Pain is a mechanism of the host defense warning system transduced through small diameter nociceptors that signal on-going damage to the body. By its predominant expression in c-fibers, the vanilloid receptor, TRPV1, plays a key role in the detection of noxious painful stimuli such as acid (low pH), noxious heat, components of the “inflammatory soup,” vanilloids such as capsaicin (the pungent component of hot chili peppers), and resiniferatoxin (an alkaloid from Ephorbia resinifera) (Caterina et al., 1997; Szallasi and Blumberg, 1999; Holzer, 2004; Szolcsányi, 2004). Capsaicin and resiniferatoxin revealed the potential of TRPV1 as a target for pain and helped to identify and clone the TRPV1 channel. Although capsaicin and resiniferatoxin cause burning pain acutely, chronic treatment was shown to produce analgesia through ablation of TRPV1-expressing neurons. These results indicated the potential of TRPV1 agonists as pain therapeutics (reviewed in Bley, 2004).

Up-regulation of TRPV1 has been shown in inflammatory pain models in preclinical species (Ji et al., 2002) and human disease conditions such as inflammatory bowel disease (Matthews et al., 2004), painful bladder syndromes (Mukerji et al., 2006), vulvodynia (Tympanidis et al., 2004), and chronic persistent cough (Groneberg et al., 2004). In addition, TRPV1 knockout mice showed reduced thermal hypersensitivity after inflammatory tissue injury (Caterina et al., 2000; Davis et al., 2000) and attenuated experimental arthritis (Szabó et al., 2005; Barton et al., 2006). Finally, TRPV1 antagonists (that block capsaicin, proton, and heat activation) representing various chemotypes acted as antihyperalgesics in inflammatory (Pomonis et al., 2003; Valenzano et al., 2003; Walker et al., 2003; Gavva et al., 2005b; Honore et al., 2005; Rami et al., 2006; Gunthorpe et al., 2007) and surgical incision (Honore et al., 2005) pain models as well as analgesics in cancer pain (Ghilardi et al., 2005) models (reviewed in Immke and Gavva, 2006; Szallasi et al., 2007). Taken together, the above data generated strong interest in TRPV1 as a therapeutic target for pain.

Most of the published TRPV1 antagonists lack optimal properties for clinical development such as selectivity, solubility, oral bioavailability, and/or reasonable pharmacokinetics. After evaluation of TRPV1 antagonists representing various chemotypes (Doherty et al., 2005, 2007; Xi et al., 2005; Ognyanov et al., 2006; Norman et al., 2007; unpublished data) for potency, selectivity, plasma half-life, effects in the capsaicin-induced flinch model, and efficacy in models of inflammatory pain, we identified AMG 517 as a potential drug for the management of pain in humans. Here, we describe the characterization of AMG 517 and the data set leading to its selection as a clinical candidate.

Materials and Methods

All animal use and husbandry were in accordance with the United States Department of Agriculture Animal Welfare Act (9 CFR, Parts 1, 2, and 3) and the Institute for Laboratory Animal Research publication, 1996). The cDNA sequences used in this study are identical or similar to AJ277028 (human TRPV1), NP_114188 (rat TRPV1), Q704Y3 (mouse TRPV1), XP_001117609 (rhesus TRPV1), EF100779 (cynomolgus TRPV1), Q9WUD2 (rat TRPV2), NP659505 (human TRPV3), NP067638 (human TRPV4), O75762 (human TRPA1), and NP_076985 (human TRPM8).

Agonist-Induced ⁴⁵Ca²⁺Uptake Assay. Two days before the assay, cells were seeded in Cytostar 96-well plates (GE Healthcare, Piscataway, NJ) at a density of 20,000 cells/well. The activation of TRPV1 and TRPV2 was followed as a function of cellular uptake of radioactive calcium (⁴⁵Ca²⁺; MP Biomedicals Inc., Costa Mesa, CA). Capsaicin (0.5 μM), pH 5, and heat (45°C) were used as agonists for TRPV1, and 2-APB (200 μM) was used as the agonist for TRPV2. All of the antagonist ⁴⁵Ca²⁺ uptake assays were conducted as reported previously (Gavva et al., 2005b) and had a final ⁴⁵Ca²⁺ concentration of 10 μCi/ml. Radioactivity was measured using a MicroBeta Jet (PerkinElmer Life and Analytical Sciences, Waltham, MA). Data were analyzed using GraphPad Prism 4.01 (GraphPad Software Inc., San Diego, CA).

