Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Vanilloid receptors are activated by capsaicin, the pungent principle of chili pepper, resiniferatoxin (RTX) isolated from Euphorbia resinifera, and some other naturally occurring or synthesized compounds (for reviews see Holzer, 1991; Szallasi and Blumberg, 1999; Caterina and Julius, 2001). Since vanilloid receptors are located primarily on mammalian unmyelinated sensory C-fibers in dorsal root ganglia, their activation by tissue injury (nociception) ultimately leads to a conscious sensation of pain. Because the initial excitation of vanilloid receptors by agonists such as capsaicin or RTX is followed by long-lasting desensitization, these compounds are, despite their pungency, therapeutically used for the treatment of chronic pain states (Cruz et al., 1997; Hautkappe et al., 1998; Szallasi and Blumberg, 1999).

The recently cloned vanilloid receptor 1 (VR1; rat: Caterina et al., 1997; human: Hayes et al., 2000) forms a Ca²⁺-permeable, nonselective cation channel that is currently believed to serve as an integrator of various painful stimuli such as capsaicin, RTX, noxious heat, acidic pH (Caterina et al., 1997; Tominaga et al., 1998; Caterina et al., 2000; Caterina and Julius, 2001; Greffrath et al., 2001; Savidge et al., 2001), and the endocannabinoid anandamide (Smart et al., 2000; Olah et al., 2001), particularly in inflamed tissue (Vyklicky et al., 1998). Furthermore, acidic conditions (Tominaga et al., 1998; McLatchie and Bevan, 2001), inflammatory mediators (bradykinin, serotonin, substance P: Vyklicky et al., 1998), cyclooxygenase and lipoxygenase products (prostaglandin E_2, Vyklicky et al., 1998; leukotriene B₄, 12-(S)-hydroperoxyeicosatetraenoic acid, Hwang et al., 2000), phosphorylation by protein kinases A and C (Premkumar and Ahern, 2000), and ATP-mediated activation of metabotropic P2Y-receptors (Tominaga et al., 2001) may act in unison to sensitize vanilloid receptors for noxious stimuli, thence lowering pain threshold and causing hyperalgesia (Davis et al., 2000).

However, only a few vanilloid receptor antagonists are available to date. Among them, capsazepine is a relatively weak competitive antagonist that has nonspecific effects at concentrations often required for antagonist activity, and ruthenium red is a weak and noncompetitive antagonist with a poorly defined mechanism of action (Bevan et al., 1992; Caterina et al., 1997; Docherty et al., 1997; Liu and Simon, 1997; Wardle et al., 1997; for reviews see Szallasi and Blumberg, 1999; Caterina and Julius, 2001). It was not until recently that three novel vanilloid receptor antagonists were described: the nonpungent capsaicin analog SDZ 249-665 (Urban et al., 2000), the highly potent compound iodo-RTX (Wahl et al., 2001), and the arginine-rich hexapeptide RRRRWW-NH₂ (R₄W₂; Planells-Cases et al., 2000). Although the latter compound noncompetitively blocks recombinant VR1 channels expressed in Xenopus oocytes with submicromolar potency, its activity in native cells has yet to be ascertained. Here, we report that in primary cultures of adult rat dorsal root ganglion neurons, the arginine-rich hexapeptide R₄W₂ competitively antagonized the effects of capsaicin and RTX. The hexapeptide consisting entirely of L-amino acids, L-R₄W₂, was more potent than the respective stereoisomer D-R₄W₂.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Chemicals. Collagenase (type A, C-9891), 7S-nerve growth factor (N-0513), ITS-supplement (insulin-transferrin-selenium, I-3146), poly-L-lysin, trypan blue, ionomycin, digitonin, capsaicin, capsazepine, resiniferatoxin, and FITC-labeled Bandeiraea simplicifolia isolectin B₄ were obtained from Sigma Chemie (Deisenhofen, Germany). Dispase and DNase-1 were obtained from Roche Diagnostics (Mannheim, Germany) and Fura-2/acetoxymethyl ester from Molecular Probes (Eugene, OR). The D-and L-enantiomers of the arginine-rich hexapeptide R₄W₂ (sequence RRRRWW-NH₂, purity >99.0% by high-performance liquid chromatography analysis) were gifts from Grünenthal (Aachen, Germany). All other chemicals were obtained at the highest available purity from Sigma or other commercial suppliers. Sterile disposable plastic ware for cell culture was by Falcon (BD Biosciences, Heidelberg, Germany), Nunc (Wiesbaden, Germany), or Nalgene (Nalge, Brussels, Belgium). Media, supplements, and trypsin for cell culture were obtained from Invitrogen (Eggenheim, Germany), and fetal bovine serum was obtained from Seromed (Berlin, Germany). Drugs were dissolved in water, methanol, or dimethyl sulfoxide and stored at 20°C. Aliquots of the stock solutions were diluted directly into the bath solution to achieve the final concentration. Solvents did not influence [Ca²⁺]_i measurements.

