Prostaglandin E2 potentiates a TTX-resistant sodium current in rat capsaicin-sensitive vagal pulmonary sensory neurones
- Corresponding author L.-Y. Lee: Department of Physiology, University of Kentucky Medical Center, 800 Rose Street, Lexington, KY 40536-0298, USA. Email: lylee{at}uky.edu
Abstract
Capsaicin-sensitive vagal pulmonary neurones (pulmonary C neurones) play an important role in regulating airway function. During airway inflammation, the level of prostaglandin E2 (PGE2) increases in the lungs and airways. PGE2 has been shown to sensitize isolated pulmonary C neurones. The somatosensory correlate of the pulmonary C neurone, the small-diameter nociceptive neurone of the dorsal root ganglion, contains a high percentage of tetrodotoxin-resistant sodium currents (TTX-R INa). Therefore, this study was carried out to determine whether these channel currents are involved in the PGE2-induced sensitization of pulmonary C neurones. We used the perforated patch-clamp technique to study the effects of PGE2 on the TTX-R INa in acutely cultured capsaicin-sensitive pulmonary neurones that were identified by retrograde labelling with a fluorescent tracer, DiI. We found that the pulmonary neurones sensitive to capsaicin had a higher percentage of TTX-R INa than that of capsaicin-insensitive pulmonary neurones. PGE2 exposure increased the evoked TTX-R INa when experiments were performed at both room temperature and at 37°C. Furthermore, stimulation of the adenylyl cyclase/protein kinase A pathway with either forskolin or Sp-5,6-DCl-cBiMPS potentiated the TTX-R INa in a manner similar to that of PGE2. We conclude that these modulatory effects of PGE2 on TTX-R INa play an important role in the sensitization of pulmonary C neurones.
Prostaglandin E2 (PGE2) is a major inflammatory mediator, and its production is found to be increased in the lungs and airways during mucosal inflammation (Liu et al. 1988; Wenzel et al. 1989; Holtzman, 1991; Jörres et al. 1995). PGE2 has been shown to have direct sensitizing effects on sensory neurones via modulation of the function of a variety of voltage-gated ion channels (England et al. 1996; Gold et al. 1996; Cantrell & Catterall, 2001). Indeed, an earlier study in our laboratory showed that PGE2 could potentiate voltage-dependent inward currents in capsaicin-sensitive vagal pulmonary neurones (pulmonary C neurones) (Kwong & Lee, 2002). However, that study did not identify the specific ion channels that were modulated by PGE2.
Afferent activities arising from the lung structures are conducted almost exclusively in vagus nerves and their branches in the rat. Morphological studies have shown that approximately 75% of the vagal pulmonary afferent fibres are unmyelinated C-fibres (Agostoni et al. 1957). It is well documented that vagal bronchopulmonary C-fibre sensory nerves play an important role in regulating airway function under various pathophysiological conditions (Coleridge & Coleridge, 1984; Lee & Pisarri, 2001). Activation of these afferents can elicit centrally mediated reflex responses, including reflex bronchoconstriction, hypersecretion of mucus, cough, tachypnoea, bradycardia and hypotension (Coleridge & Coleridge, 1984; Lee & Pisarri, 2001). Furthermore, it is known that intense stimulation of these afferent endings triggers the release of tachykinins that can produce additional local effects such as bronchoconstriction, protein extravasation and inflammatory cell chemotaxis.
Studies of nociception in dorsal root ganglion (DRG) neurones reveal that a subpopulation of neurones – small-diameter, sensitive to chemical activation, isolectin B4-positive and neurofilament-negative – express a high level of tetrodotoxin-resistant (TTX-R) sodium current (INa) and channel protein (Ogata & Tatebayashi, 1992; Michael & Priestley, 1999; Stucky & Lewin, 1999; Amaya et al. 2000). Many of the characteristics that describe nociceptive DRG neurones can also be ascribed to its vagal correlate in the lung: pulmonary C neurones. Therefore, we hypothesized that the sensitizing effect of PGE2 on vagal pulmonary C neurones could be in part mediated through voltage-gated TTX-R sodium channels. The objective in this study was to determine whether TTX-R INa existed in capsaicin-sensitive vagal pulmonary neurones and if so, whether PGE2 could modulate voltage-gated TTX-R INa in these neurones. In addition, we showed previously that activation of the EP2 prostanoid receptor, which is thought to result in an increase in intracellular cAMP, could partially mimic the potentiating effects of PGE2 (Kwong & Lee, 2002). Therefore, a second aim was to determine whether direct activation of a cAMP-dependent protein kinase A (PKA) intracellular pathway might also modulate TTX-R INa in capsaicin-sensitive vagal pulmonary neurones.
Methods
Identification of vagal pulmonary neurones
Vagal sensory neurones innervating the lungs and airways were identified by retrograde labelling from the lungs using the fluorescent tracer, 1,1′-dioctadecyl-3,3,3′,3′-tetramethylindocarbocyanine perchlorate (DiI; Molecular Probes) following a protocol reported previously (Kwong & Lee, 2002). Briefly, young adult (150–200 g) male Sprague-Dawley rats (Harlan, Indianapolis, IN, USA) were anaesthetized with an intraperitoneal injection of pentobarbital (50–60 mg kg−1) and intubated with a polyethylene tube such that the tip was positioned in the trachea above the thoracic inlet. DiI (0.25 mg ml−1) was initially dissolved and further sonicated in ethanol and then diluted in saline (1% final ethanol concentration). With the rat tilted head-up at ∼30 deg, the DiI solution was instilled into the lungs in two 0.25 ml bolus injections with a 10 min interval between them. The lungs were hyperinflated (3 × tidal volume) periodically to prevent atelectasis. After 7–11 days, an interval previously determined to be sufficient for the dye to reach the cell body, the vagal sensory ganglia, comprised of the nodose and intracranial jugular ganglia, were harvested for cell culture.
