Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Intrinsic airway ganglia provide integrated neural input controlling smooth muscle tone and glandular secretion and mediate tracheobronchial reflexes (1). In ferret tracheal plexuses, ganglia are arranged in two distinct locations: the longitudinal trunk (LT) and the superficial muscular plexus (SMP) (4, 5). Neurons in these ganglia synthesize and release neural mediators, including acetylcholine (ACh) (6, 7), vasoactive intestinal peptide (VIP) (8, 9), and nitric oxide (NO) (10). Cholinergic nerve cell bodies are preferentially located in ganglia of the LT, and VIP and neuronal NO synthase (nNOS) are preferentially located in the SMP (7). The great majority of SMP neurons contain both VIP and nNOS, and colocalization of ACh with VIP is rare.

Information about the projections and neurophysiology of LT neurons has been addressed in a few studies. Mitchell and Coburn (13) and Skoogh and Ullman (14) showed that LT neurons project to airway smooth muscle, mediating bronchial constriction and forming synapses with neurons in other ganglia of the LT. Coburn and Kalia (15) studied individual neurons in the LT by combining electrophysiologic recordings with intracellular injection of horseradish peroxidase. Two types of LT neurons were identified by electrophysiology: spiking after hyperpolarization neurons and nonspiking B-type neurons. The two types were indistinguishable based on dendritic morphology or projection patterns. Although axons of injected LT neurons ended in adjacent LT ganglia, direct neuronal contacts were not established.

Projections of neurons in SMP ganglia have not been studied. The VIP- and nNOS-containing neurons of the SMP are potentially important because of their roles as putative mediators of the inhibitory nonadrenergic noncholinergic (iNANC) relaxant innervation in airway smooth muscle (16). A recent paper demonstrated depression of the iNANC innervation in ferret airways exposed to virus (19), supporting a potential role for these neurons as active bronchodilator regulators in the airways. We have shown recently that some of the neurons in the SMP contain the neuropeptide substance P and may project fibers to the airway epithelium, suggesting a sensory function involving direct activation of airway neurons by airway irritants (14).

In this report, we examine connections between the predominantly cholinergic neurons of the LT and the VIP- and nNOS-containing iNANC neurons of the SMP. The development and application of axonal tracers have greatly facilitated the elucidation of neuronal projections. Among a variety of tracers, dextran conjugates are ideal anterograde and retrograde axonal tracers (21, 22) because of the presence of formaldehyde-fixable lysine residues. The goal of this study is to examine the projections of individual neurons in tracheal ganglia by means of anterograde and retrograde tracing. Subsequent double-labeling immunocytochemistry is used to evaluate projections of specific VIP- and nNOS-containing neurons. The findings demonstrate reciprocal projections between LT and SMP neurons, suggesting the existence of neuroregulatory mechanisms involving cholinergic and iNANC activity at the synaptic level of airway ganglia.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Animals

A total of 10 adult female ferrets (Marshall Farms, North Rose, NY) weighing 200 to 300 g were used in this study. Ferrets were housed two to four per cage with access to food and water ad libitum in the American Association for Accreditation of Laboratory Animal Care-accredited West Virginia University (WVU) Laboratory Animal Resources facility. They were killed by inhalation of carbon dioxide in a precharged chamber. All procedures were approved by the WVU Animal Care and Use Committee (ACUC) under ACUC no. 9206-05.

Experimental Procedure

Segments from the lower half of the trachea 1 to 2 cm long were removed and placed in oxygenated Krebs solution at 4°C. An excess of connective tissue was removed, the tracheal cartilages were cut in the ventral midline, the epithelial layer was removed, and the cartilages were trimmed to the insertion point of the tracheal smooth muscle. The segments were then pinned dorsal surface up in 10-cm culture dishes coated with Sylgaard (Dow Corning, Midland, MI). Using a fiberoptic darkfield illuminator (Fostec, Auburn, NY) and a dissecting microscope (Wild M8, Heerbrugg, Switzerland), connective tissue was gently removed until selected ganglia of LT were visible. Excessive microdissection was minimized to avoid disturbing or damaging axons. Extracellular iontophorectic injections of fluororuby (FR) (10,000 mol wt, 10% in phosphate-buffered saline [PBS]), a tetramethyl rhodamine-dextran-amine, or biotinylated dextran-amine (BDA, 10,000 mol wt, 2 mg/ml in PBS) were made into the capsule of LT ganglia at 7.5 mA of pulsed positive current for 20 min using a 51413 Precision Current Source (Stoelting Co., Wood Dale, IL). Both BDA and FR were purchased from Molecular Probes (Eugene, OR). Only one injection site was placed per tracheal segment to allow unambiguous identification of the origins and projections of airway neurons. In a few experiments, FR and BDA were mixed and injected to provide combined retrograde and anterograde tracing.

