Search

5-HT released by gastrointestinal mucosa and enteric interneurons has powerful effects on gut behavior. However, the targets of 5-HT-containing neurons within enteric circuits are not well characterized. We used antisera against 5-HT and selected markers of known enteric neuron types to investigate the connections made by 5-HT-containing neurons in the guinea-pig jejunum. Confocal microscopy was used to quantify the number of 5-HT-immunoreactive varicosities apposed to immunohistochemically identified cell bodies. Large numbers of varicosities were identified apposing cholinergic secretomotor neurons, immunoreactive for neuropeptide Y, in both myenteric and submucous plexuses. Subgroups of neurons identified by calretinin (ascending interneurons) and nitric oxide synthase (descending interneurons and inhibitory motor neurons) immunoreactivity were also apposed by many varicosities. Longitudinal muscle motor neurons (calretinin immunoreactive) and AH/Dogiel type II (sensory) neurons (calbindin immunoreactive) were apposed by small numbers of varicosities. Combined retrograde tracing and immunohistochemistry were used to identify excitatory circular muscle motor neurons; these were encircled by 5-HT-immunoreactive varicosities, but the appositions could not be quantified. We suggest that 5-HT-containing interneurons are involved in secretomotor pathways and pathways to subgroups of other interneurons, but not longitudinal muscle motor neurons. There also appear to be connections between 5-HT-containing interneurons and excitatory circular muscle motor neurons. Physiological evidence demonstrates a functional connection between 5-HT-containing interneurons and AH/Dogiel type II neurons, but few 5-HT-immunoreactive varicosities were observed apposing calbindin-immunoreactive cell bodies. Taken together these results suggest that neural 5-HT may have significant roles in excitatory pathways regulating both motility and secretion.

The projections, connections, morphology and electrophysiological features of the myenteric interneurons with somatostatin immunoreactivity in the guinea-pig small intestine have been established using retrograde tracing, immunohistochemistry, confocal microscopy and intracellular recording. After application of the fluorescent dye, 1,1'-didodecyl-3,3,3',3'-tetramethyl indocarbocyanine perchlorate (DiI), to the myenteric plexus, up to 900 nerve cell bodies were labelled in each preparation. Somatostatin-immunoreactive neurons accounted for 13% of all retrogradely labelled cells and were located up to 70 mm orally. When DiI was applied to the submucous ganglia, many myenteric neurons were labelled and 8% of all retrogradely labelled cells were somatostatin immunoreactive and were located up to 60 mm oral to the DiI application sites. These neurons had ovoid cell bodies, a single axon, several long filamentous dendrites and received close contacts from 40-200 somatostatin-immunoreactive varicosities. Intracellular recordings revealed that these cells had features of both S (i.e. with Synaptic inputs) and AH (i.e. neurons with After Hyperpolarization) cells, receiving fast excitatory synaptic inputs, having characteristic "sag" in their response to hyperpolarizing current pulses and sometimes a long afterhyperpolarization following soma action potentials. It is concluded that somatostatin-immunoreactive neurons have distinct electrophysiological features and form very long anally directed interneuronal chains that connect with both myenteric and submucous neurons.

Endogenous alkaline phosphatase activity has been localized histochemically on the surface of enteric neurons of the guinea-pig small intestine by both light and electron microscopy. The enzyme activity was associated with some myenteric neurons that had Dogiel type I morphology, and the histochemical reaction products typically formed a honeycomb-like structure on labelled cell bodies. No Dogiel type II neurons in the myenteric plexus or submucous neurons showed alkaline phosphatase reactivity. Nerve fibres reactive for alkaline phosphatase were present in the myenteric plexus and ran in bundles in the circular muscle and deep muscular plexus. In addition, reactive varicose axons supplied the submucous plexus and non-ganglionated plexus of the mucosa. The results of interruption of the enteric neuronal pathways demonstrated that alkaline phosphatase-reactive myenteric neurons project anally to other myenteric ganglia, to the circular muscle and to the submucous plexus. Sequential enzyme histochemistry showed that virtually all alkaline phosphatase-reactive neurons also contained nitric oxide synthase, revealed by NADPH-diaphorase reactivity. It was estimated that 14–18% of all myenteric neurons showed alkaline phosphatase reactivity. About one-third of nitric oxide synthase-containing myenteric neurons, however, did not contain alkaline phosphatase activity. At the ultrastructural level, alkaline phosphatase activity was associated specifically with the plasma membranes of nerve cell bodies, axons and dendrites of some myenteric neurons. Reactive nerve fibres made close appositions with non-reactive submucous neurons and, within myenteric ganglia, predominantly with other alkaline phosphatase-reactive neurons. In addition to its presence in neurons, alkaline phosphatase reactivity was also present in some endothelial cells in blood vessels in the submucosa and in capillary pericytes. It is concluded, on the basis of the projections and neurochemistry, that in the guinea-pig small intestine alkaline phosphatase activity is associated with nitric oxide synthase-containing neurons which include inhibitory motor neurons to the circular muscle, and anally-directed interneurons to other myenteric and submucous neurons.

The organization of synaptic connections between guinea-pig ileal submucosal neurons was examined using intracellular recordings from single or pairs of submucosal neurons. Synaptic inputs were elicited by stimulating cholinergic neurons using pressure-pulse application of 5-hydroxytryptamine (5-HT) in ganglia adjacent to those where intracellular recordings were obtained. In addition, when pairs of intracellular recordings were obtained, one neuron was activated by intracellular stimulation and synaptic responses were recorded in the other neuron. Neurobiotin-filled microelectrodes were employed to characterize cells electrophysiologically and immunohistochemically. Recordings were obtained from 176 (173 S-type and three AH-type) neurons; 81% of cells were classified as vasoactive intestinal peptide (VIP) neurons. No fast excitatory postsynaptic potentials and only rare slow excitatory postsynaptic potentials were recorded following intracellular stimulation of paired S-type neurons. However, when paired intracellular recordings were obtained from neurons within the same ganglion and 5-HT was applied to an adjacent ganglion, this stimulation evoked synchronized fast excitatory postsynaptic potentials in 94% of pairs. In contrast, when cell bodies of VIP-VIP pairs were located in different ganglia, fast synaptic activation evoked by 5-HT stimulation was not synchronized in 87% of pairs. When intracellular recordings were obtained from a single neuron and two separate ganglia were stimulated by 5-HT pressure-pulse activation, fast excitatory postsynaptic potentials originating from both sources were recorded in the same VIP neuron. Morphological study of 34 S-type and three AH-type horseradish peroxidase-labeled neurons was conducted. AH-type neurons had multiple axonal branches with dense arborization of collaterals containing numerous varicosities in three to nine ganglia, whereas axons of S-type neurons exhibited relatively rare collaterals and varicosities within adjacent ganglia. These results demonstrate that cholinergic neurons provide both diverging and converging inputs to VIP neurons, providing a mechanism to enhance activation of VIP secretomotor neurons. The axonal projections of AH-type neurons suggest they are likely candidates to provide diverging inputs to multiple VIP neurons.