Primary culture of vagal sensory neurones
The animals were anaesthetized with halothane (100%) and immediately decapitated. The head was immersed in ice-cold Hank's balanced salt solution (HBSS). Within 10 min, nodose and jugular ganglia with attached nerve trunks were extracted under a dissecting microscope and placed in ice-cold Dulbecco's minimal essential medium/F12 (DMEM/F12; Gibco/Invitrogen) solution.
Nodose ganglia were separated from jugular ganglia and cultured separately. Each ganglion was desheathed, cut into about five pieces, placed in 0.125% collagenase type IV and incubated in a humidified chamber for 1 h in 5% CO2 in air at 37°C. The ganglion suspension was centrifuged (150 g, 5 min) and the supernatant aspirated. The ganglion pellet was briefly (< 1 min) resuspended in 0.05% trypsin and 0.53 mm EDTA in HBSS, and centrifuged (150 g, 5 min). The ganglion pellet was then resuspended in a modified DMEM/F12 solution (DMEM/F12 supplemented with 10% (v/v) heat-inactivated fetal bovine serum, 100 U ml−1 penicillin and 100 μg ml−1 streptomycin, and 100 μm MEM non-essential amino acids) and gently titurated with a small-bore fire-polished Pasteur pipette. The dispersed cell suspension was centrifuged (500 g, 8 min) through a layer of 15% bovine serum albumin to separate the cells from the myelin debris. The pellets of nodose and jugular ganglion cells were resuspended in the modified DMEM/F12 solution supplemented with 50 ng ml−1 2.5S NGF (Gibco/Invitrogen) and plated onto dry poly l-lysine-coated glass coverslips. Initially, small droplets of resuspended cells were incubated to promote cell adhesion at a high density. After 3 h, the coverslips containing the cultured cells were immersed in L-15 media (Gibco/Invitrogen) supplemented with 10% fetal bovine serum and 5 mm Hepes. This media was kept in ambient atmosphere and shielded from light. The cells were used immediately for experimentation.
Perforated-patch clamp electrophysiological recording
TTX-R INa were isolated in a bath solution that contained (mm): 90 CsCl, 10 glucose, 10 Hepes, 2.8 MgCl2, 5.4 KCl, 36 NaCl, 10 tetraethylammonium-Cl (TEA-Cl), 40 4-aminopyridine (4-AP), and 100 nm tetrodotoxin citrate (TTX; Alomone Laboratories, Jerusalem, Israel); Na+ concentration was reduced in order to record the TTX-R INa with greater temporal resolution. The pH was adjusted to 7.4 with CsOH and osmolarity to ∼310 mosmol l−1 with sucrose. The pipette solution contained (mm): 1.0 CaCl2, 130 CsCl, 10 EGTA, 10 Hepes, 0.5 MgCl2, and 8 NaCl; pH was adjusted to 7.1 with CsOH and osmolarity to ∼300 mosmol l−1 with sucrose. The pipette electrodes were formed from borosilicate glass (Warner Instruments, Inc.) pulled using a micropipette puller (Sutter Instrument Co., Novato, CA, USA). The electrode tips were coated with Sylgard (Dow Corning) and fire polished to a tip resistance of 2–4 MΩ in the bath solution. The electrophysiological recordings were made using an Axopatch 200B/pCLAMP8 (Axon Instruments) low-pass filtered at 5 or 10 kHz and digitized at 50 or 100 kHz, respectively. Series resistance was compensated at ∼80% except for the experiments recorded at 37°C, which were uncompensated due to excessive oscillations. In some experiments, a −P/4 leak subtraction protocol was applied after the test pulse. Interpulse intervals for all experiments were a minimum of 8 s to avoid frequency-dependent effects (Rush et al. 1998). A 1% agar salt bridge closed the circuit between the silver chloride micropipette electrode and the silver chloride reference electrode. The liquid junction potentials and agar bridge/bath potentials were symmetrical and produced opposing junction effects. Calculations for liquid junction potential were performed using the junction potential module within the data acquisition software, based upon JPCalc software (Barry, 1994).
The whole-cell perforated patch-clamp technique was used. The stock solution of gramicidin (Sigma-Aldrich) was prepared in DMSO (100 mg ml−1) and stored desiccated in the dark at 4°C. Fresh solutions of gramicidin (50–100 μg ml−1) were made every 2 h by dissolving the stock solution in the pipette solution with brief sonication. A stable series resistance (6–15 MΩ) was achieved ∼30 min after the establishment of the gigaohm seal and maintained for up to 1–2 h; series resistances were (MΩ): 9.1 ± 0.6, 8.4 ± 0.6, 9.4 ± 0.8, 9.0 ± 0.6 and 10.0 ± 1.0 at 0, 0.5, 1.0, 1.5 and 2 h, respectively.
Current–voltage curves were fitted using a modified Boltzmann function:
where IV is the current generated at the test potential, Gmax is the maximal macroscopic conductance, V is the test potential, Erev is the reversal potential, Vh is the half-activation potential, and k is the slope factor. Conductance was calculated using the equation:
where G is conductance, I is current, Vtest is the test potential, and Erev is the reversal potential. Erev was obtained using the modified Boltzmann function. Conductance–voltage relationships were fitted with a single Boltzmann
function. All curve fitting was performed using SigmaPlot 7 (SPSS Inc., Chicago, IL, USA). Tetrodotoxin-sensitive (TTX-S)
currents were obtained by arithmetic subtraction of the TTX-resistant current from the total current using the Clampfit 8
software (Axon Instruments).
