- Email: [email protected]
Topographical distribution and immunocytochemical features of colonic neurons that project to the cranial mesenteric ganglion in the pig
Journal of the Autonomic Nervous System, 44 (1993) 119-127
119
© 1993 Elsevier Science Publishers B.V. All rights reserved 0165-1838/93/$06.00 JANS 01413
Topographical distribution and immunocytochemical features of colonic neurons that project to the cranial mesenteric ganglion in the pig Martine Barbiers, Jean-Pierre Timmermans, Dirk Adriaensen, Marie H.A. De Groodt-Lasseel and Dietrich W. Scheuermann Laboratory of Cell Biology and Histology, University of Antwerp, Antwerp, Belgium (Received 1 October 1992) (Revision received and accepted 15 February 1993)
Key words: Cranial mesenteric ganglion; Superior mesenteric ganglion; Retrograde tracing; Immunocytochemistry; Enteric nervous system; Colon; Pig Abstract Using the retrograde neuronal tracers Fast blue and Fluorogold, the topographical distribution and morphological features of porcine colonic neurons projecting to the cranial (superior) mesenteric ganglion have been investigated. Two to four weeks after injection of the tracer into the cranial mesenteric ganglion of immature pigs, labelled neurons were found throughout the colon. In the myenteric and outer submucous plexuses, they were present in ganglia situated to the side of the mesenteric attachment. The highest density of labelled neurons was observed at the end of the ascending colon, which in the pig represents 75-80% of the total colon length. The viscerofugal neurons had a multidendritic appearance and part of them were immunoreactive for calcitonin gene-related peptide or serotonin. This study has revealed similarities but also significant differences in the colono-sympathico-colonic pathways between the pig and small laboratory animals such as the guinea-pig.
Introduction
Anatomical studies and physiological experiments in the 1940s and 1950s have revealed that axons of ganglion cells located in the wall of the gastrointestinal tract extend into the prevertebral ganglia [15,24] and that reflex-inhibition is mediated through these ganglia, even after their decentralization [15]. To date, electrophysiological [1,9-12,27], immunocytochemical [3,31] and ret-
Correspondence to: D.W. Scheuermann, University of Antwerp, Laboratory of Cell Biology and Histology, Groenenborgerlaan 171, B-2020 Antwerp, Belgium.
rograde labelling techniques [2,7] have provided evidence that colonic neurons project to the caudal (inferior) mesenteric ganglion. Moreover, Kreulen and Szurszewski [13,14] demonstrated electrophysiologically that also the coeliac ganglion and cranial mesenteric ganglion receive mechanosensory input from oral and caudal regions of the colon and provided support for the existence of a colono-colonic inhibitory reflex that is mediated through pathways in prevertebral ganglia (see also Szurszewski [26]). By retrograde tracing experiments, Messenger and Furness [20,21] recently supported the presence of neurons projecting from the proximal and distal colon to the coeliac ganglion in guinea-pig.
120 Since it is known that the digestive system of the pig greatly resembles that of humans, this animal was chosen to demonstrate species differences between the enteric nervous system (ENS) of small laboratory animals and that of larger mammals. Immature pigs were preferred to mature ones because gut tissue of animals of 6 weeks of age could be optimally prepared and because previous observations had not revealed any major differences between the organization of the ENS of 6-week-old and that of 6-month-old animals. Tracing studies in the pig small intestine have shown an increase in the number of neurons projecting to the cranial mesenteric ganglion from the duodenum to the ileum [28]. The present investigation made use of the retrograde neuronal tracers Fast blue (FB) and Fluorogold (FG) to study projections from colonic neurons to the cranial mesenteric ganglion. Immunostaining for microtubule-associated protein (MAP2) and for a mixture of neurofilaments (iC8) was applied for
improved visualization of the morphology and topography of viscerofugal neurons within the ganglionic plexuses. Further immunohistochemical investigations with antisera raised against vasoactive intestinal polypeptide (VIP), calcitonin gene-related peptide (CGRP) and serotonin (5HT) were conducted to elucidate their neurochemical content.
Materials and Methods Eight just weaned, 21-23-day-old domestic pigs were anesthetized for surgical purposes with Stressnil (Janssen Pharmaceutica, Beerse, Belgium, 1 m l / 2 0 kg body weight, intramuscularly) and Hypnodil (Janssen Pharmaceutica, 1.5 m l / 2 0 kg body weight, intraperitoneally). Via a paracostal incision between the 13th and 14th rib the cranial mesenteric ganglion was exposed and different regions of the ganglion were injected with
TABLE I Detailed description of normal sera and antisera used for immunocytochemistry Normal sera
Normal goat serum (Dakopatts, Ghent, Belgium) Normal sheep serum (Nordic, Tilburg, The Netherlands) Primary antisera
Source
Host species
Dilution
MAP2 (microtubule-associated protein) iC8 (mixture of neurofilaments) CGRP (calcitonin gene-related peptide)
[see ref. 22] [see ref. 30] Amersham (Brussels, Belgium) Eurodiagnostics (Turnhout, Belgium) Incstar (Brussels, Belgium) Amersham UCB (Eigenbrakel, Belgium)
rabbit mouse rabbit
1 : 200 1 :'500 1 : 400
rabbit
1 : 50
goat
1 : 50
rabbit rabbit
1 : 200 1 : 500
5-HT (serotonin) 5-HT VIP (vasoactive intestinal polypeptide) VIP Secondary antisera and streptavidin complex
FITC-conjugated goat anti-rabbit IgG, diluted 1 : 20 (Nordic) Biotinylated sheep anti-mouse IgG, diluted 1 : 20 (Amersham) Biotinylated donkey anti-sheep IgG, diluted 1 : 20 (Amersham) Biotinylated goat anti-rabbit IgG, diluted 1 : 100 (Dakopatts) Streptavidin-Texas red complex, diluted 1 : 20 (Amersham) Streptavidin-HRP complex, diluted 1 : 100 (Amersham)
121 5 - 1 0 /xl FB (Sigma, Brussels, Belgium) (three pigs) or F G (Fluorochrome Inc., Englewood, Colorado, USA) (three pigs). In two control animals, F G was dripped on the cranial mesenteric ganglion or the serosal surface of the colon. The experiments were approved by the Ethical Medical Commission of the University of Antwerp and by the Ethical Commission of the National Fund for Scientific Research. Two to four weeks after surgery, the animals were intraperitoneally injected with colchicine (10 m g / k g body weight) and killed by decapitation 5 h later. The colon was divided into three segments of equal length, referred to as proximal, middle and distal parts. The colonic segments were ligated, filled with and immersed in 4% paraformaldehyde in 0.1 M sodium phosphate buffer containing 0.2% picric acid for 2 h at room temperature. To improve immunocytochemical conditions, the tissue was further processed as described previously [18,25]. Whole-mounts containing either the myenteric or submucous nerve plexuses were viewed for FB fluorescence in a Leitz MPV1 fluorescence microscope equipped with a Leitz Ploemopak filter block C (3 m m BG3 + S 405 excitation filter, T K 455 b e a m splitting mirror, K 460 suppression filter) and for FG fluorescence with a Leitz Ploem o p a k filter block A (BP340-380 excitation filter, R K P 400 b e a m splitting mirror, LP 430 suppression filter). Approximately every 15 cm, tissue pieces 3.5 cm in length each were investigated over the whole circumference of the colon and retrogradely labelled neurons were counted. The m e a n n u m b e r of labelled ceils per cm length was calculated for the proximal, middle and distal regions. In two of the three F G experiments, only the proximal part was examined for the presence of labelled neurons. These counts were not incorporated into the calculations for the overall distribution in the colon. For immunocytochemistry, whole-mounts were preincubated for 1 h in 10% normal serum containing 1% Triton X-100, and successively incubated in the primary (17 h) and corresponding secondary antiserum (6 h). All incubations were performed at room temperature. The fluorescent marker was conjugated either directly to the secondary antiserum or coupled via the biotin-streptavidin bridge system. More
information on the normal sera and antisera used is given in Table I. During dissection, the cranial mesenteric ganglion and the caudal mesenteric ganglion were also removed, fixed as described above and further processed for embedding in paraffin. Sections were checked for FB or F G fluorescence. For immunocytochemistry, the sections were successively incubated in the primary (17 h) and the corresponding secondary antiserum (6 h). A fluorescent marker was conjugated to the latter, or an indirect method involving the streptavidin-horseradish peroxidase complex was used (see Table I).
