Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Bradykinin, a metabolite of the kallikrein-kinin system, has multiple effects on airway function in guinea pigs, including bronchoconstriction, vasodilatation, plasma extravasation, mucus secretion, and cough. These effects are thought to be due, at least in part, to the activation of sensory nerves leading to both axonal and central reflex activity (Saria et al., 1983; Ichinose et al., 1990; Ichinose and Barnes, 1990; Barnes, 1992; Inoue et al., 1992; Miura et al., 1992; Geppetti, 1993). The afferent innervation in the guinea pig airways is predominantly vagal in origin. The vagal afferent nerve terminals subserving the trachea and bronchus arise from cell bodies located in either the nodose or jugular ganglia (Kummer et al., 1992; Riccio et al., 1996). The axons arising from these ganglia conduct action potentials mainly in the C (<1.5 m/s) or A (3-20 m/s) range (Riccio et al., 1996).

The nature of the specific afferent nerve populations in the guinea pig airways stimulated by bradykinin is not yet fully understood. The axonal reflexes in the airways can be inhibited by tachykinin receptor antagonists and thus are due to stimulation of the tachykinin-containing afferent nerve terminals (Lundberg, 1995). In guinea pig airways, tachykinin-containing afferent nerve terminals characteristically have unmyelinated (C fiber) axons, with cell bodies located in the jugular ganglion (Riccio et al., 1996). Fox et al. (1993) demonstrated that bradykinin directly stimulates C fiber afferents in vitro in guinea pig airways. Bergren (1997) noted that bradykinin can also lead to activation of A fibers in the guinea pig airways, but this effect was attributed to secondary effects on lung mechanics. Recently, we found that perhaps more important than axonal conduction velocity, the ganglionic location of the cell body is a major determinant in the physiology of vagal afferent fibers in the airways (Riccio et al., 1996; Pedersen et al., 1998). For example, nodose nerve endings in the airways are exquisitely sensitive to mechanical stimulation, whereas the nerve endings derived from jugular neurons are less sensitive to mechanical stimuli but more sensitive to hypertonic saline and irritants such as capsaicin. This is the case regardless of whether the axon conducts action potentials in the C or A range.

In the present study, our aim was to characterize the subset of afferent nerves that are directly stimulated by bradykinin. The characterization was based both on their axonal conduction velocities and on the ganglionic location of their cell body. We found that bradykinin stimulates virtually all fibers arising from the jugular ganglia, regardless of conduction velocity, whereas only C fibers from the nodose ganglia are stimulated.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Tissue Preparation. Guinea pigs (Dunkin-Hartley; male, 100-300 g) were sacrificed by CO₂ asphyxiation and exsanguinated. The trachea, mainstem bronchi, along with the right vagus nerve, and superior and recurrent laryngeal nerves with attached nodose and jugular ganglia were isolated and placed in a Plexiglas chamber and superfused with Krebs' physiological buffer solution maintained at 37°C (buffer composition: 118 mM NaCl, 5.4 mM KCl, 1.0 mM NaH₂PO₄, 1.2 mM MgSO₄, 1.9 mM CaCl₂, 25 mM NaHCO₃, 11.1 mM dextrose, gassed with a mixture of 95% oxygen and 5% carbon dioxide) at a flow rate of 6 to 8 ml/min. The larynx, trachea, and bronchus were opened longitudinally on their ventral surface and pinned to the Sylgard lining of the Plexiglas chamber. The jugular and nodose ganglia, along with attached nerves, were gently pulled through a hole leading to a separate, adjacent compartment in the same chamber where extracellular recordings were performed. Both chambers were perfused separately with buffer solution. The detailed procedure for in vitro electrophysiological recordings of stimulated sensory nerve terminals in guinea pig airways has been described previously (Riccio et al., 1996).

