INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Keywords: tachykinins; respiratory viral infection; airway innervation; sensory nerves; Substance P; Neurokinin A

Respiratory tract viral infections are frequently associated with exacerbations of asthma and increased responsiveness to tussive and bronchoconstricitve stimuli in otherwise healthy individuals (1-3). The mechanisms responsible for this remain uncertain but may involve virus-evoked airway inflammation, modulation of the immune response to allergen, and alterations in the efferent and afferent neural control of the airways (2, 4-6).

The airways of guinea pigs are innervated by phenotypically distinct groups of vagal afferent neurons whose cell bodies are located in the jugular (superior) or nodose (inferior) vagal ganglia (7). Approximately one third of the vagal afferent neurons innervating the guinea pig trachea contain tachykinins such as Substance P (SP) and Neurokinin A (NKA) (7). Unlike classic transmitters that are synthesized in nerve terminals, tachykinins are synthesized in neuronal cell bodies and carried by axonal transport to the central and peripheral terminals of primary afferent neurons. Under normal conditions, almost all of the tachykinergic fibers that innervate guinea pig airways are derived from nociceptive-like neurons whose cell bodies are located in the jugular ganglia (7). These tachykinin containing nociceptive-like neurons have characteristic features that distinguish them from nonnociceptive afferent fibers. Compared with nonnociceptive fibers, tachykinin-containing nociceptive fibers typically have small cell bodies (less than 25 µm in diameter) and axons that conduct action potentials in the C-fiber range (less than 1 m/second) (7-9). The terminals of these C-fibers in the airways are high-threshold mechanosensors that respond vigorously to capsaicin and bradykinin (7). In contrast, nonnociceptive fibers that innervate guinea pig airways have larger cell bodies (~ 30-50 µm in diameter) that reside in the nodose ganglia. These neurons are rapidly adapting low-threshold mechanosensors that conduct action potentials in the A-range (~ 5 m/second) (7-9) and are thought to supply afferent input into the brain stem necessary to regulate normal breathing and resting bronchomotor tone (10, 11).

The release of tachykinins from the peripheral terminals of nociceptive-like afferent fibers in the airways can activate neurokinin receptors in a variety of airway tissue and cell types, leading to alterations in bronchomotor tone, mucus secretion, and neurogenic inflammation (12). Respiratory infections are associated with increased levels of tachykinins in the airways (13) and are known to intensify tachykinin-evoked responses (14, 15). This appears to be at least partially due to decreased activity of neutral endopeptidase (16-18), a protease normally present within the airways that degrades tachykinins. There is accumulating evidence that irritation and inflammation of the airways is associated with the induction of tachykinin synthesis in airway afferent fibers (19). The aim of the current study was to investigate the possibility that nodose afferent neurons could be induced to express tachykinins during respiratory tract viral infection. Our findings indicate that respiratory viral infection in guinea pigs induces the expression of tachykinins in airway afferent neurons. Further, the data support the hypothesis that viral infection leads to a qualitative change in the vagal afferent innervation of the airways such that both small diameter nociceptive-like neurons and large diameter nonnociceptive neurons express tachykinins.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Male guinea pigs (150 to 250 g; Harlan Sprague-Dawley Inc., Indianapolis, IN) were used to study the effect of respiratory tract infection with Parainfluenzia Type 1 (Sendai Virus, ATCC VR-105; American Type Culture Collection, Rockville, MD) on the percentage of retrogradely labeled airway afferent nodose neuron cell bodies immunoreactive (ir) for SP/NKA. Full details of the methods can be accessed in the online data supplement.

