing peptide (GRP) (5,12,29,42), neuropeptide Y (NPY) (4,5,7,37), neurotensin (31 ), calcitonin-gene related peptide (CGRP) (35,37) and galanin (8) have been reported by us and other investigators. The tissue distribution and the identification of specific structures (e.g., islets, acinar cells, ducts, blood vessels as well as intrapancreatic ganglia) innervated by peptide-containing nerves are the base for an analysis and an understanding of sites of action of different neuropeptides. In the present investigation, we quantified the tissue contents of CGRP-, SP (or related tachykinins, TKs)-, VIP-, NPY- and GRP-IRs in different regions
METHOD
Animals and Tissue Preparations Experiments were performed on 12 male adult young Sprague-Dawley rats, weighing 250-400 g. Each animal was housed individually and had free access to water and food at all times. All experimental procedures were consistent with those stipulated in the N I H Guide for the Use of Experimental Animals and approved by the Veterans Administration Animal Research Committee.
Requests for reprints should be addressed to Catia Sternini, M.D., CURE/UCLA Digestive Disease Center, Building 115, Room 203, Wadsworth VAMC, Los Angeles, CA 90073. 91
92
DE GIORGIO ET AL.
For the radioimmunoassay study, pancreatic specimens from 7 animals, sacrificed by decapitation following halothane (Halocarbon Laboratories, Augusta, SC) anesthesia, were immediately dissected, divided into duodenal and splenic regions, frozen on dry ice, weighed and stored at - 7 0 °C until extraction. For immunohistochemistry, 5 rats were deeply anesthetized with 30% chloral hydrate and then perfused through the aorta with 100 ml of 0.9% NaC1 in 0.1 M phosphate buffer (PB), pH 7.4, followed by 600 ml of 2% paraformaldehyde and 15% saturated picric acid solution in 0.1 M PB (Zamboni's fixative). Pancreata were removed, postfixed in the same fixative for 2 hours at room temperature and subsequently placed in PB containing 25% sucrose overnight at 4°C. Serial sections (10-12 ~tm) of both duodenal and splenic regions of the pancreas were cut with a cryostat (Bright Instruments), mounted onto chromalum gelatin-coated slides and stored at - 2 0 ° C until processing for immunohistochemistry.
100
80
60
40
20
0
CGRP
NPY
VIP
GRP
SP
FIG. 1. Neuropeptide tissue levels (ng/g of tissue) in the duodenal (filled bar) and splenic (empty bar) portion of the rat pancreas. Data are expressed as mean + SD.
Tissue Extraction and Radioimmunoassay Frozen specimens of the duodenal and splenic regions of rat pancreata were placed in 0.1 N HCI, boiled and homogenized in a Polytron tissue grinder. Subsequently, extracted supernatants were used for radioimmunoassay. Diluted aliquots were assayed in duplicate. CGRP-, SP-, VIP-, NPY- and GRP-IRs were measured employing previously described radioimmunoassays which have been validated in mammalian tissues including rat (2,21,30,32,41). For all peptides measured, the intraassay variation ranged from 4% to 7%, while the interassay variation ranged from 10% to 13%. Detection limits for these assays were: 39, 10, 20, 39, and 16 pg/ml, respectively. Experiments aimed to assess the specificity of these antibodies showed that: 1) CGRP radioimmunoassay cross-reacts 100% with human or rat fl-CGRP and 50% with human or rat a-CGRP, but it does not cross-react with calcitonin; 2) SP antibody cross-reacts 100% with related TKs (neurokinin A and neurokinin B); 3) VIP radioimmunoassay does not cross-react with structurally related peptides, i.e., peptide histidine isoleucine, peptide histidine methionine, glucagon and secretin; 4) NPY antiserum displays 2% and less than 1% cross-reactivity with porcine peptide YY and human pancreatic polypeptide (PP), respectively; 5) GRP/bombesin antiserum cross-reacts 100% with G R P 14-27.
Irnmunohistochemistry The antisera used for the immunohistochemical studies are listed in Table 1. Tissue sections were processed with either the indirect immunofluorescence or the avidin-biotin-peroxidase complex (ABC) method (35,36). Briefy, tissue sections were washed in 0.1 M PB, pretreated for 30 minutes at room temperature with 5% swine or goat normal sera, and incubated in primary antibody overnight at 4°C in a humid chamber. There-
after, sections were washed in PB, incubated for two hours at room temperature in swine or goat anti-rabbit or anti-guinea pig IgG (diluted 1:50) conjugated to either fluorescein isothiocyanate (FITC) or tetramethyl rhodamine isothiocyanate (TRITC) (purchased from American Qualex International, La Mirada, CA or DAKO, DAKOPATTS, Santa Barbara, CA) for immunofuorescence. For the ABC method, sections were incubated for two hours in goat anti-rabbit biotinylated IgG (diluted 1:100) (Vector Laboratories, Burlingame, CA), followed by ABC solution and then exposed to 3,3'-diaminobenzidine with 0.01% H202 (36). To reduce nonspecific staining due to endogenous peroxidase, tissue processed with the ABC method was first dehydrated, placed in 100% methanol and then in a solution composed of 98% methanol, 1% acetic acid and 1% sodium nitroferricyanide (for 15 min) followed by 100% methanol, rehydrated and then incubated in primary antibody. Primary and secondary antibodies were diluted in 0.5% Triton Xt00 in 0.1 M PB. For double-label studies, tissue was incubated in a mixture containing guinea pig CGRP and rabbit SP antisera. The antigenantibody complexes were then visualized with a mixture of goat anti-guinea pig IgG TRITC and goat anti-rabbit IgG FITC, both used at 1:50 dilution. Sections were examined with a Leitz Orthoplan microscope (E. Leitz, Inc., Rockleigh, N J) using either bright field optics for ABC processed material, or a Ploem epi-illumination system with "I2" or "Ld' and "N2" filter cubes to visualize FITC and TRITC fluorescence, respectively.
Specificity Testsfor Immunohistochemical Techniques Specificity of the staining was assessed as follows: 1) omission of the primary antibody, 2) substitution of the primary antibody
TABLE 1 C H A R A C T E R I S T I C S O F ANTISERA U S E D F O R I M M U N O H I S T O C H E M I S T R Y
Antigen
AntibodyCode
CGRP CGRP SP VIP NPY NPY Bombesin-9
CGRP2p CGRP2f SP 7913 NPY-3 NPY-Y 8543
Host Species
Dilution
Reference
Rabbit Guinea pig Rabbit Rabbit Rabbit Rabbit Rabbit
1:5000 1:200 1:250 1:250 1:250 1:250 1:250
35 Sternini, unpublished Incstar (Minnesota) 11 33 9 40
PEPTIDERGIC NERVES IN THE PANCREAS
93
with rabbit preimmune serum diluted 1:50, and 3) incubation with the primary antibodies preabsorbed, for 12-16 hours at 4°C, with synthetic homologous or heterologous peptides (purchased from Bachem, Torrance, CA or Peninsula Laboratories, Belmont CA), i.e.: a[Tyr]rat CGRP 23-37, a-rat CGRP 1-37, somatostatin, VIP, GRP 1-27, NPY, SP, neurokinin A, CGRP 23-27, a-rat CGRP 1-37, somatostatin, VIP, GRP 1-27, NPY, SP, neurokinin A, neurokinin B, carboxyl-terminal hexapeptide of PP (PP-6) and peptide YY at concentrations of l0/~M. In addition, we performed further control experiments for double-labeling technique to verify that each of the primary antibodies did not bind to the other when mixed together and that each of the secondary antisera reacted with the appropriate antigen-antibody complex (14). These controls consisted of: l) omission of one of the primary antibodies from the first incubation, 2) omission of one of the secondary antibodies from the second incubation, and 3) preabsorption of CGRP or SP with the appropriate peptides. Since the possibility that the immunoreactive materials are structurally related to, but not identical with, the peptides under investigation, the terms CGRP, SP/TK, VIP, NPY, and GRP immunoreactive, containing or positive will be used to refer to the immunostaining observed.
Statistical Analysis Radioimmunoassay data were expressed as mean _+standard deviation (SD) of the mean. Mean values were compared by repeated measures analysis of variance (ANOVA). Computations were made on logarithmic transformation of the data. Mean comparisons were assessed for statistical significance according to Tukey's Studentized Range criteria. The level of statistical significance was set at p < 0.05.
FIG. 2. Photomicrographs of the exocrine pancreas displaying VIP- and CGRP-IR nerve fibers innervating the acini. Calibration bar" 25 ~tm. of the duodenal and splenic portion of the rat pancreas. The relative density of peptide-containing nerves in relation to each pancreatic compartment is summarized in Table 2. There are differences in the densities and patterns ofinnervation in relation to specific intrapancreatic tissue targets. Thus CGRP-, NPYand, to a lesser extent, SP/TK-containingnerves are particularly prominent in association with the vasculature and, within the exocrine pancreas, with the duct system, while VIP- and GRPIR fibers are particularly evident in intrapancreatic ganglia and endocrine pancreas. For all peptides studied, however, there are no evident differences in term of innervation patterns between the duodenal and splenic portion. Specificity for CGRP, SP/TK, VIP, NPY and GRP immunostainings was demonstrated by the lack of immunoreaction in sections in which the primary antiserum was omitted, or in
RESULTS
Radioimmunoassa£ Data Figure 1 shows the neuropeptide tissue contents in the duodenal and splenic portions of the rat pancreas. CGRP, NPY and VIP tissue levels are much higher than GRP and SP concentrations. The differences between the duodenal and splenic lobes of the rat pancreas are not significant for any of the five investigated peptides (p = 0.15).
