The presence of aromatic L-amino acid decarboxylase in certain intestinal nerve cells

Aromatic L-amino acid decarboxylase (AADC) is crucial for brain development and motor functions

Aromatic L-amino acid decarboxylase (AADC) deficiency is a rare pediatric neuro-metabolic disease in children. Due to the lack of an animal model, its pathogenetic mechanism is poorly understood. To study the role of AADC in brain development, a zebrafish model of AADC deficiency was generated. We identified an aadc gene homolog, dopa decarboxylase (ddc), in the zebrafish genome. Whole-mount in situ hybridization analysis showed that the ddc gene is expressed in the epiphysis, locus caeruleus, diencephalic catecholaminergic clusters, and raphe nuclei of 36-h post-fertilization (hpf) zebrafish embryos. Inhibition of Ddc by AADC inhibitor NSD-1015 or anti-sense morpholino oligonucleotides (MO) reduced brain volume and body length. We observed increased brain cell apoptosis and loss of dipencephalic catecholaminergic cluster neurons in ddc morphants (ddc MO-injected embryos). Seizure-like activity was also detected in ddc morphants in a dose-dependent manner. ddc morphants had less sensitive touch response and impaired swimming activity that could be rescued by injection of ddc plasmids. In addition, eye movement was also significantly impaired in ddc morphants. Collectively, loss of Ddc appears to result in similar phenotypes as that of ADCC deficiency, thus zebrafish could be a good model for investigating pathogenetic mechanisms of AADC deficiency in children.

Developmental changes of the adrenergic network in the myenteric plexus of the porcine small bowel

Knowledge about the foetal development of the normal enteric nervous system (ENS) is crucial for the understanding of congenital and acquired functional abnormalities of the gut. The ENS is the largest and most complex division of the peripheral nervous system and consists of intrinsic and extrinsic components. Although previous studies have described sympathetic innervation of the myenteric plexus, little is known regarding its age-related changes. The aim of this study was to investigate the age-related changes in the sympathetic innervation of the myenteric plexus. Whole mount and paraffin sections of the small bowel specimens from six different age groups (60 and 90 days gestation; newborn; 4 and 12 weeks old; and adult) were stained using tyrosine hydroxylase immunohistochemistry. Specimens were then analysed using fluorescence and laser scanning microscopy in detail. The tyrosine hydroxylase positive nerve fibres were first seen within the myenteric plexus at 90 days of gestation (E90). There was a significant increase in nerve fibres and varicosities observed from E90 to 12 weeks of age and stabilisation thereafter. The degree of varicosities around the ganglia, clearly seen on the whole-mount preparations, was also noted to increase up to 12 weeks of age, after which time there was no general variation noted into adulthood. Our findings show, for the first time, that sympathetic innervation of the myenteric plexus starts in the last quarter of gestation and continues till 12 weeks of age. Segmental sympathetic denervation, following bowel resection and anastomosis, during this developmental period may explain the motility dysfunction seen in newborn infants operated for necrotising enterocolitis, bowel atresia and Hirschsprung's disease.

Enteric dopaminergic neurons: definition, developmental lineage, and effects of extrinsic denervation

The existence of enteric dopaminergic neurons has been suspected; however, the innervation of the gut by sympathetic nerves, in which dopamine (DA) is the norepinephrine precursor, complicates analyses of enteric DA. We now report that transcripts encoding tyrosine hydroxylase (TH) and the DA transporter (DAT) are present in the murine bowel (small intestine > stomach or colon; proximal colon > distal colon). Because sympathetic neurons are extrinsic, transcripts encoding TH and DAT in the bowel are probably derived from intrinsic neurons. TH protein was demonstrated immunocytochemically in neuronal perikarya (submucosal >> myenteric plexus; small intestine > stomach or colon). TH, DA, and DAT immunoreactivities were coincident in subsets of neurons (submucosal > myenteric) in guinea pig and mouse intestines in situ and in cultured guinea pig enteric ganglia. Surgical ablation of sympathetic nerves by extrinsic denervation of loops of the bowel did not affect DAT immunoreactivity but actually increased numbers of TH-immunoreactive neurons, expression of mRNA encoding TH and DAT, and enteric DOPAC (the specific dopamine metabolite). The fetal gut contains transiently catecholaminergic (TC) cells. TC cells are the proliferating crest-derived precursors of mature neurons that are not catecholaminergic and, thus, disappear after embryonic day (E) 14 (mouse) or E15 (rat). TC cells appear early in ontogeny, and their development/survival is dependent on mash-1 gene expression. In contrast, the intrinsic TH-expressing neurons of the murine bowel appear late (perinatally) and are mash-1 independent. We conclude that the enteric nervous system contains intrinsic dopaminergic neurons that arise from a mash-1-independent lineage of noncatecholaminergic precursors.