Electrophysiological recordings were made in a small-volume (0.15 ml) perfusion chamber (Warner Instruments) at room temperature (21–25°C) or at 37°C. The coverslip containing the cells was centred in the perfusion chamber, which was superfused by gravity-feed with the bath, drug or vehicle control solutions at ∼1 ml min−1. PGE2 (Cayman Chemical, Ann Arbor, MI, USA) was initially dissolved in 100% ethanol, stored at −80°C, and then diluted in the bath solution to 1 μm; the final concentration of ethanol was 0.007%. Forskolin was initially dissolved in DMSO, stored at room temperature, and then diluted in the bath solution to 1 μm; the final concentration of DMSO was 0.008%. An inactive analogue for forskolin, 1,9-dideoxyforskolin, was initially dissolved in DSMO, stored at −20°C, and then diluted in the bath solution to 10 μm; the final concentration of DSMO was 0.1%. Stimulation of PKA was achieved using a membrane-permeant cAMP analogue, Sp-5,6-dichloro-1-β-d-ribofuranosylbenzimidazole-3′,5′-monophosphorothioate (cBiMPS; Axxora LLC, San Diego, CA, USA). cBiMPS was initially dissolved in 100% DMSO, stored at −20°C, and then diluted in the bath solution to 50 μm; the final concentration of DMSO was 0.5%. The chemical stimulant chosen for this study, capsaicin, is a potent stimulant of C-fibre afferents and a selective agonist of the transient receptor potential type 1 vanilloid receptor (TRPV1) (Caterina et al. 1997). It has been used extensively to preferentially stimulate pulmonary C fibres in rats (Ho et al. 2001). Confirmation of capsaicin sensitivity was carried out in a modified Ringer solution containing (mm): 1.8 CaCl2, 10 glucose, 10 Hepes, 1.0 MgCl2, 5.4 KCl and 136 NaCl. The pH was adjusted to 7.4 with NaOH and osmolarity to 310 mosmol l−1 with sucrose. Capsaicin, diluted in the modified Ringer solution to 1 μm, was delivered at the end of the experiment using a 3-channel fast-stepping perfusion system, with its tip positioned to ensure that the cell was fully within the stream of the perfusate. A cell was identified as capsaicin sensitive if a 4 s pulse of capsaicin, but not the vehicle control, elicited a transient increase (> 100 pA) in current that followed the time course of the capsaicin challenge. Data were collected from only one cell per dish to avoid possible contamination of the cells by the chemical agents. All chemicals were obtained from Sigma-Aldrich unless otherwise noted.
Experimental design
The experimental protocols were designed to answer the following questions: (1) are TTX-R INa present on capsaicin-sensitive vagal pulmonary neurones; (2) does PGE2 potentiate TTX-R INa in capsaicin-sensitive vagal pulmonary neurones; (3) can TTX-R INa also be modulated by activation of adenylyl cyclase; and (4) can direct stimulation of PKA potentiate TTX-R INa? Data are reported as mean ± s.e.m. Statistical comparisons were made using paired or unpaired t tests, where appropriate. A P value < 0.05 was considered significant.
Results
Presence of TTX-R INa in capsaicin-sensitive vagal pulmonary neurones
The family of TTX-R INa was generated with a voltage-clamp protocol that consisted of 40 ms depolarizing steps from −70 mV to 45 mV in 5 mV increments from a holding potential of −80 mV in the presence of TTX (100 nm; Fig. 1). At the holding potential of −80 mV, the TTX-R INa was not inactivated. In general, the TTX-R INa was slow to activate and inactivate, and displayed a small persistent inward current (Fig. 1B) that lasted longer than 40 ms. In contrast, TTX-sensitive (TTX-S) INa (Fig. 1C), generated from a point-for-point subtraction of the TTX-R INa from the total INa (Fig. 1A), transiently activated and inactivated with a much shorter time course and returned to baseline within 10 ms. The TTX-R INa generally activated at threshold potentials more positive than that of TTX-S INa (Fig. 1D). Similarly, the peak inward currents developed at higher potentials in TTX-R INa than in TTX-S INa (Fig. 1D). In addition, a portion of the neurones displayed a persistent outward current (e.g. Fig. 1A and B).
We measured the ratio between TTX-R INa and the total INa in various subpopulations of nodose and jugular neurones, including pulmonary (DiI-labelled), non-pulmonary (non-DiI-labelled), and capsaicin-sensitive and -insensitive neurones. A total of 126 neurones, harvested from 67 rats (∼240 g), were studied. Among the total population of neurones tested (n = 116), the peak TTX-R INa accounted for 40.0 ± 3.0% of the peak total INa. The peak current was defined as the maximum evoked current from among the entire family of evoked currents generated using the voltage-clamp protocol described above. The percentage TTX-R INa was significantly greater in capsaicin-sensitive pulmonary neurones (58.8 ± 4.9%; n = 28; Fig. 2A and B) than in capsaicin-insensitive pulmonary neurones (17.9 ± 3.9%; n = 23; P < 0.001). This increase in percentage TTX-R INa was correlated with capsaicin sensitivity: capsaicin-sensitive neurones, irrespective of DiI-labelling, had a greater percentage of TTX-R INa (56.0 ± 4.8%; n = 43; Fig. 2C) than capsaicin-insensitive neurones (18.3 ± 3.3%; n = 37; P < 0.001). No difference was detected in the percentage of TTX-R INa expressed between pulmonary (42.7 ± 3.4%; n = 79; Fig. 2D) and non-pulmonary neurones (46.0 ± 9.9%; n = 17; P > 0.05). No difference was detected in the percentage of TTX-R INa between jugular neurones (43.6 ± 6.4%; n = 38; Fig. 2E) and nodose neurones (38.2 ± 3.2%; n = 78; P > 0.05).