After the injection was completed, the tracheae were cultured for up to 2 d to provide time for transport. Culture conditions for airway neurons have been described previously (20). Briefly, the dissected tracheal segments were placed in 60 × 15-mm culture dishes containing CMRL 1066 (GIBCO, Grand Island, NY) supplemented with 0.1 µg/ml hydrocortisone hemisuccinate, 1 µg/ml recrystalized bovine insulin, 60 µg/ml penicillin G (100 U/ml), 10 µg/ml amphotericin B, 100 µg/ml streptomycin, and 5% heat-inactivated fetal calf serum, transferred to a controlled atmosphere culture chamber (Bellco Glass, Inc., Vineland, NJ), and gassed with a mixture of 45% O₂, 5% CO₂, and 50% N₂. The chamber was placed on a continuous rocker in an incubator at 37°C.

After culture, the tracheal segments were pinned out again and processed as whole mounts. Visualization of FR required no additional processing. BDA was labeled by incubating in 20 mg/ml fluorescein isothiocyanate-avidin D (Vector Laboratories, Inc., Burlingame, CA) in PBS for 3 h to overnight. To localize VIP and nNOS, immunocytochemical procedures were done on whole mounts as described previously (7). Briefly, the whole mounts were covered with sufficient primary and secondary antisera to fully cover the dorsal surface of the trachea and incubated for 24 h at 4°C. Mouse monoclonal anti-VIP was kindly provided by Dr. John Porter, University of Texas Health Science Center (Dallas, TX) and has been characterized for use in ferret airways in our previous publications (7). Rabbit polyclonal anti-nNOS was purchased from Euro-Diagnostica (Malmö, Sweden). Specificity of both the nNOS and the VIP antibody was confirmed by complete inhibition of labeling after preabsorption with purified antigen (100 µg/ml). Secondary antibodies included rhodamine-labeled goat antimouse immunoglobulin (Ig)G (Southern Biotechnology Associates, Inc., Birmingham, AL) and fluorescein- or rhodamine-labeled goat antirabbit IgG (ICN Biomedicals, Inc., Aurora, OH). All primary and secondary antibodies were diluted 1:100.

Images were recorded digitally using the BDS Multimode Cytometer (ONCOR, Inc., Gaithersburg, MD) equipped with a Zeiss Axiovert 35 inverted microscope. In some images, a montaging feature was used to combine multiple adjacent images of fields that were too large to fit in a single frame.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

The extensive neural plexus and connections of tracheal ganglia were clearly demonstrated in whole-mounted preparations by using the neural tracers FR and BDA. The injected dextran amines were taken up and transported rapidly by airway neurons. FR labeling was identified as far as 5 mm away from injected ganglia within 2 h. The two conjugates were observed traveling across the entire tissue 24 h after injection. BDA and FR both provided labeling in retrograde and anterograde directions, but a relatively higher resolution of anterograde transport to nerve terminals was observed with BDA.

BDA generated extensive labeling of numerous cell bodies at the injected ganglia (Figure 1). BDA transport was particularly useful as an anterograde tracer, identified as intense and finely labeled processes projecting away from the filled cell bodies and terminating as basketlike complexes around nerve cell bodies. Axon and cell body filling in LT and SMP ganglia was observed for distances up to several millimeters from the injection site . 

View larger version (69K): [in this window] [in a new window]   Figure 1.   Montage of a BDA injection site located in an LT ganglion (arrow). The intensity in the injected area is too bright to visualize individual cells. Anterogradely labeled fibers extend away from the injection site, projecting to a nearby LT ganglion (circle) and to trachealis muscle (M). Arrowheads show the close association between filled nerve terminals and smooth muscle. Scale bar: 300 µm.

Anterograde labeling in axons and terminals also extended from LT ganglia toward trachealis muscle where an elaborate plexus of nerves was formed (Figure 1). Fine fibers were observed either singly or in fascicles of several processes in close association with smooth muscle.

Distinctive anterograde and retrograde transport of the dextran conjugates revealed detailed morphology of projections among tracheal ganglia. Anterograde labeling of axonal fibers could be followed to adjacent LT ganglia where the majority of fibers coursed through the ganglion without forming terminals (Figure 2A). Some of the anterogradely labeled processes terminated around neurons, some of which were in LT ganglia but more extensively around neurons in SMP ganglia (Figures 2A and 2B). The terminals wrapped around the ganglion cell bodies, presenting a basketlike appearance (Figures 2A and 2B). In addition, retrograde labeling displayed filling of the nerve cell bodies of either the LT (Figure 2B) or SMP (Figure 3) ganglia.