Differential expression of catecholamine synthetic enzymes in the caudal ventral pons

The analysis of colocalization of multiple catecholamine biosynthetic enzymes within the ventrolateral part of the medulla oblongata of the rat revealed distinct subpopulations of neurons within the C1 region (Phillips et al., J Comp Neurol 2001, 432:20-34). In extending this study to include the caudal pons, it was shown for the first time that the A5 cell group could be distinguished by the presence of immunoreactivity to tyrosine hydroxylase (TH), aromatic l-amino acid decarboxylase (AADC), and dopamine beta hydroxylase (DBH). A novel cell group was also identified. The cells within this new group were immunoreactive to DBH but not TH, AADC, or phenylethanolamine N-methyltransferase (PNMT) and will be referred to as the TH-, DBH+ cell group. The TH-, DBH+ neurons were not immunoreactive for either the dopamine or noradrenaline transporters, suggesting that these neurons do not take up these transmitters. A5 neurons were immunoreactive for the noradrenaline transporter but not the dopamine transporter (as previously shown). Retrograde tracing with cholera toxin B revealed that the TH-, DBH+ neurons do not project to the thoracic spinal cord or to the rostral ventrolateral medulla, but A5 neurons do. A calbindin immunoreactive cell group is located in a region overlapping TH-, DBH+ cell group. However, only a few neurons were immunoreactive for both markers. The physiological role of the TH-, DBH+ cell group remains to be determined.

Distribution and origin of aminergic neurones in dog small intestine

In dog ileum, axons containing immunoreactivity for tyrosine hydroxylase (TH) and for DOPA decarboxylase (DDC) invest the neurones of the enteric ganglia and are present around arterioles, in the deep plexus of the circular muscle and in the laminar propria of the mucosal villi. Immunoreactivity for serotonin (5-HT) was present in mucosal enterochromaffin cells and in varicose axons investing cells of the submucous plexus, but not in axons in myenteric plexus, circular muscle or mucosa. No enteric neuronal cell bodies were immunoreactive for DDC or for 5-HT. Two weeks after extrinsic denervation of the ileum, all TH staining was absent. By contrast, the pattern of DDC staining was unchanged, except around blood vessels. We conclude that motility of dog ileum is likely to be regulated by sympathetic noradrenergic inputs and by intrinsic serotonergic and 'amine-handling' neurones. Dopaminergic innervation of dog ileum appears to be restricted to a sparse vasomotor supply.

Some parasympathetic neurons in the guinea-pig heart express aspects of the catecholaminergic phenotype in vivo

In a histochemical study of intrinsic cardiac ganglia of the guinea-pig in whole-mount preparations, it was found that some 70-80% of the neurons express aspects of the catecholaminergic phenotype. These neurons have an uptake mechanism for L-DOPA, and contain the enzymes for converting L-DOPA (but not D-DOPA) to dopamine and noradrenaline, i.e. aromatic L-aminoacid decarboxylase and dopamine beta-hydroxylase. Monoamine oxidase is also present within some of the neurons. In these respects, the neurons resemble noradrenergic neurons of sympathetic ganglia, so we refer to them as intrinsic cardiac amine-handling neurons. However, these neurons do not contain tyrosine hydroxylase and show little or no histochemically detectable uptake of alpha-methyldopa, dopamine or noradrenaline, even after depletion of endogenous stores of amines by pre-treatment with reserpine. Noradrenergic fibres from the sympathetic chain form pericellular baskets around nerve cell bodies. The uptake of L-DOPA into nerve cell bodies is not prevented by treatment with 6-hydroxydopamine sufficient to cause transmitter-depletion or degeneration of the extrinsic noradrenergic fibres. Such degeneration experiments suggest that axons of the amine-handling neurons project to cardiac muscle, blood vessels and other intrinsic neurons. The cardiac neurons do not show any immunohistochemically detectable serotonergic characteristics; there is no evidence for uptake of the precursors L-tryptophan and 5-hydroxytryptophan or 5-HT itself, whereas the extrinsic noradrenergic nerve fibres within the ganglia can take up 5-HT when it is applied in high concentrations.