Figure 2F illustrates the distribution of total INa as a function of cell capacitance. The total INa density was significantly greater in capsaicin-insensitive pulmonary neurones (−461.4 ± 57.3 pA pF−1; n = 23; Fig. 2G) than in capsaicin-sensitive pulmonary neurones (−214.6 ± 24.8 pA pF−1; n = 28; P < 0.001). The total INa density in capsaicin-insensitive neurones (−429.1 ± 42.7 pA pF−1; n = 37; Fig. 2H) was significantly greater than in capsaicin-sensitive neurones (−271.1 ± 37.6 pA pF−1; n = 43; P < 0.01). No difference in total INa was observed between pulmonary neurones (−318.6 ± 27.2 pA pF−1; n = 79; Fig. 2I) and non-pulmonary neurones (−297.5 ± 41.4 pA pF−1; n = 17; P > 0.05). The total INa in jugular neurones (−290.7 ± 38.9 pA pF−1; n = 38; Fig. 2J) was no different from that of nodose neurones (−370.2 ± 32.7 pA pF−1; n = 78; P > 0.05).
The expression of TTX-R INa appeared to correlate with the sensitivity to capsaicin, but not with the type of ganglion or pulmonary innervation: TTX-R INa density was greater in capsaicin-sensitive neurones (−130.8 ± 15.0 pA pF−1; n = 43; data not shown) than in capsaicin-insensitive neurones (−72.8 ± 14.7 pA pF−1; n = 37; P < 0.01).
PGE2 potentiates TTX-R INa in capsaicin-sensitive vagal pulmonary neurones
In capsaicin-sensitive vagal pulmonary neurones, PGE2 elicited an increase in the TTX-R INa in 7 out of the 8 neurones tested (Fig. 3A, C and D). The potentiating effects of PGE2 (1 μm) on TTX-R INa in capsaicin-sensitive vagal pulmonary neurones began to emerge within ∼2 min and reached a steady state within ∼6 min after the start of perfusion (Fig. 3B). PGE2 caused a significant increase in maximal conductance (108 ± 15.8 nS; Fig. 4A, Table 1) compared with that of control (84.1 ± 12.2 nS; P < 0.01). PGE2 shifted half-maximal activation (Vh) in a hyperpolarized direction (control: −5.8 ± 2.5 mV; PGE2: −12.9 ± 1.8 mV; P < 0.01; Fig. 4B, Table 1) and increased the voltage sensitivity (k) (control: 4.9 ± 0.6 mV per e-fold change in current; PGE2: 3.2 ± 0.4 mV per e-fold change; P < 0.05; Fig. 4B, Table 1). At peak current, voltage errors were 7.6 ± 1.3 and 11.9 ± 1.4 mV during control and PGE2 exposure, respectively. The TTX-R INa were activated at voltages more positive than −30 mV (Fig. 3D). The threshold of activation did not appear to be altered by PGE2 treatment (Fig. 4).
Steady-state inactivation was determined using a voltage-clamp protocol wherein membrane voltage was held at a series of voltages from −90 to 5 mV for 500 ms and stepped to a single depolarizing test potential of 10 mV for 20 ms (e.g. Fig. 5A). The current measured during the single depolarizing test potential, as a percentage of maximal current, reflected the relative number of channels participating, or available, in the response. We were unable to detect an effect of PGE2 on channel availability (control: −24.3 ± 1.8 mV; PGE2: −26.7 ± 1.8 mV; P > 0.05; Fig. 5B; Table 1). However, PGE2 increased voltage sensitivity (control: 4.0 ± 0.2 mV per e-fold change; PGE2: 3.3 ± 0.1 mV per e-fold change; P < 0.001) and maximal current (control: 3.8 ± 0.8 nA; PGE2: 5.9 ± 1.0 nA; P < 0.01; Fig. 5B, Table 1).
In contrast, modulation of the TTX-S INa by PGE2 was not detected. Mean peak TTX-S INa density were −413.9 ± 16.3 and −405.8 ± 67.3 pA pF−1 for control and during PGE2, respectively (n = 5; P > 0.05). Half-maximal activation (Vh) for control was −27.0 ± 1.7 mV and for during PGE2 was −29.0 ± 0.9 mV (n = 5; P > 0.05).
Intracellular mechanism underlying the potentiation of TTX-R INa
The naturally occurring diterpene, forskolin, was used to stimulate adenylyl cyclase directly and thereby increase the intracellular concentration of cAMP. In a manner similar to that of PGE2, forskolin (1 μm) potentiated TTX-R INa currents in capsaicin-sensitive vagal pulmonary neurones. Forskolin enhanced the family of transient inward currents elicited by the voltage-step protocol (e.g. Fig. 6A) in 11 out of the 12 capsaicin-sensitive pulmonary neurones studied. The currents were activated at voltages more positive than −30 mV; this activation threshold did not appear to be altered by forskolin treatment (Fig. 6). Group data of capsaicin-sensitive vagal pulmonary neurones showed that forskolin increased peak inward current (Fig. 6A), maximal conductance (control: 93.7 ± 14.0 nS; forskolin: 107.5 ± 14.9 nS; n = 12; P < 0.01; Fig. 6B; Table 2), voltage sensitivity (control: 5.5 ± 0.9 mV per e-fold change; forskolin: 4.0 ± 0.6 mV per e-fold change; P < 0.01) and modulated the half-activation potential (control: −9.0 ± 2.5 mV; forskolin: −13.8 ± 2.0 mV; P < 0.001; Fig. 6C, Table 2). At peak current, voltage errors were 8.5 ± 2.0 and 11.8 ± 2.7 mV during control and forskolin exposure, respectively. A change in the reversal potential was not detected with forskolin treatment (P > 0.05; Fig. 6A, Table 2). Treatment with an inactive analogue of forskolin, 1,9-dideoxyforskolin (10 μm), did not potentiate the TTX-R INa (Fig. 7B). Indeed, the peak current evoked at −5 mV appeared to have decreased to 83.9 ± 4.5% of control (n = 6) after the 1,9-dideoxyforskolin treatment (Fig. 7B). Likewise, no effect on the TTX-R INa was detected after vehicle treatment (98.1 ± 2.5% of control; n = 9; Fig. 7B). Responses to vehicle and 1,9-dideoxyforskolin were each significantly different from the response to forskolin (142.7 ± 11.1% of control; n = 13; P < 0.01).