Intracellular recordings were made from neurons in the submucosal ganglia of the guinea-pig distal colon. The recording electrode contained the intracellular marker biocytin, which was injected into neurons so that their electrophysiological characteristics could be correlated with their shape. Correlations of electrophysiology and shape have not been reported previously for neurons in this region. Three types of neuron were identified on electrophysiological grounds. Neurons of the first type (S neurons) had tetrodotoxin-sensitive soma action potentials, and received fast and slow excitatory synaptic inputs. They had uniaxonal morphologies and may function as secretomotor or possibly vasomotor neurons. The second type (AH neurons) received only slow synaptic input, while the soma action potential had tetrodotoxin-sensitive and -insensitive components with a shoulder on the falling phase and a long-lasting afterhyperpolarisation of the membrane potential following a single action potential. Neurons of this type had multipolar morphologies and provided dense innervation of adjacent submucosal ganglia. These neurons are similar to the submucosal intrinsic primary afferent neurons of the guinea-pig small intestine. The final type of neuron [the low-threshold (LT) neuron] had electrophysiological characteristics that set it apart from those described previously within enteric plexuses. They expressed tetrodotoxin-insensitive voltage-gated soma currents, did not have long-lasting afterhyperpolarisations and received only slow synaptic input. In addition, these neurons were very excitable: they had large input resistances and low thresholds for action potential discharge, and often fired action potentials in the absence of stimulation. Neurons with these characteristics were uniaxonal and thus are likely to be secretomotor or possibly vasomotor neurons. This study has shown that submucosal neurons of the distal colon fall into three distinct types, which can be distinguished by a combination of electrophysiological and morphological criteria.

The occurrence and distribution of neurons and nerve fibres showing vasoactive intestinal polypeptide-like and neuropeptide Y-like immunoreactivity were re-examined in the enteric nervous system of the small and large intestine of the adult rat using dual-labelling indirect immunofluorescence histochemistry to detect the co-existence of these neuropeptides. In the myenteric plexus of both small and large intestine, a population of neuropeptide Y-immunoreactive neurons that did not contain vasoactive intestinal polypeptide was noted; it accounted for 29-53% of neuropeptide Y neurons. Such neurons were also found in the submucosa but there they constituted at most 2% of neuropeptide Y-immunoreactive neurons. In both myenteric and submucous plexuses, regional variations were observed in the number of immunoreactive neurons and in the proportion of dual-labelled neurons. In the myenteric plexus, for example, the density of neurons with immunoreactivity to these two neuropeptides was constant throughout the small intestine, whereas it progressively increased distally within the colon. In addition, a distinct but small subset of immunoreactive myenteric neurons was found to have a novel soma morphology, unclassifiable according to the criteria used for porcine or guinea-pig enteric neurons. Such neurons had one or more conspicuous processes, which were much longer than the short, lamellar somal processes of typical Dogiel Type 1 neurons; moreover, these protruded from an essentially smooth soma and terminated at distances of up to two cell diameters from their point of origin. Thus, our results suggest that the organization of the enteric nervous system of the rat differs from that of other species and indicate that investigation of the co-localizations of neuropeptides and biologically active mediators in the intestinal tract would be incomplete without reference to regional differences in the incidence and distribution of such neurochemicals.

Intracellular recordings were made from 73 guinea-pig submucosal neurons using neurobiotin-filled microelectrodes; subsequently, neuropeptide immunoreactivity, morphology and nerve fibre projections were determined. Five distinct groups of cells could be distinguished: S cells with inhibitory input (61%), S cells without inhibitory input (19%), AH cells (8%), S-AH cells (5%), and glial networks. S cells with inhibitory input were immunoreactive for vasoactive intestinal polypeptide and showed Dogiel Type III morphology with the axon branching and coursing through two to 12 ganglia; varicosities and tufts of varicosities were observed surrounding other cell bodies as well as over blood vessels. S cells without inhibitory input primarily were immunoreactive for neuropeptide Y; they also showed Dogiel Type III morphology and similar, though shorter, axonal projections and varicose features surrounding other neurons. AH cells, which most likely contained substance P, lacked synaptic input and exhibited Dogiel Type II morphology; they branched more extensively than S cells and also formed varicose tufts within other ganglia. S-AH cells combined electrophysiological properties of S cells with inhibitory input and AH cells and did not show consistent morphological or histochemical characteristics. Typical glial networks were observed; in addition, on two occasions unusual networks of dye and electrical coupling between S cells without inhibitory input and a glial complex were observed. These results suggest that vasoactive intestinal polypeptide-containing S cells may act as interneurons which mediate a slow excitatory synaptic potential; that neuropeptide Y-containing S cells, which are known to be cholinergic, may play a role as cholinergic interneurons mediating the nicotinic fast excitatory synaptic potential; and that AH neurons also may provide cholinergic innervation to other submucosal neurons in addition to their previously described dual projections into mucosa and myenteric plexus.