Recording of Action Potentials. Extracellular recordings were made by positioning an alumino-silicate glass microelectrode filled with 3 M NaCl solution (electrode resistance, 2 M) near neuronal cell bodies located in either the nodose or jugular ganglion. The recorded signal was amplified (A-M Systems, Everett, WA) and displayed online on an oscilloscope (TDS 320; Tektronix, Wilsonville, OR). Data were stored on magnetic tape using a digital/audio tape recorder (DTC 59ES; sampling frequency, 22 kHz; Dagan Corporation, Minneapolis, MN). The recorded action potential discharge was digitized and analyzed off line using a customized spike discrimination and counting software program (D. M. MacGlashan, PHOCIS, Baltimore, MD). This program is designed to both quantify spike frequency and discriminate waveforms. The fiber of interest was identified after mechanical stimulation of the receptive field. For the off-line analysis, discrimination "windows" were placed such that only waves that corresponded with the predetermined amplitude and width were accepted for analysis. In most cases, only a single unit was recorded by the recording electrode, rendering window discrimination unnecessary.

Determination of Fiber Activity and Detection of Afferent Nerve Terminals. Single-fiber activity in the airways was detected by stimulating the recurrent laryngeal nerve with a suction-type electrode while the recording electrode was carefully manipulated in the jugular or nodose ganglion until single unit activity was recorded. The receptive field for the recorded fiber activity was then located using gentle mechanical probing of the surface of the trachea and right main bronchus with a blunt plastic rod (o.d. 2 mm). The exact locus of the receptive field was determined when a reproducible burst of action potentials was elicited after gentle mechanical touching of a specific area on the airway luminal surface. Only mechanically sensitive neurons were studied. The nerve fibers studied had little or no activity at rest. When the "spontaneous activity" exceeded 1 action potential/s, the fiber was not studied further.

Characterization of Airway Afferent Fibers. The conduction velocity of the fibers studied were determined by electrically stimulating the receptive field with a concentric electrode (0.5-mm tip diameter) and monitoring the time (ms) elapsed between appearance of the shock artifact and action potential. This value was divided by the distance (mm) between the receptive field and the recording electrode. Fibers were classified as C fibers if they conducted potentials at <1.3 m/s and as A fibers if they conducted action potentials at >2.0 m/s. Fibers that conducted action potentials in the intermediary range (1.3-2.0 m/s) were excluded from this study because they could fall into either category.

Chemical Stimuli. The chemosensitivity of airway afferent nerve terminals to bradykinin and capsaicin was determined by perfusing these drugs directly onto the receptive field at a rate of 6 ml/min for 3 min. The response of afferent fibers to bradykinin was recorded for 5 min. No more than three increasing concentrations of bradykinin were tested on any one tissue. In some experiments, a single concentration of capsaicin was applied at the end of the study. A washout period of at least 10 min was used between increasing bradykinin concentrations. The perfused drugs were removed with a separate suction outlet positioned 5 mm from the receptive field to avoid stimulation of surrounding receptive fields.

Chemicals. Bradykinin was obtained from Peninsula Laboratories (Belmont, CA). A stock solution (0.1 mM) was prepared in distilled water and stored in aliquots at 20°C. Further dilutions were made in Krebs' buffer solution on the day of use. Hoe 140 (D-Arg-[Hyp³,Thi⁵,D-Tic⁷,Oic⁸]bradykinin) was a generous gift from Dr. K. Wirth (Hoechst AG, Frankfurt, Germany). Methacholine and indomethacin were purchased from Sigma Chemical Co. (St. Louis, MO). A stock solution of indomethacin (10 mM) was prepared in 100% ethanol and diluted in Krebs' buffer solution on the day of use. Methacholine was prepared as a stock solution of 0.1 M in distilled water and further diluted in Krebs' buffer solution. SR-140333, SR-48968, and SR-142801 were generous gifts from Zeneca Pharmaceuticals Group (Wilmington, DE). SR-140333 was prepared as a stock solution (50 mM) in dimethyl sulfoxide. SR-48968 and SR-142801 were prepared as a stock solution (10 mM) in distilled water. All three antagonists were stored at 20°C until the day of use, when required dilutions were made in Krebs' buffer solution.