Immunohistochemistry CGRP-, SP/TK-, VIP-, NPY- and GRP-IR nerve fibers are distributed throughout the exocrine and endocrine compartment
TABLE 2 RELATIVEDENSITYOF PEPTIDE-CONTAININGNERVESIN RELATION TO DIFFERENTPANCREATICSTRUCTURES
Exocrine Pancreas Acini Septa* Ducts Vasculature Endocrine Pancreas Intrapancreatic Ganglia Fibers Cell bodies
CGRP
SP/TK
GRP
VIP
NPY
++ ++ ++/+++ +++ +/++
+ + +/++ ++ +
+ + ++
++ +/+ + + -/+ ++
++ +/+ + ++/+++ +++ +
++ -
++ -
+++ +
++ +++
++ ++/+++
Key: - = absent; - / + = occasional presence; +, ++, +++ = relative density. * Intralobular and interlobular connective tissue.
94
DE GIORGIO ET AL. other hand, absorptions of CGRP, SP, VIP, NPY or GRP antisera with structurally unrelated polypeptides did not modify the immunolabeling. Controls for double-staining experiments demonstrated that CGRP or SP antibodies did not bind to each other and also that each of the secondary antibodies bound only to the appropriate antigen-antibody complex. Exocrine Pancreas CGRP-, SP/TK-, VIP-, NPY- and GRP-IR processes, running either singly or in fascicles, are located among the acini (Figs. 2 and 3) as well as in the septa (i.e., the interlobular and intralobular connective tissue) of the duodenal and splenic portions of the rat pancreas. CGRP-, VIP- and NPY-containing processes are in general more abundant than those immunoreactive for SP/ TK or GRP. In the ductal system, CGRP-, NPY- and, to a lesser extent, SP/TK-IR fibers, running either singly or in small bundles, are particularly abundant in the connective tissue which supports the ducts (Figs. 3 and 7). Sometimes, single varicose CGRP-, NPY- and SP/TK-IR processes also can be identified underneath the surface epithelium of the ducts. VIPergic nerves are the least numerous around ducts, although constantly detectable. GRPcontaining fibers are not detected in association with the ductal system. Vasculature The vasculature receives a particularly rich NPY and CGRP innervation, especially blood vessels with the appearance of small
FIG. 3. Photomicrographs of the exocrine pancreas. The upper panel illustratesan abundant CGRP innervation of a small duct (d); the middle panel shows a dense network of NPY-IR nerves around a small artery (a), along with varicose NPY-containing fibers running among acini. In the lower panel, thin beaded VIP-IR fibers (arrows) are evident in the connective tissue supporting a large-size duct. Calibration bar: 25 urn. sections incubated with normal rabbit serum or primary antibodies preabsorbed with the appropriate peptide. Of the two rabbit NPY antibodies used, NPY-Y, but not NPY-3, in addition to neuronal elements, stained cells at the peripheral margin of the islets. The incubation of NPY-Y antiserum with PP-6 abolished the cell staining but did not affect the intensity nor the density of NPY-positive neuronal structures throughout the rat pancreas. The incubation of SP antiserum with the appropriate or related peptides, such as neurokinin A and neurokinin B, abolished the immunostaining, indicating that the antibody utilized recognized different members of the TK family. For this reason we use the terms SP/TK and SP interchangeably. On the
FIG. 4. Representative photomicrographs of the endocrine pancreas demonstrating GRP (upper panel)- and VIP-containing (lower panel) varicose processes in the islets. Calibration bar: 25 um.
PEPTIDERGIC NERVES IN THE PANCREAS
95
FIG. 5. lntrapancreatic ganglia. VIP-, NPY- and GRP-1Rs are present either in neurons (arrows) or in nerve fibers, while CGRPIR is restricted to nerves surrounding unstained ganglion cells. VIP- and NPY-IRs are present in a greater number of cell bodies than GRP-IR. Calibration bar: 25 tsm.
arteries and arterioles (Figs. 3 and 7). SP/TK-immunolabeled fibers (Fig. 7) associated with the vasculature are also abundant, even though less prominent than NPY- and CGRP-IR nerves. These three peptide-IRs appear to be localized to perivascular fibers forming dense networks at the adventitial level. Conversely, VIP-containing fibers are less often visualized in association with the vasculature, and usually they are found in paravascular position, that is, in the connective tissue located around the vessels. GRP-IR nerves are never seen in association with the vasculature.
Endocrine Pancreas Thin varicose processes containing CGRP-, SP/TK-, VIP-, NPY- or GRP-IR supply the islets with the following density: VIP- and GRP-IR fibers (Fig. 4) >t CGRP-IR > SP/TK- and NPY-IR processes. CGRP-IR is localized to endocrine-like cells identified both at the peripheral margin of the islets and scattered among the acini in the exocrine tissue. The NPY-IR visualized at the periphery of the islets with NPY-Y antibody is identical to that previously described with PP antisera (38) and is eliminated by preabsorption of this antibody with PP-6, demonstrating that these NPY-immunolabeled cells are PP-containing cells. No VIP-, GRP- or SP/TK-IR endocrine]paracrine-like cells were stained in islets or among the acini.
Intrapancreatic Ganglia VIP-IRs (Figs. 5 and 6) and, to a lesser extent, NPY-IRs are found in a good proportion of ganglion cells (Fig. 5), in addition to varicose nerve fibers. Networks of GRP-IR fibers are dense
and wrap around stained and (more often) unstained perikarya (Fig. 5). GRP-containing neuronal cell bodies are much less numerous than those positive for VIP or NPY, and they are not detected in all ganglia. Conversely, CGRP- and SP/TK-IRs are restricted to nerve fibers (Figs. 5 and 6).
Colocalization Studies The majority of fibers with CGRP-IR also contain SP/TKIR (Fig. 7). In addition, other two subpopulations of fibers are visualized: a small component of SP/TK positive nerves lacking CGRP-IR and, more frequently, CGRP positive processes without SP/TK-IR. DISCUSSION
The results of the present study, where the patterns of innervation ofCGRP-, SP/TK-, VIP-, NPY- and GRP-containing nerves in the rat pancreas are compared and their tissue contents measured, confirm and extend previous findings in the mammalian pancreas. The tissue content of each peptide in the duodenal portion is similar to that in the splenic region. Similarly, the patterns of fibers containing these peptide-IRs are identical in both pancreatic regions. However, there are differences among these peptidergic systems in relation to specific pancreatic structures, as summarized in Table 2. Thus, in the acinar and stromal compartment, CGRP-, NPY- and VIP-containing processes are > SP/TK- and GRP-IR fibers. Around ducts, CGRP-, NPYand, to a lesser extent, SP/TK-IR nerves are more abundant than those IR for VIP, while GRP-immunostained fibers are not detected. The vasculature receives prominent NPY, CGRP
96
FIG. 6. Simultaneous staining for VIP (upper panel)- and CGRP-IRs (lower panel) in an intrapancreatic ganglion. Note the absence of colocalization: VIP-IR being present in neurons and fibers, while CGRP-IR is detectable only in fibersgrouped in a bundle or single. Note a CGRPIR varicosityin closevicinityto a cellbody containingVIP-IR.Calibration bar: 25 ~zm.