;E COLON
t" !
I:.'
r.'
F.15 |-.~
CTUM
IG COLON
I Fig. I. Schematic representation of the distribution of retro-
gradely labelled neurons in the porcine colon. The ascending colonic part has the shape of a double coiled cone. Labelled cells (dots) are present in the ascending, transverse and descending colons, and are localized near the mesenteric attachment (. . . . . . ) between the taenia coil ( - - - - - - ) . The number of labelled cells per unit length is the largest at the end of the ascending colon.
122
Results In the control animal where FG was dripped on the ganglionic surface, no labelled ceils were
found in the colon. In the other control where the tracer was dripped on the serosal surface of the colon, all nerve cells in the enteric nerve networks and the outer longitudinal and circular
e~
7,
!
E
:;2-)
i
!'.: iii
GL,
~
Fig. 2a-f. Simultaneous visualization of the tracer (a, f) and the immunostaining for iC8 (b) or MAP2 (c, d, e) reveals the localization of FG-labelled neurons within the ganglionic nerve plexuses. (a, h) Labelled neuron situated in the myenteric plexus (arrows). (e-f) Submucous layer photographed at different focal plains visualizing a retrogradely traced neuron (arrows) in the outer submucous plexus (double-headed arrows). In ganglia of the inner submucous plexus (arrowheads), no retrogradely labelled neurons are observed. (e, f) Enlargement of the area indicated in (d). Scale bar = 50 /zm. Fig. 3a, b: Immunolabelling for iC8 enables one to visualize the shape and size of retrogradely labelled neurons. (a) FB-labelled neuron in the myenteric plexus. (b) The same cell after iC8 immunocytochemistry, having a large, multi(long)dendritic appearance. Scale bar = 50 /xm. Fig. 4a, b: Fluorescence micrographs of unstained whole-mount preparations showing retrogradely labelled nerve cells in the myenteric plexus. A granular appearance of the tracer can be observed in the perikarya and in some processes of the multidendritic neurons. FB: Fast blue; FG: Fluorogold. Scale bar = 50 ,~m.
123
smooth muscle layers were labelled non-specifically within a _+5 mm radius of the site of application. This non-specific distribution was clearly distinct from that seen after injection of the tracer into the cranial mesenteric ganglion. Both FB and FG yielded similar results. Two to four weeks after injection, the tracer was diffused throughout the cranial mesenteric ganglion and labelled neurons could be detected in the m y e n t e r i c and o u t e r s u b m u c o u s plexuses throughout the entire colon (Fig. 1). The inner submucous plexus appeared devoid of stained nerve cells. Retrogradely labelled cells occurred mainly in ganglia lying to the side of the mesenteric attachment, situated between the two taeniae coli (Fig. 1). In the distal part of the colon, a few were found on the antimesenteric side of the wall. A faint FB labelling of fibers was visible in mesenteric nerves extending into the myenteric plexus and in nerves following the submucosal blood vessels. MAP 2 and iC8 immunostaining of wholemounts (Figs. 2, 3) revealed one, but sometimes two to four retrogradely labelled neurons in some of the ganglia in the myenteric (Fig. 2a, b) and outer submucous plexuses (Fig. 2d, e, f) that were situated near the mesenteric attachment. MAP2 immunolabelling, which enables simultaneous vi-
sualization of the two submucous ganglionic plexuses, confirmed that the inner submucous plexus was devoid of nerve cells projecting to the cranial mesenteric ganglion (Fig. 2c). The FB or FG accumulation in the retrogradely labelled neuronal perikarya and processes was reflected in a granular appearance. For the major part, the labelled cells appeared as large multidendritic neurons, both in the myenteric and in the outer submucous plexus (Figs. 3, 4). In the latter, however, a few small smoothcontoured labelled cells could be seen. Counts were done in four animals and ranged between 710 and 2339 labelled neurons for the myenteric plexus and between 135 and 765 labelled neurons for the outer submucous plexus (see Table II), with more cells present per unit length in the myenteric plexus. Table II shows that the density of retrogradely labelled neurons is at least twice as high in the distal part as in the proximal part of the colon. Only the values for the outer submucous plexus in animal 2 are slightly deviant. In both plexuses, the viscerofugal neurons were the most numerous in the region where the ascending colon passes into the transverse colon (Fig. 5). In the pig, the transverse and descending colons account for only 20-25% of the total colon length.
40
OUTER SUBMUCOUS PLEXUS
MYEHTERIC PLEXUS 30-
30-
10"
10-
O
~e Z
|
I
:,-,-
proxlm,I
~!:
rnic~b
d~.,
i
'~
i
ASCENDING COLON
-a
TRANSVERSE +
i o..... O,N ......
pl'oxlmll
i
--- ~-
mlddlo ~
ASCENDING COLON
UU=L
dhital
a.
] TP.ANSVERSE +
, ....... i.......
i
Fig. 5a, b. Percentage distribution of labelled cells along the colon in the myenteric (a) and outer submucous plexus (b). The total colon length, indicated along the X-axis, was divided into a proximal, middle and distal part. Each part was further subdivided into three segments of equal length. Values were obtained from four animals. The n u m b e r of segments counted (3.5 cm each) totalled 54 (see Table II). In both plexuses, the highest percentage is situated near the end of the ascending colon.
124 TABLE II
Mean number of labelled neurons / cm length in the proximal (P), middle (M) and distal (D) parts of the colon Myenteric plexus
Outer submucous plexus
Animal 1 (FG) P M D t n
10.7 4.5 23.6 1 129 21
(27.6%) (11.6%) (60.8%)
2.1 1.3 4.1 209 21
(28.0%) (17.3%) (54.7%)
Animal 2 (FB) P M D t n
9.1 10.0 20.8 710 13
(22.8%) (25.1%) (52.1%)
3.6 2.7 2.8 135 13
(39.5%) (29.7%) (30.8%)
Animal 3 (FB) P M D t n
45.7 41.3 101.8 2 339 10
(24.2%) (21.9%) (53.9%)
13.0 13.9 34.4 765 10
(21.2%) (22.7%) (56.1%)
Animal 4 (FB) P M D t n
3.5 21.5 35.7 763 10
(5.8%) (35.4%) (58.8%)
0.6 4.9 5.6 136 10
(5.4%) (44.1%) (50.5%)
t: total number of labelled neurons counted; n: number of counted segments measuring 3.5 cm each; (%): proportional distribution. Since the total colon was divided into segments of equal length, these percentage values actually represent the proportional distribution within the distinct colonic segments.
