Functional effects and characteristics of cecum-projecting neurons in the dorsal motor nucleus of the vagus of rats

Preganglionic neurons in the dorsal motor nucleus of the vagus (DMV) innervate most of the gastrointestinal tract; with the stomach and the cecum/proximal colon having a greater proportion of vagal input. Cecum-projecting neurons have been thought to be distinct from other preganglionic neurons due to their location within the DMV, but it is unknown whether these neurons innervate the cecum exclusively or what effect their activation has on cecal motor activity. Therefore, we investigated the extent of coinnervation of cecum and stomach by vagal neurons, their neurochemistry, and the effect of DMV stimulation on intracecal and intragastric volumes. Fluorescent retrograde tracers injected into the serosa of the cecum and stomach revealed that in the DMV 49+/-5% CTB-labeled cecum-projecting neurons also innervated the stomach. Immunocytochemical staining for nitric oxide (NO) synthase and tyrosine hydroxylase indicated that only 3+/-1% and 4+/-1% of cecum-projecting neurons contained these markers, respectively. In anesthetized rats gastric and cecal volumes were measured by prototypic miniaturized dual barostats that were developed for use in rodents. Microinjection of l-glutamate into the DMV increased gastric contractile activity and tone, and reduced on-going cecum contractile activity (2.6+/-0.7 contractions/2 min after injection versus 8.2+/-0.4 contractions/2 min before injection, N = 5). The barostat was able to detect decreases (-0.88+/-0.13 ml) and increases (0.25+/-0.05 ml) in cecum volume in response to carbachol and sodium nitroprusside, respectively. In summary, cecum-projecting neurons are not an entirely exclusive population within the DMV because a percentage of these also innervate the stomach. Central vagal stimulation can modulate both gastric and cecum contractile activity. Together, these data support a role of the vagus in neural reflexes involving gastric and large bowel motor function, such as the immediate phase of the gastrocolonic reflex.     

Ultrastructural identification of labeled neurons in the dorsal motor nucleus of the vagus nerve following injections of horseradish peroxidase into the vagus nerve and brainstem

The efferent connections of two types of neurons in the dorsal motor nucleus of the vagus nerve (DMV) were studied in the cat by light and electron microscopy following horseradish peroxidase (HRP) injections into the cervical vagus nerve or brainstem. After injections of HRP into the vagus nerve, up to 80% of medium-sized neruons averaging 26 × 20 μm in 1-μm-thick sections were retrogradely labeled while no small neurons were labeled in the DMV. Incubation with either diaminobenzidene (DAB) or p-phenylenediamine-pyrocatechol (PPD-PC) chromogens yielded electron-dense reaction products localized mainly in lysosomes. Identification of label at the ultrastructural level was facilitated by omitting lead citrate staining and by counting numbers of lysosomes, which were higher in labeled neurons. Quantitative comparisons of the dimensions of labeled and unlabeled somata demonstrated that retrograde transport and incorporation of HRP had no effect on cell size within the 2–3-day survival times used in this study. In order to determine whether neurons in the DMV project to higher levels of the brain stem, large injections of HRP (1–3 μl) were made into the pons, mesencephalon, hypothalamus, and amygdala. After injections of HRP into the brainstem, only small neurons, measuring 17 × 10 μm, were retrogradely labeled. Approximately 90% of the small neurons remained unlabeled following the HRP injections. The ultrastructrual features of the labeled small neurons included an invaginated nucleus, low cytoplasmic/nuclear ratio, and relatively fewer organelles than the medium-sized neurons. A quantitative analysis of labeled and unlabeled small neurons demonstrated that the labeled neurons were significantly larger than the unlabeled small neurons. Thus, two populations of small neurons may exist in the DMV. One population appears to have ascending projections to higher levels of the brainstem while the other more numerous population may be interneurons or project for only short distances.     

Morphology of identified preganglionic neurons in the dorsal motor nucleus of the vagus.

