Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Epibatidine is a cholinergic agonist with affinity for neuronal nicotinic receptors (nAChRs) in the low picomolar range (Houghtling et al., 1995). Besides binding to the 42 subtype, epibatidine also binds to several other nAChRs composed of different  and  subunits (Parker et al., 1998). Epibatidine thus is a rather nonselective agonist at most nAChR subtypes. The pharmacological interest in epibatidine arises from the finding that the compound has considerable analgesic potency (Badio and Daly, 1994) and, although its unselective stimulation of nAChRs and its narrow therapeutic index make epibatidine too poisonous to be used clinically, this observation opens up the possibility of powerful analgesia without the use of opiates. The 42 nAChR subtype is the most abundant in the rat brain, and nicotine, which is slightly analgesic (Qian et al., 1993), binds with some selectivity to this receptor subtype. Accordingly, drug candidates with some analgesic potency and selectivity for this subtype have been developed (Bannon et al., 1998). However, the nonselective agonistic activity of epibatidine makes it possible that other nACh receptor subtypes also could be involved in the analgesic effect. We have previously described a receptor population in the habenula with high affinity for [³H]epibatidine but low affinity for nicotine and acetylcholine because neither 2 nor 10 µM concentrations of the respective drugs could displace the bound [³H]epibatidine (Plenge and Mellerup, 1998). This indicates that this binding site is not an 42 nACh receptor_, a result supported by recent studies with 4- as well as 2-knockout mutant mice in which [³H]epibatidine binding is preserved in the habenula but absent in 42-rich areas, such as cortex and thalamus (Zoli et al., 1998; Marubio et al., 1999).

Using autoradiographic methods, the present study was directed toward determining kinetic parameters for the "epibatidine receptor" in the habenula. Using [³H]epibatidine, K_D and B_max were estimated. By displacing [³H]epibatidine bound to the receptor with acetylcholine and nicotine, the affinity of these two substances for the receptor was determined. We studied the anatomical extension of [³H]epibatidine binding in the habenula in both coronal and horizontal sections. The "nonacetylcholine" displaceable [³H]epibatidine binding to medial habenula was sought and characterized using a considerable number of "nicotinic" compounds with selectivity for various nAChR subtypes.

Various lines of evidence suggest that the habenular complex is involved in central pain processing (Andres et al., 1999). Habenular neurones respond to noxious stimuli (Benabid and Jeaugey, 1989; Dafny and Qiao, 1990; Nagao et al., 1993), and the expression of the immediate early gene c-fos is induced in this structure after peripheral noxious stimulation (Dai et al., 1993; Smith et al., 1997; Michl et al., 2001). Analgesia can be achieved by electrical or chemical stimulation of the habenula (Cohen and Melzack, 1986, 1993; Mahieux and Benabid, 1987; Terenzi and Prado, 1990; Terenzi et al., 1990). In addition, the habenula has been shown to be involved in the modulation of acupuncture analgesia (Takeshige et al., 1993). On this background, we decided to study the analgesic effect in rats of direct bilateral stereotactic application of epibatidine to the habenular complex using the hot-plate test.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Autoradiography. Sections of rat brains containing habenula [approximately 2.6 to 4.1-mm caudal to bregma (Paxinos and Watson, 1986)] were cut in coronal orientation at 20 µm and thaw-mounted on SuperFrost/Plus slides (Menzel-Glaser, Braunschweig, Germany). On each slide, three to four sections from different rat brains were mounted. Slides were stored at 20°C until processed. Binding assays were performed in buffer A (120 mM NaCl, 5 mM KCl, 50 mM Tris, pH 7.5) at 2°C. The slides were hydrated in buffer A for 15 min before being incubated with [³H]epibatidine (Amersham Biosciences Europe GmbH, Horsholm, Denmark; 53 Ci/mmol) and other compounds for 1 h at 4°C. Nonspecific binding was determined in the presence of 1 µM nicotine. To reduce/eliminate nonspecific binding of [³H]epibatidine, slides were subsequently washed two times for 30 min at 0°C in buffer A. We have determined the half-life of [³H]epibatidine bound to rat brain membrane receptors to be 4 h at 0°C; thus, elimination of nonspecific binding by this washing procedure only results in an approximate loss of 16% of the specifically bound [³H]epibatidine. Slides were dried and exposed to Hyperfilm ³H (Amersham Biosciences Europe GmbH). The films were exposed from 3 to 6 months in autoradiography cassettes at 20°C before being developed. Developed films were analyzed and quantitated in a computer-assisted video densitometer (Scion-Image; Scion Corporation, Frederick, MD) using the standard curve generated from ³H standards. (Amersham Biosciences Europe GmbH). These standards have their radioactive concentration expressed in becquerels per milligrams wet weight, allowing calculation of the receptor density when the specific activity of the ligand is known.

