Neurovascular Interface in Porcine Small Intestine: Specific for Nitrergic rather than Nonnitrergic Neurons.

In the 1970s, by using classic histological methods, close topographical relationships between special areas of enteric ganglia and capillaries were shown in the pig. In this study, by application of double and triple immunohistochemistry, we confirmed this neurovascular interface and demonstrated that these zones are mainly confined to nitrergic neurons in the myenteric and the external submucosal plexus. In the upper small intestine of the pig, the respective neurons display type III morphology, i.e. they have long, slender and branched dendrites and a single axon. In another set of experiments, we prepared specimens for electron-microscopical analysis of these zones. Both ganglia and capillaries display continuous basement membranes, the smallest distances between them being 1,000 nm at the myenteric and 300 nm at the external submucosal level. The capillary endothelium was mostly continuous but, at the external submucosal level, scattered fenestrations were observed. This particular neurovascular relationship suggests that nitrergic neurons may require a greater amount of oxygen and/or nutrients. In guinea pig and mouse, previous ischemia/reperfusion experiments showed that nitrergic neurons are selectively damaged. Thus, a preferential blood supply of enteric nitrergic neurons may indicate that these neurons are more vulnerable in ischemia.

Ultrastructural characterization of interstitial cells of Cajal associated with the submucosal plexus in the proximal colon of the guinea pig.

Interstitial cells of Cajal (ICC) associated with the submucosal (submucous) plexus (ICC-SP) in the proximal colon of the guinea pig were studied by immunohistochemistry and electron microscopy. Whole-mount stretch preparations with c-Kit immunohistochemistry revealed that a number of ICC-SP constituted a dense cellular network around the submucosal plexus. Some of these ICC-SP were observed in the vicinity of the muscularis mucosae in sections immunostained for c-Kit and α-smooth muscle actin. Ultrastructural observation demonstrated, for the first time, that ICC-SP of the proximal colon of the guinea pig retained typical ultrastructural characteristics of ICC repeatedly reported in association with the tunica muscularis of the gastrointestinal tract: a basal lamina, caveolae, many mitochondria, abundant intermediate filaments and the formation of gap junctions with the same type of cells. The most remarkable ultrastructural finding was the presence of thick bundles composed of the processes of ICC-SP connected to each other via large gap junctions. These ICC-SP might be involved in the main mucosal functions of the proximal colon of the guinea pig, namely the transportation of water and electrolytes, possibly via their involvement in the spontaneous contractions of the muscularis mucosae.

Three-dimensional demonstration of the interstitial cells of Cajal associated with the submucosal plexus in guinea-pig caecum.

The guinea-pig caecum was studied by using immunohistochemistry for Kit receptors and nerves to clarify whether interstitial cells of Cajal (ICC) were localized in association with the submucosal plexus (ICC-SP). A large area of the guinea-pig caecum was nearly devoid of longitudinal muscles, because they were concentrated into three bundles of the taenia caeci (coli) and this allowed clear observation of the myenteric and submucosal plexus as separate networks in whole-mount stretch preparations. The myenteric plexus was observed as a loose polygonal network consisting in elongated ganglia and long connecting nerve strands, whereas the submucosal plexus was identified as smaller ovoid ganglia connected to much thinner nerve strands in different tissue layers. Three-dimensional reconstruction of confocal images revealed multipolar-shaped ICC-SP located around the submucosal ganglion in a basket formation. Bipolar ICC-SP were also observed along the connecting nerve strands of the submucosal plexus. The functional involvement of ICC-SP in mucosal activity is discussed in relation to fluid transportation. This three-dimensional study of ICC-SP thus provides a candidate for the most suitable material available for functional experiments examining the physiological significance of ICC-SP.

Morphoquantitative study of the submucous plexus (of Meissner) of the jejunum-ileum of young and old guinea pigs.