Data Analysis. All data are expressed as the arithmetic mean ± S.E.M. For extracellular recording experiments, the responses of afferent neurons to chemical stimulation are presented graphically as the total number of action potentials recorded. The peak frequency of action potentials represents the highest number of action potentials discharged in a 1-s time period. For intracellular recording experiments, the responses of C cell somata to bradykinin are presented graphically as the change in membrane potential (mV). All data were statistically analyzed using one-way ANOVA followed by Student's nonpaired t test. Differences with a value of p < .05 were considered significant.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Electrophysiological Investigations: General Characteristics of Airway Afferent Nerves. Afferent nerves with terminal projections in the trachea and right bronchus from 51 guinea pigs were studied. Of these terminals, 33 originated from cell bodies located in the jugular ganglion, and 18 originated from cell bodies located in the nodose ganglion. These nerve endings had axons subclassified as C and A fibers based on their conduction velocity. Approximately half of the axons from the jugular ganglion neurons (14 of 33) conducted action potentials in the C fiber range (mean conduction velocity, 0.75 ± 0.06 m/s), whereas the remainder of the fibers conducted in the A range (mean conduction velocity, 5.12 ± 0.65 m/s). The majority of airway afferent nerve fibers arising from the nodose ganglia (14 of 18 fibers) were A fibers (mean conduction velocity, 5.21 ± 0.72 m/s); only four fibers conducted action potentials in the C fiber range (mean conduction velocity, 1.11 ± 0.09 m/s). This is consistent with previous anatomic and electrophysiological studies that have demonstrated that relatively few cell bodies in the nodose ganglion project C-type axons to the guinea pig trachea and bronchus (Riccio et al., 1996). Axons that conducted action potentials in the intermediary range (1.3-2.0 m/s) were excluded from this study (n = 4). As the receptive fields of all fibers studied were located by gentle mechanical probing, only mechanically sensitive fibers were studied. A small number of mechanically insensitive receptive fields (n = 3) were identified using electrical search techniques (see Riccio et al., 1996), these fibers were not studied further. The respective mechanical thresholds were 2.3 ± 0.4 mN (n = 14) and 1.2 ± 0.4 mN (n = 13) for activating A and C fibers from the jugular ganglion and 0.2 ± 0.01 mN (n = 6) for activating nodose ganglion A fibers.

Responses to Bradykinin. The application of 1 µM bradykinin directly over the receptive fields of afferent fibers in the guinea pig trachea and bronchus resulted in action potential discharge from jugular A- and C-conducting neurons. By contrast, A fiber terminals in the trachea-bronchus that originated in the nodose ganglia were unresponsive to bradykinin (Fig. 1). A similar profile of responsiveness was observed when capsaicin (1 µM) was applied to the receptive fields. Thus, capsaicin stimulated jugular fiber terminals regardless of conduction velocity, whereas nodose A fiber terminals were unresponsive to this stimulus (Fig. 1). All four nodose C fiber terminals responded to bradykinin and capsaicin (1 µM; data not shown); however, due to the scarcity of nodose C fiber terminals in the guinea pig trachea and bronchus, further pharmacological manipulations were not performed on these receptive fields. Representative traces of the responses of jugular C and A fibers and nodose A fibers to bradykinin are shown in Fig. 2A. This response typically commenced within 30 s of bradykinin perfusion and often continued but never for a longer total response time than 5 min. Often, the response to bradykinin, in both C and A fiber terminals, appeared as bursts of action potentials lasting for 20 to 30 s (see jugular C fiber response in Fig. 2A). However, in some preparations, a continuous response to drug application was observed (see jugular A fiber response in Fig. 2A). The significance of the bursting response is unclear and was observed in both C- and A-conducting neurons.