To determine the degree of variation of neuronal morphology both within and between the subnuclei of the dorsal motor nucleus of the vagus (dmnX), structural features of the preganglionic neurons of each of the five primary subnuclei in the rat dmnX were characterized quantitatively. Each of the columnar subnuclei was separately labeled by application of the retrograde tracer fast blue to its corresponding subdiaphragmatic vagal branch. Fixed brain slices of 100 microns thickness were then prepared in coronal, sagittal, and horizontal orientations. Next, randomly selected fast blue labeled neurons (n = 1,256) were injected with Lucifer yellow, drawn with camera lucida, and digitized. For each cell, three features of the perikaryon and twelve of the dendritic tree were measured. Dorsal motor nucleus neurons with up to eight primary dendrites, 30 dendritic segments, and seventh order dendritic branches were observed. Throughout the dmnX, the dendrites of preganglionic neurons were preferentially oriented in the horizontal plane. Consistent with an organizing role for the columnar subnuclei, most dendrites remained within their column of origin. However, between 5 and 30% of the neurons in each of the columns projected dendrites into adjacent dmnX subnuclei or other brainstem nuclei, including the nucleus of the solitary tract (NTS). The cyto- and dendroarchitectural analyses revealed systematic gradations in morphology, although they did not support the idea that the dmnX was composed of multiple distinct preganglionic types. The most parsimonious interpretation of the data is that dmnX motorneurons are variants of a single prototype, with dendrites varying widely in length and degree of ramification. The extent of an individual preganglionic neuron's dendritic field was predicted by three factors: the cell's rostrocaudal position within the dmnX, its location within a transverse plane (i.e., its coronal position within or ectopic to the dmnX), and its subnucleus of origin. Neurons at rostral and midlongitudinal levels of each column had more extensive dendritic arbors than those at caudal levels. Ectopic neurons had more extensive dendritic fields than similar cells in the corresponding columns; in fact, of all vagal preganglionic neurons, ectopics had the most extensive dendritic fields. Somata and dendrites of celiac column neurons were more extensive than those of hepatic and gastric column cells. These differential regional distributions of vagal preganglionics suggest that their structure and function are correlated.     

Distribution and ultrastructure of dopaminergic neurons in the dorsal motor nucleus of the vagus projecting to the stomach of the rat

Almost all parasympathetic preganglionic motor neurons contain acetylcholine, whereas quite a few motor neurons in the dorsal motor nucleus of the vagus (DMV) contain dopamine. We determined the distribution and ultrastructure of these dopaminergic neurons with double-labeling immunohistochemistry for tyrosine hydroxylase (TH) and the retrograde tracer cholera toxin subunit b (CTb) following its injection into the stomach. A few TH-immunoreactive (TH-ir) neurons were found in the rostral half of the DMV, while a moderate number of these neurons were found in the caudal half. Most of the TH-ir neurons (78.4%) were double-labeled for CTb in the half of the DMV caudal to the area postrema, but only a few TH-ir neurons (5.5%) were double-labeled in the rostral half. About 20% of gastric motor neurons showed TH-immunoreactivity in the caudal half of the DMV, but only 0.3% were TH-ir in the rostral half. In all gastric motor neurons, 8.1% were double-labeled for TH. The ultrastructure of the TH-ir neurons in the caudal DMV was determined with immuno-gold-silver labeling. The TH-ir neurons were small (20.4 x 12.4 microm), round or oval, and contained numerous mitochondria, many free ribosomes, several Golgi apparatuses, a round nucleus and a few Nissl bodies. The average number of axosomatic terminals per section was 4.0. More than half of them contained round synaptic vesicles and made asymmetric synaptic contacts (Gray's type I). Most of the axodendritic terminals contacting TH-ir dendrites were Gray's type I (90%), but a few contained pleomorphic vesicles and made symmetric synaptic contacts (Gray's type II).     

Synaptic remodeling in the dorsal motor nucleus of the vagus nerve following vagal-hypoglossal nerve anastomosis in the cat

Horseradish peroxidase (HRP) retrograde tracing techniques and morphometric analyses were performed to investigate synaptic remodeling associated with neuronal and glial changes in the dorsal motor nucleus of the vagus (DMV) of cats after vagal-hypoglossal nerve anastomosis (VHA). At 25 days postoperation (dpo), in the early target-reinnervation stage, there were 50% fewer presynaptic boutons containing round vesicles (R) or round and large dense-cored synaptic vesicles (R+D) contacting HRP-labeled DMV motoneurons. The loss of R boutons was maintained throughout the remaining postoperative intervals up to 500 dpo, whereas R+D boutons were further reduced at 123 dpo but were restored at 315 dpo, so that, by 500 dpo, 71.4% of them had gained access to the DMV motoneurons. Boutons containing pleomorphic synaptic vesicles (P) were completely disconnected from the DMV motoneurons at 25 dpo and did not reappear even in the long-term reinnervation stage. Loss and recovery of presynaptic boutons occurred in parallel with changes in astroglial ensheathment of the DMV motoneurons. It is suggested that synaptic remodeling associated with astroglial ensheathment in the DMV may be influenced by some retrogradely transported factors/signals derived from the newly acquired target organ, viz. tongue skeletal musculature. Our results further suggest that the observed changes in boutonal configurations may be attributable to modified functions of the DMV motoneurons induced by VHA.     