View this table: [in this window] [in a new window]   TABLE 1 Affinity of four different nicotinic compounds to nicotinic receptors in different assays The same compound's displacement of [3H]epibatidine bound to rat brain sections containing medial habenula is shown in Fig. 3. The ligand concentrations and nicotinic receptors measured in the four assays were: [3H]cytisine, 1 nM rat cortex membranes, measuring 42 nicotinic receptors; [3H]epibatidine, 0.3 nM rat brain membranes, measuring primarily 42 nicotinic receptors, [3H]bungarotoxin, 1 nM rat cortex membranes, measuring 7 nicotinic receptors; [3H]epibatidine, 0.3 nM IMR 32 wild-type cell line, measuring primarily 34 nicotinic receptors. IC50 values are in micromolar concentrations.

Intracerebral Injections. Double guide cannulas (22-gauge, 1.2-mm distance between guiding cannulas, cut 5.0 mm below pedestal), double internal cannulas (33-gauge, 6-mm length), double dummy cannulas (5-mm length), and double connector assemblies were purchased from Plastics One, Inc. (Roanoke, VA). Approximately 3 weeks before the first hot-plate test, the rats (male, Wistar, 240-260 g; purchased from M&B Breeding Center, Lille Skensved, Denmark) were implanted with a double guiding cannula directed at the medial habenula (mHb) (anterior-posterior 3.3 mm and mediolateral-lateral ± 0.6 mm from bregma; dorsal-ventral 3.8 mm from dura). Between injections, the guiding cannulas were kept open with double dummy cannulas, covered with dust caps. At the time of testing, double injection cannulas, which protruded 1 mm into the tissue ending near the habenula, were inserted into the guide cannulas, while one experimenter gently restrained the rat. For 1 min, 2 × 1 µl of a drug solution in physiological saline were infused bilaterally with the help of two Hamilton syringes connected to the injection cannulas. The injection cannulas were left in place for 4 min more to allow diffusion into the tissue and prevent backflow into the guiding cannulas. Hot-plate testing was carried out 10 and 20 min after injection.

Hot-Plate Testing. Analgesia was tested on a hot-plate (Harvard Apparatus, Inc., Holliston, MA) set at 50°C. For each trial, the timer was started simultaneously with the rat being placed on the hot-plate. Two experimenters closely observed the rat, and the timer was stopped and the rat removed from the hot-plate at the first hind paw licking reaction. Animals that had not shown a hind paw licking reaction were removed from the hot-plate after 30 s. Each animal was injected with drug or the saline vehicle, in counterbalanced order, at a 1-week interval. The experimenters were blinded to the treatment the animals received.

Statistical Analysis. Latency to the first hind paw licking reaction after saline injection was compared with latency after drug injection using paired t tests for the two test times.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

The anatomical part of the habenula, which binds [³H]epibatidine not being displaced by 1 µM nicotine, was visualized in a series of both coronal and horizontal brain sections. Figure 1A shows representative coronal sections in frontal to caudal direction, labeled with 0.2 nM [³H]epibatidine. In the frontal part, the binding area is almost circular, whereas it becomes elongated in the more caudal sections with a ditch in the upper part. Also, in the caudal part, the beginning fasciculus retroflexus is seen protruding down. In the more caudal sections, the receptor seems to be unevenly distributed because lighter and darker patches are seen within the area with binding sites. Figure 1B shows the two parallel binding areas in horizontal sections. Taken together, the [³H]epibatidine binding area coincides with the medial habenula.

View larger version (56K): [in this window] [in a new window]   Fig. 1.   [3H]Epibatidine binding to rat brain medial habenula. A, coronal brain section incubated with 0.2 nM [3H]epibatidine and 1 µM nicotine. B, horizontal brain section incubated with 0.2 nM [3H]epibatidine without or with 1 µM nicotine.