To determine whether the forskolin-evoked potentiating effect was also present at body temperature, we recorded currents evoked from 20 ms voltage steps to −5 mV from Vhold (−80 mV) after exposure to forskolin (1 μm; 10 min) at 37°C. In capsaicin-sensitive pulmonary neurones, peak TTX-R INa evoked after exposure to forskolin was 55.6 ± 20.5% greater than after exposure to vehicle (n = 4; P < 0.05; e.g. Fig. 7C). Evoked current density before forskolin was −79.6 ± 21.4 pF pA−1 and after forskolin was −113.9 ± 23.7 pF pA−1 in capsaicin-sensitive pulmonary neurones (Fig. 7D).
To determine whether direct activation of protein kinase A could mimic the potentiating effects of PGE2, cBiMPS, a membrane-permeant analogue of cAMP, was tested in capsaicin-sensitive vagal pulmonary neurones. cBiMPS (50 μm) potentiated the peak TTX-R INa evoked by 20 ms voltage steps to −5 mV from Vhold (144.4 ± 12.7% of control; n = 6; Fig. 7E and F), whereas no effect was detected after the vehicle treatment (95.8 ± 3.2% of control; n = 6; Fig. 7F).
Lack of modulation of TTX-R INa in capsaicin-insensitive pulmonary neurones by PGE2
In stark contrast to capsaicin-sensitive pulmonary neurones, we failed to detect modulation of the TTX-R INa in capsaicin-insensitive pulmonary neurones by PGE2. Evoked TTX-R INa did not show a significant increase during PGE2 (1 μm) exposure (128.7 ± 35.0% of control; n = 4; Fig. 8A) compared with vehicle treatment (95.7 ± 1.9% of control; n = 3; P > 0.05). In a separate set of neurones, TTX-R INa evoked during forskolin exposure (109.5 ± 8.5% of control; n = 5; Fig. 8B) was no different from during vehicle exposure (99.5 ± 2.8% of control; n = 3; P > 0.05). Neurones exposed to cBiMPS (50 μm) displayed no potentiation in evoked TTX-R INa (96.5 ± 8.5% of control; n = 6; Fig. 8C), which was no different than in vehicle-treated neurones (94.5 ± 3.0% of control; n = 6; P > 0.05).
Discussion
These results provided evidence that capsaicin-sensitive vagal pulmonary neurones cultured from adult rats contained a substantial amount of TTX-R INa that was potentiated by PGE2. PGE2 increased TTX-R macroscopic conductance and affected other biophysical properties of TTX-R channels: half-activation potential was reduced and voltage sensitivity was increased. Further, both forskolin and cBiMPS potentiated TTX-R channel function, suggesting that PGE2 acted through a cAMP/PKA-mediated pathway. In addition, experiments performed at 37°C demonstrated that potentiation of the TTX-R INa by forskolin could occur at normal physiological temperatures. Lastly, in contrast to capsaicin-sensitive pulmonary neurones, TTX-R INa was not modulated by PGE2, forskolin or cBiMPS in capsaicin-insensitive pulmonary neurones.
Our data indicated a strong correlation between the ratio of TTX-R INa/total INa and sensitivity to capsaicin. The relative contribution of the TTX-R INa to the total INa was significantly higher in capsaicin-sensitive neurones, which was due to the combination of both a lower total INa and a higher TTX-R INa. It was unclear why capsaicin-sensitive neurones also expressed lower total INa density. Su and colleagues showed that capsaicin exposure decreased voltage-gated INa in DRG neurones (Su et al. 1999) and C-fibre action potentials recorded in the presence of TTX were blocked by capsaicin (Großkreutz et al. 1996). However, in our study, neurones were not exposed to capsaicin until the end of the experiment. Results of our study showed that the expression of TTX-R INa was almost 3-fold greater in capsaicin-sensitive pulmonary neurones than those insensitive to capsaicin. Hence, this finding has added additional attributes in defining the electrophysiological properties of this subset of sensory neurones in the lungs. A similar dominant distribution of TTX-R INa has been previously reported in small-size DRG neurones with nociceptive properties (Elliott & Elliott, 1993; McCleskey & Gold, 1999).
What is the extent to which PGE2 modulation of TTX-R INa could affect neuronal excitability? The answer probably depends on a number of factors including: (1) the subtype and the density of the EP prostanoid receptors expressed on the cell, (2) degree of the prostanoid receptor coupling to TTX-R sodium channels, and (3) the amount of TTX-R INa expressed. TTX-R INa is the major contributor to the upstroke of action potential in DRG nociceptor neurones (Blair & Bean, 2002). An earlier study showed that PGE2 could modulate total INa in cultured capsaicin-sensitive vagal pulmonary neurones (Kwong & Lee, 2002). Cardenas and colleagues showed that DRG neurones that express very little TTX-R INa were less sensitive to PGE2 modulation than neurones that expressed primarily TTX-R INa (Cardenas et al. 1997). Since TTX-S INa could contribute up to 100% of the inward INa, it follows that the total current in such neurones would be modulated negatively. Therefore, the total effect of PGE2 modulation would depend at least in part on whether TTX-R INa made up the majority of the INa, and whether other currents, such as potassium currents that are sensitive to PGE2 exposure, were present.