Extrinsic denervation elevates neuronal aromatic L-amino acid decarboxylase immunoreactivity in rat small intestine

In order to clarify further the neural control of digestive tract function, we have compared the neuronal localization of tyrosine hydroxylase (TH) and aromatic amino acid decarboxylase (AADC) in rat small intestine. Immunoreactivity for TH was found in numerous varicose axons associated with neurons of the enteric plexuses and in axons within the circular muscular coat and the mucosal villi. Axons with AADC immunoreactivity had a similar distribution, but were sparser in the enteric plexuses and musculature than those containing TH. Chronic extrinsic denervation of a segment of intestine removed all TH-positive nerves from that region. By contrast, the intensity of AADC immunoreactivity was enhanced and more AADC-positive axons were visible than in adjacent intact areas of intestine. The AADC-positive axons appear to represent the intrinsic 'amine-handling' neurons rather than intrinsic tryptaminergic neurons or extrinsic dopaminergic neurons, and the effect on AADC activity of removing the extrinsic nerve supply suggests that this normally exerts some restraining influence on the metabolism of the 'amine-handling' population.

Metabolic origins of 5-hydroxytryptamine in enteric neurons in a teleostean fish (Platycephalus bassensis), a toad (Bufo marinus) and the guinea-pig

1. 5-Hydroxytryptamine (5-HT) content and synthesis in mucosa-free intestine of guinea-pig, the teleost Platycephalus bassensis and the amphibian Bufo marinus was studied by HPLC with electrochemical detection or by TLC. 2. The 5-HT content of small intestine was: guinea-pig 0.58; Bufo: 1.23; Platycephalus: 26.88 nmol/g. 3. Intestine from each species synthesized 5-HT from exogenous 5-HTP. 4. Platycephalus preparations synthesized labelled 5-HT from 14C-tryptophan, but no labelled 5-HT was detected after similar incubation of guinea-pig or Bufo preparations. 5. Incubation of guinea-pig preparations with tryptophan did not increase tissue 5-HT or 5-HIAA content. 6. 5-HT in Platycephalus enteric neurons may be synthesized from tryptophan in situ; 5-HT in Bufo and guinea-pig neurons may be synthesized elsewhere, perhaps in enterochromaffin cells.

Increased gastric cholinergic activity evoked by 5-hydroxy-L-tryptophan in the rat

In gastrointestinal tissues such as rat stomach, exogenous 5-hydroxytryptamine (5HT) has little or no ability to affect nerve activity. However, endogenous 5HT might act differently, and this was investigated by stimulating 5HT synthesis using 5-hydroxy-L-tryptophan (5HTP). In longitudinal strips of rat forestomach, 5HTP (50 and 500 microM) increased cholinergically mediated contractions evoked by electrical field stimulation, probably by facilitating acetylcholine release; contractions evoked by exogenous acetylcholine were unaffected by 5HTP. The ability of 5HTP to increase electrically evoked contractions was long-lasting, required the presence of pyridoxal (a monoamine decarboxylase cofactor) and was reduced by the decarboxylase inhibitor carbidopa, but not by 6-hydroxydopamine. In the presence of paroxetine and nialamide, the 5HTP-induced increase in cholinergically mediated contractions was short-lasting. In anaesthetised rats 5HTP caused stimulation of gastric motility, which was blocked or reduced by atropine. These findings suggest that in the rat 5HTP stimulates gastric cholinergic activity, by increasing the concentration of 5HT at sites which normally synthesise 5HT.