and SP/TK nerve supply, while VIP-IR fibers are sparse and GRP-IR fibers are not visualized. In the islets, the density of peptidergic nerves is: those IR for VIP and GRP >/CGRP- > SP/TK- and NPY-IR ones. Within the intrapancreatic ganglia, VIP-, NPY- and GRP-IRs are present in neuronal cell bodies as well as in fibers, while CGRP- and SP/TK-IRs are restricted to nerves. VIP- and NPY-IR ganglion cells are much more abundant than those visualized with GRP antiserum. Within the digestive system, CGRP- and SP/TK-IRs are localized to intrinsic neurons of the gut as well as to extrinsic, capsaicin-sensitive, primary sensory fibers (13,36). Unlike SP/ TK-IR fibers, which mainly derive from intrinsic neurons and comprise only a small portion of fibers originating from extrinsic, afferent neurons, an important component of the CGRP innervation of the alimentary tract is afferent in origin (35-37). The lack of CGRP- and SP/TK-IR intrapancreatic ganglion cells observed in the present study and the virtual elimination of CGRPand SP/TK-IR fibers within the pancreas following chemical and surgical denervation (35-37) are consistent with an extrinsic origin of CGRP and SP/TK pancreatic innervation. However, the presence of occasional ganglion cells positive for SP/TK (37), but not for CGRP (35), has been reported in the pancreas of rats treated with the neurotoxin colchicine, which is commonly used to increase peptide levels in cell bodies. Therefore, a small subpopulation of SP/TK-IR fibers (i.e., those lacking CGRPIR) could derive from intrapancreatic ganglion cells. The extensive colocalization of CGRP- and SP/TK-IRs in nerve processes of the rat pancreas extends previous observations showing the
DE GIORGIO ET AL. coexpression of these two peptides in cutaneous, visceral, and vascular sensory nerves ( 13,14,26). VIP-IR has been identified in the vast majority of neuronal cell bodies of most intrapancreatic ganglia. Therefore, our results, in accordance with previously reported observations (24,37), confirm that the VIPergic innervation of the rat pancreas, as well as that of other mammalian species (5,23), is predominantly of intrinsic origin. The identification of NPY- and GRP-IR ganglion cells, the latter being much less numerous than those containing VIP- or NPY-IR, suggests that also these two peptidergic systems are, at least in part, of intrinsic origin. Whereas it is well established that the pancreas, as the gut and liver, receives extrinsic NPY-IR fibers which are mainly distributed to the vasculature and are likely to derive from sympathetic prevertebral ganglia (4,15,37), whether the GRP innervation of the alimentary tract includes an extrinsic input remains to be elucidated. Peptidergic nerves supplying the pancreas could also be derived from enteric neurons. This is based on the recent observation that there are neurons in the myenteric plexus of the stomach and proximal duodenum that send axonal projections to the pancreas (18). However, further studies are necessary to determine whether these gastroenteropancreatic neuronal elements contain peptide-lR. Some divergencies emerge between the present findings and previous observations in the pancreas of the same or different species. For example, in other studies, GRP-IR fibers could not be visualized in the rat pancreas (29) or could be detected only in the exocrine compartment but not within islets or intrapancreatic ganglia (12). In a previous investigation on NPY pancreatic innervation, NPY-IR fibers were not described in association with the islets (4). In the guinea pig pancreas, a selective distribution of SP-IR processes (lacking CGRP-IR and perhaps originating from intrapancreatic ganglion cells) to some islets and intercalated ducts has been reported (19). This finding differs from that described in the rat pancreas [present study, and (37)]. Furthermore, unlike the rat pancreas, where ducts appear to be devoid of GRP innervation, the ductal system of feline and porcine pancreata is innervated by GRP-IR processes [(5); De Giorgio and Sternini, unpublished]. These divergent findings within the same species and among several species may be explained on the basis of different sensitivities of the antibodies used in these studies, and with the existence of intraspecies variations of peptidergic nerves supplying the mammalian pancreas, respectively. The morphological relationship of peptidergic nerves with different pancreatic compartments supports the hypothesis that different peptides have distinct sites of actions. Thus the rich association of CGRP-, SP/TK- and NPY-containing fibers with the vasculature is consistent with an involvement of these vasoactive peptides (16,27,28) in the regulation of pancreatic blood flow, probably through a direct mechanism. In addition, CGRP, SP, and NPY can influence the exocrine and endocrine pancreas, which are targets of innervation of these peptidergic systems ( 1,16,17,27,28). It also emerges that the diversity of targets, including the acinar tissue, ducts, endocrine compartment and the vasculature, supplied by CGRP- and SP/TK-IR nerves, is consistent with distinct efferent roles of these fibers in addition to an afferent role, which could be speculated in view of the fact that they originate from primary sensory neurons (34-37). The association of GRP-containing fibers with acini, but not with ducts, fits the observation that intravenous infusion of the tetradecapeptide bombesin, whose amino acid sequence has a biologically active region identical to that of GRP, evokes pancreatic secretion particularly abundant in enzymes but poor in bicarbonate (39). It is likely that GRP/bombesin effects on the exocrine pancreas are mostly related to the acini. Furthermore,
P E P T I D E R G I C N E R V E S IN T H E P A N C R E A S
97
FIG. 7. Simultaneous visualization of CGRP- and SP/TK-1Rs in a subset of nerve processes (arrows) around a blood vessel (upper panel) and a large-size duct (lower panel). Arrowheads indicate CGRP-positive fibers lacking SP/TK-IR. Calibration bar: 25 ~m.
the relative abundance of G R P - and VIP-IR nerves supplying the islets supports the hypothesis that G R P and VIP are involved in the neural control of pancreatic hormone secretions (10,20). Finally, the VIP innervation of the exocrine pancreas provides the morphological basis for the stimulatory effect of VIP on exocrine pancreatic secretions (10,17). By contrast, despite the VIP vasodilatory effect (10), only occasional VIP-IR fibers are associated with blood vessels, as previously showed in other mammalian species (3,5,22). It is possible that the vascular effect of VIP is due to a related peptide not detectable by our VIP antiserum or that the VIP effect on blood flow is mediated by other substances. In summary, we can conclude that: 1) CGRP-, SP/TK-, VIP-, NPY- and G R P - I R nerve fibers represent an important supply of the pancreatic innervation; 2) they have differential patterns and densities according to specific pancreatic targets (CGRP, S P / T K and NPY innervations are particularly prominent in as-
sociation with blood vessels, ducts and, to a lesser extent, acini, while VIP- and G R P - I R fibers are more abundant in intrapancreatic ganglia and islets); 3) C G R P - and SP/TK-IRs extensively colocalize in pancreatic nerves; and 4) among the peptides investigated only C G R P appears to be expressed in endocrine-like cells. Thus, as for the gut, peptidergic nerves supplying the pancreas include an intrinsic component of fibers, originating from intrapancreatic neurons, as well as an extrinsic input of fibers, which derive from neurons located outside the pancreas. ACKNOWLEDGEMENTS This study was supported by grant DK38752 from the National Institutes of Health (C.S.) and Veteran Administration Medical Research Funds (N.C.B.). The authors wish to thank H. Wong and Drs. J. H. Walsh, P. C. Emson and J. K. McDonald for having provided GRP, NPY and VIP antisera and the RIA laboratory at Mayo Clinic (Mayo Medical School of Medicine, Rochester, MN) for RIA determinations.
REFERENCES 1. Ahrrn, B.; Pettersson, M. Calcitonin gene-related peptide (CGRP) and amylin and the endocrine pancreas. Int. J. Pancreatol. 6:1-15; 1990. 2. Benarroch, E. E.; Schmelzer, J. D.; Ward, K. K.; Nelson, D. K.; Low, P. A. Noradrenergic and neuropeptide Y mechanisms in guanethidine-sympathectomized rats. Am. J. Physiol. 259:R371-375; 1990. 3. Bishop, A. E.; Polak, J. M.; Green, I. C.; Bryant, M. G.; Bloom, S. R. The location of VIP in the pancreas of man and rat. Diabetologia 18:73-78; 1980.
4. Carlei, F.; Allen, J. M.; Bishop, A. E.; Bloom, S. R.; Polak, J. M. Occurrence, distribution and nature of neuropeptide Y in the rat pancreas. Experientia 41 : 1554-1557; 1985. 5. De Giorgio, R.; Sternini, C.; Brecha, N. C.; Widdison, A. L.; Karanjia, N. D.; Reber, H. A.; Go, V. L. W. Patterns of innervation of vasoactive intestinal polypeptide, neuropeptide Y and gastrin-releasing peptide immunoreactive nerves in the feline pancreas. Pancreas (in press). 6. Dockray, G. J. Physiology of enteric neuropeptides. In: Johnson, L. R.; Christensen, J.; Jackson, M. J.; Jacobson, E. D.; Walsh, J. H.,
98
7. 8. 9. 10. 11. 12. 13.
14.
15.
16. 17. 18. 19. 20. 21. 22. 23. 24.
25. 26.