The affinity of the non-nicotine displaceable [³H]epibatidine binding was estimated using 11 [³H]epibatidine concentrations from 5 to 5000 pM and either 1 µM nicotine to displace 42 binding or 1 µM epibatidine to determine nonspecific binding. The autoradiograms from brain sections incubated with 1 µM epibatidine were without any binding, demonstrating that the washing procedure used eliminated nonspecific binding. The binding that appeared on the autoradiograms thus was specific binding of [³H]epibatidine to nACh receptors. Saturation analysis of non-nicotine displaceable binding to mHb revealed a saturable binding with a K_D value of about 0.5 nM (K_D = 0.47 nM and 0.57 nM in two experiments). The receptor concentration, B_max value, was estimated to be about 1100 fmol/mg wet weight (930 and 1200 fmol/mg wet weight in two experiments).

Using a [³H]epibatidine concentration of 0.1 nM, the affinity of nicotine and acetylcholine for the binding site was estimated on a series of slides incubated with nicotine and acetylcholine in concentrations increased by a factor of 2 from 0.5 to 1000 µM. Figure 2 shows the binding to mHb with and without the lowest concentrations of acetylcholine. It is evident that the binding of [³H]epibatidine to cortical and thalamic receptors disappears already with 0.5 µM acetylcholine, whereas some binding in mHb persisted even with 64 µM acetylcholine; in parallel experiments, the same was found to apply for nicotine. From the displacement curves, the binding site affinities (K_i value) for nicotine and acetylcholine were estimated to be 5 and 8 µM, respectively, for displacement of [³H]epibatidine bound to the epibatidine receptor in mHb. In a similar experiment performed on sections of superior cervical ganglia, nicotine was found to displace [³H]epibatidine bound to the ganglionic 34 nACh receptor with a K_i value of 1.3 nM.

View larger version (74K): [in this window] [in a new window]   Fig. 2.   [3H]Epibatidine (0.1 nM) binding to coronal rat brain sections containing medial habenula displaced with increasing concentrations of acetylcholine.

To explore the subunit composition of the [³H]epibatidine binding receptor in mHb, about 70 different experimental compounds with known selectivity for different nACh receptor subunit combinations were obtained from various pharmaceutical companies and screened with respect to a possible affinity for the epibatidine receptor. The compounds were tested in three concentrations, 0.1, 1.0, and 10 µM, together with 0.2 nM [³H]epibatidine, with the objective of finding a compound that displaced [³H]epibatidine binding to mHb without displacing binding to cortex and thalamus. The results obtained can be illustrated with the displacement pattern of the four different compounds shown in Fig. 3. Table 1 shows the affinity as determined by the pharmaceutical companies of the four compounds to nACh receptors composed of 7, 42, and 34 subunits, respectively. To ease the discussion of the four compounds, we have named them: 7 compound, 42 compound, "mixed affinity compound", and "most selective compound". Figure 3, bottom, shows two autoradiograms of mHb-containing brain sections, which had been incubated without or with 1 µM nicotine; as expected, nicotine displaced the [³H]epibatidine binding to cortex and thalamus but not the binding to mHb. Figure 3, top, shows brain sections incubated with [³H]epibatidine and the four different compounds. The compound with selectivity for 7 receptors displaced neither the [³H]epibatidine binding to cortex and thalamus (42) nor the binding to mHb. The compound with high affinity and selectivity for nicotinic receptors composed of 42 subunits readily displaced [³H]epibatidine binding to cortex and thalamus without displacing binding to mHb, even at 10 µM. The mixed affinity compound had high affinity to both 42 and 34 nicotinic receptors (Table 1). This compound readily displaced the 42 binding to cortex and thalamus and, beginning at 1 µM and being complete at 10 µM, also displaced the binding to mHb. The most selective compound also had high affinity to receptors composed of 34 subunits but low affinity to 42 subunit containing receptors (Table 1). As seen in Fig. 3, even at 10 µM, the most selective compound only partly displaced the [³H]epibatidine binding to mHb, as well as to 42 in thalamus and cortex. This failure to displace [³H]epibatidine bound to mHb is in contrast to the result seen with the mixed affinity compound, whereas the weak affinity for an 42 subunit containing receptors reflects the binding data seen in Table 1. The remaining [³H]epibatidine labeling in thalamus and mHb, respectively, at the three concentrations (0.1, 1.0, and 10 µM) after displacement by the most selective compound was 3.9, 3.0, and 1.6 Bq/mg wet weight in the thalamus and 101, 50, and 9.7 Bq/mg wet weight in the mHb (n = 4).