Induction of apoptosis by thrombin in the cultured neurons of dorsal motor nucleus of the vagus

Background A previous study demonstrated the presence of protease‐activated receptor (PAR) 1 and 2 in the dorsal motor nucleus of vagus (DMV). The aim of this study is to characterize the effect of thrombin on the apoptosis of DMV neurons. Methods The dorsal motor nucleus of vagus neurons were isolated from neonatal rat brainstems using micro‐dissection and enzymatic digestion and cultured. Apoptosis of DMV neurons were examined in cultured neurons. Apoptotic neuron was examined by TUNEL and ELISA. Data were analyzed using anova and Student’s t‐test. Key Results Exposure of cultured DMV neurons to thrombin (0.1 to 10 U mL^?1) for 24 h significantly increased apoptosis. Pretreatment of DMV neurons with hirudin attenuated the apoptotic effect of thrombin. Similar induction of apoptosis was observed for the PAR1 receptor agonist SFLLR, but not for the PAR3 agonist TFRGAP, nor for the PAR4 agonist YAPGKF. Protease‐activated receptors 1 receptor antagonist Mpr(Cha) abolished the apoptotic effect of thrombin, while YPGKF, a specific antagonist for PAR4, demonstrated no effect. After administration of thrombin, phosphorylation of JNK and P38 occurred as early as 15 min, and remained elevated for up to 45 min. Pretreatment of DMV neurons with SP600125, a specific inhibitor for JNK, or SB203580, a specific inhibitor for P38, significantly inhibited apoptosis induced by thrombin. Conclusions & Inferences Thrombin induces apoptosis in DMV neurons through a mechanism involving the JNK and P38 signaling pathways.     

Serum-free culture of rat postnatal neurons derived from the dorsal motor nucleus of the vagus

Previous studies on dorsal motor nucleus of the vagus (DMNV) neurons have mainly used in vivo animal models and in vitro brainstem slices. Primary culture of postnatal DMNV neurons in defined serum free medium has not been reported. We report a method for culture of postnatal rat DMNV neurons using serum free medium. Cultured DMNV neurons contain both Hu positive precursor cells and mature cells staining positively for microtubule associated protein 2 (MAP2) and choline acetyltransferase. Exposure of cultured DMNV neurons to glutamate (10(-7) to 10(-3)M) induced an increase in intracellular calcium concentration ([Ca(2+)](i)) in a dose-dependent manner, indicating the functional presence of glutamate receptors. Voltage-dependent calcium currents were present in cultured DMNV neurons. Active cell proliferation was demonstrated by BrdU incorporation. Upon removal of beta FGF, the percentage of MAP2 positive mature neurons was significantly increased from 36+/-3 to 73+/-3%. Our study demonstrates that postnatal rat DMNV neurons cultured in serum free medium retain morphological and physiological characteristics of DMNV neurons in situ.     

Stimulatory action of oxytocin on neurones of the dorsal motor nucleus of the vagus nerve

Extracellular recordings were obtained from spontaneously active neurones located in the dorsal motor nucleus of vagus nerve ( DMX ) in slices of the rat brainstem. Oxytocin applied to the bath at concentrations of 10(-7) M or 10(-6) M excited 79% of these cells in a concentration-dependent, reversible manner. The remaining cells were unaffected. The stimulatory effect of oxytocin was reversibly antagonized by a synthetic structural analogue known to block the peripheral, endocrine effects of neurohypophysial peptides. A selective oxytocic agonist was as potent as oxytocin, whereas vasopressin exerted a much weaker effect. We therefore suggest that neurones located in DMX are endowed with receptors for oxytocin.     