DE G I O R G I O ET AL. eds. Physiology of the gastrointestinal tract, vol. 1. New York: Raven Press; 1987:41-66. Dunning, B. E.; Ahrrn, B.; Brttcher, G.; Sundler, F.; Taborsky, G. J., Jr. The presence and actions of NPY in the canine endocrine pancreas. Regul. Pept. 18:253-265; 1987. Dunning, B. E.; Ahrrn, B.; Veith, R. C.; Brttcher, G.; Sundler, F.; Taborsky, G. J., Jr. Galanin: A novel pancreatic neuropeptide. Am. J. Physiol. 251:E127-E133; 1986. Emson, P. C.; Corder, R.; Lowry, P. J. Demonstration of a neuropeptide Y-like immunoreactivity in human phaeochromocytoma extracts. Regul. Pept. 8:89-94; 1984. Fahrenkrug, J. Vasoactive intestinal polypeptide: Measurements, distribution and putative neurotransmitter function. Digestion 19: 149-169; 1979; Furness, J. B.; Costa, M.; Walsh, J. H. Evidence for and significance of the projection of VIP neurons from the myenteric plexus to the tenia coli in the guinea-pig. Gastroenterology 80:1557-1561; 1981. Gathei, M. A.; George, S. K.; Major, J. H.; Carlei, F.; Polak, J. M.; Bloom, S. R. Bombesin-like immunoreactivity in the pancreas of man and other mammalian species. Experientia 40:884-886; 1984. Gibbins, L. L.; Furness, J. B.; Costa, M.; Maclntyre, I.; Hillyard, C. J.; Girgis, S. Co-localization of calcitonin gene-related peptidelike immunoreactivity with substance P in cutaneous, vascular and visceral sensory neurons of guinea pigs. Neurosci. Lett. 57:125-130; 1985. Goehler, L. E.; Sternini, C.; Brecha, N. C. Calcitonin gene-related peptide immunoreactivity in the biliary pathway and liver of the guinea-pig: Distribution and colocalization with substance P. Cell Tissue Res. 253:145-150; 1988. Goehler, L. E.; Sternini, C. Neuropeptide Y immunoreactivity in the mammalian liver: Pattern of innervation and coexistence with tyrosine hydroxylase immunoreactivity. Cell Tissue Res. 265:287295; 1991. Goodman, E. C.; Iversen, L. L. Calcitonin gene-related peptide: A novel neuropeptide. Life Sci. 38:2169-2178; 1986. Hoist, J. J. Peptidergic mechanisms in the pancreas. Arch. Int. Pharmacodyn. 303:252-269; 1990. Kirchgessner, A. L.; Gershon, M. D. Innervation of the pancreas by neurons of the gut. J. Neurosci. 10:1626-1642; 1990. Kirchgessner, A. L.; Pintar, J. E. Guinea pig pancreatic ganglia: Projections, transmitter content, and the type-specific localization of monoamine oxidase. J. Comp. Neurol. 305:613-631 ; 1991. Knuhtsen, S. V.; Hoist, J. J.; Schwartz, T. W.; Jensen S. L.; Nielsen, O. V. The effect of gastrin-releasing peptide on the endocrine pancreas. Regul. Pept. 17:269-276; 1987. Koch, T. R.; Roddy, D. R.; Carney, J. A.; Telander, R. D.; Go, V. L. W. Quantitation, and origin of immunoreactive neuropeptide Y in the human gastrointestinal tract. Regul. Pept. 21:309-319; 1988. Larsson, L. I. lnnervation of the pancreas by substance P, enkephalin, vasoactive intestinal polypeptide and gastrin/CCK immunoreactive nerves. J. Histochem. Cytochem. 27:1283-1284; 1979. Larsson, L.I.; Fahrenkrug, J.; Hoist, J. J.; Schaffalitzky de Muckadell, O. B. Innervation of the pancreas by vasoactive intestinal polypeptide (VIP) immunoreactive nerves. Life Sci. 22:773-780; 1978. Larsson, L. I.; Fahrenkrug, J.; Schaffalitzky de Muckadell, O.; Sundler, F.; H~tkanson, R.; Rehfeld, J. F. Localization of vasoactive intestinal polypeptide (VIP) to central and peripheral neurons. Proc. Natl. Acad. Sci. USA 73:3197-3200; 1976. Larsson, L. I.; Rehfeld, J. F. Peptidergic and adrenergic innervation of pancreatic ganglia. Scand. J. Gastroenterol. 14:433-437; 1979. Lundberg, J. M.; Franco-Cereceda, A.; Hua, X.; Hrkfelt, T.; Fischer, J. A. Coexistence of substance P and calcitonin gene-related peptide-
27. 28.
29. 30.
31.
32. 33.
34. 35.
36.
37.
38. 39.
40. 41.
42.
like immunoreactivity in sensory nerves in relation to cardiovascular and bronchoconstrictor effects of capsaicin. Eur. J. Pharmacol. 108: 315-319; 1985. McDonald, J. K. Neuropeptide Y and related substances. Crit. Rev. Neurobiol. 4:97-135; 1988. Maggio, J. E.; Mantyh, P. W. Gut tachykinins. In: Schultz, S. G.: Makhlouf, G. M.; Rauner, B. B., eds. Handbook of physiology. A critical, comprehensive presentation of a physiological knowledge and concepts. Section 6: The gastrointestinal system. Volume II: Neural and endocrine biology. Bethesda: American Physiological Association; 1989:661-690. Moghimzadeh, E.; Ekman, R.; Hhkanson, R.; Yanaihara, N.; Sundler, F. Neuronal gastrin-releasing peptide in the mammalian gut and pancreas. Neuroscience 10:553-563; 1983. Nelson, D. K.; Service, J. E.; Studelska, D. R.; Brimijoin, S.; Go, V. L. W. Gastrointestinal neuropeptide concentrations following guanethidine sympathectomy. J. Auton. Nerv. Syst. 22:203-210; 1988. Reinecke, M. Neurotensin: Immunohistochemical localization in the central and peripheral nervous system and in endocrine cells and its functional role as neurotransmitter and endocrine hormone. Prog. Histochem. Cytochem. 16:12-172; 1985. Roddy, D. R.; Aimone, L. D.; Yaksh, T. L.; Go, V. L. W. Regional distribution and characterization of the brain-gut peptide: Calcitonin gene-related peptide in pig. Soc. Neurosci. Abstr. 15:740; 1989. Sabatino, F. D.; Murnane, J. M.; Hoffman, R. A.; McDonald, J. K. Distribution of neuropeptide Y-like immunoreactivity in the hypothalamus of the adult golden hamster. J. Comp. Neurol. 257:93104; 1987. Sharkey, K. A.; Williams, R. G.; Dockray G. J. Sensory substance P innervation of the stomach and pancreas. Gastroenterology 87: 914-921; 1984. Sternini, C.; Brecha, N. lmmunocytochemical identification of islet cells and nerve fibers containing calcitonin gene-related peptide-like immunoreactivity in the rat pancreas. Gastroenterology 90:11551163; 1986. Sternini, C.; Reeve, J. R., Jr.; Brecha, N. Distribution and characterization of calcitonin gene-related peptide immunoreactivity in the digestive system of normal and capsaicin-treated rats. Gastroenterology 93:852-862; 1987. Su, H. C.; Bishop, A. E.; Power, R. F.; Hamada, Y.; Polak, J. M. Dual intrinsic and extrinsic origins of CGRP- and NPY-immunoreactive nerves of rat gut and pancreas. J. Neurosci. 7:2674-2687; 1987. Vaillant, C.; Taylor, I. L. Demonstration of carboxyl-terminal PPlike peptides in endocrine cells and nerves. Peptides 2(Suppl. 2):3135; 1981. Walsh, J. H. Bombesin-like peptides. In: Schultz, S. G.; Makhlouf, G. M.; Rauner, B. B., eds. Handbook of physiology. A critical, comprehensive presentation of a physiological knowledge and concepts. Section 6: The gastrointestinal system. Volume II: Neural and endocrine biology. Bethesda: American Physiological Association; 1989: 587-609. Walsh, J. H.; Wong, H. C.; Dockray, G. J. Bombesin-like peptides in mammals. Fed. Proc. 38:2315-2319; 1979. Yaksh, T. L.; Michener, S. R.; Bailey, J. E.; Harty, G. J.; Lucas D. L.; Nelson D. K.; Roddy, D. R.; Go, V. L. W. Survey of distribution of substance P, vasoactive intestinal polypeptide, cholecystokinin, neurotensin, Met-enkephalin, bombesin and PHI in the spinal cord of cat, dog, sloth and monkey. Peptides 9:357-372; 1988. Yanaihara, N.; Yanaihara, C.; Mochizuki, T.; lwahara, K.; Fujita, T.; Iwanaga, T. Immunoreactive GRP. Peptides 2(Suppl. 2):185191; 1981.
0196-9781/92 $5.00 + .00 Copyright © 1992 Pergamon Press Ltd.
Printed in the USA.
Tissue Distribution and Innervation Pattern of Peptide Immunoreactivities in the Rat Pancreas ROBERTO
DE GIORGIO,*
CATIA
STERNINI,*~ ~ KARL ANDERSON,* AND VAY LIANG W. GO*~
NICHOLAS
C. B R E C H A * t ~
Center for Ulcer Research and Education/Digestive Disease Center, Departments of*Medicine and i'Anatomy & Cell Biology, ~Brain Research Institute, UCLA School of Medicine, and VAMC at Wadsworth, Los Angeles, CA 90073 R e c e i v e d 11 J u l y 1991 DE GIORGIO, R., C. STERNINI, K. ANDERSON, N. C. BRECHA AND V. L. W. GO. Tissuedistribution and innervation pattern ofpeptide immunoreactivities in the rat pancreas. PEPTIDES 13(1) 91-98, 1992.--The distribution of calcitonin generelated peptide (CGRP), substance P/tachykinin (SP/TK), vasoactive intestinal polypeptide (VIP), neuropeptide Y (NPY) and gastrin-releasing peptide (GRP) immunoreactivities (IR) in the rat pancreas was investigated using radioimmunoassay and immunohistochemistry. CGRP, NPY and VIP tissue contents are much higher than GRP and SP/TK concentrations. Peptidecontaining nerves are distributed to both the exocrine and endocrine pancreas. However, differences exist in terms of density and targets of innervation for each peptidergic system. In the acini and through the stroma, fibers IR for CGRP, NPY and VIP are > GRP- and SP/TK-containing processes. The vasculature is supplied by a prominent NPY, CGRP and, to a lesser extent, SP/ TK innervation. VIP-IR is found occasionally, and GRP-IR is never detected, in fibers associated with blood vessels. Around ducts, CGRP- and NPY-positive neurites are > SP/TK- >/VIP-1R fibers, whereas GRP-containing nerves are not visualized. In the islets, the density of peptidergic nerves is: VIP-, GRP- >/CGRP-IR > NPY or SP/TK. In intrapancreatic ganglia, VIP- and, to a lesser extent, NPY-IRs are found in numerous neuronal cell bodies and in nerve fibers; GRP-IR is present in numerous nerve processes and in few cell bodies; CGRP- and SP/TK-1Rs are detected only in fibers wrapping around unlabeled ganglion cells. The majority of CGRP-IR fibers contain SP/TK-IR. The existence of differential patterns of peptidergic nerves suggests that peptides exert their effects on pancreatic functions via different pathways. Pancreatic innervation Peptidergic nerves Calcitonin gene-related peptide Vasoactive intestinal polypeptide Neuropeptide Y Gastrin-releasing peptide
Substance P Tachykinins Intrapancreatic ganglia Rat
(duodenal versus splenic portion) of the rat pancreas, and we compared the patterns of distribution of nerve fibers containing these peptide IRs. In addition, since C G R P - I R s colocalize with SP/TK-IRs in many structures, including sensory ganglia, cutaneous, visceral and vascular sensory nerves (13,14,26), we evaluated to what extent C G R P - and SP/TK-IRs coexist in neuronal processes of the rat pancreas.