View larger version (86K): [in this window] [in a new window]   Fig. 3.   [3H]Epibatidine binding to coronal rat brain sections containing medial habenula displaced with compounds with affinity to different subtypes of nicotinic acetylcholine receptors. Table 1 shows the chemical names of the compounds and their affinity to 42, 7, and 34 nicotinic receptors. The sections were incubated with 0.2 nM [3H]epibatidine and 0.1, 1.0, or 10 µM the compounds.

To study in vivo the effect of a selective stimulation of the [³H]epibatidine binding receptors in mHb, a stereotactic method was developed by which minute amounts of epibatidine could be injected in the vicinity of habenula using guiding cannulas to direct double injection cannulas to the right position. Diffusion and washout of [³H]epibatidine from the injection site was studied by injections of 1 µCi of [³H]epibatidine in a volume of 1 µl, followed by decapitation from 5 to 40 min after injection. Figure 4A, bottom, is an autoradiogram of a coronal brain section at approximately bregma 3.30, showing the distribution of [³H]epibatidine 20 min (left) and 30 min (right) after bilateral intracerebral injection in the vicinity of the habenula. Figure 4A, top, is a Nissl-stained coronal section at approximately bregma 3.30 from the same rat, overlaid onto the autoradiogram. High concentrations of ³H radioactivity did not spread far from the injection site, and medial habenula is clearly labeled, especially 20 min after the injection. Figure 4B shows the washout curve of [³H]epibatidine from the injection area. From the slope of the curve, the half-life was estimated to be 9 min with one determination at each time point. After rats had been used in the experiments, the position of the cannulas was verified on Nissl-stained brain sections.

View larger version (31K): [in this window] [in a new window]   Fig. 4.   A, bottom, autoradiogram (bregma 3.30) of the distribution of 1 µl of [3H]epibatidine (1 µCi) 20 min (left) and 30 min (right) after a bilateral intracerebral injection in the vicinity of habenula. Top, Nissl-stained coronal section at bregma 3.30 containing medial habenula from the same rat, overlaid onto the autoradiogram. B, washout of [3H]epibatidine in vivo from the brains of rats. Rats were sacrificed from 5 to 40 min after a bilateral intracerebral injection of 1 µl of [3H]epibatidine (1 µCi). The half-life for washout was estimated to be 9 min.

Figure 5 shows hot-plate latency in two different experiments after injection of epibatidine or nicotine into the habenula. In one experiment, the animals were injected bilaterally 3 and 4 weeks after surgery with 2 × 1 µl of 2 nM epibatidine and saline in counterbalanced order, and during weeks 5 and 6, with 2 × 1 µl of 10 nM epibatidine and vehicle. In the second experiment, the animals were injected bilaterally 3 and 4 weeks after surgery with 2 × 1 µl of 10 nM epibatidine and saline in counterbalanced order, and during weeks 5 and 6 with 2 × 1 µl of 10 nM nicotine and vehicle. The two experiments included 8 and 14 rats, respectively. Latency to hind paw licking reaction was assessed 10 and 20 min after injection. Latencies during drug trial and corresponding vehicle trials were compared by paired t tests. Animals that did not respond within 30 s during the vehicle test were excluded from statistical analysis. Figure 5 shows that 10 fmol (1 µl, 10 nM) of epibatidine increased the hot-plate latency in both experiments whereas neither 2 fmol of epibatidine nor 10 fmol of nicotine had any effect.