Immunocytochemical analysis of paraffin sections of the cranial mesenteric ganglion revealed that both 5-HT- and CGRP-like immunoreactive (IR) varicose nerve fibers encircled 5-HT- or CGRP-immunonegative prevertebral neuronal perikarya (Fig. 6a, b). Despite the large number of VIP-like IR nerve fibers in the cranial mesenteric ganglion (Fig. 6c), retrogradely labelled neurons of neither the colonic myenteric nor the outer submucous 'plexus stained for this neuropeptide, although VIP-like IR nerve cell bodies were demonstrated in these nerve plexuses. Further immunocytochemical studies established that a considerable proportion, viz. 30% (155 out of 518 cells) of the labelled neurons of the myenteric plexus and 35% (53 out of 152 cells) of those of the outer submucous plexus, were IR for C G R P
(Fig. 7a, b). Simultaneous immunostaining for iC8 and C G R P showed part of these retrogradely labelled CGRP-like IR neurons to have a multidendritic appearance (Fig. 7b, c). In addition, 20% (27 out of 136 cells) of the labelled neurons of the outer submucous plexus were IR for 5-HT (Fig. 8), but retrogradely labelled neurons in the myenteric plexus did not reveal 5-HT IR. The respective numbers of CGRP- and 5-HT-like IR neurons did not differ significantly in FB- or FG-iabelled cells. Neither 5-HT-like IR nerve fibers nor tracer could be detected in fibers of the caudal mesenteric ganglion.
Discussion
When the distribution of porcine enteric neurons projecting to the cranial mesenteric ganglion in the colon is compared to that in the small intestine [28] and to the distribution of guinea-pig enteric neurons projecting to the coeliac ganglion [16,21] it can be concluded that in both species these neurons are situated predominantly near the mesenteric attachment and are present throughout the entire small intestine and colon, with the highest density in the latter region. In the guinea-pig, labelled ceils projecting to the coeliac ganglion [16,21] and to the caudal mesenteric ganglion [7] were only observed in the myenteric plexus, whereas in the pig a considerable number of neurons projecting to the cranial mesenteric ganglion were found in the outer submucous plexus as well. The existence of submucous viscerofugal neurons has so far only been mentioned for the cat [5] and, although not confirmed by Furness and collaborators, for the guinea-pig by Dalsgaard and Elfvin [2]. The fact that porcine retrogradely labelled neurons occur in the outer but not in the inner submucous plexus strengthens our assertion that these plexuses have distinct functions [29]. In the guinea-pig, denervation experiments [3,17] combined with retrograde labelling techniques [21] indicated that pathways of neurons projecting from the distal colon to the coeliac and cranial mesenteric ganglian run via the caudal mesenteric ganglion and the intermesenteric
125
? R i":~
~J
' i;.
Fig. 6a-c: Immunostaining of paraffin sections of the cranial mesenteric ganglion visualizes 5-HT (a), CGRP (b) and VIP (c) IR varicose nerve fibers encircling immunonegative prevertebral neurons. Scale bar = 50/zm. Fig. 7a-c: FB-labelled myenteric neuron (a) double-immunostained for CGRP (b) and iC8 (c) having a multidendritic appearance (arrows). Scale bar = 50/xm. Fig. 8a, b: lmmunofluorescence micrographs illustrating a FB-labelled neuron to be IR for 5-HT in the outer submucous plexus (arrows). Scale bar = 50 p,m.
nerves, while projections of neurons from the proximal colon to the coeliac ganglion run via the ileo-caeco-colonic nerves [20]. The fact that in our study no tracer could be detected in fibers of the caudal mesenteric ganglion might indicate that neuronal projections from the descending colon to the cranial mesenteric ganglion do not run via the caudal mesenteric ganglion. Additionally, although immunolabelling in nerve processes may be reduced by colchicine treatment, the inability to visualize 5-HT in fibers of the latter ganglion, unlike its clear visualization in p~rrt of the retrogradely labelled neurons, may be a further indication for this assumption. Our experiments do document, however, that input from the duodenum to the descending colon converges in the cranial mesenteric ganglion and that the vis-
cerofugal pathway of the colon does not run parallel to the well-known sympathetic efferent pathway (see Furness and Costa [6]). The divergent numbers of labelled cells in the four animals cannot be ascribed to the tracer used. The transition of ascending to transverse colon is located at a distance of 20-25% from the end of the colon. Here, the largest number of retrogradely labelled ceils is found, after which point the number sharply decreases. It might be assumed that in this region viscerofugal neurons sendihg their axons to the caudal mesenteric ganglion outnumber by far those projecting to the cranial mesenteric ganglion. The increasing gradient evidenced from duodenum to ileum [28] is also observed in the ascending colon, at least for cells of the myenteric plexus. Contrary to the
126
situation in the porcine small intestine, the colon outer submucous plexus contains fewer labelled cells than the myenteric plexus [28]. Based on calculations of the total number of myenteric nerve cells per cm 2 in sheep and monkey gastrointestinal tract [8,23], which we assumed to be of the same order of magnitude as in the porcine intestine, we estimate that 0.05% of all myenteric neurons in the first part of the ascending colon and 0.7% in the region where the ascending turns into the transverse colon project to the cranial mesenteric ganglion. Combined retrograde tracing studies and immunocytochemistry in the guinea-pig have revealed VIp a n d calbindin in neurons projecting from the colon to the caudal mesenteric ganglion [7] and the coeli~ic ganglion [21]. In the pig, no VIP-like IR is apparent in labelled colonic cells, in spite of the fact t h a t many VIP-like IR fibers are present in t h e cranial mesenteric ganglion and nerve cell bodies are demonstrated in the plexuses. Denervation experiments in guinea-pig suggest that apart from VIP IR fibers, also dynorphin [4,31], CGRP, neurotensin [31], cholecystokinin and bombesin [3] IR fibers in the caudal mesenteric ganglion originate in the colon and that VIP, cholecystokinin [3,17], dynorphin, CGRP, bombesin and enkephalin [17] IR fibers in the cranial mesenteric ganglion may also have their cell bodies in the colon. In the pig colon, only C G R P and 5-HT have thus far been detected in a proportion of neurons projecting to the cranial mesenteric ganglion. In this respect the porcine colon clearly differs from the small intestine, where C G R P has almost exclusively been localized in adendritic multiaxonal type II neurons [25]. Although their origin is still to be confirmed by further denervation experiments, the presence of C G R P and 5-HT-like IR fibers in the cranial mesenteric ganglion of the pig is supportive of the immunocytochemical data which established that at least part of the porcine viscerofugal colonic neurons contain one of these substances as neurotransmitters or neuromodulators. In the guinea-pig caudal mesenteric ganglion, nerve fibers assumed to originate from primary sensory neurons of spinal dorsal root ganglia were seen to release C G R P during colonic
distension [19]. The fact that only one third of all C G R P IR nerve fibers appeared to be affected by capsaicin treatment [19] might be a further indication that in part C G R P IR nerve fibers do not originate from dorsal root ganglion cells but rather from viscerofugal colonic neurons. The finding that in the porcine colon the retrogradely labelled neurons outnumber by far the CGRPand 5-HT-like immunolabelled ones suggests that several subpopulations with distinct neurochemical features project from the colon to the cranial mesenteric ganglion.