C O N S I D E R A B L E evidence indicates that a variety of biologically active neuropeptides contribute to the control of digestive functions, i.e., motor, absorptive and secretory activities (6). Peptide-containing nerves represent a prominent c o m p o n e n t of the innervation of the digestive system, including the pancreas. In fact, pancreatic nerves immunoreactive (IR) for vasoactive intestinal polypeptide (VIP) (3,5,22-25), substance P (SP), enkephalin, cholecystokinin (22,25,34), bombesin and its related peptide gastrin-releasing peptide (GRP) (5,12,29,42), neuropeptide Y (NPY) (4,5,7,37), neurotensin (31 ), calcitonin-gene related peptide (CGRP) (35,37) and galanin (8) have been reported by us and other investigators. The tissue distribution and the identification of specific structures (e.g., islets, acinar cells, ducts, blood vessels as well as intrapancreatic ganglia) innervated by peptide-containing nerves are the base for an analysis and an understanding of sites of action of different neuropeptides. In the present investigation, we quantified the tissue contents of CGRP-, SP (or related tachykinins, TKs)-, VIP-, NPY- and GRP-IRs in different regions
METHOD
Animals and Tissue Preparations Experiments were performed on 12 male adult young Sprague-Dawley rats, weighing 250-400 g. Each animal was housed individually and had free access to water and food at all times. All experimental procedures were consistent with those stipulated in the N I H Guide for the Use of Experimental Animals and approved by the Veterans Administration Animal Research Committee.
Requests for reprints should be addressed to Catia Sternini, M.D., CURE/UCLA Digestive Disease Center, Building 115, Room 203, Wadsworth VAMC, Los Angeles, CA 90073. 91
92
DE GIORGIO ET AL.
For the radioimmunoassay study, pancreatic specimens from 7 animals, sacrificed by decapitation following halothane (Halocarbon Laboratories, Augusta, SC) anesthesia, were immediately dissected, divided into duodenal and splenic regions, frozen on dry ice, weighed and stored at - 7 0 °C until extraction. For immunohistochemistry, 5 rats were deeply anesthetized with 30% chloral hydrate and then perfused through the aorta with 100 ml of 0.9% NaC1 in 0.1 M phosphate buffer (PB), pH 7.4, followed by 600 ml of 2% paraformaldehyde and 15% saturated picric acid solution in 0.1 M PB (Zamboni's fixative). Pancreata were removed, postfixed in the same fixative for 2 hours at room temperature and subsequently placed in PB containing 25% sucrose overnight at 4°C. Serial sections (10-12 ~tm) of both duodenal and splenic regions of the pancreas were cut with a cryostat (Bright Instruments), mounted onto chromalum gelatin-coated slides and stored at - 2 0 ° C until processing for immunohistochemistry.
100
80
60
40
20
0
CGRP
NPY
VIP
GRP
SP
FIG. 1. Neuropeptide tissue levels (ng/g of tissue) in the duodenal (filled bar) and splenic (empty bar) portion of the rat pancreas. Data are expressed as mean + SD.
Tissue Extraction and Radioimmunoassay Frozen specimens of the duodenal and splenic regions of rat pancreata were placed in 0.1 N HCI, boiled and homogenized in a Polytron tissue grinder. Subsequently, extracted supernatants were used for radioimmunoassay. Diluted aliquots were assayed in duplicate. CGRP-, SP-, VIP-, NPY- and GRP-IRs were measured employing previously described radioimmunoassays which have been validated in mammalian tissues including rat (2,21,30,32,41). For all peptides measured, the intraassay variation ranged from 4% to 7%, while the interassay variation ranged from 10% to 13%. Detection limits for these assays were: 39, 10, 20, 39, and 16 pg/ml, respectively. Experiments aimed to assess the specificity of these antibodies showed that: 1) CGRP radioimmunoassay cross-reacts 100% with human or rat fl-CGRP and 50% with human or rat a-CGRP, but it does not cross-react with calcitonin; 2) SP antibody cross-reacts 100% with related TKs (neurokinin A and neurokinin B); 3) VIP radioimmunoassay does not cross-react with structurally related peptides, i.e., peptide histidine isoleucine, peptide histidine methionine, glucagon and secretin; 4) NPY antiserum displays 2% and less than 1% cross-reactivity with porcine peptide YY and human pancreatic polypeptide (PP), respectively; 5) GRP/bombesin antiserum cross-reacts 100% with G R P 14-27.
Irnmunohistochemistry The antisera used for the immunohistochemical studies are listed in Table 1. Tissue sections were processed with either the indirect immunofluorescence or the avidin-biotin-peroxidase complex (ABC) method (35,36). Briefy, tissue sections were washed in 0.1 M PB, pretreated for 30 minutes at room temperature with 5% swine or goat normal sera, and incubated in primary antibody overnight at 4°C in a humid chamber. There-
after, sections were washed in PB, incubated for two hours at room temperature in swine or goat anti-rabbit or anti-guinea pig IgG (diluted 1:50) conjugated to either fluorescein isothiocyanate (FITC) or tetramethyl rhodamine isothiocyanate (TRITC) (purchased from American Qualex International, La Mirada, CA or DAKO, DAKOPATTS, Santa Barbara, CA) for immunofuorescence. For the ABC method, sections were incubated for two hours in goat anti-rabbit biotinylated IgG (diluted 1:100) (Vector Laboratories, Burlingame, CA), followed by ABC solution and then exposed to 3,3'-diaminobenzidine with 0.01% H202 (36). To reduce nonspecific staining due to endogenous peroxidase, tissue processed with the ABC method was first dehydrated, placed in 100% methanol and then in a solution composed of 98% methanol, 1% acetic acid and 1% sodium nitroferricyanide (for 15 min) followed by 100% methanol, rehydrated and then incubated in primary antibody. Primary and secondary antibodies were diluted in 0.5% Triton Xt00 in 0.1 M PB. For double-label studies, tissue was incubated in a mixture containing guinea pig CGRP and rabbit SP antisera. The antigenantibody complexes were then visualized with a mixture of goat anti-guinea pig IgG TRITC and goat anti-rabbit IgG FITC, both used at 1:50 dilution. Sections were examined with a Leitz Orthoplan microscope (E. Leitz, Inc., Rockleigh, N J) using either bright field optics for ABC processed material, or a Ploem epi-illumination system with "I2" or "Ld' and "N2" filter cubes to visualize FITC and TRITC fluorescence, respectively.
Specificity Testsfor Immunohistochemical Techniques Specificity of the staining was assessed as follows: 1) omission of the primary antibody, 2) substitution of the primary antibody
TABLE 1 C H A R A C T E R I S T I C S O F ANTISERA U S E D F O R I M M U N O H I S T O C H E M I S T R Y
Antigen
AntibodyCode
CGRP CGRP SP VIP NPY NPY Bombesin-9
CGRP2p CGRP2f SP 7913 NPY-3 NPY-Y 8543
Host Species
Dilution
Reference
Rabbit Guinea pig Rabbit Rabbit Rabbit Rabbit Rabbit
1:5000 1:200 1:250 1:250 1:250 1:250 1:250
35 Sternini, unpublished Incstar (Minnesota) 11 33 9 40
PEPTIDERGIC NERVES IN THE PANCREAS
93
with rabbit preimmune serum diluted 1:50, and 3) incubation with the primary antibodies preabsorbed, for 12-16 hours at 4°C, with synthetic homologous or heterologous peptides (purchased from Bachem, Torrance, CA or Peninsula Laboratories, Belmont CA), i.e.: a[Tyr]rat CGRP 23-37, a-rat CGRP 1-37, somatostatin, VIP, GRP 1-27, NPY, SP, neurokinin A, CGRP 23-27, a-rat CGRP 1-37, somatostatin, VIP, GRP 1-27, NPY, SP, neurokinin A, neurokinin B, carboxyl-terminal hexapeptide of PP (PP-6) and peptide YY at concentrations of l0/~M. In addition, we performed further control experiments for double-labeling technique to verify that each of the primary antibodies did not bind to the other when mixed together and that each of the secondary antisera reacted with the appropriate antigen-antibody complex (14). These controls consisted of: l) omission of one of the primary antibodies from the first incubation, 2) omission of one of the secondary antibodies from the second incubation, and 3) preabsorption of CGRP or SP with the appropriate peptides. Since the possibility that the immunoreactive materials are structurally related to, but not identical with, the peptides under investigation, the terms CGRP, SP/TK, VIP, NPY, and GRP immunoreactive, containing or positive will be used to refer to the immunostaining observed.