The presence of extraganglionic fluorescent neurons in the myenteric plexus of the guinea pig small intestine

Fluorescent histochemical observations of the small intestine of the guinea pig demonstrated that single fluorescent cell bodies, separate from the ganglia, were present in the myenteric plexus. These cell bodies gave rise to single processes which entered the ganglia or the interganglionic strands. They were of a very small size, and the intensity of their fluorescence increased by pretreatment with L-DOPA and nialamide. Interruption of extrinsic nerve pathways to the small intestine caused a disappearance of the meshwork of fluorescent fibers in the myenteric plexus; but in some areas a fluorescent fiber which supplied its terminal to the ganglion was seen to remain. A photograph taken from the denervated myenteric plexus revealed that a long process arising from an extraganglionic cell entered the ganglion and ramified into terminal branches.

Choline acetyltransferase- and peptide immunoreactivity of submucous neurons in the small intestine of the guinea-pig

The peptides cholecystokinin (CCK), neuropeptide Y (NPY), somatostatin (SOM), substance P (SP) and vasoactive intestinal peptide (VIP), and the synthesizing enzyme for acetylcholine, choline acetyltransferase (ChAT) were localized immunohistochemically in nerve cell bodies of the submucous ganglia in the small intestine of the guinea-pig. VIP-like immunoreactivity was found in 45% of submucous neurons. ChAT immunoreactivity was observed in a separate group of nerve cells, which made up 54% of the total population. There were three subsets of neurons immunoreactive for ChAT: (1) ChAT neurons that also contained immunoreactivity for each of the peptides CCK, SOM and NPY, representing 29% of all submucous neurons; (2) ChAT neurons that also contained SP-like immunoreactivity, representing 11% of all submucous neurons, and (3) ChAT cells that did not contain any detectable amount of the peptides that were localized in this study.

Neurons with 5-hydroxytryptamine-like immunoreactivity in the enteric nervous system: their visualization and reactions to drug treatment

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.

The fine structure of the submucous plexus of the guinea-pig ileum. I. The ganglia, neurons, Schwann cells and neuropil

A fine structural study was made of the ganglia, neurons, Schwann cells and neuropil of the submucous plexus of the guinea-pig ileum. The arrangement of the plexus as seen by light microscopy is briefly described. Submucous ganglia are small, containing an average of eight neurons per ganglion (compared with 43 in myenteric ganglia) and are connected with each other by fine nerve strands. The cell bodies of neurons and Schwann cells and a neuropil consisting of neuronal and Schwann cell processes from the ganglia. No other cell types or blood vessels are found within the ganglia. Ganglia are surrounded by a continuous basal lamina but lack a well-defined connective tissue investment. The glial investment of neurons is incomplete: many neurons lie directly beneath the basal lamina with no intervening Schwann cell processes, and the plasma membranes of adjacent neurons are often directly apposed over large areas. Other areas of apposition occur between the cell bodies and processes of neurons and Schwann cells. Desmosome-like membrane specializations may be seen between neurons and other neurons or Schwann cells. Submucous neurons could not be categorized according to size, shape, organelle content or types of processes. Processes emerging from nerve-cell bodies were placed into four broad categories on the basis of shape and microtubule content. Many bundles of closely apposed small nerve profiles lacking intervening Schwann processes are found in the neuropil in addition to a large number of vesiculated varicosities, some of which are directly apposed to the plasma membranes of nerve-cell bodies. A small proportion of vesiculated profiles from synapses with nerve cell bodies, their processes and profiles in the neuropil. From their structure, submucous neurons appear to form a more homogeneous population than myenteric neurons. Because of their incomplete investment they are more likely to be freely exposed to substances diffusing in the extraganglionic tissue than are neurons of sympathetic ganglia.

