Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

A defining feature of many nociceptive sensory fibers is their sensitivity to capsaicin (Maggi and Meli, 1988; Julius and Basbaum, 2001). Capsaicin and related compounds such as resiniferatoxin have therefore been used to great advantage as pharmacological tools with which to probe sensations and reflexes associated with nociceptive nerve activation. A major breakthrough in capsaicin pharmacology was the molecular characterization of the capsaicin receptor (Caterina et al., 1997). This receptor, referred to as vanilloid receptor 1 (VR1), is an ionotropic receptor. When capsaicin binds to the receptor the ion channel opens and cations cross the membrane, resulting in membrane depolarization and calcium influx. The membrane depolarization can lead to action potential discharge (afferent function), and the calcium flux can lead to local release of neurokinins and other transmitters stored in vesicles within the peripheral nerve terminals (efferent function) (Maggi and Meli, 1988).

In recent years, several endogenous agonists of VR1 have been identified, including anandamide, hydrogen ions, and certain lipoxygenase products of arachidonic acid (Tominaga et al., 1998; Zygmunt et al., 1999; Hwang et al., 2000). In addition, it has been shown that heat can gate the VR1 ion channel (Caterina et al., 1997). The presence of endogenous agonists supports the hypothesis that VR1 may not only be involved in the pharmacological activation of nociceptors but also play a role in the physiological activation of these fibers. To ascertain the role of VR1 in various physiological processes requires either the use of animals in which the VR1 gene has been deleted (Caterina et al., 2000), selective desensitization strategies, or the use of specific inhibitors of VR1.

Until recently, pharmacological inhibitors of VR1 have been limited to capsazepine and ruthenium red. Capsazepine is a competitive antagonist of vanilloid binding, whereas ruthenium red noncompetitively acts as a VR1 channel blocker (Amann and Maggi, 1991; Dickenson and Dray, 1991). A drawback of these two inhibitors is their nonselectivity in action. Both capsazepine and ruthenium red at concentrations only slightly greater than those needed to inhibit capsaicin responses are known to inhibit a variety of ion channels (Docherty et al., 1997; Liu and Simon, 1997). The recent discovery of a novel and very potent inhibitor of VR1 was therefore welcomed by those interested in VR1 pharmacology. The potent antagonist was discovered when it was noted that an iodinated form of the VR1 agonist resiniferatoxin competed with vanilloids for VR1 binding, but had no intrinsic efficacy (Wahl et al., 2001). Binding studies carried out in rat spinal cord membrane preparations showed that iodo-resiniferatoxin (I-RTX) was a competitive antagonist of vanilloids at VR1 with a K_i value of about 5 nM. In this preparation, I-RTX had an affinity for VR1 approximately 8000 times greater than capsazepine (Wahl et al., 2001).

Functional evidence that I-RTX acts as an antagonist for VR1 was limited to whole-cell patch-clamp studies of membrane currents in oocytes transfected to express VR1. Again, the data were impressive with a concentration of I-RTX as low as 3 nM effectively blocking the inward current evoked by 100 µM capsaicin. In this functional assay, in contrast to the binding studies, the antagonism by capsaicin was insurmountable, arguing against a purely competitive antagonist (Wahl et al., 2001).

The pharmacology of VR1 expressed in oocytes, however, may be different than that expressed in nociceptive afferents within tissues. The types and extent of glycosylation of VR1 as well as the manner with which the VR1 peptides form a multimeric ion channel may be different in different expression systems (Kedei et al., 2001). Inasmuch as I-RTX will likely become an invaluable research tool for the study of VR1 biology, experiments were designed with the purpose of more fully characterizing the pharmacology of I-RTX in inhibiting vanilloid activation of C-fiber afferent nerve endings in the guinea pig isolated airway preparation. The guinea pig isolated airway preparation is an ideal model to study both the efferent and afferent activity of vagal C-fibers. The guinea pig airway receives a relatively dense neurokinin-containing C-fiber innervation, and neurokinins are potent and effective contractile agonists of guinea pig airway smooth muscle (Ellis and Undem, 1994). Therefore, airway smooth muscle contraction in response to VR1 activation is a convenient assay for neurokinin secretion from airway sensory nerves. Action potential discharge in identified single vagal C-fibers can also be readily quantified in this model using standard electrophysiological approaches (Fox et al., 1995; Riccio et al., 1996a).