Acknowledgements The authors wish to express their sincere thanks to F. Terloo, A. Hertog, D. De Ryck, G. Svensson and J. Van Daele for technical assistance. They also are indebted to Dr. C. Triban and Prof. H. Murofushi for their respective donation of the iC8 antiserum and the MAP2 antiserum. This study was supported by grant no. 3.0005.89 from the Fund for Medical Scientific Research to D.W.S. and by a grant from the National Bank of Belgium to D.W.S.; M.B. is a Research Assistant of the Special Fund for Research (SFO) from the University of Antwerp (UIA).
References 1 Crowcroft, P.J., Holman, M.E. and Szurszewski, J.H., Excitatory input from the distal colon to the inferior mesenteric ganglion in the guinea-pig, J. Physiol., 219 (1971) 443-461 2 Dalsgaard, C.-J. and Elfvin, L.-G., Structural studies on the connectivity of the inferior mesenteric ganglion of the guinea pig, J. Auton. Nerv. Syst., 5 (1982) 265-278. 3 Dalsgaard, C.-J., H6kfelt, T., Schultzberg, M., Lundberg, J.M., Terenius, L., Dockray, G.J. and Goldstein, M., Origin of peptide-containing fibers in the inferior mesenteric ganglion of the guinea-pig: immunohistochemical studies with antisera to substance P, enkephalin, vasoactive intestinal potypeptide, cholecystokinin and bombesin, Neuroscience, 9 (1983) 191-211. 4 Dalsgaard, C.-J., Vincent, S.R., H6kfelt, T., Christensson, I. and Terenius, L., Separate origins for the dynorphin and enkephalin immunoreactive fibers in the inferior mesen-
127 teric ganglion of the guinea pig, J. Comp. Neurol., 221 (1983) 482-489. 5 F6her, E. and Vajda, J., Evidence of sensory neurons in the wall of the small intestine revealed by horseradish peroxidase technique, Z. Mikrosk.-Anat. Forsch., 96 (1982) 2-8. 6 Furness, J.B. and Costa, M., The adrenergic innervation of the gastrointestinal tract, Ergebn. Physiol. Biol. Chem. Exp. Pharmacol., 69 (1974) 1-51. 7 Furness, J.B., Kuramoto, H. and Messenger, J.P., Morphological and chemical identification of neurons that project from the colon to the inferior mesenteric ganglia in the guinea-pig, J. Auton. Nerv. Syst., 31 (1990) 203-210. 8 Gabella, G., The number of neurons in the small intestine of mice, guinea-pigs and sheep, Neuroscience, 22 (1987) 737-752. 9 JulY, Y., Krier. J. and Szurszewski, J.H., Patterns of innervation of neurones in the inferior mesenteric ganglion of the cat, J. Physiol., 344 (1983) 293-304 10 Keef, K.D. and Kreulen, D.L., Peripheral nerve pathways to neurons in the guinea pig inferior mesenteric ganglion determined electrophysiologically after chronic nerve section, J. Auton. Nerv. Syst., 29 (1990) 113-128 11 King, B.F. and Szurszewski, J.H., Mechanoreceptor pathways from the distal colon to the autonomic nervous system in the guinea-pig, J. Physiol., 350 (1984) 93-107. 12 Kreulen, D.L. and Peters, S., Non-cholinergic transmission in a sympathetic ganglion of the guinea-pig elicited by colon distention, J. Physiol. 374 (1986) 315-334. 13 Kreulen, D.L. and Szurszewski, J.H., Reflex pathways in the abdominal prevertebral ganglia: evidence for a colocolonic inhibitory reflex, J. Physiol., 295 (1979) 21-32. 14 Kreulen, D.L. and Szurszewski, J.H., Nerve pathways in celiac plexus of the guinea pig, Am. J. Physiol., 237 (1979) E90-E97. 15 Kuntz, A. and Saccomanno, G., Reflex inhibition of intestinal motility mediated through decentralized prevertebral ganglia, J. Neurophysiol., 7 (1944) 163-170. 16 Kuramoto, H. and Furness, J.B., Distribution of enteric nerve cells that project from the small intestine to the coeliac ganglion in the guinea-pig, J. Auton. Nerv. Syst., 27 (1989) 241-248. 17 Lindh, B., H6kfelt. T. and Elfvin, L.-G., Distribution and origin of peptide-containing nerve fibers in the celiac superior mesenteric ganglion of the guinea-pig, Neuroscience, 26 (1988) 1037-1071. 18 Llewellyn-Smith, I.J., Costa, M. and Furness, J.B., Light and electron microscopic immunocytochemistry of the same nerves from whole mount preparations, J. Histochem. Cytochem., 33 (1985) 857-866. 19 Ma, R.C. and Szurszewski, J.H., Release of calcjtonin gene-related peptide in guinea pig inferior mesenteric
20
21
22
23
24
25
26 27
28
29
30
31
ganglion during colonic distention, J. Auton. Nerv. Syst., 33 (1991) 118-119. Messenger, J.P. and Furness, J.B., Calbindin-immunoreactive nerve terminals in the guinea pig coeliac ganglion originate from colonic nerve cells, J. Auton. Nerv. Syst., 35 (1991) 133-142. Messenger, J.P. and Furness, J.B., Distribution of enteric nerve cells that project to the coeliac ganglion of the guinea-pig, Cell Tissue Res., 269 (1992) 119-132. Murofushi, H., Suzuki, M., Sakai, H. and Kobayashi, S., Immunohistochemical localization of microtubule-associated proteins in the nervous system of the small intestine of guinea-pig, Cell Tissue Res., 255 (1989) 315-322. Ohkubo, K., Studien fiber das intramurale Nervensystem des Verdauungskanals. III. Affe und Mensch, Jap. J. Med. Sci., I. Anat., 6 (1936) 219-247. Pilipenko, V.I., Contributions to the functional morphology of the peripheral nervous system. Part II. Function nature of Dogiel's type II cells, Bull. Exp. Biol. Med., 4l (1956) 449-452. Scheuermann, D.W., Stach, W., De Groodt-Lasseel, MH.A. and Timmermans, J.-P., Calcitonin gene-related peptide in morphologically well-defined type II neurons of the enteric nervous system in the porcine small intestine, Acta Anat., 129 (1987) 325-328. Szurszewski, J.H., Physiology of mammalian prevertebral ganglia, A. Rev. Physiol., 43 (1981) 53-68. Szurszewski, J.H. and Weems, W.A., A study of peripheral input to and its control by post-ganglionic neurones of the inferior mesenteric ganglion, J. Physiol., 256 (1976) 541556. Timmermans, J.-P., Barbiers, M., Scheuermann, D.W., Stach, W., Adriaensen, D. and De Groodt-Lasseel, M.H.A., Occurrence, distribution and neurochemical features of small intestinal neurons projecting to the cranial mesenteric ganglion in the pig, Cell Tissue Res., 272 (1993) 49-58. Timmermans, J.-P., Scheuermann, D.W., Stach, W., Adriaensen, D. and De Groodt-Lasseel, M.H.A., Distinct distribution of CGRP-, enkephalin-, galanin-, neuromedin U-, neuropeptide Y-, somatostatin-, substance P-, VIPand serotonin-containing neurons in the t.wo submucosal ganglionic neural networks of the porcine small intestine, Cell Tissue Res., 260 (1990) 367-379. Vitadello, M., Triban, C., Fabris, M., Gorio, A. and Schiaffino, S., Heterogeneity of rat neurofilament polypeptides revealed by a monoclonal antibody, J. Neurochem., 46 (1986) 665-670. Webber, R.H. and Heym, Ch., Immunohistochemistry of biogenic polypeptides in nerve cells and fibres of the guinea pig inferior mesenteric ganglion after perturbations, Histochemistry, 88 (1988) 287-297.