Airway vagal afferent neurons were retrogradely labeled by instillation of the fluorescent dye fast blue into the upper trachea of anesthetized (ketamine hydrochloride, 50 mg/kg intraperitoneally; xylazine hydrochloride, 2.5 mg/kg intraperitoneally) guinea pigs. We have previously reported that dye instilled using this technique is limited to the rostral region of the trachea (8). Seven days after retrograde labeling of tracheal afferent neurons, guinea pigs were anesthetized (ketamine hydrochloride, 50 mg/kg intraperitoneally; xylazine hydrochloride, 2.5 mg/kg intraperitoneally) and inoculated intranasally with 0.5 ml of fluid containing either 5 × 10⁵ tissue culture infective doses of Sendai virus or vehicle (diluted allantoic fluid from embryonated chicken eggs). At various days after inoculation, vehicle and virus-inoculated guinea pigs were killed by asphyxiation with CO₂, exsanguinated, and their nodose ganglia were removed, fixed (2% paraformaldehyde, 15% saturated picric acid in 0.15 M phosphate buffer), and frozen. Continuous serial cryostat sections (12-µm thick) of the nodose ganglia were thaw-mounted on gelatin-coated coverslips and prepared for estimation of neuronal cell body diameter and immunofluorescent localization of immunoreactive (ir) tachykinins as previously described (8). The primary antibody used has 100% crossreactivity with SP and 40% crossreactivity with NKA (rabbit antisubstance P; Peninsula Laboratories, Belmont, CA). The secondary antibody was a fluorescein isothiocyanate-labeled goat antirabbit IgG (Peninsula Laboratories).

Bronchoalveolar lavage of vehicle and virus-inoculated guinea pigs was performed essentially as previously described (20). Cells in collected lavage fluid were washed in 0.9% NaCl and a cytospin preparation (Shandon Centrifuge, Pittsburgh, PA) of the sample was differentially stained (Giemsa stain) and counted.

In vitro extracellular recording of action potentials from nodose vagal afferent nerve fibers that innervated the airways of vehicle or virus-inoculated guinea pigs was performed as previously described (7). Mechanically sensitive receptive fields were revealed when a burst of action potentials was recorded in response to von Frey filament stimulation of the airway mucosal surface. Conduction velocity and amplitude of the action potential were then compared with responses elicited by electrical stimulation of either the recurrent laryngeal or vagus nerve trunks to determine the trunk that supplied the fiber. Conduction velocities were calculated by electrically stimulating the receptive field and measuring the distance traveled along the nerve pathway divided by the time between the shock artifact and the recorded action potential. Mechanical responsiveness was assessed using a series of von Frey filaments to deliver forces of increasing strength to identified mechanically receptive fields. The force required to elicit half of the number of action potentials evoked in response to supramaximal mechanical stimulation was used as a measure of mechanical sensitivity. The vast majority of vagal afferent fibers that innervate the guinea pig trachea/bronchus are mechanically sensitive (7); thus, only mechanically sensitive neurons were studied, and these nerve fibers had little or no activity at rest. If spontaneous activity exceeded one action potential per second, the fiber was not studied further.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Retrograde Neuronal Tracing

Nodose neuron cell bodies were retrogradely labeled by instillation of fast blue into the trachea of guinea pigs 7 days before intranasal inoculation with Sendai virus or vehicle. A total of 663 labeled neuronal cell bodies were identified in nodose ganglia obtained from vehicle-inoculated guinea-pigs (n = 4) and a total of 478 labeled cell bodies were identified in nodose ganglia obtained from guinea pigs at 4 days after inoculation with Sendai virus (n = 4). Diameters of labeled nodose neuron cell bodies from both groups of animals were similarly distributed (Figure 1A).

View larger version (22K): [in this window] [in a new window]   Figure 1.   Diameters and tachykinin immunoreactivity of nodose neuron cell bodies labeled from the airways of vehicle (n = 4) or virus-inoculated guinea pigs (n = 4; Day 4 after inoculation). (A) The estimated diameters of nodose neuron cell bodies labeled from the airways of vehicle (open bars) or virus-inoculated (filled bars) guinea pigs were similarly distributed. (B) Viral infection was associated with a greater number of tachykinin positive, large diameter labeled nodose neuron cell bodies (filled bars) compared with the few tachykinin positive labeled nodose neuron cell bodies from vehicle-inoculated guinea pigs P (open bars).

Immunofluorescence for SP/NKA

Retrograde labeling in combination with immunofluorescence techniques was used to assess the expression of ir-SP/NKA in airway specific nodose ganglion cell bodies. Few (~ 3%) nodose neuron cell bodies labeled from the trachea of vehicle- inoculated guinea pigs (n = 4) stained positively for ir-SP/NKA (Figure 1B). In contrast, ~ 20% of cell bodies labeled from the trachea of virus-inoculated guinea pigs (Day 4 after inoculation; n = 4) stained positively for ir-SP/NKA (Figure 1B).