Luminescence Readout Assay for Measuring Intracellular Calcium. Stable CHO cell lines expressing human TRPA1, TRPM8, TRPV3, and TRPV4 were generated using tetracycline inducible T-REx expression system from Invitrogen (Carlsbad, CA). To enable a luminescence readout based on the intracellular increase in calcium (Le Poul et al., 2002), each cell line was also cotransfected with pcDNA3.1 plasmid-containing jellyfish aequorin cDNA. Twenty-four hours before the assay, cells were seeded in 96-well plates and TRP channel expression was induced with 0.5 μg/ml tetracycline. On the day of the assay, culture media were removed, and cells were incubated with assay buffer (F12 containing 30 mM HEPES for TRPA1, TRPM8, and TRPV3; F12 containing 30 mM HEPES, 1 mM CaCl₂, and 0.3% bovine serum albumin for TRPV4) containing 15 μM coelenterazine (PJK GmbH, Kleinblittersdorf, Germany) for 2 h. Antagonists were added for 2.5 min before addition of an agonist. The luminescence was measured by a charge-coupled device camera-based Flash-luminometer built by Amgen (Thousand Oaks, CA). After agonists were used to activate TRP channels: 80 μM allyl isothiocyanate for TRPA1, 1 μM icilin for TRPM8, 200 μM 2-APB for TRPV3, and 1 μM4α-PDD for TRPV4. Compound activity was calculated using GraphPad Prism 4.01.

Capsaicin-Induced Flinch Model. Male Sprague-Dawley rats (Charles River Laboratories, Wilmington, MA) weighing 190 to 220 g were allowed at least 3 days of acclimation in Amgen's Association for Assessment and Accreditation of Laboratory Animal Care approved-animal care facility before the start of the experiment. Animals (eight per group) were pretreated with vehicle (a dose volume of 5 ml/kg in an Ora-Plus-5% Tween 80) or TRPV1 antagonists (by oral gavage), 2 h before intraplantar injection of 0.5 mg of capsaicin in a volume of 25 μl of 5% ethanol in phosphate-buffered saline without Ca²⁺ and Mg²⁺ (Sigma-Aldrich, St. Louis, MO). Immediately after the injection of capsaicin, the number of flinches was recorded over a 1-min period by an individual fully blinded to treatment conditions. Blood samples were collected immediately after behavioral testing for pharmacokinetic analyses.

Radiotelemetry in Rats. Male Sprague-Dawley rats (CRL, Wilmington, MA) weighing 17 to 250 g (6–8 weeks of age) were single-housed and allowed at least a 1-week acclimation period before the start of the experiment. The temperature in the room used for the experiments was maintained at 20 ± 2°C. Rats were implanted with a radiotelemetry probe (model ER-4000 PDT; Mini Mitter, Brend, OR) and allowed at least 3 days of recovery before the drug experiment. On the day of the experiment, single-housed animals were placed on radiotelemetry receivers. Baseline body core temperature was recorded for 30 min before treatment. Rats (eight per group) were treated with vehicle (a dose volume of 5 ml/kg in an Ora-Plus-5% Tween 80) or TRPV1 antagonists (by oral gavage) with continuation of body core temperature recording for 2 h after drug administration. Blood samples were collected immediately after behavioral testing for pharmacokinetic analyses.

CFA-Induced Thermal Hyperalgesia. After multiple days of full habituation to the testing equipment and paradigm, CFA-induced thermal hyperalgesia was evaluated by measuring paw withdrawal latencies in male Sprague-Dawley rats. Twenty-one hours after CFA injection (50 μl of 0.1%), animals were dosed (p.o.) with AMG 517 or AMG8163 at a dose range of 0.001 to 30 mg/kg in a volume of 5 ml/kg. Two hours after drug dosing (23 h after CFA injection), paw withdrawal latencies were measured using modified Hargreaves hot boxes (University of California, San Diego, CA) by investigators fully blinded to treatment conditions.