View larger version (69K): [in this window] [in a new window]   Figure 2.   Labeling of neuronal cell bodies in a noninjected LT ganglion (A) and an SMP ganglion (B) after BDA injection in an LT ganglion (not shown). (A) Montaged micrograph of an LT ganglion distant from the injection site (circled in Figure 1). Anterogradely labeled fibers terminate around unfilled cell bodies seen as basketlike complexes with associated punctate terminals (arrowheads). Labeling in cell bodies of LT ganglion (asterisks) indicate retrograde filling. Labeled axons course through the ganglion (arrows). Scale bar: 120 µm. (B) BDA-labeled terminals form a basketlike network around an unlabeled SMP nerve cell body. Individual filled axons in nerve bundles are seen (arrowheads). Scale bar: 80 µm.

View larger version (69K): [in this window] [in a new window]   Figure 3.   SMP ganglion after injection of combined FR and BDA injection. Retrograde labeling by FR fills a nerve cell body in an SMP ganglion. The cell body is surrounded by basketlike fibers that are anterogradely labeled with BDA. Scale bar: 80 µm.

A substantial amount of tracer-filled nerve fibers coursed along LTs and then branched toward the SMP where they formed basket-like structures surrounding neurons in SMP ganglia (Figure 4). These data indicate that neurons in LT ganglia project to and make synaptic contacts with neurons in the SMP. Retrogradely labeled nerve cell bodies were also observed in SMP ganglia, indicating that neurons in the SMP project to and form terminals at neurons in the LT as well (Figure 3). In a few experiments where BDA and FR were simultaneously injected into one LT ganglion, retrogradely filled cells were surrounded by basketlike fiber networks, implying the presence of reciprocal connections between the LT and SMP ganglia (Figure 3).

View larger version (75K): [in this window] [in a new window]   Figure 4.   Montage of LT and nearby SMP ganglia after BDA injection. Labeled nerve fibers course along the trunk and branch to the SMP (arrow), demonstrating projection from injected LT ganglion to the SMP ganglion. Scale bar: 80 µm.

The neurochemical features of the tracheal plexus were studied by combining neural tracing and immunocytochemistry. Consistent with our previous data, VIP-immunoreactive (IR) and nNOS-IR neurons were located in neurons of the SMP. After transport of FR from injected LT ganglia, basketlike complexes were identified around nNOS-IR cell bodies in SMP (Figure 5A). BDA-labeled terminals were also observed around VIP-IR cell bodies in SMP ganglia (Figure 5B). These findings demonstrate that LT neurons project to nNOS- and VIP-IR neurons of SMP. nNOS-IR and VIP-IR were also found to coexist with transported BDA (Figures 6A and 6B) in some neurons of SMP, indicating that the nNOS-IR and VIP-IR neurons of SMP project back to the LT ganglia.

View larger version (19K): [in this window] [in a new window]   Figure 5.   (A) Neurochemical characterization of neurons in SMP ganglia along a branching nerve trunk. Two nNOS-IR nerve cell bodies (green) surrounded by basketlike terminals labeled with FR (red). Most FR fibers in the nerve trunk bypass the ganglion. Scale bar: 80 µm. (B) BDA-labeled fibers (green) terminate around four neurons in an SMP ganglion. Two of the neurons are VIP-IR (red). Scale bar: 40 µm.

View larger version (24K): [in this window] [in a new window]   Figure 6.   Localization of nNOS (A) and VIP (B) immunoreactivity anterogradely labeled with BDA in SMP nerve cell bodies. (A) Among the nNOS-IR neurons (red, arrowheads), retrogradely transported BDA filled one cell body (green, arrow). (B) Rhodamine- labeled VIP-IR neurons in SMP (red deposits in neurons at arrows) are also filled with BDA (green deposits in neurons at arrows). Scale bar: 50 µm.

Iontophoretic injections directly into parts of the LT lacking ganglia produced no labeling of axonal fibers of passage, except in one case that was associated with damage to the trunk. This finding supports our assumption that the transported BDA or FR resulted from uptake either by cell bodies or terminals in ganglia, and not by damaged fibers of passage.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

This study examined the projections of intrinsic ganglia of ferret trachea by injecting the neural tracers FR or BDA. The general anatomic features of the tracheal plexus were similar to those observed in previous studies in ferret trachea (4, 7, 23). However, because FR and BDA are compatible with immunocytochemical procedures, projections from neurochemically characterized neurons were revealed. Distinctive uptake of tracers by fibers of passage coursing through the injection sites did not appear to contribute to the tracing patterns, provided the fibers were not damaged because labeled fibers were rarely seen when injections were made into the interganglionic regions of the LTs. As suggested in previous studies (15, 23), both LT ganglia and SMP ganglia give rise to extensive projections distributed to the airways. Some neurons gave rise to axons that formed complex connections with other airway neurons and some projected to smooth muscle.