View larger version (14K): [in this window] [in a new window]   Fig. 1.   Effect of bradykinin and capsaicin on vagal afferent nerves in guinea pig airways. Bradykinin (1 µM, 3 min; i) and capsaicin (1 µM, 2 min; ii) were perfused directly over the receptive field of afferent C and A fibers. Data are expressed as the percentage of fiber originating in the jugular or nodose ganglia that respond to bradykinin and capsaicin. The total number of fibers studied are indicated in parentheses.

View larger version (37K): [in this window] [in a new window]   Fig. 2.   Response of airway afferent fibers from jugular and nodose ganglia to bradykinin. A, representative traces showing the effect of bradykinin (1 µM) on vagal afferent C and A fiber terminals originating in the jugular and nodose ganglia in guinea pig airways. Bradykinin was applied topically onto the receptive field for 3 min (beginning at the arrow, denoted by the bar), and the response was monitored for a total of 5 min. The response to mechanical stimulation of the nodose A fiber, with a von Frey filament (0.8 mN), is shown at the right of the trace. B, concentration-related effects of bradykinin on jugular afferent fibers in guinea pig airways. i, data are expressed as the total number of action potentials recorded in response to bradykinin (jugular C fibers, ; n = 14; jugular A fibers, ; n = 17). ii, peak frequency of bradykinin-induced excitation of jugular fibers. The response represents the maximum number of action potentials discharged during 1 s (jugular C fibers, ; n = 14; jugular A fibers, ; n = 17). *p < .05, **p < .001.

View larger version (22K): [in this window] [in a new window]   Fig. 3.   Dose-related effects of bradykinin (BK) on jugular ganglia neuronal soma. Data are expressed as the maximum membrane depolarization recorded to bradykinin. Increasing concentrations of bradykinin were perfused over jugular ganglion cells conducting in the C fiber range (n = 3-5). Inset, representative trace showing bradykinin (BK, 1 µM, denoted by the bar)-induced membrane depolarization of a C fiber somata. *p < .05, **p < .001 compared with vehicle control.

View larger version (20K): [in this window] [in a new window]   Fig. 4.   Effect of a B2 receptor antagonist, Hoe 140 (1 µM; hatched columns), on bradykinin-induced afferent nerve activity in guinea pig airways showing the total number of action potentials discharged in response to the application of bradykinin (BK, 0.1 and 1.0 µM) and capsaicin (CAP, 1.0 µM) directly over receptive fields of jugular ganglion C (i, n = 5) and A (ii, n = 4) fibers. The response to bradykinin alone is shown in filled columns. *p < .05, **p < .001.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

In the present study, we demonstrate that the chemical activation of guinea pig airway afferent nerve endings by bradykinin and capsaicin is dependent primarily on the location of their cell bodies. We observed that both C and A fiber terminals that originate in the jugular ganglion are activated by bradykinin. A similar response was observed for capsaicin, an agent classically used as a pharmacological tool for selective C fiber stimulation. In contrast, the application of bradykinin or capsaicin on the receptive fields of airway nodose A fibers failed to cause action potential generation. Consistent with our previous studies, relatively few C fibers with receptive fields in the guinea pig trachea and bronchus originate in the nodose ganglion (Riccio et al., 1996; Pedersen et al., 1998); hence, we are unable to make a definitive statement on the chemosensitivity of nodose C-conducting fibers. However, it is worth noting that of the few nodose C-fiber receptive fields studied (n = 4), each responded to bradykinin and capsaicin in the same manner as jugular C fibers.

Bradykinin has been previously reported to stimulate receptive fields of both C and A fibers in guinea pig lung (Bergren, 1997). The stimulation of A fibers, however, was inhibited by a -adrenoceptor agonist and thus was attributed to secondary effects on lung mechanics. This would be consistent with the observations by Fox et al. (1993) that in the guinea pig isolated tracheal preparation, bradykinin-stimulated vagal C fibers, but not A fibers, were noted. The present findings, however, indicate that a subpopulation of A fibers do indeed respond to direct application of bradykinin, as long as their cell bodies are located in the jugular ganglion. The reason for the discrepancy between the results presented here and those reported by Fox et al. (1993) is not known. It is possible that the difference in experimental design accounts for the differing results. When recording single unit activity from nerve fibers along the trunk of the vagus nerve (i.e., the method used by Fox et al.), it is not possible to consider the ganglionic source of the fiber. It is conceivable, therefore, that in the study by Fox et al., the A fibers (n = 7) that were reported not to respond to bradykinin were derived from cell bodies located in the nodose ganglion.

