We showed previously that subcutaneous injection of the injury-associated peptide mediator endothelin-1 (ET-1) into the rat plantar hindpaw produces pain behavior and selective excitation of nociceptors, both through activation of ETA receptors likely on nociceptive terminals. The potential role of ETBreceptor activation in these actions of ET-1-has not been examined. Therefore, in these experiments, we studied the effect of blocking or activating ETB receptors on ET-1-induced hindpaw flinching and excitation of nociceptors in rats. An ETBreceptor-selective antagonist, BQ-788 (3 mm), coinjected with ET-1 (200 μm) reduced the time-to-peak of flinching and significantly enhanced the average maximal flinch frequency (MFF). In contrast, coinjection of an ETB receptor selective agonist, IRL-1620 (100 or 200 μm), with ET-1 reduced the average MFF and the average total number of flinches. Interestingly, this unexpected inhibitory effect of IRL-1620 was prevented by the nonselective opioid receptor antagonist naloxone (2.75 mm). To confirm these inhibitory actions, we studied the effects of IRL-1620 on ET-1-induced spike responses in single, physiologically characterized nociceptive C-fibers. IRL-1620 suppressed spike responses to ET-1 in all (n = 12) C-units, with mean and maximum response frequencies of 0.08 ± 0.02 and 1.5 ± 0.4 impulses/sec versus 0.32 ± 0.07 and 4.17 ± 0.17 impulses/sec for ET-1 alone. In additional support of the behavioral results, coinjection of naloxone (2.75 mm) completely prevented this inhibitory action of IRL-1620. These results establish that ETB receptor activation inhibits ET-1-induced pain behavior and nociception in a naloxone-sensitive manner and point to a previously unrecognized dual modulation of acute nociceptive signaling by ETA and ETB receptors in cutaneous tissues.
Endothelin-1 (ET-1) is a potent, endogenous vasoactive peptide (Hickey at al., 1985; Yanagisawa et al., 1988) whose effects are mediated by two distinct G-protein-coupled receptors, the endothelin-A (ETA) and the endothelin-B (ETB) receptor, which usually mediate vasoconstriction and vasodilatation, respectively (Rubanyi and Polokoff, 1994). Interestingly, ET-1 has a role in pain signaling in animals and humans (Hammerman et al., 1997; Davar et al., 1998; De-Melo et al., 1998; Graido-Gonzalez et al., 1998; Jarvis et al., 2000). For example, in rodents, intraperitoneal administration of ET-1 produces an abdominal writhing response that is ETA receptor mediated and that may be behavioral evidence of acute pain (Raffa et al., 1996a,b). Intra-articular administration of ET-1 also has been shown recently to induce pain in rodents that is ETA receptor mediated, and ET-1 is known to potentiate pain states in several animal models of acute chemical- or inflammation-induced pain (Ferreira et al., 1989; Piovezan et al., 1997, 1998, 2000). In humans, the intra-arterial administration of ET-1 is reported to induce severe pain that is associated with prolonged touch-evoked allodynia in the injected limb (Dahlof et al., 1990), and, more recently, Carducci et al. (1998) have reported that antagonists of the endothelin-A receptor can reduce pain in patients with metastatic prostate cancer. Consistent with a role in cutaneous injury, ET-1 is also oversecreted after skin damage (Ahn et al., 1998) in which it might contribute to both local inflammation (Griswold et al., 1999) and pain.
Consistent with this evidence of ET-1-induced pain in animals and humans, we described recently ETAreceptor-dependent flinching behavior and selective excitation of nociceptors after subcutaneous injection of ET-1 into the rat plantar hindpaw (Gokin et al., 2001). At a cellular level, we demonstrated ETA receptor-mediated excitation of nociceptor-like sensory neurons (ND7 cells) and enhancements of tetrodotoxin-resistant sodium currents in acutely isolated rat dorsal root ganglion neurons (Chen et al., 2000; Zhou et al., 2001, 2002). These results, together with anatomic evidence of ETA (but not ETB) receptors on small-diameter DRG neurons and their axons (Pomonis et al., 2001), further strengthen the potential importance of ET-1 (and ETA receptors) in the pathogenesis of pain.