Isolation and Culture of Rat Dorsal Root Ganglion Cells. The present investigation was conducted in accordance with the principles outlined in the Declaration of Helsinki and the German law governing the Care and Use of Laboratory Animals, and was approved by the representative for animal care and use of the University of Leipzig. Adult rats (age 6-8 weeks) were killed by CO₂ and decapitation. Thoracic and abdominal organs were quickly removed, the spines were chilled at 4°C in Ca²⁺/Mg²⁺-free Hanks' balanced salt solution (HBSS), and thoracic and lumbar dorsal root ganglia were dissected and freed from connective tissue using fine tweezers and scissors under sterile conditions. Dorsal root ganglia collected in chilled HBSS were centrifuged (5 min, 124g) and resuspended in HBSS containing collagenase (0.5 mg/ml), dispase (1 mg/ml), and DNase-1 (1 mg/ml). After 40 to 50 min of incubation at 37°C in a shaking water bath, trypsin (0.6 mg/ml) was added for a further 15 min of incubation. After addition of medium 1 (Dulbecco's minimal essential medium, 35 mM total glucose, 2.5 mM L-glutamine, 15 mM HEPES, 50 µg/ml gentamicin, 5% fetal bovine serum) to deactivate enzymes, the cell suspension was mechanically triturated using fire-polished Pasteur pipettes and passed through a cell strainer (mesh size 70 µm) to remove undigested tissue fragments. After centrifugation of the cell suspension (5 min at 194g), the pellet was resuspended in medium 1 supplemented with 30 ng/ml nerve growth factor, 10 µg/ml insulin, 5.5 µg/ml transferrin, and 5 ng/ml selenium. DRG cells were plated at a density of 2 × 10⁴ cells/ml onto poly(L-lysin)-coated (25 µg/ml) glass coverslips. Cultures were maintained for 2 to 4 days in a humidified atmosphere (37°C, 5% CO₂) before experimentation.

Intracellular Calcium Measurements. DRG cell cultures from days 2 to 4 were loaded for 50 to 60 min at 37°C in the dark with the cell permeant acetoxymethyl ester of the fluorescent Ca²⁺ indicator Fura-2 (1 µM). Simultaneously, cells were exposed to the FITC-labeled Bandeiraea simplicifolia isolectin IB₄ (0.1-0.5 µg/ml). To remove excess extracellular Fura-2 and IB₄, glass coverslips were washed several times with a modified Tyrode's solution (40.0 mM NaCl, 4.5 mM KCl, 2.0 mM CaCl₂, 1.0 mM MgCl₂, 10.0 mM HEPES, 10.0 mM glucose; pH adjusted to 7.4 with NaOH) and were allowed to rest for 30 min at room temperature protected from light. Thereafter, Ca²⁺ imaging experiments were performed at room temperature in Tyrode's solution using an inverted microscope (IX-70; Olympus, Hamburg, Germany) equipped for epifluorescence and a Peltier-cooled charge-coupled device camera (IMAGO; Till Photonic Inc., Martinsried, Germany). Intracellular Fura-2 was alternately excited at 340 nm and at 380 nm, and the emitted light was measured at a wavelength of 510 nm. The TILL visION software (release 3.3; Till Photonic Inc.) was used for data acquisition, system control, and, later, off-line data analysis.