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Efferent Function of C-Fibers. Male Hartley guinea pigs (200-400 g) were killed by asphyxiation with CO₂ and exsanguinated. The trachea was removed and divided into consecutive rings, each roughly two to three cartilage rings in width. Tracheal rings were placed in tissue baths and tied with silk surgical suture to force displacement transducers (FT03C; Grass Instruments, Quincy, MA) for recording of isometric tension on a polygraph (Grass Instruments). Resting tension was set at 1 g. Tissue baths contained 10 ml of Krebs bicarbonate solution (KBS; 118 mM NaCl, 5.4 mM KCl, 1.0 mM NaH₂PO₄, 1.2 mM MgSO₄, 1.9 mM CaCl₂, 25.0 mM NaHCO₃, and 11.1 mM dextrose), which was maintained at 37°C and gassed with 95% O₂, 5% CO₂ and replaced every 15 min during a 60-min equilibration period. Indomethacin (3 µM) was added to KBS to suppress prostanoid release from the tissue. Thiorphan (1 µM) was added to the tissue bath 15 min before the completion of equilibration period to suppress the neutral endopeptidase activity (Ellis and Undem, 1990). This method has been described elsewhere (Carr et al., 2000). The experiments were approved by the Johns Hopkins Animal Care and Use Committee.

Afferent Function Studies. Male Hartley guinea pigs (200-400 g) were killed as described above. The airways with intact right-side vagal innervation were removed and placed in a dissecting dish containing KBS. Connective tissue was trimmed away, leaving the trachea, larynx, and right bronchus with intact nerves (vagus, superior laryngeal, and recurrent) and nodose and jugular vagal ganglia. A longitudinal cut was made along the ventral surface to open the larynx, trachea, and bronchus. Airways were then pinned to a Sylgard-lined Perspex chamber. The right nodose and jugular ganglia, along with the rostral-most vagus and superior laryngeal nerves, were pulled through a small hole into an adjacent compartment of the same chamber for extracellular recording. Both compartments were separately superfused with the KBS containing 3 µM indomethacin. The KBS was gassed with 95% O₂, 5% CO₂, the temperature was maintained at 37°C, and the flow rate was 6 to 8 ml min¹. This method has been described previously (Riccio et al., 1996b). The experiments were approved by the Johns Hopkins Animal Care and Use Committee.

Drugs. BaCl₂, capsaicin, citric acid, indomethacin, resiniferatoxin, and thiorphan were obtained from Sigma-Aldrich (St. Louis, MO). Bovine trypsin was obtained from Worthington Biochemicals (Freehold, NJ). Iodo-resiniferatoxin was purchased from Tocris Cookson (Ellisville, MO). Neurokinin A was obtained from Cambridge Biochemical (Wilmington, DE). SR140333 and SR48968 were used to block NK1 and NK2 receptors, respectively (kindly provided by Zeneca Pharmaceuticals, Wilmington, DE). BaCl₂, NKA, trypsin, and citric acid were dissolved in distilled water. Capsaicin, indomethacin, and thiorphan were dissolved in absolute ethanol. Resiniferatoxin and iodo-resiniferatoxin were dissolved in dimethyl sulfoxide. Subsequent dilution of all drugs except citric acid was in KBS. Citric acid was diluted in saline.

Data Analysis. Contraction in guinea pig isolated tracheal and bronchial preparations was expressed as a percentage of maximal contraction induced by 30 mM BaCl₂. Apparent dissociation constants (K_B) were calculated separately for each agonist and each concentration of I-RTX using the standard equation of (antagonist)/(dose ratio  1), converted to the negative logarithm, and expressed as pK_B. Schild plot was constructed from the series of curves using Microsoft Excel, and pA₂ and slope of the Schild plot were obtained from the line of best fit. The efferent function data were expressed as arithmetic mean ± S.E.M. and compared using Student's nonpaired t test.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Efferent Function Studies. As we and other have noted previously, capsaicin and resiniferatoxin stimulate the release of neurokinins from sensory nerves in the isolated guinea pig airway preparation, leading to smooth muscle contractions that can be abolished by blocking NK1 and NK2 receptors (Belvisi et al., 1992; Ellis and Undem, 1994). This is therefore a useful assay for the quantitative study of the efferent function of airway C-fibers. In our initial series of experiments we assessed I-RTX for any nonspecific effects on the guinea pig airway smooth muscle response to exogenously applied NKA. I-RTX (1 µM) had no effect on NKA-induced contractions of guinea pig airway smooth muscle. In five experiments the NKA concentration-response curves for tracheal contractions in the absence and presence of I-RTX were superimposable (data not shown). The log (M) ED₅₀ for NKA was 11.1 ± 0.3 and 11.3 ± 0.5 in the presence and the absence of I-RTX, respectively (P > 0.1). However, at a concentration of 10 µM, I-RTX caused a significant rightward shift in the NKA concentration-response curve that averaged 4 ± 0.2-fold (n = 5, P < 0.01; data not shown), indicating a nonspecific inhibitory effect of this large concentrations of I-RTX on airway smooth muscle contraction. Therefore, in all studies the maximum concentration of I-RTX evaluated was 1 µM. I-RTX alone, even at 10 µM, had no direct effect on airway smooth muscle tone (i.e., it revealed no partial agonist activity at VR1 in this assay).