OBJECTIVE: To study the aging of submucous plexus of the small intestine (jejunum-ileum) of the guinea pigs from the quantitative, structural and ultrastructural perspective. METHOD: Chemical preparations of membrane of the jejunum-ileum of old and young animals with the use of light and electronic microscope. RESULTS: The ganglia of young animals presented between 1 and 56 neurons and the old animals presented from 1 to 30 neurons. The mean density of the ganglia by cm(2) in the young jejunum-ileum was of 551±36.89 and in the old one 413±11.86. The density of the neurons was 5011±291.11 neurons/cm(2) average in young animals and 2918±120.70 neurons/cm(2) in the old ones. The size of the neurons varied in both age groups. The collagen fibers in the ganglia of old animals they were condensed. Degenerated mitochondrias in the interior of the cell were frequent in the old animals. CONCLUSION: In submucous plexus of the jejunum-ileum there is a loss of 38% of the neurons with aging.

Morphoquantitative study of the submucous plexus (of Meissner) of the jejunum-ileum of young and old guinea pigs / Estudo morfoquantitativo do plexo submucoso (de Meissner) do jejuno-íleo de cobaias jovens e velhas

OBJECTIVE: To study the aging of submucous plexus of the small intestine (jejunum-ileum) of the guinea pigs from the quantitative, structural and ultrastructural perspective. METHOD: Chemical preparations of membrane of the jejunum-ileum of old and young animals with the use of light and electronic microscope. RESULTS: The ganglia of young animals presented between 1 and 56 neurons and the old animals presented from 1 to 30 neurons. The mean density of the ganglia by cm² in the young jejunum-ileum was of 551±36.89 and in the old one 413±11.86. The density of the neurons was 5011±291.11 neurons/cm² average in young animals and 2918±120.70 neurons/cm² in the old ones. The size of the neurons varied in both age groups. The collagen fibers in the ganglia of old animals they were condensed. Degenerated mitochondrias in the interior of the cell were frequent in the old animals. CONCLUSION: In submucous plexus of the jejunum-ileum there is a loss of 38 percent of the neurons with aging.

OBJETIVO: Estudar o envelhecimento do plexo submucoso do intestino delgado (jejuno-íleo) das cobaias do ponto de vista quantitativo, estrutural e ultra-estrutural. MÉTODO: Preparados de membrana do jejuno-íleo de animais jovens e velhos com a utilização de microscopia de luz e eletrônica. RESULTADOS: Os gânglios de animais jovens apresentaram entre 1 e 56 neurônios e os animais velhos apresentaram de 1 a 30 neurônios. A densidade média dos gânglios por cm² no jejuno-íleo jovem foi de 551±36,89 e no velho foi de 413±11,86. A densidade dos neurônios foi de 5011±291,11 neurônios/cm² em média nos animais jovens e 2918±120,70 neurônios/cm² nos velhos. O tamanho dos neurônios variou em ambos os grupos etários. As fibras colágenas nos gânglios de animais velhos estavam mais condensadas. Mitocôndrias degeneradas no interior da célula foram freqüentes nos animais velhos. CONCLUSÃO: No plexo submucoso do jejuno-íleo há uma perda de 38 por cento dos neurônios com o envelhecimento.

Immunohistochemical analysis of neuron types in the mouse small intestine.