We have investigated the presence of c-Myc-like antigens in the enteric nervous system of the guinea-pig, rat, dog, sheep, monkey and human. c-Myc-like immunoreactivity was demonstrated by immunohistochemistry in the enteric nervous system of all animals tested, with one or more monoclonal antibodies raised against peptide sequences found in the human c-Myc protein. While in most cases the labelling was nuclear, cytoplasmic labelling was also observed. In the guinea-pig enteric nervous system, c-Myc-like immunoreactivity detected by two different antibodies remained detectable for up to 4 h in the presence of cycloheximide. The size and density of labelled nuclei in the ileal submucous plexus were consistent with exclusive neuronal labelling by one antibody and neuronal plus glial labelling by the other. Double-labelling with antiserum directed against vasoactive intestinal peptide revealed a subset of c-Myc-immunoreactive neurons that also contain this neuropeptide. Anti-c-Myc antibodies specifically immunoprecipitated proteins from guinea-pig myenteric plexus-longitudinal muscle preparations whose sizes were consistent with previous observations for c-Myc antigens and whose distribution was consistent with synthesis in the myenteric plexus. We conclude that c-Myc proteins are expressed in mammalian enteric neurons and that they have characteristics similar to those of c-Myc proteins in other non-proliferative cells.

The enteric nervous system is one of the three divisions of the autonomic nervous system, the others being the sympathetic and parasympathetic. In contrast to the other divisions, it can perform many functions independently of the central nervous system. It consists of ganglionated plexuses, their connections with each other, and nerve fibres which arise from the plexuses and supply the muscle, blood vessels and mucosa of the gastrointestinal tract. The enteric nervous system contains a large number of neurons, approximately 107 to 108. About ten or more distinct types of enteric neurons have been distinguished on electrical, pharmacological, histochemical, biochemical and ultrastructural grounds as well as on the basis of their modes of action. Both excitatory and inhibitory nerves supply the muscle and there are inhibitory and excitatory interneurons within the enteric plexuses. There are also enteric nerves which supply intestinal glands and blood vessels, but these receive less emphasis in this commentary. Correlations between groups of neurons defined on different criteria are poor and in many cases the physiological roles of the nerves are not known. The functions of noradrenergic nerves which are of extrinsic origin are reasonably well understood, but cholinergic nerves in the intestine are the only intrinsic nerves for which both the transmitter and to some extent the functions are known. In the case of non-cholinergic, non-noradrenergic enteric inhibitory nerves, the functions are understood but the transmitter is yet to be determined, both adenosine 5′-triphosphate and vasoactive intestinal polypeptide having been proposed. Other nerves have been defined pharmacologically (non-cholinergic excitatory nerves to neurons and muscle, intrinsic inhibitory inputs to neurons, and enteric, non-cholinergic vasodilator nerves) and histochemically (intrinsic amine-handling neurons and separate neurons containing peptides: substance P, somatostatin, enkephalins, vasoactive intestinal polypeptide, gastrin cholecystokinin tetrapeptide, bombesin, neurotensin and probably other peptides). Little is known of the functions of these nerves, although a number of proposals which have been made are discussed.

Intracellular recordings of membrane potential and membrane currents were made from neurones in the submucous plexus of the guinea-pig caecum in vitro. Fast and slow excitatory postsynaptic potentials and slow inhibitory postsynaptic potentials were recorded from the majority of neurones following focal stimulation of presynaptic fibres in the plexus. The slow inhibitory postsynaptic potential was associated with an increase in membrane conductance and reversed its polarity at -90 mV; it was reversibly blocked by yohimbine. The slow excitatory postsynaptic potential and its underlying current was associated with a decrease in membrane conductance. Two kinds of voltage-dependence both of the slow excitatory postsynaptic potential and current were observed; in 80% of cells, the excitatory postsynaptic potential and current became smaller with membrane hyperpolarization and reversed polarity at -90 mV (reversing type) but in 20% of cells both the excitatory postsynaptic potential and current simply disappeared when the membrane potential reached -70 mV (non-reversing type). The effects of acetylcholine, adenosine 5'-triphosphate, bombesin, 5-hydroxytryptamine, neurotensin, noradrenaline, substance P and vasoactive intestinal polypeptide were examined. The only substance which mimicked the slow inhibitory postsynaptic potential was noradrenaline; brief applications of noradrenaline caused hyperpolarizations which had the same time-course, reversal potential and sensitivity to yohimbine as the slow inhibitory postsynaptic potential. The non-reversing type of slow excitatory postsynaptic potential was mimicked only by adenosine 5'-triphosphate. The reversing type of slow excitatory postsynaptic potential was mimicked by bombesin, neurotensin, substance P and vasoactive intestinal polypeptide. 5-Hydroxytryptamine and vasoactive intestinal polypeptide (in some neurones) caused a depolarization with an increase in membrane conductance. All three synaptic potentials were reversibly depressed by superfusion of noradrenaline but noradrenaline did not affect the potential changes evoked by brief application of exogenous acetylcholine or substance P. It is concluded that, in guinea-pig submucous plexus neurones, the slow inhibitory postsynaptic potential is mediated by noradrenaline and results from a potassium conductance increase.(ABSTRACT TRUNCATED AT 400 WORDS)

The appearance and distribution of nerve endings (varicosities) containing small granular vesicles have been studied in the distal colon and rectum of the guinea-pig with the electron microscope. Two types of varicosity were recognised. They were distinguished by differences in their synaptic vesicles and in their distribution in the layers of the gut wall. The first type resembled noradrenergic nerves in having predominantly (92%) small vesicles and few (8%) large granular vesicles (90 nm diameter). This type was common in the plexuses and at the medial-adventitial border of arteries and arterioles but was sparsely distributed in the muscle coat. The second type had a lower proportion of small vesicles (69%) and a higher proportion (31%) of large granular vesicles (132 nm diameter). This type was absent in Auerbach's plexus, well represented in the muscle coat and Meissner's plexus and not associated with blood vessels. The first type was labelled with 5-hydroxydopamine, a specific marker for noradrenergic nerves, and disappeared after extrinsic denervation. a procedure which causes degeneration of noradrenergic nerves in the gut. The second type was unaffected by 5-hydroxydopamine and extrinsic denervation. It is concluded that the two types of small granular vesicle-containing varicosities belong to different axons and that the first type is noradrenergic. The second type of nerve axon has not been described in the gut before and is intrinsic to it. From the distribution and numbers of these axons in the circular muscle it would seem that they play an important role in gut motility.

