Distribution of cocaine- and amphetamine-regulated transcript (CART), neuropeptide Y (NPY) and galanin (GAL) in the pterygopalatine ganglion of the domestic duck (Anas platyrhynchos f. domestica).

INTRODUCTION: Cocaine- and amphetamine-regulated transcript (CART), neuropeptide Y (NPY) and galanin (GAL) act as neurotransmitters and neuromodulators in both the central and peripheral nervous systems. Their presence has been found in different taxonomic groups, in particular in mammals. However, only few investigators have studied these neuropeptides in the class Aves (birds). The aim of the present study was to describe the distribution of CART, NPY and GAL in the pterygopalatine ganglion (PPG) of the domestic duck (Anas platyrhynchos f. domestica). MATERIAL AND METHODS: The experiment was conducted on 16 one-year-old domestic ducks of the Pekin breed of both sexes (8 males and 8 females). Frozen sections of the PPG were subjected to immunofluorescence staining using primary mouse monoclonal antibodies directed against CART and GAL and rabbit polyclonal antibody directed against NPY. Secondary antibodies were conjugated with Cy3 and FITC fluorochromes. RESULTS: CART, NPY, and GAL were present in the PPG of the domestic duck. The highest immunoreactivity (IR) in the ganglionic cells was found for CART in the majority (83-85%) of neurons of both superior (SPPG) and inferior (IPPG) PPG. CART-IR was also found in small aggregations of neurons on the medial surface of the Harderian gland, and on the course of the palatine branch of the facial nerve. CART-IR was also observed in the nerve fibers of these neurons' aggregations; however, it was low in comparison to the immunoreactivity of the perikarya. Immunoreactivity of NPY was found in ganglionic neurons, but above all in numerous fibers of the SPPG and IPPG and within aggregations on the surface of the Harderian gland. NPY-IR cells were distributed irregularly over the cross-sections of the tested aggregations, and constituted from 36% to 43% of the SPPG and from 37% to 40% of the IPPG of all cross-sectioned neurons. GAL-immunoreactive perikarya, distributed irregularly across the sections, were observed in the SPPG, where they constituted 61-65%, and in the IPPG, where they made up 50-57% of all neurons. All immunoreactive neurons were characterized by immunopositive neuroplasm and immunonegative cell nuclei. CONCLUSIONS: The presence of CART, NPY, and GAL in the PPG of the domestic duck suggests that these peptides may contribute to the secretory innervation of the glands of the mucosa of the palate and nasal cavity, the Harderian gland, and the lacrimal gland.

Physical activity, by enhancing parasympathetic tone and activating the cholinergic anti-inflammatory pathway, is a therapeutic strategy to restrain chronic inflammation and prevent many chronic diseases.

Chronic diseases are the leading cause of death in the world and chronic inflammation is a key contributor to many chronic diseases. Accordingly, interventions that reduce inflammation may be effective in treating multiple adverse chronic conditions. In this context, physical activity is documented to reduce systemic low-grade inflammation and is acknowledged as an anti-inflammatory intervention. Furthermore, physically active individuals are at a lower risk of developing chronic diseases. However the mechanisms mediating this anti-inflammatory phenotype and range of health benefits are unknown. We hypothesize that the "cholinergic anti-inflammatory pathway" (CAP) mediates the anti-inflammatory phenotype and range of health benefits associated with physical activity. The CAP is an endogenous, physiological mechanism by which acetylcholine from the vagus nerve, interacts with the innate immune system to modulate and restrain the inflammatory cascade. Importantly, higher levels of physical activity are associated with enhanced parasympathetic (vagal) tone and lower levels of C-reactive protein, a marker of low-grade inflammation. Accordingly, physical activity, by enhancing parasympathetic tone and activating the CAP, may be a therapeutic strategy to restrain chronic inflammation and prevent many chronic diseases.

Vagal activities are involved in antigen-specific immune inflammation in the intestine.