The definition of the nerve cell types of the myenteric plexus of the mouse small intestine has become important, as more researchers turn to the use of mice with genetic mutations to analyze roles of specific genes and their products in enteric nervous system function and to investigate animal models of disease. We have used a suite of antibodies to define neurons by their shapes, sizes, and neurochemistry in the myenteric plexus. Anti-Hu antibodies were used to reveal all nerve cells, and the major subpopulations were defined in relation to the Hu-positive neurons. Morphological Type II neurons, revealed by anti-neurofilament and anti-calcitonin gene-related peptide antibodies, represented 26% of neurons. The axons of the Type II neurons projected through the circular muscle and submucosa to the mucosa. The cell bodies were immunoreactive for choline acetyltransferase (ChAT), and their terminals were immunoreactive for vesicular acetylcholine transporter (VAChT). Nitric oxide synthase (NOS) occurred in 29% of nerve cells. Most were also immunoreactive for vasoactive intestinal peptide, but they were not tachykinin (TK)-immunoreactive, and only 10% were ChAT-immunoreactive. Numerous NOS terminals occurred in the circular muscle. We deduced that 90% of NOS neurons were inhibitory motor neurons to the muscle (26% of all neurons) and 10% (3% of all neurons) were interneurons. Calretinin immunoreactivity was found in a high proportion of neurons (52%). Many of these had TK immunoreactivity. Small calretinin neurons were identified as excitatory neurons to the longitudinal muscle (about 20% of neurons, with ChAT/calretinin/+/- TK chemical coding). Excitatory neurons to the circular muscle (about 10% of neurons) had the same coding. Calretinin immunoreactivity also occurred in a proportion of Type II neurons. Thus, over 90% of neurons in the myenteric plexus of the mouse small intestine can be currently identified by their neurochemistry and shape.

Increase in stretch-induced rhythmic motor activity in the diabetic rat colon is associated with loss of ICC of the submuscular plexus.

Diabetes affects many aspects of gastrointestinal motility, in part due to changes in interstitial cells of Cajal (ICC). The effect of diabetes on the colon, however, is not well characterized, and the aim of the present study was to investigate possible relationships between altered colonic motility as a consequence of streptozotocin-induced diabetes and injury to ICC. Physiological, immunohistochemical, and ultrastructural techniques were employed. The motor pattern of the rat colon was dominated by rhythmic high-amplitude, low-frequency contractions that were primarily myogenic in origin. These rhythmic contractions were induced by stretch associated with increased tension; the amplitude of the superimposed rhythmic contractions increased with increasing applied tension. In diabetic rats, the stretch-induced rhythmic contractile activity remained robust and of similar frequency but was significantly higher in amplitude compared with that in control rats. At 700 mg of applied tension, the force of contraction in circular colonic muscle strips of the diabetic rats was 370% of control values. This robust presence of low-frequency contractions is consistent with the unaffected pacemaker, the ICC associated with Auerbach's plexus, and the increased amplitude correlates with loss of and injury to ICC of the submuscular plexus and intramuscular ICC. Loss of inhibitory nitrergic nerves does not appear to be a factor based on unaltered nNOS immunoreactivity.

Distribution and characterization of vanilloid receptors in the rat stomach.

The cloned vanilloid receptor-1 (VR1) is recognized as a common molecular target for protons, noxious heat, and vanilloids. The presence of VR1 in the dorsal root, trigeminal, and nodose ganglia has been firmly established, but it is unclear in the gut, despite this VR1 may be important for gastric mucosal homeostasis. In this study we used an antibody and a radioligand to show the distribution of vanilloid receptors (VRs) in rat stomach and to characterize it. The deafferentiation of capsaicin-sensitive nerves in rats was induced by consecutive injections of capsaicin. VR1-immunopositive nerve endings were predominantly found in the mucous neck cells of the proliferation zone, and around blood vessels in the submucosa. Radioreceptor assay using [3H]-resiniferatoxin (RTX) revealed the existence of high affinity and single-class binding site in the membrane fractions of the mucosa. Capsaicin completely inhibited the specific binding of [3H]-RTX. Both the VR1 immunoreactivity and the receptor density of [3H]-RTX binding sites significantly reduced by the application of capsaicin for prolonged periods of time in the mucosa of rats. Our results indicate that VRs are expressed in the rat stomach, and suggest that they may be involved in mucosal protection by increasing cell proliferation and blood flow.

Neurotransmitters regulating acid secretion in the proventriculus of the Houbara bustard (Chlamydotis undulata): a morphological viewpoint.