Body Temperature Evaluation in Radiotelemetred Cynomolgus Monkeys Treated with Single Doses of AMG 517. Six experimentally non-naive cynomolgus monkeys (three males and three females) weighing 2.6 to 4.8 kg were used on the study. Animals had surgically implanted radiotelemetry devices (model TL11M2-D70-PCT; Data Sciences International, St. Paul, MN) and were housed individually in stainless steel cages.

Groups 1 and 2 were treated with a single dose of vehicle (10% Pluronic F108 in Ora-Plus) via nasogastric gavage and observed for 7 days postdose. Subsequently, group 1 was treated with a single dose of 10 mg/kg AMG 517, and group 2 was treated with a single dose of 30 mg/kg AMG 517; both groups were observed for 7 days postdose. Individual body temperatures were monitored by telemetry using the Data Sciences computer system (Dataquest ART Gold data acquisition and analysis software). Body temperature data were recorded from each animal for 30 s at 5-min intervals before, during, and after dosing beginning approximately 60 min before the initiation of dosing and continuing through 72 h postdose and for 30 s at hourly intervals from 72 h through 165 h postdose. Mean body temperature across each 30-s recording period was calculated by the Dataquest software.

For pharmacokinetic analyses, blood samples were collected from all animals at predose and approximately 1, 2, 4, 24, 72, and 168 h after dosing. AMG 517 levels were quantitated in lithium-heparinized plasma samples using a 96-well protein precipitation with a reversed-phase liquid chromatography-tandem mass spectrometry analytical procedure. Individual plasma concentration-time data were analyzed by noncompartmental methods using WinNonlin (Pharsight Corporation, Mountain View, CA). After the last data collection, all animals were returned to the colony.

Body Temperature Evaluation in Naive Cynomolgus Monkeys Treated with Repeated Daily Doses of AMG 517. Cynomolgus monkeys (six per sex per group), weighing 2.0 to 3.4 kg, were treated by nasogastric gavage with 0 (vehicle control; 10% Pluronic F108 in Ora-Plus), 30, 100, or 500 mg/kg/day AMG 517 for 28 consecutive days (5 ml/kg/day). Rectal body temperatures were collected before dosing and at 3 and 24 h after dosing on days 1, 8, 15, 22, and 28 of treatment. Mean body temperatures at each time point were compared across groups with a one-way analysis of variance (ANOVA) using p ≤ 0.05 as the level of significance. If the ANOVA was significant, Dunnett's multiple comparison test was used to identify statistically significant differences between each drug-treated group and the vehicle-treated control group (at p ≤ 0.05).

Results

Characterization of AMG 517 as a TRPV1 Antagonist. In our efforts to identify a TRPV1 antagonist clinical candidate, we synthesized several novel series of molecules representing various chemotypes ((Doherty et al., 2005, 2007; Xi et al., 2005; Ognyanov et al., 2006; Norman et al., 2007; unpublished data) and evaluated their ability to block various modes of TRPV1 activation. AMG 517 inhibited capsaicin, pH 5, and heat-induced ⁴⁵Ca²⁺ uptake into cells expressing TRPV1 with IC₅₀ values of 1 to 2 nM (Table 1, Fig. 1A). AMG 517 blocked capsaicin-, proton-, and heat-induced inward currents in TRPV1-expressing cells similarly (data not shown). In addition, AMG 517 acted as a potent antagonist of all of the endogenous ligands tested (Table 1). AMG 517 inhibited native TRPV1 activation by capsaicin in rat dorsal root ganglion neurons with an IC₅₀ value of 0.68 ± 0.2 nM. AMG 517 is a competitive antagonist of both rat and human TRPV1 with dissociation constant (K_b) values of 4.2 and 6.2 nM, respectively (Fig. 1, B–E). AMG 517 did not activate TRPV1 at concentrations up to 40 μM, as measured by ⁴⁵Ca²⁺ uptake into TRPV1-expressing cells, indicating that it was not a partial agonist. Pharmacology of another TRPV1 antagonist, AMG8163 (a “tert-butyloxycarbonyl” analog of AMG 517), used in this study was reported recently (Magal, 2005; Gavva et al., 2007).