119
© 1993 Elsevier Science Publishers B.V. All rights reserved 0165-1838/93/$06.00 JANS 01413
Topographical distribution and immunocytochemical features of colonic neurons that project to the cranial mesenteric ganglion in the pig Martine Barbiers, Jean-Pierre Timmermans, Dirk Adriaensen, Marie H.A. De Groodt-Lasseel and Dietrich W. Scheuermann Laboratory of Cell Biology and Histology, University of Antwerp, Antwerp, Belgium (Received 1 October 1992) (Revision received and accepted 15 February 1993)
Key words: Cranial mesenteric ganglion; Superior mesenteric ganglion; Retrograde tracing; Immunocytochemistry; Enteric nervous system; Colon; Pig Abstract Using the retrograde neuronal tracers Fast blue and Fluorogold, the topographical distribution and morphological features of porcine colonic neurons projecting to the cranial (superior) mesenteric ganglion have been investigated. Two to four weeks after injection of the tracer into the cranial mesenteric ganglion of immature pigs, labelled neurons were found throughout the colon. In the myenteric and outer submucous plexuses, they were present in ganglia situated to the side of the mesenteric attachment. The highest density of labelled neurons was observed at the end of the ascending colon, which in the pig represents 75-80% of the total colon length. The viscerofugal neurons had a multidendritic appearance and part of them were immunoreactive for calcitonin gene-related peptide or serotonin. This study has revealed similarities but also significant differences in the colono-sympathico-colonic pathways between the pig and small laboratory animals such as the guinea-pig.
Introduction
Anatomical studies and physiological experiments in the 1940s and 1950s have revealed that axons of ganglion cells located in the wall of the gastrointestinal tract extend into the prevertebral ganglia [15,24] and that reflex-inhibition is mediated through these ganglia, even after their decentralization [15]. To date, electrophysiological [1,9-12,27], immunocytochemical [3,31] and ret-
Correspondence to: D.W. Scheuermann, University of Antwerp, Laboratory of Cell Biology and Histology, Groenenborgerlaan 171, B-2020 Antwerp, Belgium.
rograde labelling techniques [2,7] have provided evidence that colonic neurons project to the caudal (inferior) mesenteric ganglion. Moreover, Kreulen and Szurszewski [13,14] demonstrated electrophysiologically that also the coeliac ganglion and cranial mesenteric ganglion receive mechanosensory input from oral and caudal regions of the colon and provided support for the existence of a colono-colonic inhibitory reflex that is mediated through pathways in prevertebral ganglia (see also Szurszewski [26]). By retrograde tracing experiments, Messenger and Furness [20,21] recently supported the presence of neurons projecting from the proximal and distal colon to the coeliac ganglion in guinea-pig.
120 Since it is known that the digestive system of the pig greatly resembles that of humans, this animal was chosen to demonstrate species differences between the enteric nervous system (ENS) of small laboratory animals and that of larger mammals. Immature pigs were preferred to mature ones because gut tissue of animals of 6 weeks of age could be optimally prepared and because previous observations had not revealed any major differences between the organization of the ENS of 6-week-old and that of 6-month-old animals. Tracing studies in the pig small intestine have shown an increase in the number of neurons projecting to the cranial mesenteric ganglion from the duodenum to the ileum [28]. The present investigation made use of the retrograde neuronal tracers Fast blue (FB) and Fluorogold (FG) to study projections from colonic neurons to the cranial mesenteric ganglion. Immunostaining for microtubule-associated protein (MAP2) and for a mixture of neurofilaments (iC8) was applied for
improved visualization of the morphology and topography of viscerofugal neurons within the ganglionic plexuses. Further immunohistochemical investigations with antisera raised against vasoactive intestinal polypeptide (VIP), calcitonin gene-related peptide (CGRP) and serotonin (5HT) were conducted to elucidate their neurochemical content.
Materials and Methods Eight just weaned, 21-23-day-old domestic pigs were anesthetized for surgical purposes with Stressnil (Janssen Pharmaceutica, Beerse, Belgium, 1 m l / 2 0 kg body weight, intramuscularly) and Hypnodil (Janssen Pharmaceutica, 1.5 m l / 2 0 kg body weight, intraperitoneally). Via a paracostal incision between the 13th and 14th rib the cranial mesenteric ganglion was exposed and different regions of the ganglion were injected with
TABLE I Detailed description of normal sera and antisera used for immunocytochemistry Normal sera
Normal goat serum (Dakopatts, Ghent, Belgium) Normal sheep serum (Nordic, Tilburg, The Netherlands) Primary antisera
Source
Host species
Dilution
MAP2 (microtubule-associated protein) iC8 (mixture of neurofilaments) CGRP (calcitonin gene-related peptide)
[see ref. 22] [see ref. 30] Amersham (Brussels, Belgium) Eurodiagnostics (Turnhout, Belgium) Incstar (Brussels, Belgium) Amersham UCB (Eigenbrakel, Belgium)
rabbit mouse rabbit
1 : 200 1 :'500 1 : 400
rabbit
1 : 50
goat
1 : 50
rabbit rabbit
1 : 200 1 : 500
5-HT (serotonin) 5-HT VIP (vasoactive intestinal polypeptide) VIP Secondary antisera and streptavidin complex
FITC-conjugated goat anti-rabbit IgG, diluted 1 : 20 (Nordic) Biotinylated sheep anti-mouse IgG, diluted 1 : 20 (Amersham) Biotinylated donkey anti-sheep IgG, diluted 1 : 20 (Amersham) Biotinylated goat anti-rabbit IgG, diluted 1 : 100 (Dakopatts) Streptavidin-Texas red complex, diluted 1 : 20 (Amersham) Streptavidin-HRP complex, diluted 1 : 100 (Amersham)
121 5 - 1 0 /xl FB (Sigma, Brussels, Belgium) (three pigs) or F G (Fluorochrome Inc., Englewood, Colorado, USA) (three pigs). In two control animals, F G was dripped on the cranial mesenteric ganglion or the serosal surface of the colon. The experiments were approved by the Ethical Medical Commission of the University of Antwerp and by the Ethical Commission of the National Fund for Scientific Research. Two to four weeks after surgery, the animals were intraperitoneally injected with colchicine (10 m g / k g body weight) and killed by decapitation 5 h later. The colon was divided into three segments of equal length, referred to as proximal, middle and distal parts. The colonic segments were ligated, filled with and immersed in 4% paraformaldehyde in 0.1 M sodium phosphate buffer containing 0.2% picric acid for 2 h at room temperature. To improve immunocytochemical conditions, the tissue was further processed as described previously [18,25]. Whole-mounts containing either the myenteric or submucous nerve plexuses were viewed for FB fluorescence in a Leitz MPV1 fluorescence microscope equipped with a Leitz Ploemopak filter block C (3 m m BG3 + S 405 excitation filter, T K 455 b e a m splitting mirror, K 460 suppression filter) and for FG fluorescence with a Leitz Ploem o p a k filter block A (BP340-380 excitation filter, R K P 400 b e a m splitting mirror, LP 430 suppression filter). Approximately every 15 cm, tissue pieces 3.5 cm in length each were investigated over the whole circumference of the colon and retrogradely labelled neurons were counted. The m e a n n u m b e r of labelled ceils per cm length was calculated for the proximal, middle and distal regions. In two of the three F G experiments, only the proximal part was examined for the presence of labelled neurons. These counts were not incorporated into the calculations for the overall distribution in the colon. For immunocytochemistry, whole-mounts were preincubated for 1 h in 10% normal serum containing 1% Triton X-100, and successively incubated in the primary (17 h) and corresponding secondary antiserum (6 h). All incubations were performed at room temperature. The fluorescent marker was conjugated either directly to the secondary antiserum or coupled via the biotin-streptavidin bridge system. More
information on the normal sera and antisera used is given in Table I. During dissection, the cranial mesenteric ganglion and the caudal mesenteric ganglion were also removed, fixed as described above and further processed for embedding in paraffin. Sections were checked for FB or F G fluorescence. For immunocytochemistry, the sections were successively incubated in the primary (17 h) and the corresponding secondary antiserum (6 h). A fluorescent marker was conjugated to the latter, or an indirect method involving the streptavidin-horseradish peroxidase complex was used (see Table I).