The fine structure of the submucous plexus of the guinea-pig ileum. II. Description and analysis of vesiculated nerve profiles

A fine structural study has been made of the vesiculated nerve profiles of the submucous plexus of both normally innervated and extrinsically denervated segments of guinea-pig ileum. Two types of nerve profiles could be readily distinguished by their vesicular content after conventional fixation. The first type, comprising 5% of all intrinsic profiles, consisted of predominantly small vesicles containing electron dense material which usually formed a ring around the inner face of the vesicular membrane but sometimes partially or completely filled the vesicle. These profiles, termed ring-vesicle-containing profiles, remained after extrinsic denervation, and their vesicular content did not change following injection of reserpine or 5-hydroxydopamine. Thus ring-vesicle-containing profiles are not adrenergic. Profiles which were positive for the uranaffin method were similar in morphology and frequency of occurrence to ring-vesicle-containing profiles, although it is not possible to say that they are the same. The second type of profile, comprising 95% of all intrinsic profiles, contained varying proportions of large granular and small clear vesicles. These heterogeneous profiles were present in both normally innervated and extrinsically denervated tissue. Their vesicular content did not change following injection of reserpine, however, some profiles of this type in normally innervated, but not in extrinsically denervated, intestine contained electron dense deposits after injection of 5-hydroxydopamine. This means that noradrenergic profiles are a subpopulation of the heterogeneous profiles in normally innervated tissue. Analysis of intrinsic heterogeneous profiles showed that the proportion and packing density of large granular vesicles formed continuous distributions which did not provide any basis for further subdivision of this type of profile. Ring-vesicle-containing and heterogeneous profiles often formed synapses with neuronal cell bodies and processes. Two rarer types of profiles were also seen. The first type contained mainly small flattened vesicles which took up 5-hydroxydopamine and was not present in extrinsically denervated tissue. This type, like the group described above, is considered to be noradrenergic. The second rare type contained large numbers of lysosome-like dense bodies and vesicles of different sizes and content and was seen in both normally innervated and denervated tissue. This type probably represents spontaneously degenerating nerve profiles.

Branching patterns and projections of enteric neurons containing different putative transmitters

Methods have been developed for the microsurgical interruption of nerve pathways in the wall of the intestine and for the immunohistochemical localization of antigens in whole mounts. These methods have made it possible to determine accurately the distributions and projections of neurons containing different putative transmitters. In this week neurons with substance P, somatostatin and 5-hydroxytryptamine-like immunoreactivity are described. Projections of noradrenergic neurons are also demonstrated. Each substance is associated with a unique set of neurons having precise projections within the intestine.

[Effects of precursors and metabolites of catecholamines on motricity of isolated rat duodenum. Particular reference to dopamine]

Studying the effects, on the isolated rat duodenum motricity, of ten compounds precursors or metabolites of catecholamines, the following results were obtained: The direct metabolites of epinephrine and norepinephrine (metanephrine, normetanephrine), are either ineffective at concentrations below 5 X 10(-6) M, or weakly inhibitory at higher concentrations. Such inhibitory effects are prevented by alpha- and beta-blockers. 3-methoxy, 4-hydroxyphenylglycol and vanylmandelic acid have no significant effect. The catecholamine precursor, dopamine, the related compounds DOPA, 3 methoxytyramine, and to a lesser extent, 3-O methyl DOPA and homovanillic acid, have excito-motor effects at concentrations ranging mainly from 10(-7) M to 10(-5) M. At higher concentrations, the same compounds frequently exhibit inhibitory effects. The excito-motor effects might be due to a serotoninergic mechanism, since they are suppressed by the serotoninergic blocking agents methysergide and cyproheptadine. Furthermore, in the case of DOPA, we were able to establish a relationship between the excito-motor effects and duodenal serotonin stores. As for the inhibitory effects, they may be prevented by using alpha and beta blocking agents. Dihydroxyphenylacetic acid has no effect on the isolated rat duodenum motricity. The fact that dopamine and related compounds may have excitomotor effects at some concentrations, correlated with some physiopathological observations in man and animal allows some considerations about the eventual role of dopamine on intestinal motricity.

A unique population of uranaffin-positive intrinsic nerve endings in the small intestine

A positive uranaffin reaction was observed in the small (40-60 nm) diameter vesicles of some intestinal axons. There was no change in the number of reactive axons or the intensity of reaction after reserpine (5 mg/kg) or after interruption of axons reaching the intestine through the mesentery. The axons were found in the myenteric, submucous and deep muscular plexuses and in the circular muscle. Some uranaffin-positive axons formed synapses with neurons of the myenteric and submucous plexuses. It is concluded that these axons are not noradrenergic. The axons must represent one of the several nerve types which are known to be intrinsic to the intestine, but are as yet unidentified at an ultrastructural level. If, as has been postulated, the reaction localizes amine storage vesicles, the uranaffin-positive axons are probably the intrinsic amine-handling axons previously demonstrated histochemically.