Endocrine cells containing somatostatin (Som), gastrin-releasing peptide (GRP), and neuronal nitric oxide synthase (nNOS) and nerve fibers containing choline acetyl transferase (ChAT), tyrosine hydroxylase (TH), galanin (Gal), substance P (SP), and vasoactive intestinal polypeptide (VIP) were immunolocalized in the proventriculus of the Houbara bustard, Chlamydotis undulata. While GRP-immunoreactive (GRP-IR) cells occur in the inner zone, somatostatin (Som-IR) and polyclonal nNOS (nNOS-IR) immunoreactive cells were localized mainly in the peripheral zone of submucosal glands. GRP-IR, Som-IR, and nNOS-IR cells were occasionally observed in the walls of the gastric glands. Endocrine cells are of the closed variety and usually possess apical processes extending along the basal surfaces of adjacent nonreactive cells. Ultrastructural features of these cells are typical. ChAT, Gal, SP, VIP, and TH were immunolocalized in nerve fibers and terminals in the walls of arterioles and capillaries at the periphery of submucosal glands. Immunoreactivity to monoclonal nNOS occurred mainly in neuronal cell bodies in ganglia located around the submucosal glands. ChAT and TH immunoreactive cell bodies were also occasionally seen around the submucosal glands in the peripheral region. Immunoreactivity to Gal, SP, and VIP, but not ChAT or TH, was discernible around the walls of gastric glands. It was concluded that the distribution of neurotransmitters in neuronal structures is similar, but that of the endocrine cells varies from that of some avian species. The roles of these neurotransmitters in the regulation of acid secretion are discussed.

A new hypothesis on the pathogenesis of intestinal pseudo-obstruction by intestinal neuronal dysplasia (IND).

Using morphometry and image analysis, we investigated 17 patients showing intestinal pseudo-obstruction secondary to intestinal neuronal dysplasia (IND) and 20 controls. In addition to an increase in the number and size of the ganglia and the ganglionic cells, we also noted a significant increase in NO synthase-containing ganglionic cells. We found that this enzyme, responsible for the synthesis of nitrous oxide caused by oxidation of L-argynine aminoacid, is a neurotransmitter able to induce smooth muscle relaxation by activating cyclic AMP. If the increase in NO synthase-producing ganglionic cells is responsible for the increase in nitrous oxide production, one can hypothesize that an overproduction of nitrous oxide plays a role in the pathogenesis of intestinal pseudo-obstruction secondary to neuronal dysplasia. As NO synthase can be blocked, as was demonstrated by giving L-methil-arginine or N-G-nitro-L-argynine, it might be possible to treat intestinal pseudo-obstruction caused by intestinal neuronal dysplasia at the pharmacological level.

Ultrastructural study of relationships between c-kit immunoreactive interstitial cells and other cellular elements in the human colon.

C-kit immunocytochemistry was performed on ultrathin sections of human distal colon. Our attention was focused on relationships between c-kit immunoreactive interstitial cells (c-kit ICs) and muscular cells and nervous elements located in the external muscular layers of the colonic wall. C-kit ICs established membrane apposition with both nerve fibers and smooth muscle cells of, respectively, the longitudinal and circular muscle layers, the myenteric area, and the extremus submucosus plexus. C-kit ICs also surrounded the external submucosus plexus and established membrane appositions with nerve elements located inside the myenteric ganglia. These membrane appositions were observed either at the level of the c-kit IC bodies or at that of their cytoplasmic processes. In some cases, membrane appositions were observed concomitantly between the c-kit ICs, nerve fibers, and smooth muscle cells. In all the regions studied, the c-kit ICs were also found to be located in the close vicinity of blood vessels and to have established close contacts with non-immunoreactive fibroblast-like cells. The results of the present study shed essential light on the relationships of c-kit ICs with the neighboring muscle cells and nerve elements, and confirm that the intercalated c-kit ICs well fit with the so-called "interstitial cells of Cajal".