;E COLON
t" !
I:.'
r.'
F.15 |-.~
CTUM
IG COLON
I Fig. I. Schematic representation of the distribution of retro-
gradely labelled neurons in the porcine colon. The ascending colonic part has the shape of a double coiled cone. Labelled cells (dots) are present in the ascending, transverse and descending colons, and are localized near the mesenteric attachment (. . . . . . ) between the taenia coil ( - - - - - - ) . The number of labelled cells per unit length is the largest at the end of the ascending colon.
122
Results In the control animal where FG was dripped on the ganglionic surface, no labelled ceils were
found in the colon. In the other control where the tracer was dripped on the serosal surface of the colon, all nerve cells in the enteric nerve networks and the outer longitudinal and circular
e~
7,
!
E
:;2-)
i
!'.: iii
GL,
~
Fig. 2a-f. Simultaneous visualization of the tracer (a, f) and the immunostaining for iC8 (b) or MAP2 (c, d, e) reveals the localization of FG-labelled neurons within the ganglionic nerve plexuses. (a, h) Labelled neuron situated in the myenteric plexus (arrows). (e-f) Submucous layer photographed at different focal plains visualizing a retrogradely traced neuron (arrows) in the outer submucous plexus (double-headed arrows). In ganglia of the inner submucous plexus (arrowheads), no retrogradely labelled neurons are observed. (e, f) Enlargement of the area indicated in (d). Scale bar = 50 /zm. Fig. 3a, b: Immunolabelling for iC8 enables one to visualize the shape and size of retrogradely labelled neurons. (a) FB-labelled neuron in the myenteric plexus. (b) The same cell after iC8 immunocytochemistry, having a large, multi(long)dendritic appearance. Scale bar = 50 /xm. Fig. 4a, b: Fluorescence micrographs of unstained whole-mount preparations showing retrogradely labelled nerve cells in the myenteric plexus. A granular appearance of the tracer can be observed in the perikarya and in some processes of the multidendritic neurons. FB: Fast blue; FG: Fluorogold. Scale bar = 50 ,~m.
123
smooth muscle layers were labelled non-specifically within a _+5 mm radius of the site of application. This non-specific distribution was clearly distinct from that seen after injection of the tracer into the cranial mesenteric ganglion. Both FB and FG yielded similar results. Two to four weeks after injection, the tracer was diffused throughout the cranial mesenteric ganglion and labelled neurons could be detected in the m y e n t e r i c and o u t e r s u b m u c o u s plexuses throughout the entire colon (Fig. 1). The inner submucous plexus appeared devoid of stained nerve cells. Retrogradely labelled cells occurred mainly in ganglia lying to the side of the mesenteric attachment, situated between the two taeniae coli (Fig. 1). In the distal part of the colon, a few were found on the antimesenteric side of the wall. A faint FB labelling of fibers was visible in mesenteric nerves extending into the myenteric plexus and in nerves following the submucosal blood vessels. MAP 2 and iC8 immunostaining of wholemounts (Figs. 2, 3) revealed one, but sometimes two to four retrogradely labelled neurons in some of the ganglia in the myenteric (Fig. 2a, b) and outer submucous plexuses (Fig. 2d, e, f) that were situated near the mesenteric attachment. MAP2 immunolabelling, which enables simultaneous vi-
sualization of the two submucous ganglionic plexuses, confirmed that the inner submucous plexus was devoid of nerve cells projecting to the cranial mesenteric ganglion (Fig. 2c). The FB or FG accumulation in the retrogradely labelled neuronal perikarya and processes was reflected in a granular appearance. For the major part, the labelled cells appeared as large multidendritic neurons, both in the myenteric and in the outer submucous plexus (Figs. 3, 4). In the latter, however, a few small smoothcontoured labelled cells could be seen. Counts were done in four animals and ranged between 710 and 2339 labelled neurons for the myenteric plexus and between 135 and 765 labelled neurons for the outer submucous plexus (see Table II), with more cells present per unit length in the myenteric plexus. Table II shows that the density of retrogradely labelled neurons is at least twice as high in the distal part as in the proximal part of the colon. Only the values for the outer submucous plexus in animal 2 are slightly deviant. In both plexuses, the viscerofugal neurons were the most numerous in the region where the ascending colon passes into the transverse colon (Fig. 5). In the pig, the transverse and descending colons account for only 20-25% of the total colon length.
40
OUTER SUBMUCOUS PLEXUS
MYEHTERIC PLEXUS 30-
30-
10"
10-
O
~e Z
|
I
:,-,-
proxlm,I
~!:
rnic~b
d~.,
i
'~
i
ASCENDING COLON
-a
TRANSVERSE +
i o..... O,N ......
pl'oxlmll
i
--- ~-
mlddlo ~
ASCENDING COLON
UU=L
dhital
a.
] TP.ANSVERSE +
, ....... i.......
i
Fig. 5a, b. Percentage distribution of labelled cells along the colon in the myenteric (a) and outer submucous plexus (b). The total colon length, indicated along the X-axis, was divided into a proximal, middle and distal part. Each part was further subdivided into three segments of equal length. Values were obtained from four animals. The n u m b e r of segments counted (3.5 cm each) totalled 54 (see Table II). In both plexuses, the highest percentage is situated near the end of the ascending colon.
124 TABLE II
Mean number of labelled neurons / cm length in the proximal (P), middle (M) and distal (D) parts of the colon Myenteric plexus
Outer submucous plexus
Animal 1 (FG) P M D t n
10.7 4.5 23.6 1 129 21
(27.6%) (11.6%) (60.8%)
2.1 1.3 4.1 209 21
(28.0%) (17.3%) (54.7%)
Animal 2 (FB) P M D t n
9.1 10.0 20.8 710 13
(22.8%) (25.1%) (52.1%)
3.6 2.7 2.8 135 13
(39.5%) (29.7%) (30.8%)
Animal 3 (FB) P M D t n
45.7 41.3 101.8 2 339 10
(24.2%) (21.9%) (53.9%)
13.0 13.9 34.4 765 10
(21.2%) (22.7%) (56.1%)
Animal 4 (FB) P M D t n
3.5 21.5 35.7 763 10
(5.8%) (35.4%) (58.8%)
0.6 4.9 5.6 136 10
(5.4%) (44.1%) (50.5%)
t: total number of labelled neurons counted; n: number of counted segments measuring 3.5 cm each; (%): proportional distribution. Since the total colon was divided into segments of equal length, these percentage values actually represent the proportional distribution within the distinct colonic segments.