BACKGROUND AND AIMS: The mechanism of intestinal immune inflammation, such as food allergy, remains to be further understood. The present study aims to investigate the role of the vagal nerve in the pathogenesis of skewed T-helper 2 (Th2) responses in the intestine. METHODS: The expression of the immunoglobulin E (IgE) receptor on the vagus nerve in the mouse intestine was observed by immunohistochemistry. Vagus ganglion neurons (VGN) were isolated from mice and cultured in vitro. The IgE receptor/IgE complex on vagus neurons was examined by immune precipitation assay. A food allergy mouse model was developed; the effect of the partial removal of the vagal nerve (PRVn) via surgery or administration with anticholinergic agents on the suppression of Th2 inflammation was evaluated. RESULTS: The high-affinity IgE receptor was detected on the intestinal vagus nerve. An increase in the expression of the IgE receptor on the vagus nerve was observed in the intestines of mice with intestinal immune inflammation. Isolated mouse VGN express IgE receptor I, which could form complexes with IgE. Re-exposure to specific antigens activated the sensitized VGN, manifesting the release of transmitter glutamate that could activate dendritic cells by increasing the expression of CD80 and major compatibility complex class II and suppressing interleukin-12. The PRVn suppressed Th2 inflammation in the intestine. CONCLUSIONS: The intestinal vagus nerve in mice expresses a high-affinity IgE receptor. An antigen-specific immune response can activate the vagus nerve in the intestine and induces the release of transmitters to modulate dendritic cell phenotypes that facilitate the development of skewed Th2 polarization in the intestine.

Adie's pupils in paraneoplastic ganglionopathy with ANNA-1.

Autonomic disturbances are common in patients with paraneoplastic syndromes associated with type-1 antineuronal nuclear autoantibodies (ANNA-1), although pupillary disturbances are infrequent. The authors describe a patient with ANNA-1 associated paraneoplastic sensory neuronopathy and bilateral Adie's pupils.

The effect of vagal neural crest ablation on the chick embryo cloaca.

The cloaca, the caudal limit of the avian gastrointestinal tract, acts as a collecting chamber into which the gastrointestinal, urinary, and genital tracts discharge. It is intrinsically innervated by the enteric nervous system, which is derived from neural crest emigres that migrate from the vagal and sacral regions of the neural tube. Abnormal cloacal development can cause a number of anorectal anomalies, including persistent cloaca. Ablation of the vagal neural crest has previously been shown to result in an aganglionic hindgut to the extent of the colorectum. The aim of our study was to investigate the effect of vagal neural crest ablation on the cloaca, the limit of the hindgut in the developing chick embryo. Chick embryos were incubated until the 10-12 somite stage. The vagal neural tube corresponding to the level of somites 3-6 was then ablated, and eggs were incubated until harvested on embryonic day 11 (E11). Whole chick embryos were fixed, embedded in paraffin, and sectioned. Immunohistochemistry was then carried out using the HNK-1 monoclonal antibody to label neural crest cells, and results were assessed by light microscopy. Vagal neural crest ablation resulted in a dramatic decrease in the number of neural crest cells colonizing the chick embryo cloaca compared with control embryos. Ablated embryos contained only a small number of HNK-1-positive neural crest cells, which were scattered within the myenteric plexus in a disorganised pattern. Hypoganglionosis was also evident in other regions of the hindgut in ablated embryos. Ablation of the vagal neural crest results in a hypoganglionic cloaca in addition to hypoganglionosis of the hindgut. These results suggest that the cloaca is largely innervated by vagal neural crest emigres. Further studies involving quail-chick chimeras to investigate the exact contribution provided by both vagal and sacral neural crest cells to the cloaca should increase our understanding of the pathophysiology of conditions like persistent cloaca.

Autoantibodies against autonomic nervous tissues in type 2 diabetes mellitus: no association with cardiac autonomic dysfunction.

There is evidence that autoimmune factors contribute to the pathogenesis of cardiac autonomic dysfunction in Type 1 Diabetes mellitus (DM). To evaluate the presence of autoantibodies against autonomic nervous tissues in Type 2 DM, 127 patients were studied for complement-fixing sympathetic and parasympathetic ganglia (CF-SG and CF-PSG) autoantibodies with an indirect immunofluorescence technique. Five cardiac reflex tests were performed to investigate cardiac autonomic neuropathy. QTc interval was assessed in all patients. As a control group, 60 healthy non-diabetic subjects were also tested for CF-SG and CF-PSG autoantibodies. CF-SG autoantibodies were detected in 11 (9%) and CF-PSG autoantibodies were observed in 7 (6%) Type 2 DM patients, whereas in control subjects, the frequency was 1 (2%) and 0 (0%) respectively (ns vs. Type 2 DM patients). In Type 2 DM patients with cardiac autonomic neuropathy (n=31, 24%), CF-SG autoantibodies and CF-PSG autoantibodies were detected in 3 (10%) patients, respectively, compared to 8 (8%) and 4 (4%) in Type 2 DM patients without cardiac autonomic neuropathy (n=96, 76%, ns v. Type 2 DM with cardiac autonomic neuropathy). Both CF-SG autoantibodies and CF-PSG autoantibodies were observed in 2 (7%) Type 2 DM patients with cardiac autonomic neuropathy and 3 (3%) Type 2 DM patients without cardiac autonomic neuropathy. Type 2 DM patients with cardiac autonomic neuropathy demonstrated a longer QTc-interval (446+/-42 ms) than Type 2 DM patients without cardiac autonomic neuropathy (413+/-45 ms, p=0.0001). In Type 2 DM patients with a prolonged QTc-interval (>440 ms: n=29, 23%), 2 (7%) patients presented with CF-SG and 3 (10%) had CF-PSG autoantibodies. In Type 2 DM, CF-SG and CF-PSG autoantibodies are not frequently observed. The results do not give evidence, that immunological factors--like in Type 1 DM--play a role in the pathogenesis of cardiac autonomic dysfunction in Type 2 DM.