View larger version (30K): [in this window] [in a new window]   Fig. 5.   Latency to hind paw licking in a hot-plate analgesia test at 50°C. Rats received a bilateral intracerebral injection near the habenula with 2 × 1 µl of physiological saline containing epibatidine or nicotine in the stated amounts or the saline vehicle. Hot-plate latency was tested 10 and 20 min after the injection. Increase in latency was significant (P < 0.05) 20 min after injection of 2 × 10 fmol of epibatidine. , vehicle; , drug.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Nicotinic cholinergic receptors in the CNS and ganglia are composed of  and  subunits, and until now, nine different  subunits and three different  subunits have been characterized. Many of these subunits form functional acetylcholine sensitive ion channels both as homo- and heteromers when expressed in Xenopus oocytes. They can be stimulated with nicotinic agonists like epibatidine (Paterson and Nordberg, 2000), which has high affinity to several of these combinations (Parker et al., 1998). In the rat brain in vivo, neuronal cells in the medial habenula have been found to express mRNA for many of the different subunits, 3, 4, 5, 6, and 7, and 2, 3, and 4 (Sheffield et al., 2000). Thus, at least theoretically, the number of possible different nACh receptors in mHb is rather large. However, only a few of the possible combinations have until now been shown with certainty to be functional in the rat, i.e., 42, which is widespread in the CNS, and 34, which was first described in ganglia (Flores et al., 1996) and then also in the CNS (Quick et al., 1999) and the medial habenula (Sheffield et al., 2000). Many other combinations may be functional in vivo, but due to lack of compounds with selectivity for the different subunit combinations, their existence has been difficult to establish. A possible new nACh receptor with high affinity for epibatidine, and low affinity for acetylcholine and nicotine was found using autoradiographic methods (Plenge and Mellerup, 1998). The receptor has a very limited distribution in the rat brain, as it was only found in a measurable concentration in the medial habenula and a few other brain structures (nucleus interpeduncularis, fasciculus retroflexus, and the pineal gland). In the present study, we have expanded the investigation of this receptor while concentrating on the mHb. The medial habenula was chosen because of its high receptor concentration, which suggests a function for the receptor in this brain structure. Furthermore, the literature suggests a role for the habenular complex in the modulation of pain perception (Cohen and Melzack, 1993; Andres et al., 1999). A more precise outline of the epibatidine receptor distribution (Fig. 1, A and B) coincides with the shape and extension of mHb. The coronal sections of mHb labeled with [³H]epibatidine (Fig. 1A) indicate that the receptor concentration within mHb varies because lighter and darker patches are seen in the autoradiogram. The receptor concentration in mHb is high (in the presence of 1 µM nicotine, it is estimated to be about 1100 fmol/mg wet weight), with an affinity to [³H]epibatidine of about 0.5 nM. This affinity corresponds to the value for the 34 nACh receptor of 0.3 nM found by Parker et al. (1998). However, in the same study, the affinity of acetylcholine and nicotine was determined to be 0.5 and 0.3 µM, respectively, which was far from our K_i estimates of 8 and 5 µM. Sheffield et al. (2000) suggest that the 3-containing nACh receptors in the mHb consist of more additional subunits than 4 because data indicate the possibility of both 4 and 5 inclusion in an 34(2) receptor. The affinities of these combinations for epibatidine, nicotine, and acetylcholine are at present not described, and it is therefore unknown whether one of the combinations is a likely candidate for the epibatidine receptor described here. Figure 2 documents the difference between the high affinity of 42 receptors and the much lower affinity of the [³H]epibatidine binding receptor in mHb for acetylcholine because 0.5 µM acetylcholine displaced all bound [³H]epibatidine except from the mHb, where some binding persisted even in the presence of 64 µM acetylcholine.

Another way of investigating whether the epibatidine receptor in the mHb is an 34 receptor was to measure, with autoradiographic methods, the affinity of nicotine when displacing [³H]epibatidine bound to ganglionic 34 nACh receptors. This was done on sections from superior cervical ganglion, and nicotine was found to displace [³H]epibatidine with an affinity (K_i value) of 1.3 µM, which is different from the 5 µM mentioned above, thus, making it further unlikely that the epibatidine receptor in the mHb is composed of 34 subunits alone.

Several compounds with known affinity to different nicotinic receptors were tested for their ability to displace [³H]epibatidine binding to the mHb. Our objective was to search for a compound with selectivity for the epibatidine receptor in the mHb, i.e., displacement of [³H]epibatidine binding to the mHb without displacement of the binding to other brain structures. A perfect compound with this capacity was not found, but different valuable indications were obtained (see Fig. 3 and Table 1). Among the compounds were some with high selectivity and high affinity for 7 and 42 nicotinic receptors, respectively. Their systematic chemical names and affinities to different nicotinic receptors are presented in Table 1. Neither type of compound was able to displace [³H]epibatidine binding to the mHb even at 10 µM; "7" compounds did not displace [³H]epibatidine at all, whereas "42" compounds readily displaced the [³H]epibatidine binding to all other brain structures. This apparent lack of 42 receptors in mHb is consistent with the finding that [³H]epibatidine binds to the mHb in mice in which the 4- or 2-nACh subunits have been knocked out (Zoli et al., 1998; Marubio et al., 1999).

