A Novel Role for Tachykinin Neurokinin-3 Receptors in Regulation of Human Bronchial Ganglia Neurons

Investigations of the lower airways of mammals demonstrate rich networks of afferent and efferent nerves that have a broad range of neurotransmitters and functions (1). The coordinated signaling between these networks triggers normal homeostatic responses including control of bronchomotor tone, cough, mucous secretion, and increases in bronchial microvascular permeability. In pulmonary diseases, such as asthma and chronic obstructive pulmonary disease (COPD), these responses are exaggerated to the point where they contribute significantly to airway pathologies including mucus hypersecretion, bronchial edema, enhanced bronchomotor tone, and airway hyperreactivity. For example, inhaled anticholinergics are used for symptomatic relief for COPD (2), highlighting the key role of parasympathetic nerves. Communication between these nerve networks, both at the level of the central nervous system and within the trachea and bronchi, is not clearly defined, however, especially in human airways.

The tachykinins are a family of sensory neuropeptides, whose main members include substance P, neurokinin-A, and neurokinin-B, which produce their diverse effects by three distinct receptors, designated tachykinin neurokinin-1 receptors (NK-1R), NK-2R, and NK-3R. The tachykinins are found in sensory nerves in the lung and contract airway smooth muscle, mainly by interaction with NK-2Rs (3), whereas the vascular and proinflammatory effects are mediated predominantly by the NK-1R (4). Based on the elevated levels of the substance P and neurokinin-A in patients with asthma and COPD, and the myriad of effects of the tachykinins on pulmonary cells, it has been proposed that they may play a role in the pathophysiology of these diseases, by an interaction with NK-1Rs and NK-2Rs (4, 5). The presence and role of NK-3Rs in the human lung remains to be determined.

Strong evidence for an important role for tachykinins in the regulation of parasympathetic ganglion neurotransmission has been revealed recently in guinea pig airways (6). Importantly, these data show NK-3R-mediated tachykinin-induced modulation of lower airway ganglion neurotransmission (6), associated with cholinergic airway smooth muscle contraction (7), and direct activation of parasympathetic neurons associated with airway smooth muscle relaxation (8). No role for NK-3R in such responses has been identified in the human lung, however, despite the fact that there is a relatively dense neurokinin-containing innervation in the intrinsic parasympathetic ganglia in human airways (9, 10). This led us to determine whether similar receptors exist in human bronchial ganglia neurons and whether they had a potential role to play in the pathophysiology of lung disease. The present study reports for the first time the immunohistochemical detection of human airway ganglion cell NK-3Rs and their role in the modulation of peripheral reflex parasympathetic neurotransmission. The NK-3R may represent a novel therapeutic approach for the treatment of lung diseases, such as COPD, which are characterized by aberrant parasympathetic drive. Some results in this study have been previously reported in the form of an abstract (11).

METHODS

Tissue Preparation

Methods for electrophysiologic recording and visualization of bronchial ganglia neurons have been described recently (12, 13). Bronchi dissected from the lungs of 28 human organ donors were transferred to warmed (37

Confocal Microscopy Studies

Ganglia were located as previously described (12), pinned to a Sylgard block, and fixed in 4% formaldehyde (freshly prepared from paraformaldehyde) in phosphate-buffered saline (pH 7.4) for 2 hours at 4°C and then rinsed and stored in phosphate-buffered saline. The ganglia were left pinned to Sylgard for the entire protocol including viewing on the confocal microscope (14). The ganglion preparations were preincubated for 1 hour in blocking buffer (10% normal goat serum; 0.1% Triton X-100; Tris-buffered saline: 50 mM Tris-HCL, 100 mM NaCl, pH 7.4) and then exposed overnight at 4°C to both the primary antibodies diluted in blocking buffer, with or without NK-3R blocking peptide (4 mg/ml). The primary antibodies used were the polyclonal neuronspecific marker protein gene product (PGP) 9.5 (1:300; Ultraclone, Wellow-Isle of Wight, UK) and a mouse monoclonal antibody to NK-3R (1:5, Courtesy of Dr. James Krause). Primary incubation solutions were left 3 hours at room temperature before being used to allow the blocking peptide and NK-3R antibody to bind (14).