Increased expression of phosphotyrosine after axotomy in the dorsal motor nucleus of the vagus nerve and the hypoglossal nucleus

To investigate the role of tyrosine kinase underlying glial cell proliferation after axotomy, the localization of phosphotyrosine was studied immunohistochemically in the dorsal motor nucleus of the vagus nerve and the hypoglossal nucleus after nerve transection in adult rats. An anti-phosphotyrosine antibody weakly stained the cytoplasm of the neurons and some glial cells on the control side of both nuclei, while preferentially staining the plasma membrane of perineuronal microglial cells and neurons weakly on the severed side 2 days after axotomy and intensely between 3 and 7 days. Some of the microglial cells reacted positively with both anti-bromodeoxyuridine and antiphosphotyrosine antibodies, suggesting that tyrosine kinase is involved in microglial cell proliferation. Proliferation of numerous microglial cells was observed in the severed nuclei between 2 and 4 days after axotomy, while only a few were detected on days 5 and 7. These findings suggest that tyrosine kinase is involved in not only the proliferation of perineuronal microglial cells but also in some retrograde neuronal reactions such as differentiation and regeneration.     

Chemosensitivity of neurons in the dorsal motor nucleus of the vagus that responded to portal infusion of hypertonic saline in rats

Neural responses in the dorsal motor nucleus of the vagus (DMV) to topical administrations of sodium and portal infusions of hypertonic saline were investigated electrophysiologically by using multibarrel electrodes in anesthetized rats. Of 102 neurons that showed antidromic response to electrical stimulation of the ventral gastric vagus or the accessory celiac vagus, 51 neurons increased and 13 neurons decreased their discharge rates in response to the electrophoretic administration of sodium. The other 38 neurons did not respond to this stimulation. The portal infusion of hypertonic saline elicited neural responses of some DMV neurons whose axons are involved into either the ventral gastric or the accessory celiac vagus. Further, effects of the topical administration and the portal infusion of hypertonic saline were examined on 33 neurons. Typical response was characterized by an increase in discharge rate responding to both of the portal infusion and the topical administration. In conclusion, the DMV neurons receiving the afferent inputs from hepatoportal osmoreceptors may have an enteroceptor function detecting the change in osmotic pressure of their environment.     

The central organization of the vagus nerve innervating the stomach of the rat

We employed the neural tracers cholera toxin-horseradish peroxidase and wheat germ agglutinin-horseradish peroxidase to examine the organization of the afferent and efferent connections of the stomach within the medulla oblongata of the rat. The major finding of this study is that gastric motoneurons of the dorsal motor nucleus (DMN) possess numerous dendrites penetrating discrete regions of the overlying nucleus of the solitary tract (NTS). In particular, dendritic labelling was present in areas of NTS which also received terminals of gastric vagal afferent fibers such as the subnucleus gelatinosus, nucleus commissuralis, and medial nucleus of NTS. This codistribution of afferent and efferent elements of the gastric vagus may provide loci for monosynaptic vagovagal interactions. A small number of dendrites of DMN neurons penetrated the ependyma of the fourth ventricle and a few others entered the ventral aspect of the area postrema, thus making possible the direct contact of preganglionic neurons with humoral input from the cerebrospinal fluid and/or the peripheral plasma. Nucleus ambiguus neurons projecting to the stomach predominantly innervate the forestomach. The dendrites of these cells, when labelled, were generally short, and extended beyond the compact cluster of ambiguus neurons in a ventrolateral direction, parallel to the fascicles of vagal efferent fibers traversing the medulla.     

Hyperpolarizing action of enkephalin on neurons in the dorsal motor nucleus of the vagus, in vitro

Intracellular recordings were made from neurons of the dorsomotor vagal nucleus (DMV) in slices of rat medulla oblongata. [D-Ala2, D-Leu5]-enkephalin (DADLE), applied by perfusion (0.01-3 microM) or droplets, dose-dependently hyperpolarized 85% of the DMV neurons tested. The hyperpolarization, associated with a decrease in membrane resistance, persisted after elimination of synaptic activity by perfusion with Ca2(+)-free/high-Mg2+ solution or with 1 microM TTX solution. The opioid antagonist, naloxone, reversibly inhibited DADLE-induced hyperpolarization. The hyperpolarization depended on extracellular K+ concentration and reversed at about -90 mV. DADLE also decreased Ca2(+)-dependent spike duration and after-hyperpolarization (AHP). DAGO (a selective mu-receptor agonist), but not DPLPE (a selective delta-receptor agonist), mimicked DADLE's effects on membrane potential, Ca2(+)-dependent spike duration, and AHP. It is concluded that DADLE, through postsynaptic mu-type opioid receptors, hyperpolarized DMV neurons by increasing K+ conductance, which may have an inhibitory effect on DMV output. DADLE-induced decrease of spike duration and AHP was also mediated by mu-receptors and could have additional effects on functions of the DMV neuron by virtue of reduction in Ca2+ entry.     