AMG 517 was found to be selective for TRPV1 among the recombinant TRP family members that we have tested (Fig. 2, A–E). The IC₅₀ value for AMG 517 was >20 μM against 2-APB-activated TRPV2 and TRPV3, 4-αPDD-activated TRPV4, allyl isothiocyanate-activated TRPA1, and icilin-activated TRPM8 in cell-based assays that measure agonist-induced increases in intracellular calcium in CHO cells recombinantly expressing the appropriate TRP channel. In a selectivity screen conducted by NovaScreen, 10 μM concentration of AMG 517 showed significant binding (≥45% inhibition) to only 3 of 90 evaluated targets that included receptors, enzymes, and ion channels (see Supplementary Material). AMG 517 showed binding to peripheral monoamine oxidase B (88%) and site 2 of the sodium ion channel (47%) and showed 79% inhibition of anandamide transport in a functional assay. IC₅₀ values of 3.4, 11.3, and 19.4 μM were subsequently determined for the monoamine oxidase B (peripheral), anandamide transporter, and sodium channel (site 2) targets, respectively.

Pharmacokinetic analyses showed that plasma half-life values of AMG 517 were 31, 41, and 62 h in rats, dogs, and monkeys, respectively. Oral bioavailability was 47 to 51% in rats, 23 to 29% in dogs, and 42 to 52% in monkeys (Doherty et al., 2007).

TRPV1 Antagonists Block Capsaicin-Induced Flinch in Rats. Because the pharmacokinetic properties of AMG 517 and its closely related analog, AMG8163, were amenable to in vivo studies, they were evaluated in a rat capsaicin-induced flinch model as a measure of TRPV1 antagonism (Seabrook et al., 2002). Although oral administration of both AMG 517 and AMG8163 produced a dose-dependent increase in plasma concentrations, they also produced a dose-dependent decrease in the number of flinches induced by capsaicin treatment (Fig. 3, A and B) (Doherty et al., 2007; Gavva et al., 2007). The minimally effective dose (MED), based on a statistically significant difference in number of flinches from the vehicle versus capsaicin-administered group, was 0.3 mg/kg (p < 0.05) for AMG 517 and 0.03 mg/kg (p < 0.05) for AMG8163. The corresponding plasma concentrations were 90 to 100 ng/ml for AMG 517 and approximately 10 ng/ml for AMG8163. Correlating with its long plasma half-life in rats, the time course of AMG 517 (3 mg/kg) exhibited significant reductions in capsaicin-induced flinch up to 24 h after dosing (Fig. 3C). At these time points, plasma levels of AMG 517 were ≥300 ng/ml. There was no difference in flinch response between the vehicle group and groups treated with AMG 517 by 48 h.

To evaluate the effect of repeated dosing with TRPV1 antagonists on capsaicin-induced flinch, 3 mg/kg AMG8163 was administered twice (p.o.) daily for 4 days. Complete inhibition of the capsaicin-induced flinch response on days 2, 3, and 4 indicated continuous antagonism of TRPV1 by AMG8163 in vivo (Fig. 3D).

TRPV1 Antagonists Act as Antihyperalgesics in Rat Models of Inflammatory Pain and Cause Hyperthermia in Rats. Next, we evaluated the ability of AMG 517 and AMG8163 to block pain behavior (thermal hyperalgesia) under inflammatory conditions in rats. Intraplantar injections of CFA produced a dose-dependent thermal hyperalgesia indicated by a decrease in paw withdrawal latency in rats, 24 h postinjection. TRPV1 antagonists such as AMG9810, A-425619, and N-(4-tertiarybutylphenyl)-4-(3-cholorphyridin-2-yl)tetrahydropyrazine-1(2H)-carboxy-amide were shown to block hyperalgesia induced by CFA (Pomonis et al., 2003; Gavva et al., 2005b; Honore et al., 2005). As expected, oral administration of both AMG 517 and AMG8163 reversed established thermal hyperalgesia in a dose-dependent manner at 21 h after CFA injection (Fig. 4, A and B) (Doherty et al., 2007). Associated plasma concentrations increased in a dose-dependent manner for both drugs up to 10 mg/kg. The MEDs were 1 mg/kg (p.o.) for AMG 517 and 0.3 mg/kg (p.o.) for AMG8163. The minimally effective concentration (MEC) was approximately 300 ng/ml in plasma for AMG 517 (with a maximal reversal of ∼40%) and 10 to 30 ng/ml in plasma for AMG8163 (with a maximal reversal of ∼50%). In the same experiment, the positive control, ibuprofen (30 mg/kg p.o.), produced a comparable (∼55%) reversal of CFA-induced thermal hyperalgesia. In addition, TRPV1 antagonists showed efficacy in carrageenan-induced (42% reversal by 3 mg/kg dose of AMG 517) and surgical incision-induced (48% reversal by 3 mg/kg dose of AMG8163) thermal hyperalgesia (Supplemental Fig. 1). Because most TRPV1 antagonists reported to date, including AMG 517 and AMG8163, showed only 40 to 50% block of hyperalgesia in rodent models of pain (reviewed in Immke and Gavva, 2006), we concluded that TRPV1 partially contributes to inflammation- and incision-induced hyperalgesia.