Whereas a role for ETA receptors in ET-1-induced pain has been established, the importance of ETBreceptors for this pain is not known. Although antihyperalgesic effects of ETB receptor activation have been described in some models of exogenous ET-1 delivery (Piovezan et al., 2000), the mechanism of this effect has not been determined. Furthermore, despite evidence to support a role for the ETB receptor in cutaneous inflammation (Griswold et al., 1999), the role of the ETB receptor in acute pain produced by ET-1 administration remains unknown. Therefore, in these experiments, we examined the effects of either a locally administered ETB receptor antagonist or an ETB receptor agonist on hindpaw flinching and excitation of nociceptors induced by the acute subcutaneous application of ET-1. The potential role of opioid receptors in ETB receptor-mediated inhibition of ET-1-induced flinching and nociceptive firing that we observed during the course of these experiments was also examined.
MATERIALS AND METHODS
General. Experiments were performed on 122 adult (175–225 gm) male Sprague Dawley rats (Harlan Sprague Dawley, Indianapolis, IN). All procedures were approved by the Standing Committee on Animals at Harvard Medical School. Animals were treated and cared for according to the ethical standards and guidelines for investigators of experimental pain in animals prescribed by the Committee for Research and Ethical Issues of the International Association for the Study of Pain (Zimmermann, 1983). All rats were housed in a viral antibody-free facility (three per cage) on a 12 hr light/dark cycle with food and water available ad libitum. Before beginning experiments, animals were handled for 1–2 d and were thereby acclimated to both the testing environment and the experimenters.
Drugs. All drugs were diluted in PBS (Invitrogen, Rockville, MD), pH 7.4, and kept on ice during experiments. Synthetic ET-1 (98% pure peptide content), the ETBreceptor-selective antagonist BQ-788 (N-cis- 2,6-dimethylpiperidinocarbonyl-l-γ-methylleucyl-d-1-methoxycarbonyltrptophanyl-d-Nle) (reported IC50 of 1.2 nm at the ETB receptor and 1.3 μm at the ETA receptor) (Ishikawa et al., 1992), and the ETBreceptor-selective agonist IRL-1620 (Suc-Asp-Glu-Glu-Ala-Val-Tyr-Phe-Ala-His-Leu-Asp-lle-lle-Trp) (reportedK I of 16 pm at the ETB receptor and 1.9 μm at the ETA receptor) (Takai et al., 1992) were supplied by American Peptides (Sunnyvale, CA). The concentration of ET-1 used (200 μm) gives less than a saturating response for flinching behavior, as described by Fareed et al. (2000). Therefore, we assume that we are providing an effective concentration of ET-1 to target receptors that is less than four times the K D (80% occupancy), where the K D for both the ETA and ETB receptors are in the high picomolar to low nanomolar range (Rubanyi and Polokoff, 1994). BQ-788 was used at concentrations well in excess of the K I (∼100 nm) (Webber et al., 1998) to ensure complete blockade of the ETB receptor. Naloxone was obtained from Sigma (St. Louis, MO) and dissolved in PBS. The dose of naloxone used for local injection was based on previously described reports of efficacy in rat models of cutaneous pain (Stein et al., 1990a; Eisenberg et al., 1996).
Injection procedures. For injections, naive rats were briefly anesthetized with the rapidly reversible inhalational anesthetic sevoflurane (3–4 min of inhalation). The method of ET-1 administration that we described previously (Gokin et al., 2001) was modified as follows to minimize the volume injected and to reduce the number of injections. Control and experimental subcutaneous injections were always performed as either a single 20 μl injection (for the 4 nmol dose of ET-1 alone) or as two sequential 10 μl injections. The first injection always contained one of the following: vehicle (when examining the effects of ET-1 alone), an ETB receptor agonist or antagonist, or naloxone. The second injection contained ET-1 mixed with one of these agents. Agonists, antagonists, and naloxone were always injected at the same concentration for both injections and were given in the first injection to optimize binding to the target. The first injection was made 40 sec after the onset of limb cooling as described by Gokin et al. (2001). The hindlimb was cooled with a small amount of packed ice in a 15 ml polypropylene centrifuge tube (Corning, Corning, NY) placed beneath the limb; a small amount of ice in a small (∼25 ml) plastic bag was also placed on top of the hindlimb. The second injection was performed 1.5–2 min after the first injection. Both injections were delivered subcutaneously into the midplantar paw. After the first injection, the area was lightly outlined with a felt tip marker, and the second injection was made along the same needle track into the marked area, usually ∼1 cm distal to the heel.