Evidence for specific autoimmunity against sympathetic and parasympathetic nervous tissues in Type 1 diabetes mellitus and the relation to cardiac autonomic dysfunction.

There is growing evidence for the involvement of immunological factors in the pathogenesis of cardiac autonomic dysfunction in Type 1 diabetes mellitus (DM). To evaluate the presence of autoantibodies against autonomic nervous tissues and their relationship with tests of autonomic function, 64 newly diagnosed and 142 long duration Type 1 DM patients were investigated for sympathetic and parasympathetic ganglia (CF-SG and CF-PSG) autoantibodies with a complement-fixing indirect immunofluorescence technique. Five cardiac reflex tests were performed to assess autonomic function. Fifty-seven patients with neurological diseases other than diabetic neuropathy and 131 healthy control subjects were also tested for CF-SG and CF-PSG autoantibodies. CF-SG autoantibodies were observed in 47 (23%) and CF-PSG autoantibodies in 21 (10%) of 206 Type 1 DM patients (p < 0.001). In contrast, these autoantibodies were detected in 3 (5%) and 1 (2%) of patients with non-diabetic neurological diseases and 3 (2%) and 4 (3%) of control subjects (p < 0.01, p < 0.05, p < 0.0001, p < 0.05 vs Type 1 DM patients). All except two Type 1 DM patients with CF-PSG autoantibodies also presented with CF-SG autoantibodies. In diabetic patients with long duration, CF-SG autoantibodies were more frequent in patients with ECG-based cardiac autonomic neuropathy (CAN; > or =2 of 5 cardiac reflex tests abnormal) compared to patients without CAN although this did not reach statistical significance (29% vs 17%, p = 0.06). However, 4 (80%) of 5 newly diagnosed and 23 (32%) of 73 established Type 1 DM patients with abnormalities in heart rate variation during deep breathing and/or standing from lying presented with CF-SG autoantibodies compared to 12 (25%) of 58 newly diagnosed (p < 0.05) and 7 (11%) of 63 established Type 1 DM patients (p < 0.01), in whom both tests were normal. The results suggest that autoimmune factors contribute to the pathogenesis of cardiac autonomic dysfunction in Type 1 DM and that autoantibodies against sympathetic and parasympathetic nervous tissues are relatively specific for Type 1 DM.

Influence of antigen on membrane properties of guinea pig bronchial ganglion neurons.

The bronchus was isolated from actively sensitized guinea pigs, and the effect of antigen challenge on the excitability of bronchial parasympathetic ganglion neurons was examined with standard intracellular recording techniques. Based on histological examination, we found that mast cells were located near parasympathetic ganglia neurons. Antigen challenge resulted in a loss of mast cell staining and the release of the mast cell-associated mediators, histamine (38 ng/g, approximately 14% of total content) and prostaglandin D2 (PGD2, 118 ng/g wet weight of tissue). Challenging the isolated bronchus with the sensitizing antigen resulted in a transient depolarization (mean 6 mV) of the resting membrane potential of the neurons. Antigen challenge also had a dramatic effect on the accommodative properties of the neurons. Before antigen challenge, two subpopulations of neurons could be differentiated by their response to cathodal current steps: 60% of the cells responded in a "phasic" manner, firing one to six spikes and then accommodated, whereas the balance fired spikes repetitively throughout the current pulse. In phasic firing cells, ovalbumin challenge produced a decrease in accommodation. This was evidenced by a fivefold increase in the number of action potentials elicited during a 500-ms suprathreshold current pulse. The antigen-induced depolarization could be mimicked by histamine, whereas the decrease in accommodation was mimicked by application of PGD2. Leukotriene C4, another mast cell-associated mediator, had no effect on these neuronal properties. These results provide evidence that the immediate hypersensitivity response in guinea pig airways may involve changes in membrane characteristics of bronchial parasympathetic ganglia neurons.