Statistical Analysis Radioimmunoassay data were expressed as mean _+standard deviation (SD) of the mean. Mean values were compared by repeated measures analysis of variance (ANOVA). Computations were made on logarithmic transformation of the data. Mean comparisons were assessed for statistical significance according to Tukey's Studentized Range criteria. The level of statistical significance was set at p < 0.05.
FIG. 2. Photomicrographs of the exocrine pancreas displaying VIP- and CGRP-IR nerve fibers innervating the acini. Calibration bar" 25 ~tm. of the duodenal and splenic portion of the rat pancreas. The relative density of peptide-containing nerves in relation to each pancreatic compartment is summarized in Table 2. There are differences in the densities and patterns ofinnervation in relation to specific intrapancreatic tissue targets. Thus CGRP-, NPYand, to a lesser extent, SP/TK-containingnerves are particularly prominent in association with the vasculature and, within the exocrine pancreas, with the duct system, while VIP- and GRPIR fibers are particularly evident in intrapancreatic ganglia and endocrine pancreas. For all peptides studied, however, there are no evident differences in term of innervation patterns between the duodenal and splenic portion. Specificity for CGRP, SP/TK, VIP, NPY and GRP immunostainings was demonstrated by the lack of immunoreaction in sections in which the primary antiserum was omitted, or in
RESULTS
Radioimmunoassa£ Data Figure 1 shows the neuropeptide tissue contents in the duodenal and splenic portions of the rat pancreas. CGRP, NPY and VIP tissue levels are much higher than GRP and SP concentrations. The differences between the duodenal and splenic lobes of the rat pancreas are not significant for any of the five investigated peptides (p = 0.15).
Immunohistochemistry CGRP-, SP/TK-, VIP-, NPY- and GRP-IR nerve fibers are distributed throughout the exocrine and endocrine compartment
TABLE 2 RELATIVEDENSITYOF PEPTIDE-CONTAININGNERVESIN RELATION TO DIFFERENTPANCREATICSTRUCTURES
Exocrine Pancreas Acini Septa* Ducts Vasculature Endocrine Pancreas Intrapancreatic Ganglia Fibers Cell bodies
CGRP
SP/TK
GRP
VIP
NPY
++ ++ ++/+++ +++ +/++
+ + +/++ ++ +
+ + ++
++ +/+ + + -/+ ++
++ +/+ + ++/+++ +++ +
++ -
++ -
+++ +
++ +++
++ ++/+++
Key: - = absent; - / + = occasional presence; +, ++, +++ = relative density. * Intralobular and interlobular connective tissue.
94
DE GIORGIO ET AL. other hand, absorptions of CGRP, SP, VIP, NPY or GRP antisera with structurally unrelated polypeptides did not modify the immunolabeling. Controls for double-staining experiments demonstrated that CGRP or SP antibodies did not bind to each other and also that each of the secondary antibodies bound only to the appropriate antigen-antibody complex. Exocrine Pancreas CGRP-, SP/TK-, VIP-, NPY- and GRP-IR processes, running either singly or in fascicles, are located among the acini (Figs. 2 and 3) as well as in the septa (i.e., the interlobular and intralobular connective tissue) of the duodenal and splenic portions of the rat pancreas. CGRP-, VIP- and NPY-containing processes are in general more abundant than those immunoreactive for SP/ TK or GRP. In the ductal system, CGRP-, NPY- and, to a lesser extent, SP/TK-IR fibers, running either singly or in small bundles, are particularly abundant in the connective tissue which supports the ducts (Figs. 3 and 7). Sometimes, single varicose CGRP-, NPY- and SP/TK-IR processes also can be identified underneath the surface epithelium of the ducts. VIPergic nerves are the least numerous around ducts, although constantly detectable. GRPcontaining fibers are not detected in association with the ductal system. Vasculature The vasculature receives a particularly rich NPY and CGRP innervation, especially blood vessels with the appearance of small
FIG. 3. Photomicrographs of the exocrine pancreas. The upper panel illustratesan abundant CGRP innervation of a small duct (d); the middle panel shows a dense network of NPY-IR nerves around a small artery (a), along with varicose NPY-containing fibers running among acini. In the lower panel, thin beaded VIP-IR fibers (arrows) are evident in the connective tissue supporting a large-size duct. Calibration bar: 25 urn. sections incubated with normal rabbit serum or primary antibodies preabsorbed with the appropriate peptide. Of the two rabbit NPY antibodies used, NPY-Y, but not NPY-3, in addition to neuronal elements, stained cells at the peripheral margin of the islets. The incubation of NPY-Y antiserum with PP-6 abolished the cell staining but did not affect the intensity nor the density of NPY-positive neuronal structures throughout the rat pancreas. The incubation of SP antiserum with the appropriate or related peptides, such as neurokinin A and neurokinin B, abolished the immunostaining, indicating that the antibody utilized recognized different members of the TK family. For this reason we use the terms SP/TK and SP interchangeably. On the
FIG. 4. Representative photomicrographs of the endocrine pancreas demonstrating GRP (upper panel)- and VIP-containing (lower panel) varicose processes in the islets. Calibration bar: 25 um.
PEPTIDERGIC NERVES IN THE PANCREAS
95
FIG. 5. lntrapancreatic ganglia. VIP-, NPY- and GRP-1Rs are present either in neurons (arrows) or in nerve fibers, while CGRPIR is restricted to nerves surrounding unstained ganglion cells. VIP- and NPY-IRs are present in a greater number of cell bodies than GRP-IR. Calibration bar: 25 tsm.
arteries and arterioles (Figs. 3 and 7). SP/TK-immunolabeled fibers (Fig. 7) associated with the vasculature are also abundant, even though less prominent than NPY- and CGRP-IR nerves. These three peptide-IRs appear to be localized to perivascular fibers forming dense networks at the adventitial level. Conversely, VIP-containing fibers are less often visualized in association with the vasculature, and usually they are found in paravascular position, that is, in the connective tissue located around the vessels. GRP-IR nerves are never seen in association with the vasculature.
Endocrine Pancreas Thin varicose processes containing CGRP-, SP/TK-, VIP-, NPY- or GRP-IR supply the islets with the following density: VIP- and GRP-IR fibers (Fig. 4) >t CGRP-IR > SP/TK- and NPY-IR processes. CGRP-IR is localized to endocrine-like cells identified both at the peripheral margin of the islets and scattered among the acini in the exocrine tissue. The NPY-IR visualized at the periphery of the islets with NPY-Y antibody is identical to that previously described with PP antisera (38) and is eliminated by preabsorption of this antibody with PP-6, demonstrating that these NPY-immunolabeled cells are PP-containing cells. No VIP-, GRP- or SP/TK-IR endocrine]paracrine-like cells were stained in islets or among the acini.
Intrapancreatic Ganglia VIP-IRs (Figs. 5 and 6) and, to a lesser extent, NPY-IRs are found in a good proportion of ganglion cells (Fig. 5), in addition to varicose nerve fibers. Networks of GRP-IR fibers are dense
and wrap around stained and (more often) unstained perikarya (Fig. 5). GRP-containing neuronal cell bodies are much less numerous than those positive for VIP or NPY, and they are not detected in all ganglia. Conversely, CGRP- and SP/TK-IRs are restricted to nerve fibers (Figs. 5 and 6).