We estimated the diameters of ir-tachykinin-positive nodose neuron cell bodies labeled from the airways of vehicle and virus-inoculated guinea pigs. The few labeled nodose neuron cell bodies from vehicle-inoculated guinea pigs that did stain positively for ir-tachykinins had relatively small cell diameters (mode = 16-20 µm; Figure 1B). In contrast, the majority of ir-SP/NKA-positive nodose neurons that were labeled from the trachea of virus-inoculated guinea pigs (Day 4 after inoculation) had relatively large cell body diameters (mode = 36-40 µm; Figure 1B).

Time Course of Virus-induced Expression of ir-SP/NKA

Respiratory viral infections in guinea pig are associated with airway inflammation that develops over the first 2 days after inoculation, peaks within 3-4 days, and usually resolves within 3 weeks (20). Consistent with this time course, there were significantly fewer inflammatory cells in bronchoalveolar lavage fluid from virus-inoculated guinea pigs at Day 28 after inoculation compared with Day 4 after inoculation (data not shown).

To evaluate the time course of virus-induced changes of ir-tachykinin expression in nonnociceptive neurons, we estimated the percentage of large diameter (> 25 µm) ir-SP/ NKA-positive labeled nodose neuron cell bodies from vehicle and virus-inoculated animals at Day 1, Day 4, and Day 28 after inoculation. Consistent with the known time course of virus-induced airway inflammation, the number of large diameter labeled nodose neurons expressing ir-SP/NKA was significantly greater at Day 4 than at Day 1 or Day 28 after inoculation with virus (Figure 2).

View larger version (20K): [in this window] [in a new window]   Figure 2.   Time course of virus-induced expression of ir-SP/NKA in large diameter (> 25 µm) nodose neuron cell bodies labeled from the airways of vehicle-inoculated (n = 4; combined data from two guinea pigs at Day 4 after inoculation and two guinea pigs at Day 28 after inoculation) or virus-inoculated guinea pigs at Day 1 (n = 4), Day 4 (n = 4), or Day 28 (n = 4) after inoculation. Data are presented as the mean ± SEM. * p < 0.05 compared with vehicle.

Electrophysiologic Studies

In vitro extracellular recording was used to estimate the conduction velocity and mechanical sensitivity of vagal afferent fibers whose cell bodies resided in nodose ganglia and whose receptive fields were located in the trachea of control or virus-inoculated guinea pigs (Day 4 after inoculation). The conduction velocity of identified nodose fibers innervating the airways of control guinea pigs (4.4 ± 0.7 m/second, n = 4) was similar to that of nodose fibers innervating the airways of virus-inoculated guinea pigs (3.2 ± 0.8 m/second, n = 5). Nodose fiber receptive fields located in the guinea pig trachea/ bronchus are exquisitely sensitive to mechanical stimuli, to which they respond with a characteristic rapidly adapting pattern of action potential discharge (7, 21). Nodose fiber receptive fields in the trachea/bronchus of control or virus-inoculated guinea pigs were similarly sensitive to mechanical stimulation (Figure 3A), to which they responded with a characteristic rapidly adapting response (Figure 3B). The force required to evoke a half maximal response to mechanical stimulation of receptive fields in the tracheas of control guinea pigs was 180 ± 50 mg (n = four fibers from two guinea pigs), whereas that of virus-inoculated guinea pigs was 200 ± 30 mg (n = five fibers from three guinea pigs).

View larger version (20K): [in this window] [in a new window]   Figure 3.   In vitro electrophysiologic recording of afferent nodose fibers innervating the trachea/bronchus of control and virus-infected guinea pigs. (A) Force-response relationship of mechanically-sensitive vagal afferent fibers whose cell bodies were located in nodose ganglia and whose receptive fields were located in trachea/bronchus isolated from control (open circles, n = four fibers from two guinea pigs) or virus- inoculated guinea pigs (filled circles, n = 5 fibers from four guinea pigs at Day 4 after inoculation). Data are presented as mean ± SEM. (B) Extacellular recording trace (inset) and histogram showing the pattern of action potential discharge recorded in response to application (for ~ 2 seconds) of a punctate mechanical stimuli to the receptive field of a nodose fiber located in the trachea of a guinea pig at Day 4 after inoculation with Sendai virus. The thick horizontal line in the trace is baseline noise and the vertical lines represent the arrival of action potentials at the extracellular recording electrode placed near the neuronal cell body in the nodose ganglion. Application of punctate mechanical force with a von Frey filament (calibrated to deliver 445 mg) to the receptive field located in the trachea evoked an initial high-frequency burst of action potentials ("on-response"), followed by a period in which no action potentials were recorded, i.e., the fiber adapted rapidly to maintained mechanical stimulation. Note the burst of action potentials recorded in response to removal of the von Frey filament from the receptive field ("off response"). This pattern of "on response," rapid adaptation and "off response" is typical of nodose fibers innervating the guinea pig trachea (21).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