Behavioral assessments. Behavioral assessments were performed as described previously (Davar et al., 1998), with animals freely moving on a flat surface that was enclosed by an inverted, large clear Plexiglas cage. Repetitive and spontaneous flinching of the ipsilateral hindpaw (rapid lifting of the entire hindlimb that begins with hip flexion and includes dorsiflexion of the toes) and the number of events and duration of biting or licking were counted beginning 5 min after ET-1 injection and continuing every 5 min for 75 min of observation. Blinding was performed in initial studies by providing experimenters with unlabeled tubes containing ET-1 or control solutions. However, the very clear and robust effect of ET-1 when compared with PBS made it difficult for the experimenters to remain blind. Similar results (effect readily discerned) were obtained when comparing ET-1 with BQ-788, IRL-1620, and naloxone, reducing the usefulness of blinding.
Neurophysiological experiments. Using methods we described previously in detail (Gokin et al., 2001), single-unit nerve activity was recorded from the right sciatic nerve before and after injection of ET-1, IRL-1620 together with ET-1, or IRL-1620 together with ET-1 and naloxone. Briefly, 12 adult male Sprague Dawley rats weighing 250–300 gm (Harlan Sprague Dawley) were initially anesthetized with intraperitoneal urethane (1.3 gm/kg; Sigma) or sodium pentobarbital (50–60 mg/kg). The right jugular vein was cannulated to permit intravenous administration of additional doses of sodium pentobarbital to maintain general anesthesia, titrated to the absence of corneal reflexes, accelerated heart rate, and withdrawal reflexes to noxious stimuli. Heart rate was monitored with a Tektronix (Beaverton, OR) 498 EKG monitor. Tracheotomy was performed for artificial respiration. During recordings, rats were immobilized with pancuronium bromide (1 mg · kg−1 · h−1, i.v.; Sigma) and artificially ventilated via a respirator (RSP1002; Kent Scientific, Torrington, CT). End-tidal CO2was continuously monitored with an end-tidal CO2analyzer (IITC Life Science, Woodland Hills, CA) and maintained at 4–4.5%. Core body temperature was monitored by a rectal thermometer and maintained at 36–37.5°C with a circulating water heating pad and heating lamps. At the end of an experiment, rats were killed with an overdose of sodium pentobarbital (100–200 mg/kg, i.v.).
To record single-unit activity, a restricted skin incision was made over the posterior hindlimb, and the skin and muscle were opened to expose the middle and distal part of the sciatic nerve. Rats were then placed in a stereotaxic frame (David Kopf Instruments, Tujunga, CA) to immobilize the lower spine and pelvis. The skin at the incision was sewn to a metal ring to form a pool. The fascia and sheath overlying the sciatic nerve were carefully removed, and the nerve was placed on a platform and covered with warm mineral oil. Small nerve filaments, transected proximally, were teased gently from the sciatic under a dissecting microscope (Zeiss, Thornwood, NY). Isolated fine filaments were then wrapped around a silver wire recording electrode, which was connected to a high-impedance probe. One or two such electrodes were used for recording from one or two separate microfilaments. Reference electrodes were placed in the surrounding tissues.
The action potential of an isolated afferent fiber was amplified 1000×, filtered with a bandwidth of 300–3000 Hz with a DAM8 amplifier (World Precision Instruments, Sarasota FL). Filtered signals were visualized on an oscilloscope (model 5301; Tektronix), with parallel audio monitoring, and recorded and stored on computer disc using the CED1401 Plus interface (Cambridge Electronics Design, Cambridge, UK) coupled to a Pentium processor-based personal computer. Signals were analyzed with Spike-3 software (Cambridge Electronics Design). Discharge frequency was counted by using spike shape discrimination(Spike-2; Cambridge Electronics Design), and a histogram was created for each fiber. The nerve activity is presented here as both native records and bin histograms.