A small outward current persisted despite efforts to eliminate potassium currents (i.e. use of caesium, TEA and 4-AP in the bath or pipette solutions). In order to determine whether this residual outward current contributed to the modulation of the TTX-R INa, we examined whether this outward component was sensitive to PGE2 or forskolin. At the reversal potential for sodium, by definition, there was no net driving force for the flow of sodium current; therefore currents measured at this potential reflect other ionic fluxes. We found that the steady-state evoked currents measured at the calculated reversal potential for sodium were insensitive to either PGE2 or forskolin treatment (data not shown). Therefore, the modulatory effects of PGE2 and forskolin on the inward current were probably affecting the TTX-R INa. However, it should be noted that this protocol does not eliminate the possibility that PGE2 could have modulated a transient outward current. Indeed it has been well demonstrated that PGE2, cAMP and PKA can modulate potassium currents in a variety of cell types (Akins & McCleskey, 1993; Chung & Kaczmarek, 1995; England et al. 1996; Evans et al. 1999).
In this study, we have focused on the effect of a single inflammatory mediator, PGE2. However, it is well known that other inflammatory mediators can sensitize sensory neurones. For example, PGI2, an arachidonic acid metabolite, has also been shown to sensitize sensory neurones to a substantial degree (Smith et al. 1998). Like the receptors for PGE2, EP2 and EP4, the cognate receptor for PGI2, IP, is a G-protein-coupled receptor that interacts with Gs. Whether PGI2 can modulate TTX-R INa remains to be determined, but given that both forskolin and cBiMPS are able to modulate TTX-R INa, it is highly likely that receptors that utilize this intracellular pathway would also be able to modulate TTX-R INa as has been shown for DRG neurones (Gold et al. 1998). In addition, other Gs-mediated, non-cyclooxygenase-derived inflammatory agents may also play a role in the sensitization of airway neurones. For example, Gold et al. (1996) demonstrated that serotonin and adenosine could sensitize DRG neurones. It is unclear whether metabolites of the lipoxygenase pathway, which have been implicated in asthma (Drazen et al. 1999), could also act through an increased sensitivity in pulmonary sensory neurones.
Activation of PKA, either directly via cBiMPS, or indirectly via forskolin mimicked the potentiating effects of PGE2. The likely sequence of events is that the PGE2-elicited activation of adenylyl cyclase increases intracellular cAMP, which in turn activates PKA, which ultimately phosphorylated serine or threonine residues on ion channels (Murphy et al. 1993; Frohnwieser et al. 1997; Smith & Goldin, 1997). Whether the PKA molecule directly phosphorylated the ion channel responsible for the TTX-R INa detected in these capsaicin-sensitive vagal pulmonary neurones is not known, but subtypes of TTX-R sodium channels are known to be phosphorylated by PKA (Sangameswaran et al. 1996; Fitzgerald et al. 1999). Work with the cloned Nav1.8 channel reveals that elimination of PKA phosphorylation consensus sites in the intracellular loop linking transmembrane domains I and II can abrogate sensitivity to PGE2 treatment (Fitzgerald et al. 1999). However, recent studies from Rush and coworkers showed that in native neurones, Nav1.9 but not Nav1.8 can be modulated by PGE2 (Rush & Waxman, 2004). Results from our experiments demonstrating similar potentiating effects of forskolin and cBiMPS suggested an involvement of the cAMP/PKA signalling pathway.
Capsaicin-sensitive vagal pulmonary neurones correspond to vagal pulmonary C-fibres, which are distinguished from the other vagal pulmonary sensory afferents (i.e. slowly and rapidly adapting receptors) primarily by its sensitivity to chemical irritants such as capsaicin, H+, bradykinin and phenylbiguanide, a 5-HT3 agonist (Coleridge & Coleridge, 1984; Lee & Pisarri, 2001; Ho et al. 2001). In fact, it has been clearly demonstrated that capsaicin sensitivity is a distinct characteristic feature of C-fibre afferents innervating the lungs and airways in rats, guinea pigs and dogs (Bergren, 1997; Ho et al. 2001; Lee & Pisarri, 2001; Undem et al. 2004). In mice, Kollarik and colleagues observed that among sensory afferents innervating the trachea, C-fibres, which were defined as fibres with conduction velocities < 1.5 m s−1, were a heterogeneous population with respect to capsaicin- and bradykinin-sensitivity; only the lowest conduction-velocity fibres were sensitive to these stimulants (Kollarik et al. 2003). On the basis of the tight negative correlation between capsaicin sensitivity and conduction velocity (Ho et al. 2001), it is very likely that the capsaicin-sensitive vagal pulmonary neurones studied here are vagal pulmonary C-fibres, although more definitive evidence in addition to the capsaicin sensitivity, such as neurofilament-negative immunoreactivity, would be more conclusive information about their properties.
The main advantage in using a perforated patch technique is that perturbations to intracellular milieu are kept to a minimum and diffusible transduction components such as cAMP, ATP and GTP are not dialysed out of the cell over time (Horn & Marty, 1988). In particular, some of these experiments lasted for a relatively long duration (> 60 min). Gramicidin, a polypeptide antibiotic, is thought to insert ionophores into the patch membrane that are permeable to monovalent cations such as Na+ and K+, but not to anions or divalent cations, such as Ca2+ (Korn & Horn, 1989; Akaike & Harata, 1994; Kyrozis & Reichling, 1995; Akaike, 1996). The disadvantage of using perforated patch is, of course, that series resistance, in general, is greater. Whereas traditional whole-cell recordings are performed with a series resistance ranging from 2 to 10 MΩ, our values ranged from 6 to 15 MΩ, before series compensation. The slightly elevated series resistance, with 80% series resistance compensation, resulted in voltage errors ranging from 7.6 to 11.9 mV with a mean TTX-R INa of 2.4 nA. However, series resistance remained stable throughout the course of the recordings. Additionally, we could not rule out the possibility that the change in the slope of activation curve, k, following PGE2 exposure may have been due to a loss of voltage control.