We found that respiratory infection with Sendai virus in guinea pigs is associated with the presence of ir-SP/NKA in a population of airway-specific, large-diameter nodose afferent neurons. Given their distinct physiologic characteristics, the induction of tachykinin expression in nodose neurons that innervate the airways may contribute to the abnormal physiology associated with respiratory viral infections.

Under normal conditions the vagal tachykinergic innervation of guinea pig airways is derived almost exclusively from nociceptive-like C-fibers (22, 23), whose small diameter cell bodies reside in the jugular ganglia (7, 9, 24). The nodose afferent fibers that specifically innervate guinea pig airways are typically faster conducting A-fibers, with characteristic large diameter cell bodies and receptive fields in the airways that do not respond to classic nociceptive fiber-selective stimuli such as bradykinin and capsaicin (7, 9, 24). These fibers are exquisitely sensitive to mechanical stimuli, to which they respond with a characteristic rapidly adapting pattern of action potential discharge (7, 21). The nodose fibers whose receptive fields we identified in the airways of control and virus-inoculated guinea pigs conducted action potentials in the A range and were similarly highly sensitive to punctate mechanical stimuli, to which they responded with a rapidly adapting pattern of action potential discharge. Thus, respiratory viral infection did not appear to overtly influence the general physiologic characteristics of nodose nerve endings in the airways.

Few (~ 3%) nodose neuron cell bodies labeled from the airways of vehicle-inoculated guinea pigs stained positively for tachykinins. In contrast, ~ 20% of nodose neuron cell bodies labeled from the airways of virus-inoculated guinea pigs (Day 4 after inoculation) stained positively for SP/NKA, suggesting that viral infection, like allergen-induced inflammation (24), is associated with the induction of tachykinin expression in nodose ganglion neurons. This induction appears to be reversible, as we found significantly fewer tachykinin-positive neurons in nodose ganglia from virus-infected guinea pigs at Day 28 after inoculation, a time point when virus-induced airway inflammation, assessed from bronchoalveolar lavage fluid, had all but resolved. An interesting finding of this study was that the majority nodose neuron cell bodies that contained ir-SP/NKA during viral infection had large diameters (> 25 µm), indicative of nonnociceptive neurons, whereas the few tachykinin-positive cell bodies from vehicle-inoculated guinea pigs had small diameters, indicative of nociceptive-like C-fibers. Thus, it would appear that although there were a few tachykinergic nodose fibers innervating the airways of vehicle-inoculated guinea pigs, they were probably nociceptive-like neurons; however, during viral infection, tachykinin expression was induced in a significant number of airway-specific nonnociceptive nodose fibers.

As visceral inflammation has been associated with an increase in the cell body diameter of afferent neurons located in dorsal root ganglia (25), an alternative interpretation of our data is that viral infection led to an increased cell body diameter of SP/NKA-expressing nociceptive neurons in the nodose ganglia. However, this hypothesis is unlikely to explain our findings, as only ~ 3% of the airway-specific nodose neurons in vehicle-inoculated animals had small diameter cell bodies; thus, even if all of these increased their diameter, it would not explain our observation that ir-SP/NKA was found in ~ 20% of the labeled nodose neurons during viral infection. In any case, we found that the frequency histogram of cell diameters of airway-specific neurons in the nodose ganglia was not different between virus and vehicle-inoculated guinea pigs. Thus, the more likely explanation for our findings is that viral infection induced the expression of tachykinins in a population of nodose neurons with large cell bodies.