Ganglia were then washed repeatedly at 4°C over a 24-hour period in wash buffer (0.1% Triton X-100; Tris-buffered saline: 50 mM TrisHCL, 100 mM NaCl, pH 7.4). Incubation with secondary antibodies (goat antimouse Alexa Fluor 488 and goat antirabbit Alexa Fluor 546; 1:250; Molecular Probes, Eugene, OR) was conducted overnight at 4°C in the dark. Preparations were again washed over a 24-hour period in wash buffer and then examined with a BioRad MRC1024 laser scanning confocal microscope (Hemmel Hempstead, UK) using sequential imaging for dual emission protocols (14).

Data Presentation and Statistical Analysis

Unless otherwise stated, all data are summarized as a mean ± the SEM of n experiments, where n is the number of neurons in electrophysiologic preparations. The effects of the NK-3R antagonists on nerve-stimulated or capsaicin-induced depolarizations were compared using paired and nonpaired Student's t test. In all cases, p values less than 0.05 were considered significant.

Drugs and Reagents

Hexamethonium, ASM-substance P (Arg-Pro-Lys-Pro-Gln-Gln-Phe-PheSar-Leu-Met [O2]-NH^sub 2^), ßAla-8-neurokinin A-4-10, atropine sulfate, and capsaicin (8-methyl-N-vanillyl-trans-6-nonenamide) were obtained from Sigma (St. Louis, MO). CP 99994 ([+], [2S, 3S]-3-[2-methoxybenzyl-amino]-2-phenyl-piperidine) was a gift from Merck Frost (Kirkland, PQ, Canada). SR 48968 ([S]-N-methyl-N-[4-acetylamino-4-phenylpiperidino2- (3,4-dichlorophenyl) butyl]benzamide), SB 223412 ([S]-[-]-N- [a-ethylbenzyl]-3-hydroxy-2-phenylquinoline-4-carboxamide), and SB 235375 ([-]-[S]-N- [a-ethybenzyl]-3-[carboxymethoxy]-2-phenylquinoline-4-carboxamidehydrochloride) were synthesized by the Department of Medicinal Chemistry at GlaxoSmithKline (King of Prussia, PA). Senktide analogue ([Asp^sup 6^, Asp^sup 7^, MePhe^sup 8^]-SP6-11) was obtained from Peninsula Laboratories (Belmont, CA). Stock solutions of hexamethonium (100 mM), SB 235375 (5 mM), ASM-substance P (1 mM), ßAla-8-neurokinin-A-4-10 (1 mM), atropine sulfate (10 mM), and senktide analogue (1 mM) were dissolved in water. SR 48968 (1 mM), CP 99994 (10 mM), and SB 223412 (5 mM) were dissolved in dimethyl sulfoxide, and capsaicin (10 mM) in ethanol. Vehicle controls at the dilutions shown in the results section had no effect on the resting potential or input resistance of these neurons (data not shown).

RESULTS

Confocal Microscopy

Immunofluorescent staining of NK-3R in whole mount preparations of human isolated bronchial parasympathetic ganglia (n = 4) was examined by confocal microscopy. An antibody to PGP 9.5 was used as a marker of neurons including ganglion cell bodies. As expected, cell bodies and their processes were positive for PGP 9.5 staining (red) (Figure 1A). NK-3R staining (green) was also seen in these ganglion structures (Figure 1B). Instead of staining the entire structure as observed with PGP 9.5, however, NK-3R appeared as localized, punctate staining, associated with some, but not all, of the neuronal cell bodies. When the images for PGP 9.5 and NK-3R staining were computer-superimposed, areas of colocalized PGP 9.5 and NK-3R appear in yellow (Figure 1C). When the NK-3R antibody was incubated with its blocking peptide, there was no staining even though the PGP 9.5 staining was still visible in red (data not shown).