Our findings using perforated-patch recordings are, in general, consistent with data from other investigators examining PGE2 modulation of TTX-R INa in non-pulmonary sensory neurones using a whole-cell patch clamp. For example, England and colleagues (England et al. 1996) reported that in neonatal rat DRG neurones, PGE2 caused a hyperpolarizing shift in the steady-state inactivation curve as well as a ∼6 mV hyperpolarizing shift in activation potential, which is similar to the ∼7 mV hyperpolarizing shift we detected in this study. Gold and colleagues (Gold et al. 1996) reported a ∼5 mV hyperpolarizing shift in activation potential in a subset of DRG neurones and a ∼30% increase in voltage sensitivity, which was similar to the ∼34% increase in voltage sensitivity we detected in our pulmonary neurones. Similar PGE2-elicited modulation of TTX-R INa was observed in colonic DRG neurones (Gold et al. 2002).
Currently, the molecular identity of two neuronal TTX-R channels have been identified in sensory neurones: Nav1.8 (SNS, PN3) and Nav1.9 (NaN, SNS2) (Akopian et al. 1996; Sangameswaran et al. 1996; Dib-Hajj et al. 1998; Tate et al. 1998). Both are distributed among nociceptive neurones of the DRG (Tate et al. 1998; Cummins et al. 1999; Amaya et al. 2000; Benn et al. 2001; Fang et al. 2002). Nav1.8 recorded from DRG neurones revealed half-activation potential of −17 mV, midpoint of inactivation of −35 mV, and a slope factor of 4 mV per e-fold change (Saab et al. 2003). Nav1.9 also recorded from DRG neurones displayed a half-activation potential of −32 mV, midpoint of inactivation of −43 mV, and slope factor of 9 mV per e-fold change (Maruyama et al. 2004). Whether these two TTX-R sodium channel isoforms do in fact exist in capsaicin-sensitive vagal pulmonary neurones and whether they are differentially regulated by PGE2 remains to be explored. However, the TTX-R INa recorded in our study under similar ionic conditions suggests that the TTX-R INa species recorded here most closely resembles Nav1.8.
In conclusion, these experiments are consistent with and extend prior experiments in vivo, showing that PGE2 caused hypersensitivity of the airways via its action on pulmonary C-fibres (Ho et al. 2000). One important implication of this study was that PGE2-mediated modulation of TTX-R INa could play an important role in sensitization of capsaicin-sensitive vagal pulmonary neurones. Hence, selective targeting of this channel or its coupling with the PGE2 receptor signalling pathway with therapeutic treatments may be potentially effective for alleviating the airway hypersensitivity associated with mucosal inflammation.
A representative capsaicin-sensitive vagal pulmonary neurone illustrating that total inward INa is comprised of both TTX-S and TTX-R components A, a family of total inward currents was generated from a voltage-clamp protocol in which 40 ms duration voltage steps from −70 to 45 mV in 5 mV increments (Vhold = −80 mV) were applied to a nodose ganglion pulmonary neurone (22.3 pF). Traces shown are the first 10 ms of each tracing of the 40 ms voltage pulse. B, the family of current tracings in the same neurone as A in the presence of 100 nm TTX. C, point-for-point subtraction of the TTX-R INa (B) from the total INa (A) yielded the TTX-S INa. D, current–voltage relationship showing differences in the peak current profiles among the INa species.
Capsaicin-sensitive vagal pulmonary neurones display a high percentage of TTX-R current A–J, the TTX-R INa and total INa were obtained from the peak current of the family of evoked currents generated using the voltage step protocol described in Fig. 1. A, the subset of capsaicin-sensitive vagal pulmonary neurones showed a different percentage TTX-R INa–capacitance profile than that of capsaicin-insensitive vagal pulmonary neurones. B, comparison of percentage TTX-R INa between capsaicin-sensitive, DiI-labelled neurones (n = 28), which were presumed to be those that had innervated pulmonary structures, and capsaicin-insensitive pulmonary neurones (n = 23). Comparison of percentage TTX-R INa in the total populations of capsaicin-sensitive neurones (n = 43) and capsaicin-insensitive neurones (n = 37) in C, pulmonary neurones (n = 79) and non-pulmonary neurones (n = 17) in D, and jugular neurones (n = 38) and nodose neurones (n = 78) in E. F, a plot of the total INa as a function of cell capacitance. Comparisons of the total INa density between capsaicin-sensitive pulmonary neurones (n = 28) and capsaicin-insensitive pulmonary neurones (n = 23) in G, capsaicin-sensitive (n = 43) and capsaicin-insensitive neurones (n = 37) in H, pulmonary (n = 79) and non-pulmonary neurones (n = 17) in I, and jugular (n = 38) and nodose ganglion neurones (n = 78) in J. White circles, capsaicin-insensitive pulmonary neurones; grey circles, capsaicin-sensitive pulmonary neurones. Statistical comparisons were made using unpaired t tests. †P < 0.01; ‡P < 0.001.
PGE2 increases a voltage-dependent TTX-R INa A, the representative experimental record depicts the rate at which the effects of PGE2 are manifested in a capsaicin-sensitive vagal pulmonary neurone (25.6 pF) from the nodose ganglion. The currents were evoked by a single 20 ms voltage step to −5 mV from Vhold (−80 mV; top). Raw tracings show an increase in the amplitude of evoked current during PGE2 perfusion (1 μm; bottom). The arrows indicate current traces recorded before (a) and after PGE2 treatment (b). For clarity, only the first 10 ms of the 20 ms voltage step tracings are shown. B, the plot of the peak TTX-R INa during PGE2 perfusion as a function of time is shown. The peak currents developed from a baseline level before PGE2 (a) to a maximal steady-state level during PGE2 perfusion (b) over the course of ∼5 min. Bar, duration of PGE2 perfusion. C, the experimental record was generated using the voltage-clamp protocol (top) to evoke a family of current tracings recorded before (middle) and after PGE2 treatment (1 μm; bottom) in a capsaicin-sensitive vagal pulmonary neurone (25.6 pF) harvested from the nodose ganglia. The full 40 ms voltage pulse was truncated for clarity. D, current density–voltage relationship showing that PGE2 pretreatment (1 μm; •) increased the TTX-R INa density over that of control (○) in capsaicin-sensitive vagal pulmonary neurones (n = 8).