Intracellular recordings were made from neurons of the guinea pig submucosal plexus and the effects of substance P and the substance P analogue [D-Arg1,D-Pro2,D-Trp7,9,Leu11]substance P were examined. Substance P (20-200 nM) depolarized all submucosal neurons; these depolarizations were shown to be due to a decrease in the resting (or "leak") potassium conductance of the membrane. In approximately 50% of the 46 neurons tested, superfusion with [D-Arg1,D-Pro2,D-Trp7,9,Leu11]substance P (0.2-20 microM) produced a dose-dependent membrane hyperpolarization. This hyperpolarization was prevented by the alpha 2-adrenoceptor antagonist idazoxan (300 nM) or by concentrations of cobalt which abolished all spontaneous and evoked synaptic potentials, indicating that it resulted from release of noradrenaline from sympathetic nerve terminals. [D-Arg1,D-Pro2,D-Trp7,9,Leu11]substance P depressed the amplitude of the three synaptic potentials recorded from submucosal neurons; the concentrations that caused 50% of the maximal inhibition of the fast excitatory postsynaptic potential, the inhibitory postsynaptic potential, and slow excitatory postsynaptic potential were 40 microM, 600 nM and 20 microM, respectively. When idazoxan was present, the substance P analogue was less effective in depressing the amplitudes of the fast and slow excitatory synaptic potentials suggesting that much of its presynaptic inhibition also resulted from release of noradrenaline. These results provide evidence that [D-Arg1,D-Pro2,D-Trp7,9,Leu11]substance P releases noradrenaline from sympathetic nerves in the submucosal plexus. One effect of this is a membrane hyperpolarization; another is a presynaptic inhibition of transmitter release. These actions much limit the usefulness of this "substance P antagonist" in efforts to show that synaptic potentials, such as the slow excitatory synaptic potential, are mediated by substance P.

Using antisera raised against neurofilaments and the glial fibrillary acidic protein (GFAP) we have examined the appearance and distribution of neurofilament- and GFAP-like immunoreactivity in the enteric nervous system of rat, mouse and guinea-pig. In whole mounts of the external circular and longitudinal muscle layers, including the myenteric plexus, a high number of neurofilament-positive perikarya were visualized both in the ganglia and in the circularly running interconnecting strands in all three species. These cells were large, usually with eccentrically placed nuclei and single, relatively thick neurofilament-positive processes. In addition, in guinea-pig myenteric plexus a small number of cells with multiple processes could be seen. Both in the longitudinal and circular interconnecting strands a large number of thin, smooth, neurofilament-positive fibres were observed. This regular network of ganglia and strands was superimposed on a sparse system of thin, usually individual neurofilament-positive fibres in the underlying circular muscle layer. Cryostat sections revealed neurofilament-positive cell bodies in the submucous plexus, whereas fibres showing neurofilament-like immunoreactivity were observed in all layers of the gut wall, with the exception of the epithelium. In whole mounts including rat and mouse myenteric plexus, a large number of cells and fibres showing GFAP-like immunoreactivity were visualized. The GFAP-positive cells were smaller and more numerous than the neurofilament-positive ones. They were present both within the ganglia and in the interconnecting strands. Several short fluorescent processes could frequently be seen emanating from the cell body. Both the strands and the ganglia contained a high number of thin, GFAP-positive fibres. Fluorescent fibres and cells were also observed in the circular muscle layer. In sections of rat and mouse small intestine, cells were observed throughout the gut wall, with the exception of the epithelium. Double labelling experiments clearly showed that neurofilament- and GFAP-positive cells represented separate cell populations. Furthermore, GFAP-positive cells and fibres outlined the neurofilament-positive perikarya. It is thus likely that the GFAP-positive cells represent enteric glial cells. The pre- and postnatal development of neurofilament- and GFAP-like immunoreactivity was studied in whole mounts from rat embryos and pups. Furthermore, the presence of neurofilament and GFAP-positive fibres was observed in whole mount preparations of rat and mouse mesenterium.(ABSTRACT TRUNCATED AT 400 WORDS)

The distribution of peptide-containing neurons in the oesophagus, stomach and small and large intestine of the rat and the guinea-pig has been studied with the indirect immunofluorescence technique of Coons & Co-workers (1958) using antisera to substance P, vasoactive intestinal polypeptide (VIP), enkephalin, somatostatin, gastrin and neurotensin. (The gastrin antiserum is to the C-terminal portion and consequently reacts also with cholecystokinin (CCK)-like peptides.) For comparison, the noradrenergic innervation was visualized with antiserum to dopamine β-hydroxylase. For improved visualization of peptide-containing cell bodies, a mitotic inhibitor (colchicine or vinblastine) was applied locally on the different parts of the gastro-intestinal tract of several animals. Substance P-, VIP-, enkephalin- and somatostatin-like immunoreactivity was observed in all parts of the gastro-intestinal tract studied. Gastrin/CCK had a more limited distribution, especially in the guinea-pig and neurotensin was seen only in certain regions and layers of the rat gastro-intestinal tract. Immunoreactivity to all peptides except neurotensin was observed both in cell bodies and fibres; immunoreactivity to neurotensin has so far only been seen in nerve fibres. Substance P and enkephalin immunoreactive cells were often numerous in the myenteric plexus, whereas VIP and somatostatin immunoreactive cells were preferentially located in the submucous plexus. Some VIP immunoreactive cells were observed in the lamina propria. Large numbers of especially substance P-, VIP- and enkephalin-containing fibres were often seen in the circular muscle layer and in the two ganglionic plexuses. Substance P immunoreactive fibres formed the densest network in the ganglionic plexuses, whereas VIP immunoreactive fibres constituted the most impressive network in the lamina propria and often extended into the most superficial parts of the mucosa. Enkephalin immunoreactive structures were mainly confined to the circular and longitudinal muscle layers and the myenteric plexus. Somatostatin immunoreactive fibres were mainly found in the ganglionic plexuses. Peptide-containing fibres, particularly these containing substance P and VIP were often seen along blood vessels, but never with such a density as the noradrenergic (dopamine β-hydroxylase immunoreactive) fibres. No somatostatin or neurotensin immunoreactive fibres were observed in relation to clearly identifiable blood vessels. The possible coexistence of two peptides in one neuron was studied. For this part of the study the proximal colon and five antisera, namely substance P, VIP, enkephalin. somatostatin and gastrin/CCK antisera were selected. Evidence was obtained for the occurrence of a somatostatin-like and a gastrin/ CCK-like peptide in the same neurons. This may indicate a common precursor for the two peptides in these particular neurons. Each of the substance P-, VIP- and enkephalin-like peptides. on the other hand, seem to be present in different neuronal populations, which were themselves distinct from the somatostatin-gastrin/CCK immunoreactive neurons. In addition, somatostatin immunoreactive neurons different from the gastrin/CCK immunoreactive ones seem to exist. The gastrin/CCK immunoreactive fibres around blood vessels may represent a further, separate population of fibres, since no somatostatin immunoreactive fibres were seen at this location. The findings indicate the existence of numerous subpopulations of enteric neurons, each characterized by its content of a certain peptide (or peptides). The axons of most of these neurons probably terminate in the wall of the gastro-intestinal tract, but some seem to project to other organs. In addition, some peptide-containing fibres in the gastro-intestinal wall may have an extrinsic origin. The relationship between these peptide-containing neurons and the cholinergic enteric neurons and any of the other non-cholinergic. non-adrenergic inhibitory and excitatory neurons present in the enteric nervous system is not known. It is, however, noteworthy that a somatostatin-like peptide seems to be present in noradrenergic neurons of prevertebral ganglia that project to the intestine. The possibility must be kept in mind that one or more of the peptides in the gut could be localized in neurons that contain other potential transmitters, e.g. acetylcholine. The wide variety of pharmacological actions of these neuronal peptides on smooth muscle and neurons in the gut and on its blood vessels raises the possibility that some of them may be neurotransmitters.