Colocalization Studies The majority of fibers with CGRP-IR also contain SP/TKIR (Fig. 7). In addition, other two subpopulations of fibers are visualized: a small component of SP/TK positive nerves lacking CGRP-IR and, more frequently, CGRP positive processes without SP/TK-IR. DISCUSSION
The results of the present study, where the patterns of innervation ofCGRP-, SP/TK-, VIP-, NPY- and GRP-containing nerves in the rat pancreas are compared and their tissue contents measured, confirm and extend previous findings in the mammalian pancreas. The tissue content of each peptide in the duodenal portion is similar to that in the splenic region. Similarly, the patterns of fibers containing these peptide-IRs are identical in both pancreatic regions. However, there are differences among these peptidergic systems in relation to specific pancreatic structures, as summarized in Table 2. Thus, in the acinar and stromal compartment, CGRP-, NPY- and VIP-containing processes are > SP/TK- and GRP-IR fibers. Around ducts, CGRP-, NPYand, to a lesser extent, SP/TK-IR nerves are more abundant than those IR for VIP, while GRP-immunostained fibers are not detected. The vasculature receives prominent NPY, CGRP
96
FIG. 6. Simultaneous staining for VIP (upper panel)- and CGRP-IRs (lower panel) in an intrapancreatic ganglion. Note the absence of colocalization: VIP-IR being present in neurons and fibers, while CGRP-IR is detectable only in fibersgrouped in a bundle or single. Note a CGRPIR varicosityin closevicinityto a cellbody containingVIP-IR.Calibration bar: 25 ~zm.
and SP/TK nerve supply, while VIP-IR fibers are sparse and GRP-IR fibers are not visualized. In the islets, the density of peptidergic nerves is: those IR for VIP and GRP >/CGRP- > SP/TK- and NPY-IR ones. Within the intrapancreatic ganglia, VIP-, NPY- and GRP-IRs are present in neuronal cell bodies as well as in fibers, while CGRP- and SP/TK-IRs are restricted to nerves. VIP- and NPY-IR ganglion cells are much more abundant than those visualized with GRP antiserum. Within the digestive system, CGRP- and SP/TK-IRs are localized to intrinsic neurons of the gut as well as to extrinsic, capsaicin-sensitive, primary sensory fibers (13,36). Unlike SP/ TK-IR fibers, which mainly derive from intrinsic neurons and comprise only a small portion of fibers originating from extrinsic, afferent neurons, an important component of the CGRP innervation of the alimentary tract is afferent in origin (35-37). The lack of CGRP- and SP/TK-IR intrapancreatic ganglion cells observed in the present study and the virtual elimination of CGRPand SP/TK-IR fibers within the pancreas following chemical and surgical denervation (35-37) are consistent with an extrinsic origin of CGRP and SP/TK pancreatic innervation. However, the presence of occasional ganglion cells positive for SP/TK (37), but not for CGRP (35), has been reported in the pancreas of rats treated with the neurotoxin colchicine, which is commonly used to increase peptide levels in cell bodies. Therefore, a small subpopulation of SP/TK-IR fibers (i.e., those lacking CGRPIR) could derive from intrapancreatic ganglion cells. The extensive colocalization of CGRP- and SP/TK-IRs in nerve processes of the rat pancreas extends previous observations showing the
DE GIORGIO ET AL. coexpression of these two peptides in cutaneous, visceral, and vascular sensory nerves ( 13,14,26). VIP-IR has been identified in the vast majority of neuronal cell bodies of most intrapancreatic ganglia. Therefore, our results, in accordance with previously reported observations (24,37), confirm that the VIPergic innervation of the rat pancreas, as well as that of other mammalian species (5,23), is predominantly of intrinsic origin. The identification of NPY- and GRP-IR ganglion cells, the latter being much less numerous than those containing VIP- or NPY-IR, suggests that also these two peptidergic systems are, at least in part, of intrinsic origin. Whereas it is well established that the pancreas, as the gut and liver, receives extrinsic NPY-IR fibers which are mainly distributed to the vasculature and are likely to derive from sympathetic prevertebral ganglia (4,15,37), whether the GRP innervation of the alimentary tract includes an extrinsic input remains to be elucidated. Peptidergic nerves supplying the pancreas could also be derived from enteric neurons. This is based on the recent observation that there are neurons in the myenteric plexus of the stomach and proximal duodenum that send axonal projections to the pancreas (18). However, further studies are necessary to determine whether these gastroenteropancreatic neuronal elements contain peptide-lR. Some divergencies emerge between the present findings and previous observations in the pancreas of the same or different species. For example, in other studies, GRP-IR fibers could not be visualized in the rat pancreas (29) or could be detected only in the exocrine compartment but not within islets or intrapancreatic ganglia (12). In a previous investigation on NPY pancreatic innervation, NPY-IR fibers were not described in association with the islets (4). In the guinea pig pancreas, a selective distribution of SP-IR processes (lacking CGRP-IR and perhaps originating from intrapancreatic ganglion cells) to some islets and intercalated ducts has been reported (19). This finding differs from that described in the rat pancreas [present study, and (37)]. Furthermore, unlike the rat pancreas, where ducts appear to be devoid of GRP innervation, the ductal system of feline and porcine pancreata is innervated by GRP-IR processes [(5); De Giorgio and Sternini, unpublished]. These divergent findings within the same species and among several species may be explained on the basis of different sensitivities of the antibodies used in these studies, and with the existence of intraspecies variations of peptidergic nerves supplying the mammalian pancreas, respectively. The morphological relationship of peptidergic nerves with different pancreatic compartments supports the hypothesis that different peptides have distinct sites of actions. Thus the rich association of CGRP-, SP/TK- and NPY-containing fibers with the vasculature is consistent with an involvement of these vasoactive peptides (16,27,28) in the regulation of pancreatic blood flow, probably through a direct mechanism. In addition, CGRP, SP, and NPY can influence the exocrine and endocrine pancreas, which are targets of innervation of these peptidergic systems ( 1,16,17,27,28). It also emerges that the diversity of targets, including the acinar tissue, ducts, endocrine compartment and the vasculature, supplied by CGRP- and SP/TK-IR nerves, is consistent with distinct efferent roles of these fibers in addition to an afferent role, which could be speculated in view of the fact that they originate from primary sensory neurons (34-37). The association of GRP-containing fibers with acini, but not with ducts, fits the observation that intravenous infusion of the tetradecapeptide bombesin, whose amino acid sequence has a biologically active region identical to that of GRP, evokes pancreatic secretion particularly abundant in enzymes but poor in bicarbonate (39). It is likely that GRP/bombesin effects on the exocrine pancreas are mostly related to the acini. Furthermore,
P E P T I D E R G I C N E R V E S IN T H E P A N C R E A S
97
FIG. 7. Simultaneous visualization of CGRP- and SP/TK-1Rs in a subset of nerve processes (arrows) around a blood vessel (upper panel) and a large-size duct (lower panel). Arrowheads indicate CGRP-positive fibers lacking SP/TK-IR. Calibration bar: 25 ~m.
the relative abundance of G R P - and VIP-IR nerves supplying the islets supports the hypothesis that G R P and VIP are involved in the neural control of pancreatic hormone secretions (10,20). Finally, the VIP innervation of the exocrine pancreas provides the morphological basis for the stimulatory effect of VIP on exocrine pancreatic secretions (10,17). By contrast, despite the VIP vasodilatory effect (10), only occasional VIP-IR fibers are associated with blood vessels, as previously showed in other mammalian species (3,5,22). It is possible that the vascular effect of VIP is due to a related peptide not detectable by our VIP antiserum or that the VIP effect on blood flow is mediated by other substances. In summary, we can conclude that: 1) CGRP-, SP/TK-, VIP-, NPY- and G R P - I R nerve fibers represent an important supply of the pancreatic innervation; 2) they have differential patterns and densities according to specific pancreatic targets (CGRP, S P / T K and NPY innervations are particularly prominent in as-
sociation with blood vessels, ducts and, to a lesser extent, acini, while VIP- and G R P - I R fibers are more abundant in intrapancreatic ganglia and islets); 3) C G R P - and SP/TK-IRs extensively colocalize in pancreatic nerves; and 4) among the peptides investigated only C G R P appears to be expressed in endocrine-like cells. Thus, as for the gut, peptidergic nerves supplying the pancreas include an intrinsic component of fibers, originating from intrapancreatic neurons, as well as an extrinsic input of fibers, which derive from neurons located outside the pancreas. ACKNOWLEDGEMENTS This study was supported by grant DK38752 from the National Institutes of Health (C.S.) and Veteran Administration Medical Research Funds (N.C.B.). The authors wish to thank H. Wong and Drs. J. H. Walsh, P. C. Emson and J. K. McDonald for having provided GRP, NPY and VIP antisera and the RIA laboratory at Mayo Clinic (Mayo Medical School of Medicine, Rochester, MN) for RIA determinations.
REFERENCES 1. Ahrrn, B.; Pettersson, M. Calcitonin gene-related peptide (CGRP) and amylin and the endocrine pancreas. Int. J. Pancreatol. 6:1-15; 1990. 2. Benarroch, E. E.; Schmelzer, J. D.; Ward, K. K.; Nelson, D. K.; Low, P. A. Noradrenergic and neuropeptide Y mechanisms in guanethidine-sympathectomized rats. Am. J. Physiol. 259:R371-375; 1990. 3. Bishop, A. E.; Polak, J. M.; Green, I. C.; Bryant, M. G.; Bloom, S. R. The location of VIP in the pancreas of man and rat. Diabetologia 18:73-78; 1980.
4. Carlei, F.; Allen, J. M.; Bishop, A. E.; Bloom, S. R.; Polak, J. M. Occurrence, distribution and nature of neuropeptide Y in the rat pancreas. Experientia 41 : 1554-1557; 1985. 5. De Giorgio, R.; Sternini, C.; Brecha, N. C.; Widdison, A. L.; Karanjia, N. D.; Reber, H. A.; Go, V. L. W. Patterns of innervation of vasoactive intestinal polypeptide, neuropeptide Y and gastrin-releasing peptide immunoreactive nerves in the feline pancreas. Pancreas (in press). 6. Dockray, G. J. Physiology of enteric neuropeptides. In: Johnson, L. R.; Christensen, J.; Jackson, M. J.; Jacobson, E. D.; Walsh, J. H.,
98
7. 8. 9. 10. 11. 12. 13.