Immunocytochemical analysis of paraffin sections of the cranial mesenteric ganglion revealed that both 5-HT- and CGRP-like immunoreactive (IR) varicose nerve fibers encircled 5-HT- or CGRP-immunonegative prevertebral neuronal perikarya (Fig. 6a, b). Despite the large number of VIP-like IR nerve fibers in the cranial mesenteric ganglion (Fig. 6c), retrogradely labelled neurons of neither the colonic myenteric nor the outer submucous 'plexus stained for this neuropeptide, although VIP-like IR nerve cell bodies were demonstrated in these nerve plexuses. Further immunocytochemical studies established that a considerable proportion, viz. 30% (155 out of 518 cells) of the labelled neurons of the myenteric plexus and 35% (53 out of 152 cells) of those of the outer submucous plexus, were IR for C G R P
(Fig. 7a, b). Simultaneous immunostaining for iC8 and C G R P showed part of these retrogradely labelled CGRP-like IR neurons to have a multidendritic appearance (Fig. 7b, c). In addition, 20% (27 out of 136 cells) of the labelled neurons of the outer submucous plexus were IR for 5-HT (Fig. 8), but retrogradely labelled neurons in the myenteric plexus did not reveal 5-HT IR. The respective numbers of CGRP- and 5-HT-like IR neurons did not differ significantly in FB- or FG-iabelled cells. Neither 5-HT-like IR nerve fibers nor tracer could be detected in fibers of the caudal mesenteric ganglion.
Discussion
When the distribution of porcine enteric neurons projecting to the cranial mesenteric ganglion in the colon is compared to that in the small intestine [28] and to the distribution of guinea-pig enteric neurons projecting to the coeliac ganglion [16,21] it can be concluded that in both species these neurons are situated predominantly near the mesenteric attachment and are present throughout the entire small intestine and colon, with the highest density in the latter region. In the guinea-pig, labelled ceils projecting to the coeliac ganglion [16,21] and to the caudal mesenteric ganglion [7] were only observed in the myenteric plexus, whereas in the pig a considerable number of neurons projecting to the cranial mesenteric ganglion were found in the outer submucous plexus as well. The existence of submucous viscerofugal neurons has so far only been mentioned for the cat [5] and, although not confirmed by Furness and collaborators, for the guinea-pig by Dalsgaard and Elfvin [2]. The fact that porcine retrogradely labelled neurons occur in the outer but not in the inner submucous plexus strengthens our assertion that these plexuses have distinct functions [29]. In the guinea-pig, denervation experiments [3,17] combined with retrograde labelling techniques [21] indicated that pathways of neurons projecting from the distal colon to the coeliac and cranial mesenteric ganglian run via the caudal mesenteric ganglion and the intermesenteric
125
? R i":~
~J
' i;.
Fig. 6a-c: Immunostaining of paraffin sections of the cranial mesenteric ganglion visualizes 5-HT (a), CGRP (b) and VIP (c) IR varicose nerve fibers encircling immunonegative prevertebral neurons. Scale bar = 50/zm. Fig. 7a-c: FB-labelled myenteric neuron (a) double-immunostained for CGRP (b) and iC8 (c) having a multidendritic appearance (arrows). Scale bar = 50/xm. Fig. 8a, b: lmmunofluorescence micrographs illustrating a FB-labelled neuron to be IR for 5-HT in the outer submucous plexus (arrows). Scale bar = 50 p,m.
nerves, while projections of neurons from the proximal colon to the coeliac ganglion run via the ileo-caeco-colonic nerves [20]. The fact that in our study no tracer could be detected in fibers of the caudal mesenteric ganglion might indicate that neuronal projections from the descending colon to the cranial mesenteric ganglion do not run via the caudal mesenteric ganglion. Additionally, although immunolabelling in nerve processes may be reduced by colchicine treatment, the inability to visualize 5-HT in fibers of the latter ganglion, unlike its clear visualization in p~rrt of the retrogradely labelled neurons, may be a further indication for this assumption. Our experiments do document, however, that input from the duodenum to the descending colon converges in the cranial mesenteric ganglion and that the vis-
cerofugal pathway of the colon does not run parallel to the well-known sympathetic efferent pathway (see Furness and Costa [6]). The divergent numbers of labelled cells in the four animals cannot be ascribed to the tracer used. The transition of ascending to transverse colon is located at a distance of 20-25% from the end of the colon. Here, the largest number of retrogradely labelled ceils is found, after which point the number sharply decreases. It might be assumed that in this region viscerofugal neurons sendihg their axons to the caudal mesenteric ganglion outnumber by far those projecting to the cranial mesenteric ganglion. The increasing gradient evidenced from duodenum to ileum [28] is also observed in the ascending colon, at least for cells of the myenteric plexus. Contrary to the
126
situation in the porcine small intestine, the colon outer submucous plexus contains fewer labelled cells than the myenteric plexus [28]. Based on calculations of the total number of myenteric nerve cells per cm 2 in sheep and monkey gastrointestinal tract [8,23], which we assumed to be of the same order of magnitude as in the porcine intestine, we estimate that 0.05% of all myenteric neurons in the first part of the ascending colon and 0.7% in the region where the ascending turns into the transverse colon project to the cranial mesenteric ganglion. Combined retrograde tracing studies and immunocytochemistry in the guinea-pig have revealed VIp a n d calbindin in neurons projecting from the colon to the caudal mesenteric ganglion [7] and the coeli~ic ganglion [21]. In the pig, no VIP-like IR is apparent in labelled colonic cells, in spite of the fact t h a t many VIP-like IR fibers are present in t h e cranial mesenteric ganglion and nerve cell bodies are demonstrated in the plexuses. Denervation experiments in guinea-pig suggest that apart from VIP IR fibers, also dynorphin [4,31], CGRP, neurotensin [31], cholecystokinin and bombesin [3] IR fibers in the caudal mesenteric ganglion originate in the colon and that VIP, cholecystokinin [3,17], dynorphin, CGRP, bombesin and enkephalin [17] IR fibers in the cranial mesenteric ganglion may also have their cell bodies in the colon. In the pig colon, only C G R P and 5-HT have thus far been detected in a proportion of neurons projecting to the cranial mesenteric ganglion. In this respect the porcine colon clearly differs from the small intestine, where C G R P has almost exclusively been localized in adendritic multiaxonal type II neurons [25]. Although their origin is still to be confirmed by further denervation experiments, the presence of C G R P and 5-HT-like IR fibers in the cranial mesenteric ganglion of the pig is supportive of the immunocytochemical data which established that at least part of the porcine viscerofugal colonic neurons contain one of these substances as neurotransmitters or neuromodulators. In the guinea-pig caudal mesenteric ganglion, nerve fibers assumed to originate from primary sensory neurons of spinal dorsal root ganglia were seen to release C G R P during colonic
distension [19]. The fact that only one third of all C G R P IR nerve fibers appeared to be affected by capsaicin treatment [19] might be a further indication that in part C G R P IR nerve fibers do not originate from dorsal root ganglion cells but rather from viscerofugal colonic neurons. The finding that in the porcine colon the retrogradely labelled neurons outnumber by far the CGRPand 5-HT-like immunolabelled ones suggests that several subpopulations with distinct neurochemical features project from the colon to the cranial mesenteric ganglion.
Acknowledgements The authors wish to express their sincere thanks to F. Terloo, A. Hertog, D. De Ryck, G. Svensson and J. Van Daele for technical assistance. They also are indebted to Dr. C. Triban and Prof. H. Murofushi for their respective donation of the iC8 antiserum and the MAP2 antiserum. This study was supported by grant no. 3.0005.89 from the Fund for Medical Scientific Research to D.W.S. and by a grant from the National Bank of Belgium to D.W.S.; M.B. is a Research Assistant of the Special Fund for Research (SFO) from the University of Antwerp (UIA).