The nervous system of the chicken proventriculus: an immunocytochemical and ultrastructural study.

The proventriculus constitutes the glandular region of the chicken stomach. This organ is innervated by two parasympathetic networks, the myenteric and submucous plexus, and here we present a systematic study of this system by immunohistochemistry and electron microscopy. All the neurons and fibres were positive for the neural markers, protein gene product 9.5 and the amidating enzymes. Immunoreactivities for the constitutive neuronal isoform of the enzyme nitric oxide synthase and the vasoactive intestinal peptide were present in neuronal bodies suggesting an intrinsic origin for the similarly immunoreactive fibres found in the proventriculus. On the other hand, immunoreactivity to gastric inhibitory peptide was only found in varicose fibres making contact with the blood vessels and the glandular epithelium, but never in the neuronal somas, suggesting that this substance may be provided by an extrinsic nervous system whose neuronal bodies are located elsewhere. Electron microscopy revealed frequent neuromuscular and neuroepithelial connections in the muscle layers, the wall of the blood vessels and the epithelium. In addition, synapsis-like structures were identified in the proximity of cells belonging to the diffuse endocrine system, providing a new example of neuroendocrine contacts. No positivity was found for antibodies against other neural substances including somatostatin, peptide histidine-isoleucine, peptide tyrosine-tyrosine, neuropeptide tyrosine, bombesin, met-enkephalin, serotonin, substance P, galanin, calcitonin gene-related peptide and S-100 protein.

Neuronal distribution and subcellular localization of HCNP-like immunoreactivity in rat small intestine.

A novel peptide, hippocampal cholinergic neurostimulating peptide (HCNP), originally purified from young rat hippocampus, affects the development of specific cholinergic neurons of the central nervous system in vitro. In this study, HCNP-like-immunoreactive nerve processes and nerve cell bodies were identified by electron microscopic immunocytochemistry in the rat small intestine. Labeled nerve processes were numerous in the circular muscle layer and around the submucosal blood vessels. In the submucosal and myenteric plexuses, some HCNP-like-immunopositive nerve cell bodies and nerve fibers were present. The reaction product was deposited on the membranes of various subcellular organelles, including the rough endoplasmic reticulum, Golgi saccules, ovoid electron-lucent synaptic vesicles in axon terminals associated with submucosal and myenteric plexuses, and the outer membranes of a few mitochondria. The synaptic vesicles of HCNP-like positive terminals were 60-85 nm in diameter. The present data provide immunocytochemical evidence that HCNP-like-positive nerve cell bodies and nerve fibers are present in the submucosal and myenteric plexuses of the rat small intestine. An immunohistochemical light microscopic study using mirror-image sections revealed that in both the submucosal and myenteric ganglia, almost all choline acetyltransferase (ChAT)-immunoreactive neurons were also immunoreactive for HCNP. These observations suggest (i) that HCNP proper and/or HCNP precursor protein is a membrane-associated protein with a widespread subcellular distribution, (ii) that HCNP precursor protein may be biosynthesized within neurons localized in the rat enteric nervous system, and (iii) that HCNP proper and/or HCNP precursor protein are probably stored in axon terminals.

Sensory nerve endings in the hard palate and papilla incisiva of the rhesus monkey.

The sensory innervation of the hard palate of the rhesus monkey was studied by light and electron microscopy. The mucosa of the hard palate is subject to a particularly heavy mechanical load requiring functional specialisation of the horny epithelium in the form of thickenings - the papilla incisiva and eight pairs of rugae palatinae. A thin layer of firm connective tissue (lamina propria) attaches the mucosa to the periost of the hard palate. Sensory nerve fibres were found most abundantly in the papilla incisiva and first rugae palatinae. Their number decreases in an aboral direction. Five types of sensory nerve endings were found. Free nerve endings were ubiquitous in the epithelium and lamina propria. Merkel nerve endings were found in the basal layer of the epithelium of the papilla incisiva and rugae palatinae. Meissner corpuscles were located in the connective tissue between epithelial pegs, while lamellated corpuscles were seen below the epithelial pegs. Ruffini corpuscles were found in the deeper layer of the lamina propria. Thus, a variety of sensory nerve endings were found in the hard palate, especially in those areas that are in close contact with the tongue during chewing of food. This rich innervation suggests an important role in monitoring the mechanical properties of food and the position of the tongue.