Electrophysiology

Electrophysiologic current clamp recordings were made from the somas of intrinsic ganglia neurons as previously described (12); the active and passive membrane properties (e.g., resting membrane potentials and membrane resistance) were similar to those reported in that study, with ranges of -31 to -68 mV, and 10 to 70 MO, respectively.

Effects of nerve stimulation. Sensory innervation of human bronchial ganglia neurons was demonstrated by antidromic stimulation of sensory nerves. Stimulation of afferent nerves (1 millisecond, 40 V, 30 Hz, 5-second train) evoked a 4.5 ± 0.8 mV (n = 5) depolarization of the resting membrane potential in a subpopulation (14 of 21) bronchial ganglia neurons (Figure 2A). At the peak of the depolarization, there was a 32 ± 4% decrease in membrane resistance (Figure 2A). In preliminary experiments, the depolarization (3.9 ± 0.8 mV; n = 4) evoked by an initial antidromic nerve stimulation was followed by a 30-minute period and a second nerve stimulation evoked a response that was not different than the first (p > 0.05).

The NK-3R-selective antagonists SB 235375 (6, 15) or SB 223412 (6, 16) essentially abolished nerve-stimulated membrane depolarization. In paired preparations, in the presence of SB 235375 (1 µM, n = 4) or SB 223412 (1 µM, n = 3), antidromic stimulation of sensory nerves evoked a 0.5 ± 1 mV depolarization (Figure 2B) of the resting potential (p < 0.05) when compared with predrug control responses (4.4 ± 0.9 mV for SB 235375 and 3.8 ± 1.1 mV for SB 2223412; Figure 2C) with no change in the membrane resistance. In separate, unpaired experiments, ganglia were perfused with 0.1 µM SB 223412 for greater than or equal to 1 hour and nerve stimulation had little effect on four neurons; depolarization in the presence of SB 223412 was 0.5 ± 0.5 mV (n = 4; not shown). Similar observations were made with 0.1 µM SB 235375 (n = 2; not shown).

In three preparations examined, continuous superfusion (= 30 minutes) with hexamethonium (100 µM), atropine (1 µM), and NK-1R antagonist CP 99994 (0.1 µM) had no effect on the antidromic nerve-evoked depolarization: control response, 4.2 ± 1.1 mV; response after treatment, 4 ± 1.2 mV (n = 3; p > 0.05; not shown). In four preparations, pretreatment (= 1 hour) with the NK-2R antagonist SR48968 (0.1 µM), hexamethonium (100 µM), and atropine (1 µM) had no effect on nerve-evoked depolararization (3.7 ±1.5 mV).

Effects of capsaicin. Further evidence for sensory innervation of human bronchial ganglia neurons was obtained from examination of the response to capsaicin, an irritant that releases neuropeptides from sensory C fibers. Bath application of capsaicin (10-20 ml, 1 µM) depolarized the resting potential of four ganglion neurons (8.2 ± 2 mV; Figure 3A). One neuron depolarized to threshold, resulting in generation of high-frequency action potentials (Figure 3B). Repeated applications of capsaicin at this concentration (1 µM) evoked depolarizations of the resting potential (n = 2). In paired preparations the NK-3R antagonists SB 235375 (1 µM, n = 3) or SB 223412 (1 µM, n = 3) markedly inhibited the depolarization response to capsaicin: 1 ± 1 m V depolarization of the resting potential (p < 0.05 when compared with predrug control responses; Figure 3C) with no change in the input resistance. One neuron hyperpolarized 2 mV in response to capsaicin (1 µM) in the presence of SB 235375 (1 µM; not shown). Repeated application of higher concentrations of capsaicin (10 or 30 µM) in responsive neurons caused tachyphylaxis of the second response (n = 2 for each concentration; not shown). In seven neurons that were unresponsive to antidromic nerve stimulation (see above), capsaicin (1-10 µM) either had no effect on the resting membrane potential (n = 4) or caused a small depolarization (1 ± 1 mV, n = 3) of the resting potential (not shown).