View larger version (18K): [in this window] [in a new window]   Fig. 1.   Concentration-response curves for resiniferatoxin-induced tachykinergic contractions of guinea pig tracheal rings in the absence and presence of different concentrations of I-RTX. Each point represents the mean ± S.E.M. of eight experiments. Inset, Schild plot of the effects of I-RTX on resiniferatoxin response, pA2 = 7.8 obtained from the line of best fit.

View larger version (19K): [in this window] [in a new window]   Fig. 2.   Concentration-response curves for capsaicin-induced tachykinergic contractions of guinea pig tracheal rings in the absence and presence of different concentrations of I-RTX. Each point represents the mean ± S.E.M. of eight experiments. The numbers in parentheses indicate the time to onset of contraction to the first effective concentration of capsaicin. The onset of contraction was significantly increased by each concentration of I-RTX compared with vehicle (P < 0.05). There was no significant difference in the time to onset of the capsaicin response in tissues treated with 0.3 µM versus 1 µM I-RTX. Inset, traces of the smooth muscle tension recordings showing an example of the delay in the onset of the response to capsaicin in the presence of I-RTX (30 nM). Arrows indicate addition of capsaicin (30 nM) into the tissue bath. Top trace, onset of the contraction of control tissue treated with vehicle was <1 min, the response was 70% of the maximal contraction. Bottom trace, onset of contraction in the presence of I-RTX (30 nM) was ~7 min, the response was 48% of the maximal contraction.

View larger version (17K): [in this window] [in a new window]   Fig. 3.   A, concentration-response curves for anandamide-induced tracheal contractions in the absence and presence of 1 µM I-RTX. Each point represents the mean ± S.E.M. of six experiments. I-RTX significantly (P < 0.05) inhibited the response to 105 and 104 M, but not 3 × 10 4 M, anandamide. B, concentration-response curves for trypsin-induced contraction of guinea pig tracheal rings in the absence and presence of I-RTX (1 µM). Each point represents the mean ± S.E.M. of seven experiments. I-RTX had no significant effect on the trypsin-induced contractions (P > 0.1). In two experiments, SR140333 and SR4896 (1 µM each) rapidly and completely reversed the trypsin-induced contraction when added after the 100 µg/ml concentration.

Afferent Function Studies. In 10 of 10 experiments, a 500-µl aliquot of capsaicin (0.3 µM) applied to the superfusion solution over the receptive field of C-fibers innervating the isolated trachea/bronchus preparation caused a robust action potential discharge (Fig. 4A, left; Table 1). Capsaicin desensitizes VR1 so the effect of I-RTX was studied in separate C-fibers. I-RTX (1 µM) did not evoke action potential discharge in any fiber studied. When the airway was superfused with buffer containing 30 nM I-RTX for 30 min, a 500-µl aliquot of 0.3 µM capsaicin evoked only about seven action potentials (Fig. 4A, right; Table 1). This potent inhibitory effect of I-RTX was likely due to the delay in onset of the capsaicin response as described in the efferent studies above. When 0.3 µM capsaicin was continuously applied (superfused over the receptive field) for a more extended period (>10-min) so that it had time to equilibrate with VR1, it was capable of surmounting the inhibition of I-RTX. An example of this is shown in Fig. 4B.