We and others have previously shown that TRPV1 antagonists cause transient hyperthermia in multiple species (rats, dogs, and cynomolgus monkeys). These observations suggested that TRPV1 is tonically activated in naive rats, dogs, and primates and that tonic TRPV1 activation regulates body temperature (Bannon, 2004; Swanson et al., 2005; Gavva et al., 2007). As expected, AMG 517 caused transient hyperthermia in rodents, dogs, and monkeys (Fig. 4, C and D) (data not shown). AMG 517 induced hyperthermia in a steep dose-dependent manner, with 0.3, 1, and 3 mg/kg associated with 0.5, 0.6, and 1.6°C increases in body temperature, respectively. Body temperatures of rats treated with all doses of AMG 517 returned to baseline within 10 to 20 h. In rats, body temperatures in the vehicle group varied from 37.4 ± 0.05°C during the light cycle to 37.8 ± 0.05°C during the dark cycle, suggesting that body temperature is a dynamically regulated process. In monkeys, single doses of AMG 517 at 10 and 30 mg/kg caused an approximate 1 to 1.5°C increase in body temperature that lasted beyond 24 h. Mean temperatures in both drug-treated groups differed from those of the vehicle treated group in a statistically significant manner (ANOVA/Dunnett's) at the majority of evaluated postdose time points (Fig. 4D).

Toxicological Evaluation of AMG 517. To support initial clinical investigations, AMG 517 was evaluated in a package of nonclinical toxicology studies that were designed in accordance with relevant regulatory guidelines [ICH (2000) Topic M3 (R1): Nonclinical safety studies for the conduct of human clinical trials for pharmaceuticals, 2000, http://www.fda.gov/cder/guidance/1855fnl.pdf]. Pivotal toxicology studies were conducted in compliance with global good laboratory practice regulations, including those defined by the U.S. Food and Drug Administration (21 CFR Part 58). Individual studies were designed in consideration of relevant regulatory guidelines (refer to the following cited (selected) guidelines for detail). The package included single-dose studies in mice and rats [U.S. Food and Drug Administration (1996) Guidance for industry: single dose acute toxicity testing for pharmaceuticals, http://www.fda.gov/cder/guidance/pt1.pdf], in vitro and in vivo safety pharmacology studies [ICH (2001) Topic S7A: Safety pharmacology studies for human pharmaceuticals, http://www.fda.gov/cder/guidance/4461fnl.pdf], repeated dose studies in rats, monkeys, and dogs [ICH (1995) Topic S3A: Toxicokinetics: a guidance for assessing systemic exposure in toxicology studies, http://www.emea.europa.eu/pdfs/human/ich/038495en.pdf; and European Agency for the Evaluation of Medicinal Products-Committee for Proprietary Medicinal Products (2000) Note for guidance on repeated dose toxicity, http://www.emea.europa.eu/pdfs/human/swp/104299en.pdf], and a standard battery of in vitro and in vivo genetic toxicity studies [ICH (1997) Topic S2B: Genotoxicity: a standard battery for genotoxicity testing of pharmaceuticals, http://www.fda.gov/cder/guidance/1856fnl.pdf]. AMG 517 was administered p.o. in all in vivo studies. Definitive 28-day repeated-dose studies evaluated daily oral doses up to 500 mg/kg/day in rats and monkeys, producing end-of-treatment C_max levels of 13,500 and 9160 ng/ml, respectively, at 500 mg/kg/day. Notably, no AMG 517-related adverse effects were observed in any of the above-listed toxicology studies. However, transient drug-related body temperature increases (see below) were noted in mice, rats, dogs, and monkeys.