View larger version (16K): [in this window] [in a new window]   Fig. 1.   Effect of vanilloid receptor agonist capsaicin on [Ca2+]i in adult rat dorsal root ganglion neurons in primary culture. Time course of [Ca2+]i was expressed as Fura-2 fluorescence ratio in a representative experiment. The neuron was depolarized by means of KCl (50 mM) and was exposed to capsaicin in concentrations of 0.1, 0.3, 1, 3, 10, and 30 µM; the exposure time was 2 s each and is denoted by the dashes below the [Ca2+]i trace. Rmax and Rmin were determined at the end of the experiment as defined under Materials and Methods.

Statistical Analysis. Results were expressed as mean values ± S.E.M. of n experiments if not indicated otherwise. Concentration-response curves were fitted to individual data sets using the following standard four-parameter logistic equation
with minimal and maximal effect (min, max), half-maximal excitatory drug concentration (EC₅₀), drug concentration [D], and Hill slope n_H. The minimal effect was treated as a constant (= 0) throughout, whereas the other parameters were variables. Statistical differences were analyzed by means of ANOVA followed by Dunnett's test versus the respective control group and considered significant at p < 0.05. If a nonparametric test was required, ANOVA on ranks followed by Dunn's test versus the control group was applied.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

In Fig. 1, the time course of [Ca²⁺]_i, expressed as Fura-2 fluorescence ratio, is shown in a representative adult rat dorsal root ganglion neuron in primary culture. Elevating extracellular KCl concentration from 4.5 mM to 50 mM shifts the K⁺ equilibrium potential from approximately 90 mV to about 29 mV (Nernst equation, [K⁺]_i = 155 mM, room temperature). The ensuing depolarization of the neuron is associated with opening of voltage-dependent Ca²⁺ channels; the resulting [Ca²⁺]_i transient has a rapid onset, and elevated [Ca²⁺]_i returns to baseline level within seconds, demonstrating neuronal excitability and viability (Fig. 1). On average, KCl-induced depolarization increased the fluorescence ratio in neurons (as defined by their ellipsoid-ovoid-spherical shape) by 0.359 ± 0.013 over baseline levels of 0.862 ± 0.008 (n = 397, p < 0.05), whereas in non-neuronal cells (as defined by their polygonal-flat-spindle-like morphology), KCl-induced depolarization had a significantly smaller effect (ratio increase by 0.081 ± 0.016, basal ratio 0.750 ± 0.019, n = 60, p < 0.05). These results are consistent with the presence of voltage-gated Ca²⁺ channels in neurons and their near absence in the accompanying non-neuronal cells.

The experiment continued by exposing the neuron to increasing concentrations of capsaicin (Fig. 1). Although rapid in onset, the VR-induced increase in [Ca²⁺]_i was sustained whether or not the agonists capsaicin (Fig. 1) or RTX (not shown) continued to be present, thus allowing the establishment of cumulative agonist concentration-response curves. The slow decline of agonist-induced elevated [Ca²⁺]_i levels toward baseline values is illustrated in the right part of Fig. 1. Even after prolonged washout (10 min) of either agonist, [Ca²⁺]_i signals remained elevated at approximately 50% of the maximum agonist effect as described before (Cholewinski et al., 1993). Both VR agonists, capsaicin and RTX, produced concentration-dependent increases in [Ca²⁺]_i at micromolar and nanomolar concentrations, respectively (Fig. 1, representative experiment; Fig. 2, average). Maximal increases in fluorescence ratio above baseline values were similar for capsaicin (0.511 ± 0.074, n = 145) and for RTX (0.476 ± 0.080, n = 77; not significant). Half-maximal effects were reached with 0.25 µM capsaicin and with 0.091 nM RTX (compare Table 1 for complete list of parameters of concentration-response curves); these values are in line with published data (Jerman et al., 2000; McIntyre et al., 2001; for reviews see Szallasi and Blumberg, 1999; Caterina and Julius, 2001).