View larger version (42K): [in this window] [in a new window]   Fig. 4.   Representative electrophysiological recordings showing activation of airway C-fibers by capsaicin in the absence and presence of I-RTX. Each trace was obtained from a different fiber because of the desensitizing effect of capsaicin. In all traces the horizontal line under the recording denotes the duration of capsaicin administration. A, left, transient exposure to 0.3 µM capsaicin induced robust activation of an airway C-fiber, which was virtually abolished by I-RTX (A, right; see Table 1 for mean data). B, when capsaicin (0.3 µM) was superfused over the receptive field for 15 min, I-RTX (30 nM) delayed the onset but did not abolished the response. In the recording (B), a low-amplitude multiunit activity could be noted, but only the large-amplitude fiber (amplitude extends the range of the recorder) had a mechanically sensitive receptive field in the airway wall. *, P < 0.05.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

I-RTX has recently been reported to reversibly bind to VR1 in rat spinal cord membrane preparations or human embryonic kidney 293 cells stably expressing human VR1 with an affinity constant (K_i of ~1-5 nM) more than 1000-fold lower than that of capsazepine (K_i of ~8 µM) (McDonnell et al., 2002; Wahl et al., 2001). These observations, considered with the finding that I-RTX had no intrinsic efficacy, predicted that I-RTX would be a potent competitive antagonist at VR1. Functional studies in which the I-RTX was used to inhibit capsaicin-induced inward current in VR1-expressing oocytes revealed that I-RTX is an extremely potent antagonist in this system. Concentrations of I-RTX as low as 10 nM nearly abolished capsaicin-induced inward current in the voltage-clamped oocytes (Wahl et al., 2001). The results presented herein are in agreement with the conclusion that I-RTX is a selective VR1 antagonist. As was observed in the oocyte preparation, we found no evidence of partial agonist activity with I-RTX in either the efferent or afferent assays of airway C-fiber function. In studies on airway C-fibers, however, the potency of I-RTX at inhibiting vanilloid-induced responses is significantly lower than that observed in VR1-expressing oocytes or VR1-expressing human embryonic kidney 293 cells. In the isolated airway preparation, I-RTX had a K_B value of about 0.1 µM that in paired experiments was only about 30-fold more potent than capsazepine.

In binding studies, I-RTX behaved as a competitive antagonist (Wahl et al., 2001; McDonnell et al., 2002). Consistent with competitive antagonism, we noted that the I-RTX shifted the vanilloid agonist concentration-response curves to the right but did not inhibit the maximum response. Thus, the antagonism by even large concentrations of I-RTX (1 µM) was completely surmountable by increasing the agonist concentration in both the afferent and efferent assays. This is in contrast to the functional studies in VR1-expressing oocytes. In the oocyte system, the characteristics of inhibition were consistent with a noncompetitive mode of action. For example, 3 nM I-RTX had apparently no effect on the EC₅₀ for capsaicin, but inhibited the maximum response by more than 50% (Wahl et al., 2001). The explanation for this discrepancy between the effect of I-RTX on C-fiber terminals in guinea pig airway tissue and that observed in transfected cells is not clear but may be related to the kinetics of I-RTX. We found that in addition to causing a parallel rightward shift in the capsaicin concentration-response curves, I-RTX caused an obvious and profound increase in the latency of the agonist response. This was observed when the effects of capsaicin on either the afferent function (action potential discharge) or efferent function (contractions) were studied. The mechanism underlying this delay is not known but may involve the relatively slow rate at which I-RTX dissociates from the receptor (Wahl et al., 2001). Regardless of the mechanism, this effect of I-RTX needs to be considered when interpreting concentration-response data. For example, if capsaicin is not allowed time to equilibrate with the receptor (i.e., when the agonist is given in a transient manner, as is often the case in electrophysiological experiments), this effect of I-RTX will lead to an apparent noncompetitive antagonism, as well as an overestimation of the antagonist affinity. This was observed in the action potential discharge study, when it was noted that when capsaicin was applied over a 3-s period, a concentration of I-RTX as low as 30 nM was capable of essentially abolishing the response of a maximally effective concentration of capsaicin. When the same concentration of capsaicin was allowed to superfuse over the C-fiber receptive field for >10 min, the antagonism afforded by I-RTX was less apparent.