Monoclonal antibodies made to chick mesencephalic neural crest cells and to ciliary ganglion neurons identify a common antigen on the neurons and a neural crest subpopulation.

We previously reported the production of monoclonal antibodies (Mabs) that identified cell surface components of cultured chick and quail ciliary ganglion (CG) neurons and of a subpopulation of neural crest (NC) cells from 31-hr chick embryos (stage 9). Here we demonstrate that another Mab, CG-14, which was prepared to nitrocellulose-immobilized, lightly fixed (0.125% paraformaldehyde) mesencephalic NC cells from 31-hr (stage 9) chick embryos, labels the same antigen(s) recognized by CG-1 and CG-4 on both the CG neurons and the subpopulation of NC cells. All three Mabs label a polypeptide of 75 kD on Western blots of one-dimensional SDS-polyacrylamide gels. CG-14 blocked the binding of CG-1 and/or CG-4 to the 75 kD band on Western blots and blocked the binding of CG-1 and CG-4 to CG and NC cells. CG-1 and/or CG-4 antibodies, in turn, blocked the binding of CG-14 to Western blots, as well as NC and CG cells. We had previously shown that antibodies CG-1 and CG-4 were synergistically cytotoxic for the majority (95%) of cultured CG neurons in vitro in the presence of guinea pig complement. Here we show that the antibodies, which are both of the gamma 2a subclass, are also cytotoxic for the NC cells that they label in vitro. After the cells are ablated in culture, no other cells bearing the antigen(s) recognized by any of the three Mabs appear over a 2.5-week period. CG-14, however, is not cytotoxic for either the CG or NC cell populations alone or in combinations with CG-1 or CG-4. These results confirm our original observation that cultured CG neurons and NC cells share cell-surface antigen(s). The antigen recognized by all three Mabs appears to be the same whether the immunogen used to produce the antibodies was CG neurons or NC cells. This finding encourages us to continue tests of the hypothesis that the subpopulation of mesencephalic neural crest cells contributes to the formation of the ciliary ganglion in the embryo. Further characterization of the antigen appears in the accompanying paper.

Antigen recognized by monoclonal antibodies to mesencephalic neural crest and to ciliary ganglion neurons is involved in the high affinity choline uptake mechanism in these cells.

High-affinity choline uptake mechanisms are among the characteristics of cholinergic neurons such as the ciliary and choroid subpopulations in the ciliary ganglion (Barald and Berg, 1979). We have produced three monoclonal antibodies (Mabs), two of which were made to 8-day embryonic chick ciliary ganglion (CG) neurons (CG-1, CG-4) (Barald, 1982) and one of which was made to cultured mesencephalic neural crest (NC) cells (CG-14) removed from the embryo 31 hr after incubation. We have shown that all three Mabs label a common 75 kD antigen present on the cell surface of both CG neurons and NC cells (Barald, 1988). Here we report that the CG-1 and CG-4 antibodies, used in the same ratios in which they are synergistically cytotoxic for both the CG and NC cells (Barald, 1988), and Mab CG-14 alone, have specific effects on the high-affinity choline uptake mechanism (HACU) of CG neurons and isolated antigen-positive NC cells in the absence of complement. CG-1 and CG-4 in ratios of 8/1 (the same ratios that are used to kill the CG and the NC subpopulation), but neither singly, inhibit the HACU of CG neurons by 40% and that of isolated antigen-positive NC cells by 75%. However, CG-14 alone, at 1 microgram/ml, inhibits the HACU of both CG neurons and isolated NC cells by 95%. None of the antibodies had an effect on numbers of ouabain binding sites (a measure of the Na+/K+ ATPase) or cell surface acetylcholinesterase (AChE) of CG neurons or NC cells isolated by "no-flow" fluorescence cytometry with a Meridian Instruments ACAS470 cytometer. CG or NC cells grown in the presence of the antibodies without complement grow and remain healthy for many weeks. They exhibit no difference in morphology, protein content, lactate dehydrogenase activity (LDH), or division time from untreated sister cultures. Therefore, the antigen recognized by all three Mabs may be involved in a high-affinity choline uptake mechanism, a common characteristic of cholinergic neurons. The Mabs themselves may possibly label some element of the high-affinity transporter or a proximal membrane component. This implies that such a high-affinity uptake mechanism is present in the subpopulation of NC cells at early times in development. If these cells in fact are destined to contribute to the avian CG, these characteristics are present in the subpopulation before the NC cells take on a neuronal morphology.