To determine the state of development of mechanoreceptors during the period in which central neurol regions can be altered by peripheral lesions, the distribution of terminal neurites and specialized receptors was examined in the glabrous forepaw skin of a developmental series of pouch young opossums, using the silver stain of Sevier & Munger (1965) and electron microscopic observations. At 10 days after birth bundles of neurites approach the dermal-epidermal junction, but neither neurites nor receptors are seen in the epidermis. Some neuntes enter the epidermis by 20 days; by 25 days developing Merkel cells are seen in the epidermis; at 30 days mature Merkel cells and primordial Pacinian corpuscles are present, respectively, in the epidermis and deep dermis; at 42 postnatal days dermal papillae containing neurites were observed at the dermal-epidermal junction, which may be developing Meissner corpuscles. Numbers of neurites, and of their branchings, increase up to 60 days; as do the size, number and degree of differentiation of the Pacinian corpuscles. The critical period for the effects of peripheral lesions upon the morphology of the cuneate-gracile nuclear complex, as well as the times of initial massive synaptogenesis, cell migration and differentiation in this first and second order mechanosensory synaptic region, are completed prior to the innervation of the epidermis and the appearance of specialized mechanoreceptors. Thus input from epidermal and specialized receptors cannot be responsible for guidance of these developmental processes.

A histochemical technique was used to stain the myenteric neurons in the intact wall of the small intestine of mice, guinea-pigs and sheep. The length and diameter of the small intestine and the total serosal surface area were also obtained. Myenteric neurons were counted on large whole-mount preparations of the muscularis externa. Counts were carried out also on the submucosal plexus, on a more limited scale. In the mouse a spatial density of 10,600 myenteric neurons per cm2 was found. The small intestine was 33 cm long and measured on average 11.5 mm in circumference, the total outer surface (serosal surface) amounting to about 38.0 cm2. The total number of myenteric neurons in the small intestine was calculated as about 403,000. In the guinea-pig the length of the small intestine was 145 cm, the average circumference 22 mm and the total outer surface area about 319 cm2. The neuronal packing density was 8600/cm2, and the total number of myenteric neurons about 2,750,000. In the sheep the small intestine was about 2100 cm long with an average circumference of 60 mm and a total surface area of about 12,600 cm2. The ganglion neuron density was about 2500/cm2, and the total number of myenteric neurons in the small intestine was calculated as about 31,500,000. Thus, in the sheep the small intestine contained about 11 times as many myenteric neurons as the guinea-pig and about 80 times as many as the mouse. The differences are in the same direction as, but not proportional to, the differences in body weight and in the length of the intestine. The neuronal spatial density was highest in the mouse and lowest in the sheep, and in the sheep the neurons were markedly larger, and gathered in ganglia that were larger and further apart from one another, than in the mouse, while they had intermediate values in the guinea-pig. A new way of expressing neuronal packing densities is presented together with the proposal of an arbitrary but reproducible unit of intestinal length (a segment whose length is equal to its diameter). In the submucosal plexus the neuronal density was about 3000/cm2 in the guinea-pig, about 8700/cm2 in the mouse and about 4500/cm2 in the sheep. In the mouse the submucosal neuron density decreased gradually along the length of the small intestine.

Intracellular recordings were made from neurones in the myenteric and submucous plexuses of the guinea-pig ileum. All myenteric neurones that were hyperpolarized by [Met5]enkephalin (or normorphine) were also hyperpolarized by noradrenaline (or clonidine); neurones unaffected by opioids were unaffected by noradrenaline. The hyperpolarizations resulted from an increase in potassium conductance of the membrane and were blocked by the respective antagonists naloxone and idazoxan. Neurones of the submucous plexus were hyperpolarized by noradrenaline but not by normorphine. The results suggest that myenteric neurones possesses both mu-opioid receptors and alpha 2-adrenoceptors whereas submucous neurones have alpha 2-adrenoceptors but not mu-opioid receptors.