TRPV1 Antagonist-Induced Hyperthermia Attenuates after Repeated Dosing. To evaluate the effects of repeated dosing on TRPV1 antagonist-induced hyperthermia, we administered AMG8163 twice daily for 4 days to rats implanted with radiotelemetry probes and monitored their body temperature (Fig. 5A). Because body temperature measurements were gathered by continuous monitoring we compared the effect of AMG8163 on subsequent days by averaging the temperature from each treatment group for each 24 h period and ran four separate t tests comparing the AMG81630-treated group to the vehicle-treated group. AMG8163 induced a 0.5 to 0.8°C increase in body temperature on the first day of treatment that was not apparent by day 4 (Fig. 5, A and B), suggesting that TRPV1 antagonist-induced hyperthermia attenuates with repeat dosing. Statistically significant differences (2₁₁ = 2.9; p < 0.05) were only apparent on day 1, with average temperature in the vehicle treated group of 37.6 ± 0.12°C and average temperature in the AMG8163-treated group of 38.1 ± 0.07°C (Fig. 5B).

AMG 517 produced an approximate 1°C increase in body temperature in both rats and monkeys (Fig. 4, C and D) (data not shown). When administered as a single dose, the post-treatment temperature increase was transient, peaking within a few hours and lasting up to 20 h. The AMG 517-induced hyperthermic response diminished with repeat dosing in rats (data not shown), dogs (data not shown), and monkeys (Fig. 5C), as seen with AMG8163 in rats (Fig. 5, A and B). In monkeys treated daily for 28 consecutive days, transient increases (up to approximately 1.1°C greater than controls) in mean body temperature, relative to predose, were observed on day 1 in all drug-treated groups (30, 100, and 500 mg/kg/day), with mean temperatures at 3 h postdose statistically significantly (p < 0.05) higher than those in the control group. After continued repeated dosing, clear drug-related patterns in an effect on body temperature (on days 8, 15, 22, and 28 of the treatment period) were not readily apparent (no other statistically significant differences from controls were observed). This observation suggests that a physiological tolerance develops with repeat dosing in rodents, canines, and nonhuman primates.

TRPV1 Antagonist-Induced Hyperthermia Can Be Suppressed by the Antipyretic, Acetaminophen. Because we predicted that TRPV1 antagonists might produce hyperthermia in humans, we tested whether anti-pyretic acetaminophen could suppress TRPV1 antagonist-induced hyperthermia. We administered acetaminophen, before and after AMG8163 administration, to rats surgically implanted with radiotelemetry probes. Acetaminophen (300 mg/kg) suppressed the established TRPV1 antagonist-induced hyperthermia, as evidenced by a drop of body temperature from 38.5 to 37°C (Fig. 6, A and B). During baseline telemetry recordings, there was no significant difference among groups (F_3,8 = 3.5, p > 0.05) with an average temperature across groups of 36.8 ± 0.03°C. AMG8163-treated rats showed an average temperature of 38.0 ± 0.02°C, whereas the average temperature in vehicle-treated rats was 37.2 ± 0.03°C, during the 30- to 120-min postdosing phase (t₂₉₈ = 21.2, p < 0.01). During the 30 to 60 min after acetaminophen administration, average body temperature dropped to 36.4 ± 0.04°C, whereas the vehicle-treated group showed an average temperature of 37. 2 ± 0.03°C (t₁₄₈ = 18.3; p < 0.01) (Fig. 6B). Acetaminophen reduced the AMG8163-induced body temperature to 37.3 ± 0.05°C, whereas rats treated with AMG8163 alone had an average temperature of 37.8 ± 0.03°C (t₁₄₇ = 9.1; p < 0.01). This average temperature in acetaminophen/AMG8163-treated rats is within 0.1°C of the average temperature of the vehicle-treated group (not statistically significant; t₁₄₈ = 1.5; p > 0.05), but approximately 0.5°C greater than the average temperature in the vehicle/acetaminophen-treated group (t₁₅₈ = 15.4; p < 0.01) (Fig. 6B). Low doses of acetaminophen (100 or 150 mg/kg) that did not drop body temperature by themselves were not effective in reversing AMG8163-induced hyperthermia (data not shown). The reversal by acetaminophen (Fig. 6, A and B) of AMG8163-induced hyperthermia suggests that TRPV1 antagonist-induced hyperthermia could potentially be managed in humans with antipyretics, if required.