The expression of tachykinins in large diameter nodose fibers during viral infection may be expected to impact airway physiology. The release of tachykinins from the peripheral terminals of afferent C-fibers in the airways is known to result in increased blood flow through the microvascular bed of the airways, increased microvascular permeability, and inflammatory cell recruitment, collectively know as neurogenic inflammation (12). Like nociceptive C-fibers in other tissues, airway C-fibers are not thought to release tachykinins under normal physiologic conditions, but rather do so only in response to noxious stimuli. In sharp contrast, large diameter nodose A-fibers that innervate guinea pig airways are low threshold mechanosensors (7) that may be active under normal conditions where they play a role in the reflex regulation of normal breathing and resting airway caliber (10, 11). Thus, if the nodose A-fibers that are induced to express tachykinins are able to release these tachykinins from their peripheral terminal in the airways, nonnoxious stimuli sufficient to activate A-fibers may now evoke the release of tachykinins in the airways during viral infection. Respiratory infections are associated with enhanced responsiveness to tachykinins (14-16) and increased levels of tachykinins in the airways (13) due, at least in part, to decreased enzymatic degradation. Consistent with this, the administration of neurokinin receptor antagonists to guinea pigs can reduce virus-induced airway inflammation (26), although the neuronal source of the tachykinins responsible is unknown. Our current observations suggest that nonnociceptive nodose A-fibers as well as nociceptive-like jugular C-fibers are potential sources of tachykinins in the airways during viral infection.

Although much less studied, tachykinin release from the central terminals of primary afferent neurons in nucleus of the solitary tract of the brain stem may have more influence on airway physiology than tachykinin release from their peripheral terminals in the airway wall. In the somato-sensory system, tachykinins are released from the central terminals of primary afferent neurons and act as neuromodulators to enhance the excitability of secondary neurons in the spinal cord, thereby enhancing the perception of pain and/or reflex responses to noxious stimuli (27). Relatively little work has been published on the mechanisms by which tachykinins modulate vagal afferent neuronal input to the brain stem from the airways. Nevertheless, pharmacologic studies have demonstrated that the neurokinin receptors NK₁, NK₂, and NK₃ are present in secondary neurons within the nucleus of the solitary tract, and stimulation of these receptors can profoundly affect airway reflex physiology (28, 29). Also consistent with a role for tachykinins in modulating airway reflexes, the NK₁ receptor-selective antagonist CP-99,994 and SR-48968, an NK₂ receptor-selective antagonist, were able to inhibit mechanically-induced cough in the guinea pig and cat by an action in the central nervous system (30). As tachykinin release in the brain stem may enhance airway reflexes, inflammatory stimuli that lead to the expression of tachykinins in low threshold nodose A-fibers that are active during normal breathing, may cause the release of tachykinins from their central terminals, where they may alter airway afferent neuronal input into the brain stem.

The mechanism by which viral infection leads to the expression of tachykinins in large diameter nodose A-fibers is unknown, but may involve the neurotrophins, a family of peptides known to act on peripheral nerve endings, and send signals to their cell bodies contained in remotely located ganglia. Potential cellular sources of neurotrophins in the airways include the respiratory epithelium, T lymphocytes, alveolar macrophages, and mast cells (31-33). Neurotrophins initiate their effects in vagal afferent neurons, in part, by binding to high-affinity receptors of the tyrosine kinase family, followed by uptake and retrograde transport to the cell body in the vagal ganglia (34). Among the family of neurotrophins most attention has been given to nerve growth factor (NGF) as a potential mediator of neuronal plasticity in inflamed airways. NGF is found in human airways and the NGF content of airway lavage fluid is increased after allergen challenge (31, 35, 36). NGF is capable of stimulating the expression of mRNAs encoding the precursors of SP and calcitonin gene-related peptide in mature afferent neurons (37). Moreover, instillation of NGF into the trachea of guinea pigs was associated with a greater percentage of large diameter airway-specific afferent neurons that express ir-SP (9). Combined, these findings are consistent with the hypothesis that neurotrophins may contribute to tachykinin expression is airway specific large diameter nodose ganglion neurons during viral infection.

In summary, the major finding of this study was that respiratory infection of guinea pigs with Sendai virus is associated with the expression of ir-SP/NKA in a population of vagal afferent neurons whose large diameter cell bodies reside in the nodose ganglia. The data support the hypothesis that these neurons project low-threshold nonnociceptive fibers to the airways. This sets up a potential condition in which the release of tachykinins from the peripheral and central terminals of airway afferent nerves could occur independently of nociceptive C-fiber stimulation.