To search for units, we used electrical stimulation of nerve fibers at a site between the recording site and the receptive field (RF) of the fiber, and we used mechanical stimulation of their hindpaw RFs. Stimuli from a Grass Instruments (Quincy, MA) model 88 stimulator were delivered via bipolar sharp needle electrodes with duration of 0.5–0.75 msec and amplitude of 30–100 V to activate C-fibers. The main criteria used for physiological characterization and classification of fibers were responses to natural stimulation of their RFs and their conduction velocities, as described previously (Gokin et al., 2001). Conduction velocities (CVs) were calculated by dividing the distance between the stimulating and recording electrodes by the latency of the electrically evoked spike. Units with CVs <2 m/sec were identified as C-fibers (Sanders and Zimmermann, 1986; Handwerker et al., 1991; Leem et al., 1993; Huang et al., 1997). Physiological characterization of C-fibers was based on their responses to various mechanical (light and strong) and thermal (heat and cold) stimulation. They were classified as high-threshold mechanoreceptors (HTMs) if they fired predominantly in response to a strong pinch of the skin with forceps, or von Frey monofilaments (15–52 gm). The thermal responsiveness of mechanically activatable units was determined by applying a heated metal spatula (∼52°C) and a piece of ice (cold stimulation) to their cutaneous RFs. On the basis of both their responses to these stimuli and their CVs, C-units were classified as either (1) polymodal (mechano-heat) nociceptors (C-PMNs) or (2) high-threshold mechanoresponsive (HTMr) C-fibers, several of which also responded to thermal stimuli (heat or cold). To confirm that we were recording from the same electrically and physiologically activated unit, a modification of the “blocking” method of Iggo (1958)was used, as described recently by Gokin et al. (2001).
Drug injections were performed in an identical manner to those in the neurobehavioral experiments, except that limb cooling was not used because we showed previously that spike responses are easily obtained with subcutaneous injection of ET-1 without cooling (Gokin et al., 2001). Recording continued in most instances for 40 min to 2 hr after administration of drugs.
Data analysis. The maximal (peak) number of flinches per 5 min epoch that occurred during the observation period (75 min) for each animal was defined as the maximal flinch frequency (MFF) occurring within a 5 min block and was scored independent of the time of occurrence. The MFF, time to reach MFF, total number of flinches occurring within the 75 min observation period, number of biting or licking events, and total duration of biting or licking behavior were also determined for each animal. Data are reported as means ± SEM. To establish significant differences between the effects of different ET-1 doses or between injected agents, an unpaired, two-tailed Student's t test was applied (Origin 5.1; Origin Lab, Northampton, MA), with p < 0.05 considered significant.
For neurophysiological experiments, a response was defined as firing that occurred after completion of the second of two injections and needle withdrawal. The latency of spike response was measured as the time from the end of ET-1 injection to the onset of the first nonelectrically evoked response. The duration of responses was measured from the onset of response until afferent activity returned to baseline. Mean response frequency (MRF) (in impulses per second) was calculated as the number of spikes divided by the duration of the entire ET-1-induced response. Maximum frequency (MxF) was determined, to characterize responses within bursting patterns, as the number of spikes within a brief (1 sec) interval of rapid firing. Duration and MRF were used as quantitative parameters for comparing the magnitudes of responses to different doses of ET-1. All results are presented as means ± SEM. One-way ANOVA was used to evaluate the significance of the difference of means. Differences were considered statistically significant at p < 0.05.
General and behavioral effects of ET-1 injection
Similar to results described recently by Gokin et al. (2001), subcutaneous administration of 2 (10 μl of 200 μm), 4 (20 μl of 200 μm), and 6 (10 μl of 600 μm) nmol of ET-1 into the rat plantar hindpaw induced immediate blanching at the injection site that began 2–5 min after injection and that reached a maximum size of 2.5 mm2. This blanching was followed 5–15 min later by the development of local erythema of the plantar surface and at 20–50 min by diffuse rubor of the hindpaw below the knee that lasted for 60–70 min before resolving.
ET-1-induced hindpaw flinching
When ET-1 was injected subcutaneously as a single bolus into the rat plantar paw, ipsilateral hindpaw flinching, a behavioral response that indicates pain behavior in the rat, was observed in 100% of animals (Davar et al., 1998; Fareed et al., 2000; Gokin et al., 2001). Biting or licking of the hindpaw was also observed in most animals and is reported at the end of Results. Flinching began 5–10 min after observations began, increasing with time until a MFF was reached at ∼30–50 min and resolving to near baseline by 75 min (Gokin et al., 2001). For single injections of ET-1 (4 nmol, 20 μl of 200 μm), the averaged MFF of 40 ± 4 flinches/5 min (n = 12) occurred at 41 ± 3 min after observations began, whereas the averaged total number of flinches was 178 ± 29 over the 75 min observation period.