Effects of tachykinin receptor agonists. The responses to nerve stimulation and capsaicin application were mimicked by the NK-3R-selective agonist, senktide analogue. Senktide analogue (0.1 µM, 8-10 ml, 1 minute) caused a 4.3 ± 2 mV (n = 3, range of 3-6 mV) depolarization of human bronchial ganglion neurons (Figure 4A); a second application of senktide analogue (0.1 µM, 8-10 ml), 30 minutes after the first application, elicited a 3 ± 1 mV depolarization (p > 0.05 when compared with the first application).

In a separate series of experiments, a single application of senktide analogue (0.1 µM, 8-10 ml, 1 minute) caused a 6.4 ± 2 mV membrane depolarization (range of 5-9 mV) in six neurons; three other neurons did not depolarize in response to initial application of senktide analogue and were not further studied (not shown). SB 235375 (1 µM; 30-minute perfusion pretreatment), essentially abolished the response to senktide analogue (0.1 µM, 8-20 ml, 1-2 minutes) in four of the responding neurons; depolarization in the presence of SB 235375 was 1 ± 1 mV (p < 0.05 compared with first application; n = 4). SB 223412 (1 µM) had a similar effect on senktide analogue (0.1 µM, 8-20 ml, 1-2 minutes) depolarizations in two neurons (not shown). In separate, unpaired experiments, ganglia were perfused with 0.1 µM SB 223412 for greater than or equal to 60 minutes and senktide analogue (0.1 µM, 8-20 ml, 1-2 minutes) had little or no effect on four neurons; depolarization in the presence of SB 223412 was 0.7 ± 0.5 mV (n = 4; not shown). Similar observations were made with 0.1 µM SB 235375 (n = 2; not shown).

The NK-1R-selective agonist ASM-substance P (0.01 ± 1 µM, 8-10 ml, 1 minute) did not depolarize the resting potential, but at concentrations of 0.1 and 1 µM increased the membrane resistance 23 ± 11% (range of 12-44%) in three of six neurons (Figure 4C); this effect was blocked with the NK-1 receptor antagonist CP 99994 (1 µM; not shown). Neurokinin-A (0.1 µM, 8-10 ml, 1 minute; n = 4) or the NK-2R-selective agonist ßAla-80-neurokinin-A-4-10 (1 µM, 8-10 ml; n = 2) had no effect on the resting membrane potential or input resistance (not shown).

DISCUSSION

Although the effects and potential pathophysiologic roles of tachykinins and the NK-1Rs and NK-2Rs in the mammalian lung have been extensively studied (4, 5, 17), there is limited information on NK-3Rs, in particular, in human lung. Indeed, only very low levels of NK-3R have ever been reported in human lung (18). The present study is the first to identify the location and define the function of NK-3Rs in a subpopulation of human bronchial parasympathetic ganglion neurons. These data provide evidence for a unique neuromodulatory pathway with the potential to significantly influence airway caliber in disease, and may represent a novel therapeutic approach (i.e., NK-3R antagonism) for lung disease characterized by dysfunction in neuronal inputs (e.g., COPD) (2).

The size and location of these NK-3R-containing cells are very similar to principal neurons in human bronchial parasympathetic ganglia (12). Intracellular electrophysiologic recordings provided convincing evidence for peripheral reflex regulation of parasympathetic nerve activity in human bronchus. Given that the parasympathetic autonomic nervous system provides the dominant neural efferent drive to critical submucosal tissues, including airway smooth muscle, the negative impact of upregulated NK-3R-mediated enhancement of a reflex is likely to be significant, particularly in airway diseases involving airflow limitation, such as asthma and COPD. The predominant role of the parasympathetic system in controlling lung function is highlighted by the use of inhaled anticholinergics, such as ipratropium or tiotropium, for symptomatic relief in COPD (2).