The sites of action of 5-hydroxytryptamine in nerve-muscle preparations from the guinea-pig small intestine and colon

1 The sites of action of 5-hydroxytryptamine (5-HT) were examined in isolated segments of guinea-pig intestine. Mechanical records were taken from the longitudinal muscle of the ileum and proximal colon and from the circular muscle of the ileum and distal colon.2 In order to examine direct actions of 5-HT, nerve-mediated responses were blocked with tetrodotoxin (0.2 mug/ml). There was a gradient in the responsiveness of the longitudinal muscle of the ileum; in the proximal ileum it was usually unresponsive, whereas in the distal ileum about 30% of the amplitude of contraction was caused by a direct effect on the muscle. In the circular muscle from all parts of the ileum, direct effects on the muscle were weak or absent. In the distal colon, the circular muscle was almost always unresponsive to direct effects of 5-HT even when concentrations of 5-HT as great as 100 mug/ml were used. All direct actions of 5-HT on intestinal muscle were blocked by methysergide (1 mug/ml), which itself did not affect nerve-mediated responses.3 Excitatory cholinergic nerves and excitatory and inhibitory nerves which released unidentified substances were all stimulated by 5-HT. The contractions mediated through cholinergic nerves were blocked by hyoscine (0.6 mug/ml).4 Tachyphylaxis to the action of 5-HT occurred both for effects mediated through nerves and for direct effects on the muscle. Responses returned promptly after 5-HT was washed from the organ bath.5 While 5-HT blocked its own action on neural receptors, it did not antagonize the stimulation of nicotinic receptors on cholinergic neurones by 1-1 dimethyl-4-phenylpiperazinium iodide (DMPP). Moreover, pentolinium markedly reduced contractions caused by DMPP without significantly affecting responses to 5-HT. In contrast, (+)-tubocurarine, another nicotinic receptor antagonist, was effective in reducing contractions caused by 5-HT.6 Phenyldiguanide, which has been reported to antagonize the stimulant action of 5-HT on cholinergic neurones in the mouse small intestine, did not cause any significant reduction in the action of 5-HT on cholinergic neurones in the guinea-pig ileum unless a concentration of 1 mg/ml was used. However, contractions elicited by carbachol and DMPP were antagonized to a similar extent by phenyldiguanide at this concentration. Antagonism of the action of 5-HT at neural receptors by bromolysergic acid and by tryptamine was found but it was not specific, these drugs causing comparable decreases in responses to 5-HT, carbachol and DMPP.7 The present results, which show that 5-HT has little or no direct effect on the circular muscle of the ileum and colon, imply that, if 5-HT is a transmitter in enteric reflexes, it must be released from interneurones.

The innervation and fine structure of paraneuronic cells in an amphibian pulmonary artery

The pulmonary artery of Bufo marinus contains large numbers of bipolar cells situated in the tunica adventitia and in the outer layers of the media. These cells show a bright green-yellow fluorescence (emission spectra 485 nm) after formaldehyde pre-treatment suggesting that they contain a primary monoamine. The most characteristic fine-structural feature of these cells is the presence of numerous dense-cored vesicles (80--300 nm diameter) in their cytopalsm. The cells are in close contact (20 nm gap) with both agranular and granular nerve fibres. Both EM-cytochemical and formaldehyde-induced fluorescence tests indicate that the granule-containing nerve fibres are adrenergic. The agranular nerve fibres form discrete synaptic contacts with pre and post-synaptic membrane thickenings on the cells. This was never observed with respect to the adrenergic fibres. Each process of the cells is about 45 micrometer long. The processes do not bear any special relationship to either vessels of the arterial vasa vasorum or medial smooth muscle cells. Their location in the wall of the artery suggests that they are functionally significant with respect to activity of the arterial media.