Dose dependence of the effects of ET-1
Flinching frequency was increased significantly over the 5–25 min period (Fig. 1), as was the averaged MFF (42 ± 9 flinches/5 min; n = 8) when 6 nmol was injected compared with 2 nmol (24 ± 2 flinches/5 min;n = 12; p < 0.05). The mean time to reach MFF (22 ± 4 min for 6 nmol vs 52 ± 3 min for 2 nmol) was also different for these two doses (p < 0.0001), as was the mean total number of flinches (253 ± 66 vs 123 ± 13 for 6 vs 2 nmol of ET-1, respectively; p< 0.05). A dose of 2 nmol of ET-1 was therefore selected as reliably submaximal, to further study the effects of agents modulating the ETB receptor.
BQ-788 enhances ET-1-induced flinching
To demonstrate a contribution of ETBreceptors to the pain-inducing actions of ET-1, we next examined the behavioral consequences of administration of a selective ETB receptor antagonist (BQ-788) before and then together with ET-1. As seen in Fig.2 A, BQ-788 (3 mm, 60 nmol) accelerated the development of ET-1-induced hindpaw flinching. Flinching frequency was increased significantly in the presence of BQ-788 at the 15, 20, and 25 min time points (Fig. 2 A), as was the averaged MFF when compared with 2 nmol of ET-1 alone (p < 0.05) in 9 of 12 tested rats (Fig.3 A). Three of 12 rats treated with BQ-788 plus ET-1 showed signs of toxicity (red tears and reduced exploratory behavior) similar to what we observed previously with high doses of ET-1 (Gokin et al., 2001) and, despite evidence of hindpaw rubor, had practically no flinching behavior. These animals were not included in the data analysis.
The averaged MFF also occurred earlier in BQ-788-treated rats [20 ± 1 (n = 9) vs 52 ± 3 min (n = 12) for 2 nmol of ET-1 alone; p < 0.0001]. Although BQ-788 increased the mean total number of flinches produced by ET-1 by the same proportion (1.27-fold) as was observed for its effects on MFF, this difference did not reach statistical significance. When administered alone in two subsequent injections of 10 μl, BQ-788 did not evoke flinching that was different from PBS for either MFF or total number.
IRL-1620 inhibits ET-1-induced flinching
To further investigate the role of ETBreceptors, we next studied the effect of the selective ETB receptor agonist IRL-1620 on ET-1-induced flinching. IRL-1620 (200 μm, 4 nmol) decreased the frequency of flinching at later times (significance reached at 45 and 65 min; p < 0.05) (Fig. 2 B) and reduced the averaged MFF (n = 12) when compared with ET-1 alone (n = 12; p < 0.05) (Fig.3 A). The mean total number of flinches was also decreased by IRL-1620 (p < 0.02) (Fig. 3 B). Although IRL-1620 administered alone in two 10 μl injections also induced flinching behavior (22 ± 3 flinches; n = 9) that was slightly higher than observed for PBS alone (13 ± 2;n = 8; p < 0.05), this was not unexpected because some activation of the ETAreceptor may occur at this concentration of the agonist (Takai et al., 1992).
To help determine whether the actions of IRL-1620 were mediated at a single class of receptors, we next determined the relationship between IRL-1620 dose and flinching behavior. Coadministration of 2 nmol (20 μl total volume, 100 μm) of IRL-1620 with ET-1 had similar effects to 4 nmol of IRL-1620 on ET-1-induced hindpaw flinching (Fig. 2 B). The averaged MFF was reduced (n = 12; p < 0.05) (Fig.3 A), as was the mean total number of flinches (p < 0.002) (Fig. 3 B) when compared with ET-1 alone. At the lowest concentration of IRL-1620 (20 μm, total dose of 0.4 nmol) that we used, neither MFF nor total flinches were different from ET-1 alone (Fig. 3). However, MFF did occur earlier (30.4 ± 5.8 vs 51.7 ± 3.1 min for ET-1 alone; p < 0.000005), at the same time point observed with higher concentrations of IRL-1620. To reduce any possible actions of IRL-1620 at the ETA receptor, while maintaining its inhibitory effects, IRL-1620 was used at the 100 μmconcentration in subsequent experiments.