14.
15.
16. 17. 18. 19. 20. 21. 22. 23. 24.
25. 26.
DE G I O R G I O ET AL. eds. Physiology of the gastrointestinal tract, vol. 1. New York: Raven Press; 1987:41-66. Dunning, B. E.; Ahrrn, B.; Brttcher, G.; Sundler, F.; Taborsky, G. J., Jr. The presence and actions of NPY in the canine endocrine pancreas. Regul. Pept. 18:253-265; 1987. Dunning, B. E.; Ahrrn, B.; Veith, R. C.; Brttcher, G.; Sundler, F.; Taborsky, G. J., Jr. Galanin: A novel pancreatic neuropeptide. Am. J. Physiol. 251:E127-E133; 1986. Emson, P. C.; Corder, R.; Lowry, P. J. Demonstration of a neuropeptide Y-like immunoreactivity in human phaeochromocytoma extracts. Regul. Pept. 8:89-94; 1984. Fahrenkrug, J. Vasoactive intestinal polypeptide: Measurements, distribution and putative neurotransmitter function. Digestion 19: 149-169; 1979; Furness, J. B.; Costa, M.; Walsh, J. H. Evidence for and significance of the projection of VIP neurons from the myenteric plexus to the tenia coli in the guinea-pig. Gastroenterology 80:1557-1561; 1981. Gathei, M. A.; George, S. K.; Major, J. H.; Carlei, F.; Polak, J. M.; Bloom, S. R. Bombesin-like immunoreactivity in the pancreas of man and other mammalian species. Experientia 40:884-886; 1984. Gibbins, L. L.; Furness, J. B.; Costa, M.; Maclntyre, I.; Hillyard, C. J.; Girgis, S. Co-localization of calcitonin gene-related peptidelike immunoreactivity with substance P in cutaneous, vascular and visceral sensory neurons of guinea pigs. Neurosci. Lett. 57:125-130; 1985. Goehler, L. E.; Sternini, C.; Brecha, N. C. Calcitonin gene-related peptide immunoreactivity in the biliary pathway and liver of the guinea-pig: Distribution and colocalization with substance P. Cell Tissue Res. 253:145-150; 1988. Goehler, L. E.; Sternini, C. Neuropeptide Y immunoreactivity in the mammalian liver: Pattern of innervation and coexistence with tyrosine hydroxylase immunoreactivity. Cell Tissue Res. 265:287295; 1991. Goodman, E. C.; Iversen, L. L. Calcitonin gene-related peptide: A novel neuropeptide. Life Sci. 38:2169-2178; 1986. Hoist, J. J. Peptidergic mechanisms in the pancreas. Arch. Int. Pharmacodyn. 303:252-269; 1990. Kirchgessner, A. L.; Gershon, M. D. Innervation of the pancreas by neurons of the gut. J. Neurosci. 10:1626-1642; 1990. Kirchgessner, A. L.; Pintar, J. E. Guinea pig pancreatic ganglia: Projections, transmitter content, and the type-specific localization of monoamine oxidase. J. Comp. Neurol. 305:613-631 ; 1991. Knuhtsen, S. V.; Hoist, J. J.; Schwartz, T. W.; Jensen S. L.; Nielsen, O. V. The effect of gastrin-releasing peptide on the endocrine pancreas. Regul. Pept. 17:269-276; 1987. Koch, T. R.; Roddy, D. R.; Carney, J. A.; Telander, R. D.; Go, V. L. W. Quantitation, and origin of immunoreactive neuropeptide Y in the human gastrointestinal tract. Regul. Pept. 21:309-319; 1988. Larsson, L. I. lnnervation of the pancreas by substance P, enkephalin, vasoactive intestinal polypeptide and gastrin/CCK immunoreactive nerves. J. Histochem. Cytochem. 27:1283-1284; 1979. Larsson, L.I.; Fahrenkrug, J.; Hoist, J. J.; Schaffalitzky de Muckadell, O. B. Innervation of the pancreas by vasoactive intestinal polypeptide (VIP) immunoreactive nerves. Life Sci. 22:773-780; 1978. Larsson, L. I.; Fahrenkrug, J.; Schaffalitzky de Muckadell, O.; Sundler, F.; H~tkanson, R.; Rehfeld, J. F. Localization of vasoactive intestinal polypeptide (VIP) to central and peripheral neurons. Proc. Natl. Acad. Sci. USA 73:3197-3200; 1976. Larsson, L. I.; Rehfeld, J. F. Peptidergic and adrenergic innervation of pancreatic ganglia. Scand. J. Gastroenterol. 14:433-437; 1979. Lundberg, J. M.; Franco-Cereceda, A.; Hua, X.; Hrkfelt, T.; Fischer, J. A. Coexistence of substance P and calcitonin gene-related peptide-
27. 28.
29. 30.
31.
32. 33.
34. 35.
36.
37.
38. 39.
40. 41.
42.
like immunoreactivity in sensory nerves in relation to cardiovascular and bronchoconstrictor effects of capsaicin. Eur. J. Pharmacol. 108: 315-319; 1985. McDonald, J. K. Neuropeptide Y and related substances. Crit. Rev. Neurobiol. 4:97-135; 1988. Maggio, J. E.; Mantyh, P. W. Gut tachykinins. In: Schultz, S. G.: Makhlouf, G. M.; Rauner, B. B., eds. Handbook of physiology. A critical, comprehensive presentation of a physiological knowledge and concepts. Section 6: The gastrointestinal system. Volume II: Neural and endocrine biology. Bethesda: American Physiological Association; 1989:661-690. Moghimzadeh, E.; Ekman, R.; Hhkanson, R.; Yanaihara, N.; Sundler, F. Neuronal gastrin-releasing peptide in the mammalian gut and pancreas. Neuroscience 10:553-563; 1983. Nelson, D. K.; Service, J. E.; Studelska, D. R.; Brimijoin, S.; Go, V. L. W. Gastrointestinal neuropeptide concentrations following guanethidine sympathectomy. J. Auton. Nerv. Syst. 22:203-210; 1988. Reinecke, M. Neurotensin: Immunohistochemical localization in the central and peripheral nervous system and in endocrine cells and its functional role as neurotransmitter and endocrine hormone. Prog. Histochem. Cytochem. 16:12-172; 1985. Roddy, D. R.; Aimone, L. D.; Yaksh, T. L.; Go, V. L. W. Regional distribution and characterization of the brain-gut peptide: Calcitonin gene-related peptide in pig. Soc. Neurosci. Abstr. 15:740; 1989. Sabatino, F. D.; Murnane, J. M.; Hoffman, R. A.; McDonald, J. K. Distribution of neuropeptide Y-like immunoreactivity in the hypothalamus of the adult golden hamster. J. Comp. Neurol. 257:93104; 1987. Sharkey, K. A.; Williams, R. G.; Dockray G. J. Sensory substance P innervation of the stomach and pancreas. Gastroenterology 87: 914-921; 1984. Sternini, C.; Brecha, N. lmmunocytochemical identification of islet cells and nerve fibers containing calcitonin gene-related peptide-like immunoreactivity in the rat pancreas. Gastroenterology 90:11551163; 1986. Sternini, C.; Reeve, J. R., Jr.; Brecha, N. Distribution and characterization of calcitonin gene-related peptide immunoreactivity in the digestive system of normal and capsaicin-treated rats. Gastroenterology 93:852-862; 1987. Su, H. C.; Bishop, A. E.; Power, R. F.; Hamada, Y.; Polak, J. M. Dual intrinsic and extrinsic origins of CGRP- and NPY-immunoreactive nerves of rat gut and pancreas. J. Neurosci. 7:2674-2687; 1987. Vaillant, C.; Taylor, I. L. Demonstration of carboxyl-terminal PPlike peptides in endocrine cells and nerves. Peptides 2(Suppl. 2):3135; 1981. Walsh, J. H. Bombesin-like peptides. In: Schultz, S. G.; Makhlouf, G. M.; Rauner, B. B., eds. Handbook of physiology. A critical, comprehensive presentation of a physiological knowledge and concepts. Section 6: The gastrointestinal system. Volume II: Neural and endocrine biology. Bethesda: American Physiological Association; 1989: 587-609. Walsh, J. H.; Wong, H. C.; Dockray, G. J. Bombesin-like peptides in mammals. Fed. Proc. 38:2315-2319; 1979. Yaksh, T. L.; Michener, S. R.; Bailey, J. E.; Harty, G. J.; Lucas D. L.; Nelson D. K.; Roddy, D. R.; Go, V. L. W. Survey of distribution of substance P, vasoactive intestinal polypeptide, cholecystokinin, neurotensin, Met-enkephalin, bombesin and PHI in the spinal cord of cat, dog, sloth and monkey. Peptides 9:357-372; 1988. Yanaihara, N.; Yanaihara, C.; Mochizuki, T.; lwahara, K.; Fujita, T.; Iwanaga, T. Immunoreactive GRP. Peptides 2(Suppl. 2):185191; 1981.