It is unlikely that the stimulation of jugular A fiber terminals is an indirect effect. That bradykinin stimulated A fibers in the isolated tissue rules out any effects secondary to lung mechanics or vascular events. We evaluated the effect of bradykinin on isometric tension of the smooth muscle in guinea pig trachea during concomitant electrophysiological recordings of jugular A fiber excitation. Bradykinin had negligible effects on smooth muscle tone at the concentrations that effectively activated jugular nerve fibers. Bradykinin is known to increase the excitability of guinea pig vagal sensory neurons by inhibiting a slow calcium-dependent hyperpolarizing current (Undem and Weinreich, 1993). This effect is secondary to the production of prostacyclin (Weinreich et al., 1995). In the present study, however, we noted that the stimulation of jugular A fiber terminals by bradykinin was not dependent on the production of prostaglandins because pretreatment of the tissue with indomethacin did not inhibit the response in these fibers. A similar lack of involvement of prostaglandins in the C fiber-mediated responses to bradykinin was observed in this and previous studies (Fox et al., 1993). We also addressed the hypothesis that bradykinin selectively stimulates C fibers causing the release of sensory neuropeptides, which in turn stimulates jugular A fibers. However, the bradykinin-induced activation of jugular A fibers was unaltered in the presence of neurokinin receptor antagonists. Furthermore, neither Substance P nor calcitonin gene-related peptide stimulates guinea pig airway afferent fibers in vitro (data not shown). Taken together, these data suggest that the response to bradykinin observed in jugular A fibers is mediated by the same direct mechanism as that for jugular C fibers.

Consistent with the observation of Fox et al. (1993), the results support the hypothesis that bradykinin-induced stimulation of airway C fibers is secondary to activation of B₂ type receptors. This also holds true for stimulation of jugular A fibers in the guinea pig airway. Thus, the selective B₂ receptor antagonist Hoe 140 was effective in inhibiting the response of both jugular C and A fibers to bradykinin. Challenge with a B₁ receptor agonist did not lead to action potential discharge in any afferent fiber type in the guinea pig trachea.

Concentrations of bradykinin of more than 1 nM were required to stimulate the jugular nerve endings in this experimental design. To establish whether the effectiveness of bradykinin in this preparation was reduced by epithelium-derived peptidases or by access barriers, we studied the effect of bradykinin applied directly onto the somata of jugular neurons. Using intracellular recording techniques, we observed that the effective dose range of bradykinin to cause membrane depolarization of jugular C cell somata was similar to that required to excite the nerve terminals in the trachea-bronchus of guinea pigs. These results suggest that the access of bradykinin at afferent terminals in the trachea-bronchus was not hampered by barriers or epithelium-derived peptidases.

In conclusion, these data demonstrate that in the guinea pig trachea, bradykinin activates both sensory C fibers and A fibers. The stimulation of A fibers, however, is strictly dependent on the location of the cell body. The jugular A fibers are responsive to bradykinin, whereas A fibers arising from nodose neurons are not. Neurons in the jugular ganglion are thought to be derived from different embryological structures than those in the nodose ganglia (Altman and Bayer, 1982), and it is likely that they are linked to different reflex physiology. Knowledge of the nature of the primary afferent nerves stimulated by bradykinin is essential to the eventual understanding of the mechanism underlying bradykinin-induced neuronal reflexes in the airways. The present findings, considered together with previous observations, indicate that bradykinin stimulates the same subset of primary sensory fibers in the guinea pig trachea that are activated by capsaicin and hypertonic saline (Riccio et al., 1996; Pedersen et al., 1998).