View larger version (27K): [in this window] [in a new window]   Fig. 2.   Effect of vanilloid receptor agonists and antagonists on [Ca2+]i in adult rat dorsal root ganglion neurons in primary culture. Concentration-response curves of resiniferatoxin (RTX)- and capsaicin (Caps)-induced increases in [Ca2+]i under control conditions (RTX, ; Caps, ) and in the presence of 0.1 µM arginine-rich hexapeptide L-R4W2 () and D-R4W2 (), respectively. Values are mean ± S.E.M., and number of cells is given in parentheses. Curves were generated using the averages of parameters resulting from individual curve fittings. Mean values ± S.E.M. for the curve parameters maximal effect, log EC50, and Hill slope are listed in Table 1.

Concentration-response curves obtained with either VR agonist, capsaicin or RTX, were shifted to the right in the presence of the well characterized VR antagonist capsazepine (Table 1). The magnitude of this shift was significantly dependent on the concentration of capsazepine, whereas maximum effects of capsaicin and the Hill slopes of the respective concentration-response curves were not affected (Table 1), indicating a competitive antagonism of capsazepine at vanilloid receptors. Both RTX and capsaicin concentration-response curves were also shifted to the right by the arginine-rich hexapeptides L-R₄W₂ and D-R₄W₂ in a concentration-dependent manner (Table 1). The antagonist effect of L-R₄W₂ was already significant at 0.1 µM, the lowest concentration tested, whereas the same concentration of D-R₄W₂ did not yet shift agonist concentration-response curves (Table 1; Fig. 2). This effect suggests a stereoselectivity of VR block by the hexapeptides, the L-enantiomer being a more potent antagonist than D-R₄W₂. Since neither the maximum agonist effects nor the Hill slopes of concentration-response curves were affected by either hexapeptide enantiomer, the mechanism of antagonist action of L-R₄W₂ and D-R₄W₂ may be a competitive one, too.

To characterize the mechanism of antagonist action of the hexapeptide enantiomers in more detail, a Schild analysis was performed, the results of which are depicted in Fig. 3. Because only one concentration-response curve could be established in each set of neurons due to the incomplete reversibility of agonist effects within reasonable times, concentration ratios for the Schild analysis were estimated based on the mean EC₅₀ values for the individual experimental groups listed in Table 1. The concentration ratio data so obtained were subject to linear regression analysis with the exception of D-R₄W₂ and RTX, where a linear relation between log(CR  1) and antagonist concentration was apparently not obvious (Fig. 3A). Linear regression analysis of the data obtained with L-R₄W₂ and RTX or capsaicin yielded similar results. The apparent dissociation constants of L-R₄W₂ were 4.0 nM and 3.7 nM with RTX and capsaicin, respectively, as agonists. However, the regression lines were shallow, possessing slopes of 0.38 ± 0.01 (RTX) and 0.42 ± 0.04 (capsaicin) that were significantly smaller than unity (p < 0.05). With capsaicin as agonist, regression analysis yielded apparent antagonist dissociation constants of 153 nM for D-R₄W₂ and 4.1 µM for capsazepine, and slopes of 0.67 ± 0.06 (D-R₄W₂) and 1.19 ± 0.28 (capsazepine), which were significantly smaller than and indistinguishable from unity, respectively. Although these results support the notion of a stereoselectivity of block of vanilloid receptors by the hexapeptide enantiomers, with L-R₄W₂ being more potent than D-R₄W₂, they also suggest that the interaction between arginine-rich hexapeptides and the vanilloid receptor may not be as stoichiometric as that between capsazepine and VR.