Table 1 and Fig. 3 also show results obtained with two other substances, named by us mixed affinity compound and most selective compound. Both compounds have the same high affinity to the 34 receptor subtype, whereas their affinity to the 42 receptor differs; the mixed affinity compound has very high affinity, whereas the most selective compound has low affinity. Autoradiograms obtained with the mixed affinity compound (Fig. 3) showed that already at 0.1 µM the [³H]epibatidine binding to 42 receptors was displaced as expected. At 1 µM, the binding to the mHb faints, disappearing completely at 10 µM. Autoradiograms obtained with the most selective compound showed a completely different picture. Even at 10 µM, not all the [³H]epibatidine bound to 42 receptors in the thalamus and cortex had vanished, but neither had the binding to mHb. Because the two compounds in question have about the same affinity to 34 receptors (the most selective compound actually having slightly higher affinity; see Table 1), this result must be interpreted as showing that the [³H]epibatidine binding receptor in mHb most likely is not an 34 receptor. The partial displacement of the [³H]epibatidine bound to the mHb by compounds with high affinity for 34 may indicate that the epibatidine receptor contains 3 subunits because epibatidine binds to this subunit (Warpman et al., 1998). However, when seen in the light of the results discussed above, the receptor in the mHb most likely is composed of more nACh subunits than 34. A suggestion could be 325 or 345 receptors, which have been expressed in a cell line (Wang et al., 1998).

The strong analgesic effect of epibatidine has spurred interest in analgesia obtained via stimulation of nACh receptors, and the 42 receptor has been proposed to be responsible for this effect. The drug ABT-594 is a fairly selective 42 ligand and has analgesic properties in different pain models after systemic administration (Bannon et al., 1998). Also, an injection of minute amounts of ABT-594 into nucleus raphe magnus has an antinociceptive effect on rats in a hot-plate assay, probably by gating transmission of afferent nociceptive inputs from reaching higher centers. Following this idea, and because various lines of evidence suggest an involvement of the habenular complex in pain processing (Cohen and Melzack, 1993), a role for the habenular epibatidine receptors in signal gating might be suggested.

A rat model was developed in which the effect of minute amounts of epibatidine injected bilaterally into brain areas near the mHb could be studied. Figure 4A confirms that epibatidine injected near mHb labels the epibatidine receptors in the mHb. In preliminary experiments, amounts of epibatidine up to 500 fmol were injected bilaterally near the habenula to evaluate the toxicity of epibatidine after this form of administration. None of the doses tested seemed to affect the spontaneous behavior of the rats, e.g., no signs of sedation, hyperactivity, or motor disturbances were observed. Based on the affinity of epibatidine (0.5 nM at 4°C) for the epibatidine receptor in the mHb and the washout half-life (9 min) of [³H]epibatidine injected locally into the brain, it was decided to test the analgesic effect of about 10 fmol (1 µl, 10 nM) of epibatidine injected near the mHb. As seen in Fig. 5, 10 fmol of epibatidine indeed proved to have some analgesic effect, 10 and 20 min after injection, whereas neither 2 fmol of epibatidine nor 10 fmol of nicotine (which we considered would stimulate 42 receptors) had any effect.

Compounds with selectivity for 34 receptors are not likely to be useful analgesics because ganglionic nACh receptors primarily are of this subtype and, therefore, both agonists and antagonists would have profound physiological effects. Compounds with selectivity for 42 receptors have analgesic effects as shown by Bannon et al. (1998), but due to the widespread distribution of the 42 receptor in the CNS, it may be anticipated that other, maybe less desirable effects may appear, when stimulating this receptor. The moderate analgesic effect observed in our experiments is most likely due to stimulation of the epibatidine receptors in the mHb because the estimated concentration of epibatidine 10 and 20 min after an injection of 10 fmol is in the right concentration range compared with the receptor K_D of 0.5 nM found in the saturation measurements. The effect is not likely to be caused by stimulation of 42 receptors because injection of 10 fmol of nicotine was without analgesic effect. The epibatidine receptor in the mHb thus may represent a possibility for obtaining analgesia selectively. Due to the very limited distribution of the receptor in the CNS and probable absence from ganglia, it could be a pharmacologically interesting target. The development of a selective drug with the right balance of agonistic activity remains, however, a challenge for future research.