Discussion

In the present study, we characterized AMG 517 as a potent TRPV1 antagonist and compared its pharmacology to that of a closely related analog, AMG8163, in various in vitro and in vivo assays. Both molecules are potent antagonists of TRPV1 activation by chemical ligands (capsaicin, proposed endogenous agonists, and protons) and heat. Because both AMG 517 and AMG8163 bind the same pocket as capsaicin and block all modes of TRPV1 activation, they represent “group A” antagonists (Gavva et al., 2005a) (data not shown) that lock the channel conformation in a “closed state.” We demonstrated the ability of AMG 517 and AMG8163 to block TRPV1 in vivo by showing antagonism of capsaicin-induced flinch in rats and their antihyperalgesic properties through reduction of thermal hyperalgesia in inflammatory pain models (Doherty et al., 2007; this study). Based on its pharmacological activity, favorable pharmacokinetic properties (Doherty et al., 2007), and favorable nonclinical safety profile in toxicology studies, AMG 517 was chosen for clinical development.

We and others previously showed that TRPV1 antagonists that act as antihyperalgesics can cause transient hyperthermia in multiple species (Bannon, 2004; Swanson et al., 2005; Gavva et al., 2007). As expected, AMG 517 caused hyperthermia in rodents, dogs, and monkeys. Importantly, as little as 0.3 mg/kg AMG 517 (MED) or its associated 100 ng/ml plasma concentration (MEC), induced transient hyperthermia. Notably, this hyperthermia-inducing MEC/MED is comparable to the MEC/MED required to demonstrate on-target coverage in the capsaicin-induced flinch model. The MEC/MED required for antihyperalgesia in the CFA-induced pain model is higher for both antagonists compared with the MEC/MED for causing hyperthermia. This finding suggests that antagonism of a small pool of tonic TRPV1 activation is sufficient to trigger hyperthermia, whereas antagonism of a large pool of TRPV1 is required for antihyperalgesia in models of inflammation. Because AMG 517 is a selective TRPV1 antagonist (IC₅₀ against closely related TRP channels is >5000-fold), the hyperthermia and antihyperalgesia were believed to be caused by TRPV1 antagonism alone. Both AMG 517 and AMG8163 are brain-penetrant compounds with brain/plasma ratios of 1.01 and 0.58, respectively. As suggested recently (Cui et al., 2006), the site of action for antihyperalgesia by TRPV1 antagonists may be present in the central nervous system, thus requiring higher brain concentrations of antagonists for significant effects.

Although TRPV1 agonists such as capsaicin and resiniferatoxin cause a drop in body temperature (hypothermia), until recently it was not known whether TRPV1 is tonically activated in vivo or even if it is involved in body temperature regulation. We recently demonstrated that TRPV1 is tonically activated in vivo on the basis of the fact that antagonists representing a wide variety of chemotypes that block TRPV1 activation by capsaicin, heat, and protons (as well as those that block only capsaicin and heat activation), caused transient hyperthermia through a site of action outside of the blood-brain barrier (Gavva et al., 2007). Selective antagonists of TRPV1 that cause hyperthermia in multiple species [AMG9810, AMG8163, and several others in rats; JYL1421 in dogs and monkeys (Gavva et al., 2007); or JNJ-17203212 in rats (Swanson et al., 2005) and AMG 517 in multiple species (this study)] further demonstrate that TRPV1 is tonically activated in vivo and is involved in body temperature regulation. The fact that AMG 517 caused transient hyperthermia in rats, dogs, and monkeys, further confirms that the involvement of TRPV1 in thermoregulation is conserved across rodents to primates.