View larger version (16K): [in this window] [in a new window]   Fig. 3.   Schild plots for L-R4W2 (), D-R4W2 (), and capsazepine () with resiniferatoxin (A) and capsaicin (B) as agonists. Note that mean values and ranges (min-max) are depicted. Apparent dissociation constants and respective slopes were estimated by means of linear regression analysis of the data.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

We have demonstrated for the first time in native rat dorsal root ganglion neurons, which are the most important peripheral integration site for noxious stimuli, that the positively charged arginine-rich hexapeptide R₄W₂ antagonizes vanilloid receptor-mediated effects in a stereoselective manner, with the peptide consisting of L-amino acids, L-R₄W₂, being more potent than D-R₄W₂. Both stereoisomers antagonized capsaicin- and RTX-induced [Ca²⁺]_i increases in an apparently competitive manner; however, the significant deviation from unity of Hill coefficients of concentration-response curves and of Schild plots is incompatible with the presence of a single binding site.

Thus, the present investigation confirms in native dorsal root ganglion neurons the potent (submicromolar) vanilloid receptor antagonist activity of the arginine-rich hexapeptide R₄W₂ that has been described recently in a Xenopus expression system (Planells-Cases et al., 2000), although the mechanism of antagonist action and its putative stereoselectivity are controversial. R₄W₂ blocked capsaicin-induced currents mediated by heterologously expressed VR1 channels in an apparently noncompetitive manner because of virtually identical capsaicin EC₅₀ concentrations in the absence and in the presence of the hexapeptide (Planells-Cases et al., 2000). In contrast, our results rather suggest the presence of a competitive antagonism since concentration-response curves for RTX and capsaicin were simply shifted along the concentration axis in the presence of R₄W₂ without maximum response or slope being affected (Fig. 2; Table 1).

Notwithstanding the possibility of more or less subtle differences in the properties of recombinant vanilloid receptors in a nonmammalian expression system versus native vanilloid receptor in DRG neurons in terms of VR1 homo-/hetero-oligomers, glycosylation state, or interacting modulatory proteins (Planells-Cases et al., 2000; Jahnel et al., 2001; Kedei et al., 2001), the present shift to the right of VR agonist concentration-response curves without a change in the maximum effect could not only be attributed to competitive antagonism (Clements and Westbrook, 1994), but might also reflect an allosteric interaction (Li et al., 1997, 1998). If R₄W₂ acted competitively at the capsaicin or RTX binding site, increasing hexapeptide concentrations would be expected to shift the agonist concentration-response curve to the right in a progressive and (almost) indefinite manner. Such a progressive shift was indeed observed with R₄W₂ concentrations over two orders of magnitude before limited drug supplies and elevated solvent concentrations precluded further experiments. If R₄W₂ acts at an allosteric site, the hexapeptide should cease to shift agonist concentration-response curves when its allosteric site(s) of action is(are) saturated; however, the corresponding curvilinear Schild plot reaching a limiting value at high concentrations of the antagonist R₄W₂ was not observed. Another method to discriminate between competitive and allosteric interaction relies on measuring activation and inactivation rates of vanilloid receptor channels, where a competitive antagonist will decrease the activation rate without changing the deactivation rate (Clements and Westbrook, 1994), whereas an increased deactivation rate and an unaffected activation rate are characteristic of an allosteric interaction (Li et al., 1997, 1998). However, this approach was unsuccessful since experimental conditions without desensitization of capsaicin-activated currents could not be satisfactorily established (data not shown).