The cellular origin of peptides derived from preprovasoactive intestinal peptide has been studied in rat stomach and ileum. Antisera specific for the C-terminal regions of the N-terminal flanking peptide (preprovasoactive intestinal peptide 22–80), bridging peptide (preprovasoactive intestinal peptide 111–124), C-terminal flanking peptide (preprovasoactive intestinal peptide 156–170) and vasoactive intestinal peptide were used in immunohistochemical studies on sections and whole mounts. All four antisera stained nerve fibres and cell bodies in the stomach and intestine. However, there were distinct differences in the pattern of colocalization of peptides derived from provasoactive intestinal peptide. In the sub-mucous plexus of the ileum virtually 100% of neurons reacting with vasoactive intestinal peptide antibodies also reacted with antibodies to the other three peptides. In contrast, in the stomach, while all vasoactive intestinal peptide-immunoreactive neurons of the myenteric plexus contained C-terminal flanking peptide- and bridging peptide-like immunoreactivity, only 50% of these cells reacted with the antiserum to N-terminal flanking peptide. The data indicate that in a population of neurons in the myenteric plexus of the rat stomach, preprovasoactive intestinal peptide is processed in such a way that the antigenic determinant of the N-terminal flanking peptide is not produced. In a second population of enteric neurons in the stomach and in the intestine, it appears that processing of preprovasoactive intestinal peptide results in the production of peptides reacting with antibodies to vasoactive intestinal peptide, the flanking and bridging peptides.

Intracellular recordings were made from neurons in the submucous plexus of the guinea-pig caecum. Muscarinic agonists (acetylcholine, bethanechol and muscarine) depolarized about 70%, and hyperpolarized about 30% of the submucous plexus neurons. Low concentrations of pirenzepine reversibly antagonized both responses. The measured dissociation constants (KD) of 10-30 nM for the depolarizations and 1-3 nM for the hyperpolarizations suggest that each response was mediated by muscarinic M1 cholinoceptors. The muscarinic depolarization and hyperpolarization were associated with a decreased and an increased conductance, respectively, and the reversal potential for the muscarinic responses varied as the potassium concentration varied, always being around the potassium equilibrium potential. In cells depolarized by muscarinic agonists these agents appeared to decrease a potassium conductance that could also be inactivated by substance P. In approximately 30% of the submucous neurons, the slow inhibitory postsynaptic potential, elicited in response to single or repetitive focal stimuli (1-10 pulses at 20-40 Hz), appeared to consist of a large component which was sensitive to the blocking action of idazoxan (100-300 nM) and a small component which was idazoxan-insensitive. The latter (muscarinic slow inhibitory postsynaptic potential) was completely abolished by pirenzepine. The concentrations of pirenzepine which caused a 50% depression ranged from 5 to 20 nM. The muscarinic slow inhibitory postsynaptic potential was increased in amplitude and duration by physostigmine (100-300 nM). The muscarinic slow inhibitory postsynaptic potential was accompanied by a decrease in membrane input resistance, and was reversed in polarity near the potassium equilibrium potential. When muscarine induced a hyperpolarization and/or focal stimulation elicited a muscarinic slow inhibitory postsynaptic potential in the presence of idazoxan (100-300 nM), the intracellular injection of guanosine 5'-O-(3-thiotriphosphate) produced a progressive membrane hyperpolarization during which the muscarinic hyperpolarizing responses were attenuated. It is concluded that the muscarine-induced reduction in potassium conductance is mediated through a muscarinic M1 receptor which has a relatively low affinity for pirenzepine. The muscarine-induced increase in potassium conductance is probably produced by the association of a guanine nucleotide-binding regulatory protein with another muscarinic M1 receptor that has a relatively high affinity for pirenzepine.

Intracellular recordings were made from submucous plexus neurons of the guinea-pig cecum maintained in vitro. Histamine (0.3-10 microM) produced a dose-dependent membrane depolarization (congruent to 13 mV with 3 microM) in about 28% of the cells tested; most of these cells showed a prominent calcium-activated potassium conductance (AH cells). The depolarization was due primarily to an inactivation of potassium conductance which is available at the resting membrane potential of -60 mV. Peak amplitude of the fast excitatory postsynaptic potential was depressed by histamine (0.1-10 microM) in a dose-dependent manner (congruent to 62% depression with 1 microM). This was observed even in those cells in which histamine did not produce any membrane depolarizations (mostly S cells). The depression of the fast excitatory postsynaptic potential resulted from the presynaptic inhibition of acetylcholine release. Histamine also reduced the amplitude of the non-cholinergic, presumably peptidergic, slow excitatory postsynaptic potential by suppressing peptide release from presynaptic nerve terminals. Peak amplitude of the adrenergic inhibitory synaptic potential was not depressed by histamine suggesting that histamine receptors are not present on presynaptic terminals of sympathetic nerve fibres. Both postsynaptic and presynaptic actions of histamine were blocked by cimetidine or ranitidine but not by pyrilamine implying that H2 receptors are involved.

Galanin, a 29 amino acid peptide, was recently isolated from the porcine gut. Immunocytochemistry revealed a dense network of galanin-immunoreactive nerve fibers in the submucosa, smooth muscle layers and intramural ganglia throughout the rat gastrointestinal tract. In the smooth muscle the density of innervation was lower in the colon than in the small intestine. In the mucosa galanin-immunoreactive fibers were quite numerous in the small intestine, less numerous in the large intestine and rare in the stomach. A moderate number of galanin-immunoreactive nerve cell bodies could be detected in the myenteric ganglia throughout the digestive tract and in the submucous ganglia of the intestines. Enteric galanin-immunoreactive fibers seem to be intrinsic to the gastrointestinal tract since their distribution and frequency were unaffected after extrinsic denervation (chemical sympathectomy, vagal denervation or clamping of nerves running within the mesenterium). Myectomy at the mid-jejunal level resulted in a total loss of galanin-immunoreactive nerve fibers 5 mm anally to the lesion with a gradual return of galanin-immunoreactive fibers up to 15-20 mm further anally; there was no overt loss of fibers orally. The findings indicate that myenteric galanin-immunoreactive neurones issue long descending projections that terminate either within the myenteric ganglia or in the smooth muscle.