References 1 Crowcroft, P.J., Holman, M.E. and Szurszewski, J.H., Excitatory input from the distal colon to the inferior mesenteric ganglion in the guinea-pig, J. Physiol., 219 (1971) 443-461 2 Dalsgaard, C.-J. and Elfvin, L.-G., Structural studies on the connectivity of the inferior mesenteric ganglion of the guinea pig, J. Auton. Nerv. Syst., 5 (1982) 265-278. 3 Dalsgaard, C.-J., H6kfelt, T., Schultzberg, M., Lundberg, J.M., Terenius, L., Dockray, G.J. and Goldstein, M., Origin of peptide-containing fibers in the inferior mesenteric ganglion of the guinea-pig: immunohistochemical studies with antisera to substance P, enkephalin, vasoactive intestinal potypeptide, cholecystokinin and bombesin, Neuroscience, 9 (1983) 191-211. 4 Dalsgaard, C.-J., Vincent, S.R., H6kfelt, T., Christensson, I. and Terenius, L., Separate origins for the dynorphin and enkephalin immunoreactive fibers in the inferior mesen-
127 teric ganglion of the guinea pig, J. Comp. Neurol., 221 (1983) 482-489. 5 F6her, E. and Vajda, J., Evidence of sensory neurons in the wall of the small intestine revealed by horseradish peroxidase technique, Z. Mikrosk.-Anat. Forsch., 96 (1982) 2-8. 6 Furness, J.B. and Costa, M., The adrenergic innervation of the gastrointestinal tract, Ergebn. Physiol. Biol. Chem. Exp. Pharmacol., 69 (1974) 1-51. 7 Furness, J.B., Kuramoto, H. and Messenger, J.P., Morphological and chemical identification of neurons that project from the colon to the inferior mesenteric ganglia in the guinea-pig, J. Auton. Nerv. Syst., 31 (1990) 203-210. 8 Gabella, G., The number of neurons in the small intestine of mice, guinea-pigs and sheep, Neuroscience, 22 (1987) 737-752. 9 JulY, Y., Krier. J. and Szurszewski, J.H., Patterns of innervation of neurones in the inferior mesenteric ganglion of the cat, J. Physiol., 344 (1983) 293-304 10 Keef, K.D. and Kreulen, D.L., Peripheral nerve pathways to neurons in the guinea pig inferior mesenteric ganglion determined electrophysiologically after chronic nerve section, J. Auton. Nerv. Syst., 29 (1990) 113-128 11 King, B.F. and Szurszewski, J.H., Mechanoreceptor pathways from the distal colon to the autonomic nervous system in the guinea-pig, J. Physiol., 350 (1984) 93-107. 12 Kreulen, D.L. and Peters, S., Non-cholinergic transmission in a sympathetic ganglion of the guinea-pig elicited by colon distention, J. Physiol. 374 (1986) 315-334. 13 Kreulen, D.L. and Szurszewski, J.H., Reflex pathways in the abdominal prevertebral ganglia: evidence for a colocolonic inhibitory reflex, J. Physiol., 295 (1979) 21-32. 14 Kreulen, D.L. and Szurszewski, J.H., Nerve pathways in celiac plexus of the guinea pig, Am. J. Physiol., 237 (1979) E90-E97. 15 Kuntz, A. and Saccomanno, G., Reflex inhibition of intestinal motility mediated through decentralized prevertebral ganglia, J. Neurophysiol., 7 (1944) 163-170. 16 Kuramoto, H. and Furness, J.B., Distribution of enteric nerve cells that project from the small intestine to the coeliac ganglion in the guinea-pig, J. Auton. Nerv. Syst., 27 (1989) 241-248. 17 Lindh, B., H6kfelt. T. and Elfvin, L.-G., Distribution and origin of peptide-containing nerve fibers in the celiac superior mesenteric ganglion of the guinea-pig, Neuroscience, 26 (1988) 1037-1071. 18 Llewellyn-Smith, I.J., Costa, M. and Furness, J.B., Light and electron microscopic immunocytochemistry of the same nerves from whole mount preparations, J. Histochem. Cytochem., 33 (1985) 857-866. 19 Ma, R.C. and Szurszewski, J.H., Release of calcjtonin gene-related peptide in guinea pig inferior mesenteric
20
21
22
23
24
25
26 27
28
29
30
31
ganglion during colonic distention, J. Auton. Nerv. Syst., 33 (1991) 118-119. Messenger, J.P. and Furness, J.B., Calbindin-immunoreactive nerve terminals in the guinea pig coeliac ganglion originate from colonic nerve cells, J. Auton. Nerv. Syst., 35 (1991) 133-142. Messenger, J.P. and Furness, J.B., Distribution of enteric nerve cells that project to the coeliac ganglion of the guinea-pig, Cell Tissue Res., 269 (1992) 119-132. Murofushi, H., Suzuki, M., Sakai, H. and Kobayashi, S., Immunohistochemical localization of microtubule-associated proteins in the nervous system of the small intestine of guinea-pig, Cell Tissue Res., 255 (1989) 315-322. Ohkubo, K., Studien fiber das intramurale Nervensystem des Verdauungskanals. III. Affe und Mensch, Jap. J. Med. Sci., I. Anat., 6 (1936) 219-247. Pilipenko, V.I., Contributions to the functional morphology of the peripheral nervous system. Part II. Function nature of Dogiel's type II cells, Bull. Exp. Biol. Med., 4l (1956) 449-452. Scheuermann, D.W., Stach, W., De Groodt-Lasseel, MH.A. and Timmermans, J.-P., Calcitonin gene-related peptide in morphologically well-defined type II neurons of the enteric nervous system in the porcine small intestine, Acta Anat., 129 (1987) 325-328. Szurszewski, J.H., Physiology of mammalian prevertebral ganglia, A. Rev. Physiol., 43 (1981) 53-68. Szurszewski, J.H. and Weems, W.A., A study of peripheral input to and its control by post-ganglionic neurones of the inferior mesenteric ganglion, J. Physiol., 256 (1976) 541556. Timmermans, J.-P., Barbiers, M., Scheuermann, D.W., Stach, W., Adriaensen, D. and De Groodt-Lasseel, M.H.A., Occurrence, distribution and neurochemical features of small intestinal neurons projecting to the cranial mesenteric ganglion in the pig, Cell Tissue Res., 272 (1993) 49-58. Timmermans, J.-P., Scheuermann, D.W., Stach, W., Adriaensen, D. and De Groodt-Lasseel, M.H.A., Distinct distribution of CGRP-, enkephalin-, galanin-, neuromedin U-, neuropeptide Y-, somatostatin-, substance P-, VIPand serotonin-containing neurons in the t.wo submucosal ganglionic neural networks of the porcine small intestine, Cell Tissue Res., 260 (1990) 367-379. Vitadello, M., Triban, C., Fabris, M., Gorio, A. and Schiaffino, S., Heterogeneity of rat neurofilament polypeptides revealed by a monoclonal antibody, J. Neurochem., 46 (1986) 665-670. Webber, R.H. and Heym, Ch., Immunohistochemistry of biogenic polypeptides in nerve cells and fibres of the guinea pig inferior mesenteric ganglion after perturbations, Histochemistry, 88 (1988) 287-297.