The mechanisms of TRPV1 agonist-induced hypothermia include skin vasodilation and reduction in metabolic heat production [reviewed in (Hori, 1984)], whereas the mechanisms of TRPV1 blockade elicited hyperthermia include skin vasoconstriction that reduces heat loss and an increase in metabolic heat production as measured by increased oxygen consumption in rats (Steiner et al., 2007). In addition, agonist-induced hypothermia and antagonist-induced hyperthermia are absent in TRPV1 knockout mice (Caterina et al., 2000; Steiner et al., 2007), suggesting that the effects of agonists and antagonists on body temperature are exclusively mediated by TRPV1 channels. Even though TRPV1 is activated by heat, it is not known whether TRPV1 itself acts as a thermosensor for the body. It seems that TRPV1 blockade does not affect thermosensation because the thermal preference of TRPV1 antagonist-administered rats was not altered in a thermotaxis gradient assay (Steiner et al., 2007), suggesting that TRPV1 may not act as a thermosensor in vivo. Interestingly, TRPV3 knockout mice display strong deficits in responses to innocuous and noxious heat in thermotaxis assays, suggesting that TRPV3 acts as a thermosensor (Moqrich et al., 2005; Dhaka et al., 2006).

Because body temperature maintenance is a tightly regulated physiological process, which may have multiple mechanisms to bring the body temperature toward normal range in the event of a significant change (a drop or an increase), we evaluated whether hyperthermia persists with continuous blockade of TRPV1 by repeatedly administering TRPV1 antagonists (to maintain a constant plasma concentration). Because the hyperthermia elicited by TRPV1 blockade attenuated by day 2 to 4 of repeated administration of antagonists, we hypothesize that other mechanisms suppressed TRPV1 blockade elicited hyperthermia. Both AMG 517 and AMG8163 blocked capsaicin-induced flinch response in rats during this period, demonstrating continuous blockade of TRPV1 in vivo. We believe that other mechanisms may have compensated for the role of TRPV1 in body temperature maintenance during days 2 through 4. Similar attenuation of TRPV1 blockade elicited hyperthermia after repeated administration of another TRPV1 antagonist was recently reported (Cortright, 2006)).

Because TRPV1 antagonist-induced hyperthermia is an undesirable on-target effect, we evaluated the potential utility of antipyretics in managing such hyperthermia. We found that the antipyretic, acetaminophen, blocked development of hyperthermia when given before TRPV1 antagonists and reversed ongoing hyperthermia when given after TRPV1 antagonists. This finding suggests that antipyretics might be effective clinically should TRPV1 antagonists cause hyperthermia in humans. The antipyretic reduction of hyperthermia is not mediated by a direct modulation of TRPV1, because acetaminophen neither blocks nor activates the TRPV1 channel (unpublished data). It should be noted here that doses of acetaminophen (100 or 150 mg/kg) that do not drop body temperature by themselves did not block or reverse the TRPV1 antagonist-induced hyperthermia. It was reported that TRPV1 antagonists cause the same magnitude of hyperthermia in normal rats as well as in rats with ongoing inflammation, suggesting that inflammation-induced fever and TRPV1 antagonist-induced hyperthermia act only additively (Cortright, 2006). Consequently, antagonist-induced hyperthermia may pose an issue for clinical development of TRPV1 antagonists as therapeutic agents in human subjects prone to infections if temperature increases of 2–3°C result.

Among the results presented here and elsewhere (Cortright, 2006), the most promising is the apparent attenuation of antagonist-induced hyperthermia with repeated dosing, suggesting that TRPV1 blockade-elicited hyperthermia may be overcome. Because peripherally restricted TRPV1 antagonists cause hyperthermia (Gavva et al., 2007) and central nervous system-penetrant antagonists seem to be more effective antihyperalgesics than peripherally restricted compounds (Cui et al., 2006), the best option seems to be the repeated dosing-mediated attenuation of TRPV1 antagonist-induced hyperthermia. Hence, carefully planned repeated dosing paradigms of TRPV1 antagonists in humans to attenuate hyperthermia will be a critical and necessary step for developing them as therapeutic agents. Still to be determined are 1) whether TRPV1 antagonists will be efficacious in human disease, 2) whether TRPV1 antagonists affect human body core temperature, and 3) whether TRPV1 blockadeelicited hyperthermia will be more pronounced during infections or injuries. The impact of TRPV1 antagonist-induced hyperthermia on their clinical utility is still unclear.