Naloxone enhances ET-1-induced flinching
Opioids are known to have analgesic actions in both the CNS and peripheral nervous system. To examine the potential role of endogenous opioids in the expression of ET-1-induced pain behavior and in ETB receptor-mediated antinociception (see below), we next administered the nonselective opioid receptor antagonist naloxone (2.75 mm, total dose of 55 nmol) before and then together with ET-1 (2 nmol). The averaged MFF was ∼50% higher for naloxone plus ET-1 than for ET-1 alone (p < 0.01) (Fig. 3 A). Averaged MFF also occurred earlier in naloxone-treated rats (41 ± 2 vs 52 ± 3 min for ET-1 alone; p < 0.01). However, naloxone did not alter the mean total number of flinches observed (Fig.3 B), nor did naloxone alone induce flinching behavior that was different from PBS (p > 0.05).
Naloxone prevents the inhibition of ET-1-induced flinching by IRL-1620
To more specifically examine the role of opioids in ETB receptor-induced antinociception, naloxone was injected together with IRL-1620 (100 μm, 2 nmol) before and then together with ET-1 (2 nmol). The averaged MFF for IRL-1620 plus ET-1 plus naloxone (n = 12) was twice that for IRL-1620 plus ET-1 without naloxone (n = 12;p < 0.05), which is evidence that IRL-1620 has no effect in the presence of naloxone (Fig. 3 A). Averaged MFF in this group occurred at 40 ± 3 min and was not different from the averaged MFF for ET-1 plus naloxone (p > 0.05). The mean total number of flinches evoked by IRL-1620 plus ET-1 plus naloxone was identical to that evoked by ET-1 plus naloxone (Fig.3 B) and, in both cases, was higher than that evoked by IRL-1620 plus ET-1 (p < 0.002). In support of these actions of naloxone at opioid receptors, we have recently observed, in a preliminary manner, that a μ-opioid receptor-selective antagonist (CTOP) can also prevent the actions of IRL-1620 (data not shown).
Biting and licking behavior induced by ET-1 injection
Biting or licking events were also observed after ET-1 administration in 92% (2 nmol) and 75% (6 nmol) of rats compared with 13% for PBS injections and ranged from 1 to 39 events per rat. Neither the mean number nor the duration of biting or licking events were different between different doses of ET-1 or when ET-1 was compared with BQ-788, IRL-1620, or naloxone administered together with ET-1.
To obtain more direct evidence of a naloxone-sensitive inhibitory effect of IRL-1620 on ET-1 induced pain, impulse activity was recorded from 21 physiologically characterized nociceptive C-fibers. Recordings were made before and after subcutaneous injection of ET-1 (200 μm, 2 nmol; n = 6), IRL-1620 (100 μm, 1 nmol) followed 1–2 min later by ET-1 (200 μm) plus IRL-1620 (n= 12), and before and after subcutaneous injection of IRL-1620 (100 μm) plus naloxone (2.75 mm) followed 1–2 min later by ET-1 (200 μm) plus IRL-1620 plus naloxone (n = 3) into the cutaneous RFs of these units. The RFs of these units were located on the glabrous hindpaw within territory innervated by the plantar and sural nerves and were 1–2 mm in size (Fig. 4). The conduction velocities of C-units ranged from 0.63 to 1.1 (mean of 0.86 ± 0.03) m/sec and were not significantly different between experiments. Most fibers had no ongoing spontaneous activity.
ET-1 induces spike responses in C-nociceptors
Four of the six units studied here responded only to strong (15–52 gm von Frey hair, numbers 5.18–5.88) mechanical stimulation of their RFs (HTMr units). Noxious pinch gave a maximal response of 11–29 impulses/sec (mean of 17.5 ± 5.4). The remaining two units also responded to heat (C-PMNs). All six fibers responded to ET-1 (200 μm). Five responded in a nearly identical manner, showing the typical bursting discharge pattern (Fig. 4 A) that we described previously (Gokin et al., 2001). The latency (0.315 ± 0.08 min), MRF (0.318 ± 0.07 impulses/sec), and MxF (4.14 ± 0.17 impulses/sec) of these six units were also similar to those reported previously by us (Gokin et al., 2001). The duration of responses for these units was 24.13 ± 3.76 min; most of these units were not observed beyond 40 min because we never detected a resumption of spiking, under these conditions, once the activity of a unit had returned to baseline.