Another putative explanation for the discrepancy between the results of Planells-Cases et al. (2000) and those of the present paper may become possible when considering the particular properties of cells (Xenopus expression system versus native rat dorsal root ganglion neurons) and signal detection methods used (voltage-clamp versus Ca²⁺ imaging). In the Xenopus expression system, the major, if not only, inward current (directly measured by means of the voltage-clamp technique) is mediated by recombinant vanilloid receptors, the block of which attenuates the maximal inward current flow, thus resembling a noncompetitive antagonism. Dorsal root ganglion neurons, on the other hand, possess a variety of ligand- and voltage-gated cation channels, and by means of Ca²⁺ imaging, the contribution from all participating Ca²⁺ sources (active Ca²⁺ conduits) is integrated. Na⁺ and Ca²⁺ influx via capsaicin- or RTX-activated vanilloid receptors is assumed to depolarize the neuron, thus serving as the trigger for opening of voltage-dependent Ca²⁺ channels. This is consistent with observations that VR-mediated [Ca²⁺]_i signals were depressed by removal of extracellular Na⁺ or by dihydropyridine block of voltage-dependent Ca²⁺ channels (data not shown; Greffrath et al., 2001). Since a very small inward current of only a few picoamperes may suffice as trigger for a large depolarization and the ensuing full-size Ca²⁺ influx, the maximal signal detected by Ca²⁺ imaging may be preserved even when the trigger, i.e., the cation current via vanilloid receptor channels, is markedly reduced, thus resembling a competitive antagonism.

In terms of a putative stereoselective mechanism of action, it has been reported that equal concentrations of L-R₄W₂ and D-R₄W₂, blocked heterologously expressed VR1 channels with similar efficacy, suggesting lack of stereoselectivity of block, whereas only the D-hexapeptide significantly decreased capsaicin-induced ocular irritation in mice, seemingly because it is proteolysis-resistant (Planells-Cases et al., 2000). In contrast, the present results clearly support the conclusion of a stereoselective antagonism with L-R₄W₂ being more potent than D-R₄W₂ (Fig. 2; Table 1). Membrane-bound extracellularly oriented peptidases acting on a broad range of substrates occur throughout the nervous system in and ex vivo (Konkoy and Davis, 1996) and also in mixed neuronal/glial cell cultures from dorsal root ganglia (Berger et al., 1995). However, an ostensible stereoselectivity due to preferential proteolytic degradation of one of the hexapeptide enantiomers is discounted, because the "physiological" L-R₄W₂ is expected to be more susceptible to proteolysis than D-R₄W₂, whereas the latter was clearly less potent than the former.

Furthermore, Hill coefficients around unity are consistent with the occurrence of a single binding site in heterologously expressed vanilloid receptors (Planells-Cases et al., 2000), whereas in rat dorsal root ganglion neurons the results from Schild analysis are incompatible with a single binding site, suggesting, instead, the presence of several binding sites (Fig. 3). One of these several putative binding sites is most likely located on the pore loop of the vanilloid receptor, because the amino acid sequence of the pore loop contains four negatively charged residues (E636, D646, E648, and E651) that may constitute a binding site for positively charged molecules. Although this very binding site apparently does not discriminate between L-R₄W₂ and D-R₄W₂ (Planells-Cases et al., 2000), other arginine-rich hexapeptide binding sites do so, e.g., N-methyl-D-aspartate (NMDA) receptors where D-R₄W₂ has been reported to be twice as potent as L-R₄W₂ (Ferrer-Montiel et al., 1998). Thus, one might speculate that the large number of NMDA receptors present in rat dorsal root ganglion neurons preferentially bind D-R₄W₂, thus acting as a sink for the D-hexapeptide, thereby resulting in the observed apparent stereoselective antagonism at vanilloid receptors.

In conclusion, both synthetic (e.g., R₄W₂) and naturally occurring (e.g., dynorphin A) arginine-rich hexapeptides have been demonstrated to possess antagonist activity at NMDA and vanilloid receptors (Ferrer-Montiel et al., 1998; Planells-Cases et al., 2000; present paper), which constitute important sites within the pain pathway for integration of noxious stimuli. The present data are consistent with R₄W₂ acting as a stereoselective and competitive antagonist at native vanilloid receptors in dorsal root ganglion neurons, although an allosteric interaction cannot be totally excluded. Future work will show whether it will be possible to develop a low molecular weight molecule that mimics the antagonist action of arginine-rich hexapeptides at NMDA and vanilloid receptors and may prove a useful analgesic for the treatment of chronic pain.