Target specificity of neuropeptide Y-immunoreactive cranial parasympathetic neurons.

We recently showed that neuropeptide Y (NPY)-like immunoreactivity occurs in subpopulations of neurons in 3 cranial parasympathetic ganglia: the otic, sphenopalatine, and ciliary. The present work identifies the target tissues innervated by cranial parasympathetic NPY-immunoreactive neurons. Plexuses of NPY-immunoreactive fibers were observed in the parotid gland, the target of the otic ganglion, and in the intraorbital lacrimal gland and palate, targets of the sphenopalatine ganglion. NPY-immunoreactive fibers of apparent parasympathetic origin innervated glandular acini in all 3 structures and were also present around small blood vessels in the parotid and intraorbital lacrimal glands. These fibers were presumed to be parasympathetic because they were not affected by removal of the superior cervical ganglion and because their distribution was coextensive with that of vasoactive intestinal polypeptide (VIP) immunoreactivity, which we have previously shown to be colocalized with NPY in the cell bodies of otic and sphenopalatine ganglion neurons. In contrast, no NPY-immunoreactive fibers were observed in the iris or ciliary body of acutely sympathectomized rats, suggesting that NPY-immunoreactive neurons in the ciliary ganglion do not normally transport detectable levels of NPY to their terminals. The target specificities of cranial parasympathetic NPY-immunoreactive neurons are different from those of sympathetic NPY-immunoreactive neurons. Sympathetic NPY-immunoreactive fibers innervated the iris and ciliary body, and the blood vessels but not the parenchymal cells of all the glands examined. In contrast, parasympathetic NPY-immunoreactive fibers primarily innervated glandular acini. NPY-immunoreactive neurons in the sphenopalatine ganglion displayed an additional level of specificity in their projection pattern in that they innervated only a subset of the ganglion's array of target glands: they innervated the intraorbital lacrimal gland and the seromucous glands of the palate but not the exorbital lacrimal gland or the glands of the nasal mucosa. The finding that NPY immunoreactivity is present in the parasympathetic innervation of secretory acini in several craniofacial glands raises the possibility that NPY plays a role in the parasympathetic control of glandular secretion. The observed overlap in the distributions of NPY- and VIP-immunoreactive fibers in these glands further suggests that NPY may interact with VIP to stimulate secretion.

Functional blockade of neuronal acetylcholine receptors by antisera to a putative receptor from brain.

Antisera to a putative acetylcholine receptor purified from chick brain specifically inhibit the acetylcholine response of chick ciliary ganglion neurons in cell culture. The putative brain receptor and a similar membrane component previously identified on ciliary ganglion neurons appear to be functional nicotinic acetylcholine receptors in the nervous system and are clearly distinct from membrane components in the tissues that bind alpha-bungarotoxin.

Cholinergic modulation of an acetylcholine receptor-like antigen on the surface of chick ciliary ganglion neurons in cell culture.

Chick ciliary ganglion neurons have a membrane component that shares an antigenic determinant with the "main immunogenic region" of the alpha subunits in nicotinic ACh receptors from skeletal muscle and electric organ. Ultrastructural studies on antibody binding in the ganglion have shown that the cross-reacting antigen on the neuron surface is located predominantly in synaptic membrane. Biochemical studies have shown that the cross-reacting component has a number of other properties expected for the ganglionic nicotonic ACh receptor and that it is distinct from the alpha-bungarotoxin binding component in the tissue. Here we show that ciliary ganglion neurons grown in dissociated cell culture express a similar component that cross-reacts with monoclonal antibodies to ACh receptors, and that the number of antibody-binding sites on the neurons can be modulated by exposure to cholinergic agonists and a protein neurotoxin that reversibly inhibits ACh receptors on the neurons. In most, though not all, cases, levels of ACh sensitivity associated with the neurons are specifically comodulated in parallel with the changes in number of antibody binding sites. The results suggest that at least a portion of the cross-reacting sites on the surface of ciliary ganglion neurons is likely to represent nicotinic ACh receptors. The fact that in some instances levels of ACh sensitivity can be altered without changing the number of cross-reacting sites, however, leaves open the possibility that not all of the sites are associated with receptors or that the neurons can alter the proportion of receptors that is functional.