IRL-1620 inhibits ET-1-induced spike responses
Twelve C-units were studied. Six were classified as HTMr, two as C-PMNs, two as HTMr that also responded to cold, and two remaining units, one an HTMr that also responded to low-intensity stimuli and the other an HTMr that also responded to low-intensity stimuli and cold. Overall, IRL-1620 (100 μm) suppressed ET-1 (200 μm) induced spike responses in all units (Table1). Complete suppression was observed in four of 12 units, whereas late responses of reduced maximal and mean frequency were observed in two of 12 units, and weak responses, showing similar latency to onset to the results with ET-1 alone, were observed in the remaining six units. A typical example of this inhibitory effect of IRL 1620 is seen in Figure 4 B. Insignificant, or occasionally very short duration (≤20 sec), responses were observed in the injection of IRL-1620 alone, which preceded the injection of IRL-1620 plus ET-1 (Fig. 4 B).
Naloxone reverses IRL-1620 inhibition of ET-1-induced spike responses
Three units were recorded (two C-PMNs and one HTMr) while injecting ET-1 and IRL-1620 together with naloxone to their RFs. The typical bursting pattern seen with ET-1 alone was observed in all three units (Fig. 4 C). MRF and MxF of responses determined for these three units were not different from those obtained for ET-1 alone but higher than those obtained with ET-1 plus IRL-1620-treated units (Table 1).
These results are evidence that ETBreceptors are important modulators of ETAreceptor-mediated pain in cutaneous tissues. ET-1 delivered as a single bolus subcutaneous injection to the rat plantar hindpaw induces hindpaw flinching that is greater and occurs earlier when it is coadministered with a selective ETB receptor antagonist, similar to the pattern observed with increasing ET-1 dose. Conversely, coadministration of a selective ETB receptor agonist reduces hindpaw flinching and inhibits ET-1-induced spike responses in nociceptors, both of which are prevented by local injection of the nonselective opioid receptor antagonist naloxone.
Although relatively high concentrations of ET-1 (and BQ-788 and IRL-1620) were used in these experiments, the local concentrations of these agents at the receptor sites are unknown, and their diffusion to the receptors is limited by closely opposed dermal and epidermal cells, the vascular flow that rapidly removes many small molecules, and by the actions of many degradative enzymes. Analogously high concentrations of relatively hydrophobic compounds, such as local anesthetics (e.g., 30–50× the IC50 on isolated neurons) are often needed to obtain in vivo efficacy from subcutaneous delivery (Khodorova and Strichartz, 2000). The endogenous concentration of ET-1 released by cutaneous tissue injury might indeed be low (Hara et al., 1995; Tsuboi et al., 1995) yet sufficient to induce flinching behavior in rats (Gokin et al., 2001). Increasing the concentration of ET-1 increased flinching, which peaked earlier. This enhanced effect may be the result of increased actions at the ETAreceptor as a consequence of increased ET-1 dose (Fareed et al., 2000). However, increased availability of ET-1 that may occur secondary to its reduced internalization and inactivation by the ETB receptor, a process described in heterologous expression systems (Bremnes et al., 2000; Paasche et al., 2001), might also contribute to this result (Fig. 5). The earlier onset of flinching is also consistent with reduced actions at the ETB receptor, as we observed with ETB receptor blockade, which may lead to enhanced and accelerated actions of ET-1 at the ETAreceptor (Rubanyi and Polokoff, 1994).
Consistent with the proposed role for ETBreceptors in the modulation of the ETAreceptor-mediated response, blockade of the ETBreceptor also increases flinching and accelerates its onset. This suggests the possibility that ETB receptor activation inhibits and delays the effects of ETAreceptor activation. The substantially reduced (50%) flinching observed with an ETB receptor agonist further supports the idea that ETB receptors can suppress ETA receptor-mediated pain in cutaneous tissues. Lowering the dose of IRL-1620 by a factor of 10 eliminated its inhibition of flinching behavior, pointing to a moderately steep dose–response relationship that is consistent with actions at a single class of receptors.