Studies on the intrinsic nervous system of the wild rodent Calomys callosus digestive tract. II. The submucous plexus.

The submucous plexus of the normal small and large intestine of Calomys callosus was studied by NADH and AChE histochemical techniques and by transmission and scanning electron microscopy. The plexus contains (X +/- SD) 7,488 +/- 293 neurons/cm2 in the duodenum, 5,611 +/- 836 in the jejunum, 2,741 +/- 360 in the ileum, 3,067 +/- 179 in the cecum, and 3,817 +/- 256 in the proximal colon. No ganglia or nerve cell bodies were seen in the esophagus, stomach, distal colon or rectum. The neurons are pear-shaped with a round or oval nucleus and the neuronal cell profile areas were larger in the large intestine than in the small intestine. Most of the neurons display intense AChE activity in the cytoplasm. AChE-positive nerve fibers are present in a primary meshwork of large nerve bundles and in a secondary meshwork of finer nerve bundles. At the ultrastructural level, the ganglia are irregular in shape and covered with fibroblast-like cells. The nucleoplasm of the neurons is finely granular with a few condensations of chromatin attached to the nuclear envelope. In the neuropil numerous varicosities filled with vesicles of different size and electron densities are seen. The pre- and post-synaptic membrane thickenings are asymmetric. Characteristic glial cells with oval nuclei and few organelles are numerous. These data provide a detailed description of this submucosal meshwork.

Studies on the intrinsic nervous system of the wild rodent Calomys callosus digestive tract. II. The submucous plexus

The submucous plexus of the normal small and large intestine of Calomys callosus was studied by NADH and AChE histochemical techniques and by transmission and scanning electron microscopy. The plexus contains (X + SD) 7,488 + 293 neurons/cm2 in the duodenum, 5,611 + 836 in the jejunum, 2,741 + 360 in the ileum, 3,067 + 179 in the cecum, and 3,817 + 256 in the proximal colon. No ganglia or nerve cell bodies were seen in the esophagus, stomach, distal colon or rectum. The neurons are pear-shaped with a round or oval nucleus and the neuronal cell profile areas were larger in the large intestine than in the small intestine. Most of the neurons display intense AChE activity in the cytoplasm. AChE-positive nerve fibers are present in a primary meshwork of large nerve bundles and in a secondary meshwork of finer nerve bundles. At the ultrastructural level, the ganglia are irregular in shape and covered with fibroblast-like cells. The nucleoplasm of the neurons is finely granular with a few condensations of chromatin attached to the nuclear envelope. In the neuropil numerous varicosities filled with vesicles of different size and electron densities are seen. The pre- and post-synaptic membrane thickenings are asymmetric. Characteristic glial cells with oval nuclei and few organelles are numerous. These data provide a detailed description of this submucosal meshwork.

Characterization of the interstitial cells associated with the submuscular plexus of the guinea-pig colon.

Interstitial cells associated with the submuscular plexus of the guinea pig colon were studied by electron microscopy and by light microscopic wholemount stretch preparations. Their cytoplasmic features are similar to those of fibroblasts and they contain a well-developed Golgi apparatus, granular endoplasmic reticulum and many mitochondria. Intermediate filaments are abundantly distributed throughout the perinuclear region and processes. Numerous caveolae, a basal lamina and subsurface cisterns are observed on the cell membrane as in smooth muscle cells. The most characteristic feature of this cell type is the existence of many large gap junctions that interconnect these cells to each other and with the smooth muscle cells. Nerve varicosities containing synaptic vesicles are observed in close apposition with cells of this type. Whole-mount preparations stained by the zinc iodide-osmic acid method and by vimentin immunohistochemistry clearly demonstrated the stellate form of these gap junction-rich cells and suggested that they correspond to the interstitial cells of Cajal.