Access to gastric tissue promotes the survival of axotomized neurons in the dorsal motor nucleus of the vagus in neonatal rats.

Lesioning the vagus nerve in the neck (cervical vagotomy) results in a rapid and virtually complete loss of motoneurons in the dorsal motor nucleus of the vagus in neonatal rats. The present study sought to determine whether access to gastric target tissue will promote the survival of these motoneurons after axotomy. Quantitative analysis demonstrates that subdiaphragmatic vagotomy, which leaves the cut vagal axons in close proximity to their normal gastric targets, results in significantly less motoneuron loss than cervical vagotomy. Furthermore, the loss of motoneurons after cervical vagotomy can be significantly reduced by transplanting embryonic gastric tissue to the neck of vagotomized neonatal host rats, in the vicinity of the cut axons. The survival effect of transplanted gastric tissue appears specific because control transplants of embryonic bladder tissue fail to reduce motoneuron death after cervical vagotomy. Injections of the neural tracers Fluoro-Gold and cholera toxin-horseradish peroxidase into gastric transplants labeled surviving motoneurons in cervically vagotomized rats, whereas tracer injections into bladder transplants or into host cervical tissues did not. These results indicate that neonatal vagal motoneurons are capable of making the adjustments necessary to survive axotomy if they have access to gastric target cells. The apparent dependence of injured neonatal vagal motoneurons on gastric tissue offers a new system in which to examine in vivo the trophic interactions between neurons and their targets.     

Oxytocinergic inputs to the nucleus of the solitary tract and dorsal motor nucleus of the vagus in neonatal rats

The paraventricular nucleus of the hypothalamus (PVN) modulates vagal digestive motor functions via oxytocinergic projections to the nucleus of the solitary tract (NST) and dorsal motor nucleus of the vagus (DMV) in adult rats. Little is known regarding the structural or functional maturation of these projections. The present study examines the postnatal development of immunocytochemically identified oxytocinergic fibers in gastric subregions of the medial NST-DMV. For this purpose, a monoclonal antibody (PS36) that recognizes both oxytocin (OT)-neurophysin and its prohormone was used to identify oxytocinergic fibers. PS36-positive fibers already were present within the NST-DMV in rats on the day of birth. Retrograde transport of cholera toxin neural tracer from the NST-DMV in newborn rats confirmed that PVN neurons were the sole source of these oxytocinergic fibers. The cumulative length of PS36-positive fibers in sampled subregions of the medial NST and DMV increased approximately 23-fold and 94-fold, respectively, between birth and adulthood. The observed postnatal increases in PS36 immunolabeling could reflect increased delivery of immunoreactive antigen from hypothalamic perikarya to distal axons and/or increasing oxytocinergic innervation of the NST-DMV. Additional work will be needed to address these questions and to determine the time course during which central oxytocinergic pathways become mature in their ability to influence vagally mediated digestive functions. J. Comp. Neurol. 399:101–109, 1998. © 1998 Wiley-Liss, Inc.     

Electrophysiologic identification of preganglionic neurons in rat dorsal motor nucleus and analysis of vagus afferent projections

Electrophysiological studies on preganglionic neurons (PGNs) in the dorsal motor nucleus (DMN) of the vagus nerve has been hampered by technical limitations. Conventional electrical stimulation of the vagus nerve with cathodal square-wave pulses activates both preganglionic and afferent fibers. Thus, some PGNs cannot be identified because the anticipated antidromic responses would be blocked due to collision with those orthodromic responses evoked with shorter latencies by activation of fast-conducting afferent fibers. Projections of vagus afferent fibers to PGNs are difficult to analyse with conventional methods because a preceding antidromic response may affect an orthodromic response which has a slightly longer latency. A new stimulation method was designed, consisting of anodal triangular pulse stimulation and spontaneous-spike triggered stimulation. In chlorase-urethane-anesthetized rats, unitary responses of the DMN to electrical stimulation of the ipsilateral vagus nerve were recorded. When only an orthodromic response by a DMN neuron was recorded with conventional stimulation, application of anodal triangular pulse stimulation revealed an antidromic response, so that the cell in question could be identified as a PGN. Some neurons that produced only an antidromic response to conventional stimulation, revealed an orthodromic response on spontaneous-spike triggered stimulation, which blocked the antidromic response due to collision. With these procedures, orthodromic responses due to vagus afferent projections were recorded in 35% of the identified PGNs, mostly due to C and partly, A fiber activations. All these projections were polysynaptic in nature. In conclusion, one-third of the PGNs of the DMN are involved in vagovagal reflexes, which occur through multisynaptic pathways.     