The inhibitory effect of ETB receptor activation is unlikely to be through sensory fibers themselves, whereon ETB receptors have not been detected (Pomonis et al., 2001), and instead is more likely the result of an indirect action mediated by another cell that leads to the suppression of spike activity in nociceptors. The prevention of IRL-1620-induced inhibition of flinching by locally injected naloxone supports this hypothesis and suggests the possibility that ETB receptor activation on adjacent cutaneous cells leads to the release of an endogenous opioid peptide. The source of this peptide could either be a supportive glial cell [e.g., a non-ensheathing Schwann cell (Pomonis et al., 2001)] found adjacent to cutaneous sensory axons or a resident cutaneous cell (Tada et al., 1998). Among cutaneous cells, keratinocytes are closely associated with sensory axons (Haberberger and Bodenbenner, 2000) and modulate the activity of cutaneous nociceptors (Lewin and Mendell, 1993). They also possess ETB receptors and express and secrete opioid peptides in a regulated manner (Wintzen et al., 1996, 2000; Zanello et al., 1999), making them reasonable candidates to mediate the ETB receptor effects we observed here. Other cutaneous cells that possess ETB receptors and express pro-opiomelanocortin, such as fibroblasts, might also be candidates for a cellular source of ETBreceptor-induced opioid secretion in plantar hindpaw (Teofoli et al., 1999; Shraga-Levine and Sokolovsky, 2000). In addition, inflammatory cells such as lymphocytes recruited to sites of tissue injury could release opioid peptides (Cabot et al., 1997) and might respond to ET-1, but the rapid inhibitory effects we observed here do not support this slower process that is dependent on cellular migration.
Impulse activity in nociceptors after the subcutaneous injection of ET-1, identical in pattern and intensity to what was reported previously (Gokin et al., 2001), was inhibited by coinjection of the ETB receptor agonist. Spike responses in identified C-nociceptors were completely abolished in several units, whereas the response latency and the duration of response was prolonged in the remaining units. These results provide a mechanism for the inhibitory actions of ETB receptor activation on ET-1-induced pain behavior. More importantly, the prevention of this effect by locally administered naloxone supports the possibility of an opioid receptor-mediated action of IRL-1620 and helps validate our model of ETB receptor-induced modulation of ETA receptor actions on nociceptors in skin (Fig.5). Incomplete suppression of spike responses in some units might be the result of partial actions of IRL-1620 at the ETA receptor, as described above, or secondary to competition with ET-1 for ETB receptor binding sites, leading to increased actions at ETAreceptors. The prolonged latency with low-level firing in responding units is consistent with the selective actions of this potent ETB receptor agonist, which would be expected to delay the onset of ETA receptor-mediated firing.
These results suggest a dual level of control over the pain-related actions of ET-1 in cutaneous tissues, much like its divergent actions of vasoconstriction and vasodilatation in vascular tissue, respectively, mediated by ETA and ETB receptors found on different vascular cells (Rubanyi and Polokoff, 1994). Thus, in cutaneous tissues, ETA receptors on nociceptors can directly activate pain responses, whereas ETB receptors on supportive cells inhibit these effects of ET-1 in a naloxone-sensitive manner. The source and type of opioid peptide that mediates this effect is unknown but likely originates locally from cutaneous or supportive cells. Based on these results, we surmise that ETB receptor activation in cutaneous tissues leads to the local release of an endogenous opioid peptide that hyperpolarizes nociceptors, in the face of ETAreceptor-dependent excitation, thereby suppressing impulse generation and inhibiting the effects of ET-1.
↵* A.K., M.U.F., and A.G. contributed equally to this work.
This work was supported by United States Public Health Service Grant CA 80153. We acknowledge the technical assistance of Jamie Bell and consultative help from Dr. Guy Hans.
Correspondence should be addressed to Dr. Gudarz Davar, Molecular Neurobiology of Pain Laboratory, Department of Anesthesiology, Perioperative and Pain Medicine, Brigham and Women's Hospital, 75 Francis Street, Boston, MA 02115. E-mail:.