Direct recording of cellular electrical activity revealed that activation of the capsaicin-sensitive nerves innervating the bronchus evoked a marked membrane depolarization of airway ganglion neurons. The cells that responded to nerve stimulation were most likely principal neurons and not interneurons, because fast excitatory postsynaptic potentials were recorded in these cells (12). That the observed response is caused by antidromic stimulation of sensory nerves came from evidence that (1) the response is associated with tachykinins; (2) the response is not inhibited by a cholinergic receptor antagonist (e.g., hexamethonium or atropine); and (3) we can mimic the response with capsaicin, a compound known to cause the release of the transmitter from sensory nerves. Evidence that these responses were caused by NK-3R activation, and were not secondary to NK-1, NK-2, nicotinic, or muscarinic receptor activation (19), was clearly demonstrated by the lack of effect of NK-1R or NK-2R antagonism, hexamethonium, or atropine, respectively, on nerve-stimulated depolarizations. In addition, this observation is supported by the lack of significant effect of selective NK-1R or NK-2R agonists. Furthermore, antidromic sensory nerve stimulation and capsaicin-induced membrane depolarization were mimicked by the NK-3R-selective agonist senktide analogue. Importantly, all of these responses were attenuated by the NK-3R antagonist SB 223412 (16) and abolished by the closely related analogue SB 235375 (15). The time course and changes in membrane resistance associated with the depolarizations of human bronchial ganglia neurons by senktide analogue were similar to those reported for guinea pig bronchial neurons (20). Unlike guinea pig bronchial neurons, however, activation of NK-3R in human ganglia depolarized the neurons to action potential threshold (Figure 3). In the human airway ganglia, only a subpopulation of neurons responded to nerve stimulation, capsaicin, or senktide analogue, correlating with the anatomic observation that only a subpopulation of neurons was immunoreactive for the NK-3R. Such results suggest either that the sensory nerve fibers have degraded or, possibly, heterogeneity within the population of neurons within a ganglion.

Previous studies have demonstrated that activation of tachykinin receptors alters cholinergic contractions of the lower airways of rabbits (21) and guinea pigs (7, 22), with such effects usually associated with postganglionic mechanisms. The postganglionic effects of neurokinins on both the cholinergic nerves and on the airway smooth muscle have been shown to be caused exclusively by activation of NK-1R and NK-2R (21, 22). A unique role for NK-3R activation has been recently reported for guinea pig lower airway parasympathetic ganglia where NK-3R activation does not elicit action potentials in ganglia neurons but does potentiate synaptic transmission and cholinergic contractions (6). The results in the present study are entirely distinct from these reported effects of tachykinins on guinea pig airway nerves and describe an additional, novel pathway for peripheral reflex regulation of smooth muscle tone in the lower airway. Conceptually, the peripheral reflex may be a mechanism for local regulation of airflow to the lung, such that when a stimulus is localized to one bronchial segment, parasympathetic tone is altered only in that segment.

The present study provides morphologic and electrophysiologic evidence for NK-3R on principal neurons in human bronchial parasympathetic ganglia and also provides the first reported evidence of a tachykinin-mediated peripheral reflex in any human autonomic ganglion. Based on the present studies, tachykinins released from sensory nerves in airway ganglia may activate NK-3R and have an important role in localized neural regulation of airflow to the lungs. That NK-3R are restricted to parasympathetic nerves indicates their potential to mediate tachykinin-induced transmission within the ganglion (23), without affecting normal function of sensory nerves or neurotransmitters released from parasympathetic ganglion nerve terminals. These findings demonstrate an unrealized role for NK-3Rs in the modulation of aberrant parasympathetic drive in the lower airways.