Immunoreactivity for vasoactive intestinal polypeptide has been localized in neurons in the guinea-pig ileum, colon and stomach. In the ileum, 2.5% of the nerve cell bodies of the myenteric plexus and 45% of those of the submucous plexus showed vasoactive intestinal polypeptide-like immunoreactivity. Varicose axons containing vasoactive intestinal polypeptide ramified amongst the nerve cell bodies of both plexuses and in some cases formed rings of varicosities around non-reactive nerve cells. Axons were traced from the myenteric plexus to the circular muscle and deep muscular plexus. There were numerous positive axons running in fine strands within the circular muscle, parallel to the muscle bundles. Axons containing vasoactive intestinal polypeptide were associated with mucosal blood vessels, but few supplied the vascular network of the submucosa; some immunoreactive axons also contributed to the periglandular plexus of the mucosa. There were no changes in the distribution of axons in the ileum after extrinsic denervation. The results are discussed in relation to the possible functional roles of neurons that contain vasoactive intestinal polypeptide in the intestine: the distribution of such nerve cells in the myenteric plexus and of axons in the circular muscle and sphincters is consistent with this polypeptide being a transmitter of enteric inhibitory neurons; it is also possible that vasoactive intestinal polypeptide is the enteric vasodilator transmitter.

The activity of tyrosine hydroxylase and the localization of dopamine β-hydroxylase were determined in the myenteric and submucous plexuses of the normal guinea-pig ileum and in these plexuses after extrinsic denervation. In the normal ileum, the distribution of axons showing immunoreactivity for dopamine β-hydroxylase was not distinguishable from the distribution of noradrenergic axons determined by the fluorescence histochemical localization of catecholamines. The distribution of tyrosine hydroxylase in the different layers of the intestine correlated well with the distribution of dopamine β-hydroxylase and noradrenaline, tyrosine hydroxylase activity being most concentrated in the myenteric and submucous plexuses. Extrinsic denervation resulted in the complete disappearance of both biochemically detectable tyrosine hydroxylase and immunohistochemically demonstrable dopamine β-hydroxylase. It is concluded that if the amine-handling neurons which are known to be intrinsic to the intestine synthesize an aromatic amine, it is almost certainly not a catecholamine, and is probably an indoleamine.

We have analyzed changes in the distributions of terminals with vasoactive intestinal polypeptide (VIP)-like immunoreactivity, and accumulations in severed processes, that occur after lesions of intrinsic and extrinsic nerve pathways of the guinea-pig small intestine. The observations indicate that enteric vasoactive intestinal polypeptide immunoreactive neurons have the following projections. Nerve cell bodies in the myenteric plexus provide varicose processes to the underlying circular muscle; the majority of these pathways, if they extend at all in the anal or oral directions, do so for distances of less than 1 mm. Nerve cell bodies of the myenteric plexus also project anally to provide terminals to other myenteric ganglia. The lengths of the majority of these projections are between 2 and 10 mm, with an average length of about 6 mm. Processes of myenteric neurons also run anally in the myenteric plexus and then penetrate the circular muscle to provide varicose processes in the submucous ganglia at distances of up to 15 mm, the average length being 9-12 mm. In addition, there is an intestinofugal projection of myenteric neurons whose processes end around nerve cell bodies of the coeliac ganglia. A similar projection from the colon supplies the inferior mesenteric ganglia. The nerve cell bodies in submucous ganglia give rise to a subepithelial network of fibres in the mucosa and also supply terminals to submucous arterioles. It is concluded that vasoactive intestinal polypeptide is contained in neurons of a number of intrinsic nerve pathways, influencing motility, blood flow and mucosal transport. The myenteric neurons that project to prevertebral sympathetic ganglia may be involved in intestino-intestinal reflexes.

Changes in the distribution of 5-hydroxytryptamine-like immunoreactivity have been examined in enteric neurons at various times after microsurgical lesions of the enteric plexuses. In the myenteric plexus, varicose immunoreactive nerve fibres disappeared or were reduced in number in ganglia anal to an interruption of the myenteric plexus. Up to about 2 mm on the anal side, all varicose immunoreactive fibers disappeared from the ganglia. At about 14-16 mm below an interruption, there were about 50% of the normal number of fibres in the myenteric ganglia and at about 24 mm the innervation was normal. In the submucosa, fibres immunoreactive for 5-hydroxytryptamine were absent from an area on the anal side following interruption of the myenteric plexus. From consideration of the pattern of disappearance, it is deduced that some myenteric nerve cell bodies send immunoreactive axons in an anal direction to supply submucous ganglia. The axons run for about 8 mm in the myenteric plexus, enter the submucosa and then run for a further 4 mm approximately. Thus, varicose fibres immunoreactive for 5-hydroxytryptamine, which occur around the enteric ganglion cells of both plexuses arise from nerve cell bodies in myenteric ganglia than send their axons in an anal direction.

Intracellular recordings were made from neurons of the guinea-pig submucous plexus and the actions of 5-hydroxytryptamine on the postsynaptic membrane and on evoked synaptic potentials were examined. 5-Hydroxytryptamine produced two types of direct postsynaptic responses: (1) A depolarization associated with a fall in input resistance was observed in all cells. Voltage-clamp and ion substitutions showed that this depolarization resulted primarily from an inward sodium current. This response could be as brief as 30 ms; it showed desensitization and was selectively abolished by 0.2-2 microM ICS 205-930. (2) A depolarization (or inward current) associated with a decreased conductance was observed in about 50% of neurons, usually after the first response was blocked by ICS 205-930. This response was due to a decreased potassium conductance; the minimum time course of this response was 8-10 s. It did not show desensitization and was not sensitive to blockade by currently available antagonists of 5-hydroxytryptamine, nicotinic and/or muscarinic receptors. Higher concentrations of 5-hydroxytryptamine were required to produce the sodium conductance increase than the potassium conductance decrease; 2-methyl-5-hydroxytryptamine was equally effective in producing these responses. 5-Hydroxytryptamine also caused a barrage of "spontaneous" nicotinic excitatory post-synaptic potentials which were sensitive to tetrodotoxin. This response desensitized, was blocked by ICS 205-930 and is presumed to reflect excitation of other cholinergic cell bodies in the plexus by the sodium conductance increase mechanism described. The evoked nicotinic excitatory postsynaptic potential and the adrenergic inhibitory postsynaptic potential were decreased by 5-hydroxytryptamine; a portion of this inhibition showed desensitization and was blocked by ICS 205-930 as well as by the muscarinic receptor antagonists, atropine and pirenzepine. The ICS 205-930-insensitive portion of this inhibition could not be attributed to activation of 5-hydroxytryptamine-1 or 5-hydroxytryptamine-2 receptors. Thus, the following conclusions are drawn: 5-hydroxytryptamine excites submucous plexus neurons by activating two distinct 5-hydroxytryptamine receptors. Activation of the 5-hydroxytryptamine-3 receptor (sensitive to ICS 205-930) produces a depolarization mediated by an increased sodium conductance. The same effect occurring in other cholinergic cell bodies initiates action potentials which are responsible for the 5-hydroxytryptamine-induced release of acetylcholine.(ABSTRACT TRUNCATED AT 400 WORDS)