A light and electron microscopic study of the dorsal motor nucleus of the vagus nerve in the cat

The morphology of the dorsal motor nucleus of the vagus nerve (DMV) in the cat was studied with the aid of light and electron microscopy. In frozen sections stained by the Kluver-Barrera method or stained to show retrograde labeling in the DMV following injections of horseradish peroxidase (HRP) in the cervical vagus nerve and the stomach wall a range of sizes of DMV neurons was observed but it was not possible to distinguish separate types. In contrast, two distinct types of neurons, one medium-sized and the other small, were identified with the light microscope in Golgi-Cox and 1-μm Epon sections and with the electron microscope in ultrathin sections. The medium-sized neurons had a range of sizes but generally measured 18 × 25 μm and possessed three to four proximal dendrites which branched two or three times. Spines were observed occasionally on the soma and on dendrites. These neurons contained a well-developed cytoplasm and a noninvaginated round to oval nucleus. The small neurons generally measured 9 × 14 μm and were round or slightly elongated in shape. Their dendritic processes were fewer and thinner than those of the medium-sized neurons and extended for shorter lengths. Their soma contained scanty cytoplasm and an invaginated nucleus. The medium-sized neurons outnumbered the small neurons by more than three to one but both neuronal types were distributed evenly throughout the nucleus. The medium-sized neurons seemed to correspond in size to the parasympathetic efferent neurons of the viscera as indicated by the HRP studies. Axosomatic synapses on both types of neurons and axodendritic synapses were observed in the DMV. Terminals containing mainly small clear round vesicles and making asymmetrical contact with the postsynaptic membrane were involved in the majority of synapses on both the soma and dendrites. Terminals containing predominantly pleomorphic vesicles and making symmetrical contact with the postsynaptic membrane were also common, comprising up to one-third of all synapses observed. Serial sections revealed that most synaptic terminals contained varying numbers of large (75–110 nm) dense-cored vesicles. Smaller dense-cored vesicles (45–55 nm) were sometimes observed, often close to the area of synaptic contact. Terminals 1–2 μm in diameter which contacted dendrites 1–3 μm in diameter formed the most common synaptic combination throughout the rostral to caudal extent of the DMV. No distinct regional differences were observed with respect to distribution of synaptic types.     

Direct synaptic contacts on the myenteric ganglia of the rat stomach from the dorsal motor nucleus of the vagus

The myenteric ganglia regulate not only gastric motility but also secretion, because a submucous plexus is sparsely developed in the rodent stomach. We have examined whether the neurons of the dorsal motor nucleus of the vagus (DMV) have direct synaptic contacts on the myenteric ganglia and the ultrastructure of the vagal efferent terminals by using wheat germ agglutinin conjugated to horseradish peroxidase (WGA-HRP). The myenteric ganglia of the rat were composed of four types of neurons, i.e., small, medium-sized, large, and elongated neurons. The average numbers of axosomatic terminals per profile were 2.0 on the small neurons, 3.1 on the medium-sized neurons, 1.2 on the large neurons, and 4.2 on the elongated neuron. More than half of the axosomatic terminals contained round vesicles and formed asymmetric synaptic contacts on the small, medium-sized, and large neurons. About 80% of the axosomatic terminals on the elongated neurons contained pleomorphic vesicles and formed asymmetric synaptic contacts. When WGA-HRP was injected into the DMV, many anterogradely labeled terminals were found around the myenteric neurons. The labeled terminals were large (3.16 +/- 0.10 microm) and contacted exclusively the somata. Most of them (about 90%) contained round vesicles and formed asymmetric synaptic contacts. Serial ultrathin sections revealed that almost all neurons in a ganglion received projections from the DMV. The vagal axon terminals generally contacted the medium-sized or the elongated neurons, whereas a few labeled terminals contacted the small and the large neurons. The present results indicate that the DMV projects to all types of neurons and that their axon terminals contain mostly round synaptic vesicles and form asymmetric synaptic contacts.