An ultrastructural study of the submucous plexus of guinea pig intestine after unilateral vagotomy.

This study describes the degenerative changes of the axon terminals making synaptic contacts with the neurons in the submucous ganglia of guinea pig small and large intestines following left or right cervical vagotomy. There were no noticeable ultrastructural changes 1 d after the operation. Beginning at the 3rd postoperative day for the small intestines and the 5th day for the large intestines, some of the axon terminals presynaptic to the submucous neurons displayed different stages of degeneration. The most obvious feature of the degenerating terminals was the swelling and vacuolation of mitochondria with disrupted cristae; another change was the clumping of synaptic vesicles. In the animals killed 7 d after vagotomy, besides the above changes, the degenerating axon terminals also showed an accumulation of a variable amount of glycogen-like particles. The changes were most severe in the guinea pigs killed on the 10th postoperative day. The synaptic vesicles in some degenerating terminals were depleted. There were occasional degenerating nerve cell bodies in the submucous ganglia in the period studied. A variable number of lamellated bodies were found in some of the submucous neurons of 10 and 15 d postoperative animals. By 30 d after vagotomy, all the terminals presynaptic to the submucous neurons appeared normal. There was no significant difference in the total number of axon terminals between 10 d and 30 d after operation (P > 0.05). This suggests that either there had been successful reinnervation or new synaptic contacts had been established by sprouting from the unoperated vagus or other intrinsic neurons. The present quantitative studies also showed that the number of degenerating axon terminals in the submucous ganglia of duodenum after left vagotomy was higher than right vagotomy (P < 0.05). This feature was particularly evident in animals killed 10 d after the operation. The larger number of degenerating terminals in the submucous ganglia of duodenum after left vagotomy suggests that the presynaptic axon terminals in the submucous ganglia of different gut regions are not evenly distributed. Finally, the present study showed that the degenerative changes in the small intestine preceded the large intestine.

Ultrastructural localisation of substance P, vasoactive intestinal peptide and somatostatin immunoreactivities in the submucous plexus of guinea pig ileum.

The submucous neurons, especially those related to the lymphatic vessels, together with their associated synapses, were studied ultrastructurally with respect to their immunoreactivities for 3 types of neuropeptides, namely substance P (SP), vasoactive intestinal peptide (VIP) and somatostatin (SOM). With the antibodies directed against the 3 types of neuropeptides, a variable number of submucous neurons including those contacting the lymphatic vessels were immunostained. Based on the immunoreactivities and synaptic relations with the submucous neurons contacting the lymphatic vessels, at least 4 types of synaptic relations with the submucous neurons contacting the lymphatic vessels, at least 4 types of synaptic configurations were observed: immunopositive terminals with positive neurons, immunopositive terminals with negative neurons, immunonegative terminals with positive neurons and immunonegative terminals with negative neurons. All 4 types of synaptic configurations were observed in SP and VIP-immunostained specimens, with the exception of type 3 which was not encountered in samples immunoreacted for SOM. When the proportions of all 4 types of peptidergic immunopositive terminals contacting the lymphatic vessel-associated neurons were totalled, the value exceeded 100%, suggesting the coexistence of 2 or more neuropeptides in the same terminals. Furthermore, some immunoreactive axon terminals made direct synaptic contacts with positive neurons suggesting the formation of the so-called 'peptide neuron chain'. It is speculated from this study that the submucous neurons receive multiple peptidergic inputs. The various synaptic contacts would imply a complicated reflex pathway in the submucous plexus.