Immunofluorescence methods have been used to determine the detailed distribution of vasoactive intestinal polypeptide (VIP), substance-P and enkephalin nerve fibres in fixed cryostat sections from guinea-pig duodenum, jejunum, ileum, caecum at the site of the taenia coli, and proximal and distal colon. A novel method is used involving immunostaining of tissue culture preparations of both myenteric and submucous plexuses. These preparations allow each plexus to be studied in isolation from all axonal input for the first time, since they provide unequivocal extrinsic denervation together with severance of any intrinsic connections between the plexuses. In tissue sections the most prominent sites of VIP and substance-P immunoreactive fibres are the ganglia of the myenteric and submucous plexuses, the circular muscle layer and the longitudinal muscle of the taenia coli. In addition, VIP is prominent in the lamina propria of the submucosa except in the caecum. Enkephalin-immunopositive fibres are restricted to the ganglia of the myenteric plexus, the circular muscle layer and the longitudinal layer of the taenia coli. The culture preparations reveal that intrinsic ‘VIP neurons’ are common in the submucous plexus of the caecum and colon. They are also present, but in much lower numbers, in the myenteric plexus of the small intestine and colon but are not found in the myenteric plexus of the caecum. Intrinsic ‘substance-P neurons’ are present in the myenteric plexus from the small intestine, caecum and colon as well as in the submucous plexus of the colon; intrinsic ‘substance-P neurons’ are not found in the submucous plexus of the caecum. ‘Enkephalin neurons’ are numerous in the myenteric plexus of the small intestine, caecum and colon but are absent from the submucous plexus. Immunoreactivity is compared in the normal and denervated caecum by both the histochemical method and by radioimmunoassay of tissue extracts. In conjunction with the studies on tissue cultures, the results provide evidence for intrinsic reciprocal connections between the myenteric and submucous plexus of the caecum by neurons containing VIP and substance-P. An extensive comparison of these results with data from functional studies shows that the distribution of VIP, substance-P and enkephalin fibres in the gut is broadly in agreement with present knowledge of the action of these peptides on gut tissue, if it is assumed that they function as neurotransmitters or neuromodulators. In some instances, however, peptide-containing fibres and pathways are found which do not correlate with present knowledge obtained from functional studies. These observations provide new clues to the role of peptide neurons in gut function.

Substance P-like immunoreactivity has been localized in whole mount preparations of the isolated layers of the guinea-pig ileum. Axons containing substance P formed dense networks around the nerve cells and ran in the primary, secondary and tertiary nerve bundles of the myenteric plexus. 3.6% of the nerve cell bodies of the myenteric plexus and 11.3% of the cell bodies in the submucous plexus showed immunoreactivity for substance P. Axons ran in fine nerve bundles parallel to the longitudinal muscle, between this muscle and the myenteric plexus. Axons containing substance P also ran in small nerve trunks parallel to the circular muscle throughout its thickness and in the deep muscular plexus at the base of this muscle coat. In the submucosa, these axons ramified amongst ganglion cells of the plexus and ran in the internodal strands. In addition they formed a perivascular network around submucous arteries and contributed to the paravascular nerves following these arteries. Axons containing substance P formed a delicate plexus in the mucosa. After extrinsic denervation the nerves containing substance P that were associated with submucous arteries, and some in the submucous plexus, disappeared. The nerves in the other areas were not detectably different from normal. Comparison with the distribution of somatostatin, enkephalin and vasoactive intestinal polypeptide indicated the neurons containing substance P constitute a separate population within the enteric nervous system.

Immunoreactive nerve cell bodies and fibres in the intestine have been examined using three antibody preparations raised against 5-hydroxytryptamine. Cross reactivity studies indicate that the substance localized was an hydroxylated indoleamine. In the guinea-pig small intestine, nerve cell bodies were located in the myenteric plexus and varicose fibres were found in the ganglia of the myenteric and submucous plexus. The nerve cell bodies had prominent short, broad processes and a single long process. Similar nerve cells and fibres were found in the guinea-pig stomach and large intestine and areas of intestine that were examined in mice, rabbits and rats. Properties of the neurons were examined in the small intestine of the guinea-pig. The immunoreactive material was depleted by treatment with reserpine, but not by guanethidine or 6-hydroxydopamine in dose sufficient to deplete noradrenaline stores in axons in the intestine. No depletion of 5-hydroxytryptamine by the neurotoxin 5, 7-dihydroxytryptamine was observed. After depletion by reserpine, immunoreactivity of the neurons could be restored by application in vitro of 5-hydroxytryptamine, 5,7-dihydroxytryptamine or 5-hydroxytryptophan. The restoration by 5-hydroxytryptophan was prevented by the inhibitor of L-aminoacid decarboxylase, benserazide. After reserpine treatment, immunoreactivity was not restored by tryptophan. Uptake of 5, 7-dihydroxytryptamine into the nerves was antagonized by fluoxetine. The distribution of neurons with 5-hydroxytryptamine-like immunoreactivity was compared with the distribution of enteric amine-handling neurons that take up and decarboxylate L-dopa. This comparison indicated that there are two classes of aromatic amine neuron in the guinea-pig small intestine, the enteric 5-HT neurons and enteric, non-5-HT, amine handling neurons.