Characterization of a component in chick ciliary ganglia that cross-reacts with monoclonal antibodies to muscle and electric organ acetylcholine receptor.

Chick ciliary ganglion neurons have previously been shown to contain a component that shares an antigenic determinant with the "main immunogenic region" of the alpha-subunit in nicotinic acetylcholine receptor from skeletal muscle and electric organ. Ultrastructural studies of antibody binding in the ganglion have shown that the cross-reacting antigen exposed on the surface of the neurons is located predominantly in synaptic membrane. Here we show that the neuronal antigen can be identified in detergent extracts of ciliary and sympathetic ganglia, but not in extracts of heart, liver, spinal cord, retina, or dorsal root ganglia. In the ciliary ganglion the component is present as an integral membrane constituent, and, when detergent solubilized, it sediments as a 10 S species and binds to concanavalin A. The component is distinct from the alpha-bungarotoxin-binding site on the neurons since toxin-binding sites and antibody-binding sites can be precipitated separately in ganglion extracts. The component reaches peak levels per ganglionic protein between embryonic days 8 and 12. These are some of the properties expected for the nicotinic acetylcholine receptor on ciliary ganglion neurons.

Origin and distribution of capsaicin-sensitive substance P-immunoreactive nerves in the nasal mucosa.

In immunohistochemical studies, substance P-immunoreactivity (SP-IR) was found in a population of trigeminal ganglion cells in guinea pig, rat and cat. SP-IR nerve endings were found in the spinal trigeminal nucleus, around sphenopalatine ganglion cells, around blood vessels, as well as under and within the epithelium of the nasal mucosa. Ligation and denervation experiments in the cat indicated that the SP-IR nerves in the sphenopalatine ganglion and the nasal mucosa are of trigeminal origin. Capsaicin pretreatment of guinea pigs and rats resulted in a selective loss of the SP-IR nerves in the nasal mucosa and sphenopalatine ganglion, while the parasympathetic and sympathetic nerves were still present.

A cell surface determinant expressed early on migrating avian neural crest cells.

We have produced monoclonal antibodies against quail ciliary ganglion in an attempt to identify specific markers of this neural crest derivative. One of these antibodies, NC/1, recognizes supportive and neuronal cells of the peripheral nervous system and also most, if not all migrating neural crest cells. We report herein the use of NC/1 to identify crest cells during their migration to their site of final localization. In addition, this antibody may shed light on how the neural crest derived mesectoderm and the peripheral nervous system segregate from one another since the NC/1-defined antigen becomes restricted to the cells of the latter.

Simultaneous expression of neuronal and glial properties by chick ciliary ganglion cells during development.

Autoradiographic methods were used to show that non-neuronal cells from dissociated chick ciliary ganglia grown in cell culture for 1 day exhibit high affinity uptake for norepinephrine (NE) and/or specific receptors for nerve growth factor (NGF). Using immunofluorescence procedures, it was demonstrated that these cells reacted neither with the neuron-specific marker tetanus toxin nor with antibodies to the fibroblast marker fibronectin. The cells were, however, positive for 04 antigen, which is present on Schwann cells and oligodendrocytes and is recognized by a monoclonal antibody (Schachner, M., S. K. KIm, and R. Zehnle (1981) Dev. Biol. 83: 328-338). At all stages studied between embryonic day 6 (E6) and embryonic day 14 (E14), about 80% of the non-neuronal cells were positive for 04 antigen, the other non-neuronal cells being identified as fibroblasts or fibroblast-like cells by staining with antibodies to fibronectin. The proportion of cells with NGF receptors and cells with NE uptake decreased during development between E6 and E14. The percentage of 04-positive cells which have NGF receptors decreased from about 95% at E6 to about 35% at E14. The proportion of 04-positive cells with NE uptake decreased from about 57% at E6 to about 15% at E14. Thus, a considerable proportion of the non-neuronal cell population in embryonic ciliary ganglia displays neuronal properties. We suggest that those cells exhibiting biochemical properties of both differentiated glial cells and neurons are precursor cells which have the potential to develop either into glial cells or neurons.