PGE2 treatment potentiates the maximal conductance, voltage dependence and half-activation potential of the TTX-R INa in capsaicin-sensitive vagal pulmonary neurones A, in capsaicin-sensitive vagal pulmonary neurones (n = 8), mean data show that treatment with PGE2 (1 μm; •) increased the maximal conductance compared with that of before treatment (○). The threshold of current activation was not different between experimental and control conditions. B, the conductance was normalized to the maximal conductance in order to better visualize the difference in voltage sensitivity and half-activation potential. PGE2 treatment (1 μm) appears to increase the voltage dependence and cause a hyperpolarizing shift in the half-activation potential compared with that of control. Each set of data was fitted to a single-term Boltzmann function.
Experimental record showing PGE2 modulation of steady state TTX-R INa in a capsaicin-sensitive vagal pulmonary neurone A, the steady-state inactivation protocol consisted of holding the membrane at various prepulse potentials ranging from −90 mV to −5 mV for 500 ms and then stepping to a single depolarizing test potential of 10 mV for 20 ms (top). Raw traces from a nodose ganglion neurone (25.6 pF) document the changes recorded before (middle) and after PGE2 exposure (1 μm; bottom). B, group data showing that PGE2 increases steady-state TTX-R INa in capsaicin-sensitive vagal pulmonary neurones. Average data in capsaicin-sensitive vagal pulmonary neurones (n = 8) from nodose and jugular ganglia in which the steady-state TTX-R INa was plotted against the prepulse potential (top). The data illustrate that the TTX-R INa was dramatically potentiated after PGE2 treatment (•) compared with the currents evoked before treatment (○). In the expanded view of the steady-state inactivation curve that has been normalized to the maximal conductance (bottom), treatment with PGE2 (1 μm) appears to cause little increase in voltage dependence.
Group data illustrating the modulating effects of forskolin on capsaicin-sensitive vagal pulmonary neurones A, mean data plotting the TTX-R INa density as a function of membrane voltage in capsaicin-sensitive vagal pulmonary neurones (n = 12). The plot compares the TTX-R INa density before (○) and after forskolin treatment (1 μm; •). B, forskolin increases maximal conductance of TTX-R INa in capsaicin-sensitive vagal pulmonary neurones. Group data plotting conductance as a function of membrane voltage before and after treating capsaicin-sensitive vagal pulmonary neurones with forskolin. Each set of data was fitted to a single-term Boltzmann function. C, forskolin increases half-activation potential and voltage dependence of TTX-R INa. Averaged data in the 12 capsaicin-sensitive vagal pulmonary neurones show that treatment with forskolin causes a steeper dependence on voltage compared with before treatment. Further, forskolin treatment shifted the half-activation potential in a hyperpolarized direction. Statistical analysis of A–C is represented in Table 2.
Modulation of TTX-R INa by forskolin or cBiMPS at room temperature (∼21°C) and 37°C Currents were evoked by 20 ms steps to −5 mV from Vhold (−80 mV). A, raw tracings from the same neurone illustrating the modulation of evoked current during control and forskolin treatment (1 μm). B, mean data showing that peak evoked current at room temperature during forskolin treatment (n = 13) was greater than that during 1,9-dideoxyforskolin (10 μm; 1,9-ddf; P < 0.01; n = 6) and vehicle treatment (P < 0.01; n = 9). Statistical comparisons were evaluated using unpaired t tests. †Significantly different from forskolin treatment. C, peak evoked current recorded at 37°C was compared before and after a ∼10 min exposure to forskolin (1 μm) or vehicle. D, group data illustrating that at 37°C, forskolin significantly modulated TTX-R INa (n = 4) compared with that of vehicle treatment (P < 0.05; n = 4). Statistical comparisons were evaluated using unpaired t tests. *Significantly different from forskolin treatment. E, experimental record showing the potentiating effect of cBiMPS treatment on evoked TTX-R INa at 21°C compared with that of vehicle treatment. F, mean data (n = 6) showing that evoked currents were greater during cBiMPS (50 μm) than during vehicle treatment. Statistical comparisons were made using paired t tests. *Significantly different from cBiMPS treatment (P < 0.05).
Lack of modulation of TTX-R INa in capsaicin-insensitive pulmonary neurones Currents evoked by 20 ms steps to −5 mV from Vhold (−80 mV). No difference (P > 0.05) was detected in evoked TTX-R INa during exposure to PGE2 (1 μm; n = 4) compared with vehicle (n = 3) in A, to forskolin (1 μm; n = 5) compared with vehicle (n = 3) in B, and to cBiMPS (50 μm; n = 6) compared with vehicle (n = 6) in C. Statistical comparisons were made using unpaired t tests.
Effect of PGE2 on the biophysical properties of TTX-R INa in capsaicin-sensitive vagal pulmonary neurons
Effect of forskolin on the activation kinetics of TTX-R INa in capsaicin-sensitive vagal pulmonary neurons
Acknowledgements
The authors thank Dr Jonathan Satin for his critical comments and advice in this study. The authors also thank Dr Qihai Gu for his many helpful suggestions. Finally, the authors thank Mr Wenbin Yang and Ms Lifang Zhang for their technical expertise.
This study was supported by National Institutes of Health Grants HL-67379 (L.-Y.L.) and NIH-NIDCD T32-DC-00065 (K.K.).
Footnotes
-
- Accepted February 10, 2005.
- Received November 6, 2004.
- Revision received February 3, 2005.
- The Physiological society 2005