By combining tracing methods with immunocytochemistry, intrinsic neural pathways could be correlated with neurochemical content. Special attention was given to VIP and nNOS because VIP and NO are putative transmitters mediating iNANC bronchodilation (12, 17, 18). The absence of sympathetic innervation in human and guinea pig airway smooth muscle makes iNANC innervation the only relaxation mechanism (24, 25), and thus, of particular importance. Although one study was unable to demonstrate iNANC innervation in the adult ferret trachea (26), another study showed that iNANC relaxation was present in ferrets 8 and 24 wk of age (19). The ferrets used in our experiments were 8 to 12 wk old.

Our observations that BDA and FR are retrogradely transported after injection into LT ganglia suggest that the SMP neurons project to and synapse on LT neurons. Electron microscopic studies have demonstrated synapses on LT neurons suggestive of peptidergic innervation (23). The present study also shows that the presumed synapses, seen as immunoreactive boutons closely associated with the cell membrane of LT neurons, originate from SMP neurons that contain nNOS and VIP. Sekizawa and coworkers (27) showed that VIP enhances cholinergic neurotransmission in ferret trachea at low concentrations and attenuates cholinergic transmission at higher concentrations. They suggested that synaptic modulation probably occurred at nerve terminals, but modulation at the nerve cell body would be equally likely. Thus, VIP from SMP nerve terminals released at cholinergic neurons in LT ganglia of ferrets may modulate cholinergic contractile mechanisms, depending on the amount released. Interestingly, VIP at concentrations that enhanced cholinergic transmission in ferrets caused attenuated cholinergic release (28). The possible role of NO in modulating neurotransmission at airway ganglia is not clear.

Previous studies have demonstrated that VIP- and nNOS-containing neurons provide innervation to airway smooth muscle, glands, and blood vessels (29). However, the origin of the preganglionic connections to noncholinergic airway neurons is not well characterized. In the guinea pig trachea, the main part of the vagus nerve has no direct synaptic connection with either nNOS- or VIP-containing iNANC neurons (30). The projections from LT ganglia to VIP-IR and nNOS-IR nerve cell bodies in SMP ganglia suggest that some neurons in LT ganglia provide presynaptic contacts to VIP- and nNOS-containing neurons in SMP ganglia. This supports the observation that ACh, stored and released from LT neurons, can modulate iNANC responses (31). Based on the prominent projections of LT ganglia to neurons in the SMP, it seems likely that presynaptic modulation of the iNANC system by intrinsic cholinergic neurons may be an important regulatory mechanism of airway or vascular smooth muscle tone or mucus gland secretion in the airways.

The anatomic organization and neurochemical expression of neurons in tracheal and bronchial neural plexuses vary considerably among humans and various animal species. The LT and the SMP are well characterized in ferrets (4). Some similarities in neural organization have been noted between ferret and human trachea (23), particularly the existence in human trachea of two layers of neural plexuses similar to the LT and SMP described in ferret trachea. Furthermore, ganglionated plexuses have been identified in both the adventitial and submucosal regions of bronchi from the airways of many mammals, including human bronchi (2). There is extensive variation in the location of VIP- and nNOS-containing neurons in the airways of different mammals. Recent studies demonstrated that the VIP- and nNOS-containing airway neurons in guinea pig are located in the myenteric plexus of the esophagus (33). In human airways, VIP- and nNOS-containing neurons are present within the airway wall (8, 34), and most of the VIP-containing nerves in the human airway are separate from cholinergic nerves (35), both features that are similar in ferret trachea. These findings suggest that innervation of human airways may be similar to that of the ferret. However, the pattern of connectivity between differing groups of neurons in human airways has not been determined.

In conclusion, we have demonstrated that cholinergic neurons of LT ganglia provide innervation to smooth muscle and to neurons of other LT ganglia as well as to SMP ganglia. It also appears that VIP- and nNOS-containing neurons of SMP project to LT neurons. The study verifies the complex neuronal circuitry revealed in previous studies. The projections may represent the neuroanatomic basis for modulation of neurotransmission in airway ganglia.