A more critical analysis of the vanilloid agonist concentration-response curves in the absence and presence of different concentrations of I-RTX would seem to indicate that something more than simple competitive antagonism at a single binding site is occurring between I-RTX and the agonists. The slope of the Schild plot was significantly less than unity when resiniferatoxin was used as the agonist. That something other than simple competitive antagonism was occurring was even more evident when capsaicin was used as the agonist. In fact, the capsaicin concentration-response curves essentially became stationary in the presence of 0.3 and 1 µM I-RTX. These results are consistent with the speculation that capsaicin may be acting on at least two receptors (or two isoforms of the same receptor), one with high affinity the other with a lower affinity for capsaicin, to evoke tachykinin release from the C-fibers. When the concentration-response curve is shifted sufficiently to the right (e.g., by 0.3 µm I-RTX) capsaicin may act on the lower affinity form of the receptor to evoke the measured response. In this scenario, I-RTX has an affinity for the high-affinity capsaicin binding site of about 0.1 µM (K_B value estimated with 30 nM I-RTX) but not for the lower affinity binding state. A more reduced experimental design than the isolated tissue is required to directly address this hypothesis.

We have previously shown that trypsin contracts guinea pig airways by evoking the release of tachykinins from C-fibers (Carr et al., 2000). Although trypsin induces the efferent activity in C-fibers, it does not evoke action potential discharge. The mechanism by which trypsin causes tachykinin release is not known but likely involves some type of protease-activated receptors. That I-RTX (1 µM) had no effect on the trypsin-induced tachykinergic responses indicates that trypsin may release tachykinins from C-fibers independently of VR1 (or at least independently of the vanilloid binding sites). These data also serve as a testament to the selectivity of I-RTX as a VR1 antagonist. A lack of effect on the trypsin-induced tachykinergic contractions means that I-RTX (1 µM) did not nonselectively interfere with tachykinin secretion, or airway smooth muscle contraction in response to the endogenously released tachykinins. This is consistent with the lack of effect of I-RTX (1 µM) on NKA concentration-response curve. However, the observation that 10 µm I-RTX had significant inhibitory effects on the NKA-mediated tracheal contractions indicates that at this concentration of I-RTX acts on cellular processes other than VR1.

The antagonism by I-RTX of VR1 activation was not selective for vanilloid agonists. Anandamide is known to activate VR1 in isolated neurons and in VR1 expression systems (Zygmunt et al., 1999). Anandamide stimulates the release of tachykinins from guinea pig airway C-fibers by a mechanism that can be inhibited with capsazepine (Tucker et al., 2001). The rightward shift caused by I-RTX in the anandamide concentration-response curve is consistent with the hypothesis that anandamide activates VR1 by binding to the vanilloid binding site(s).

Acid is another nonvanilloid that can activate VR1. The finding that I-RTX inhibited acid-induced action potential discharge is similar what has been reported with capsazepine (Fox et al., 1995). How this occurs is not clear. Protons have been shown to bind to VR1 and increase the channel's open probability (Tominaga et al., 1998). This, however, is not thought to occur by directly affecting the vanilloid binding site. It may be argued that upon specific binding to the vanilloid binding site, I-RTX (or capsazepine) nonspecifically blocks the channel pore. However, capsazepine is relatively ineffective in inhibiting inward current or calcium entry through VR1 that is induced by heat (Liu and Simon, 2000; Savidge et al., 2001). A possibility remains that acid application leads to production of an endogenous agonist that acts at the vanilloid binding site. In this light it is interesting to note that there are several endogenous molecules that can act as VR1 agonists, including anandamide, and lipoxygenase products of arachidonic acid (Zygmunt et al., 1999; Hwang et al., 2000). The question of whether acid leads to VR1 activation in airway C-fibers by a direct or indirect mechanism cannot be further resolved within the framework of a complex tissue. This issue requires a determination of whether I-RTX can inhibit acid-induced VR1 channel activity in excised membranes, at the single channel level.

In summary, I-RTX is a selective VR1 antagonist that can be used to inhibit VR1-mediated afferent and efferent activity in airway C-fiber terminals. Under conditions in which the antagonists and agonist are allowed time to equilibrate with the receptor, I-RTX provides a moderately potent, surmountable antagonism of VR1 with a K_B value of around 0.1 µM (about 30 time more potent than capsazepine in this system). Concentrations of I-RTX as low as 30 nM are effective at abolishing the response to relatively large concentrations of capsaicin, however, if capsaicin is added to the tissue in a transient manner. This is because I-RTX effectively and potently delays the onset of capsaicin-induced responses.