Activation of peripheral nociceptors in humans by capsaicin results in burning pain (Park et al., 1995). Capsaicin, and its ultrapotent analog resiniferatoxin, aided the identification and characterization of the vanilloid receptor 1 (aka VR1 and TRPV1). TRPV1 is a nonselective cation channel with high permeability to calcium (Caterina et al., 1997) and belongs to a superfamily of ion channels known as the transient receptor potential channels or TRPs (Clapham et al., 2001). In addition to activation by exogenous agonists such as capsaicin and resiniferatoxin, TRPV1 can be activated by physical stimuli, such as heat (>42°C) and protons (pH 5). Based on their structural similarity to capsaicin, several endogenous ligands have been proposed that include anandamide (AEA), 12-hydroperoxy-5,8,10,14-eicosatetraenoic acid,N-arachidonyl dopamine (NADA), N-oleoyldopamine (OLDA), and products of lipoxygenases (Hwang et al., 2000; Olah et al., 2001; Huang et al., 2002; Chu et al., 2003). TRPV1 is up-regulated during inflammation (Ji et al., 2002), and channel activity is modulated by the action of inflammatory mediators, such as prostaglandins, and bradykinin. These mediators act presumably via protein kinase C- or protein kinase A-mediated TRPV1 phosphorylation (Premkumar and Ahern, 2000; Cortright and Szallasi, 2004). Disruption of the TRPV1 gene in mice causes reduction in both thermal nociception and inflammation-induced hyperalgesia (Caterina et al., 2000, Davis et al., 2000). Together, these data suggest a role for TRPV1 as an integrator of multiple pain-producing stimuli. Consequently, TRPV1 may represent a novel target for the discovery of analgesics.

Early research on structure-activity relationships around exogenous activators led to the identification of several antagonists. First, the capsaicin analog capsazepine was synthesized and characterized extensively (Bevan et al., 1992; Walpole et al., 1994; Walker et al., 2003). As a research tool, capsazepine has provided interesting insights into the pharmacology of TRPV1. For example, capsazepine blocks both human and guinea pig, but not rat, TRPV1 activation by protons (McIntyre et al., 2001; Savidge et al., 2002). Also, capsazepine produced significant antihyperalgesic effects in a model of inflammatory and nerve injury-related pain in guinea pigs, but not in rats or mice (Walker et al., 2003). These species differences indicate that in vivo efficacy correlates with capsazepine potency for blocking proton activation of TRPV1. In addition, simplified analogs of resiniferatoxin (Lee et al., 1999), and iodinated analogs of resiniferatoxin (I-RTX; Wahl et al., 2001; Seabrook et al., 2002) and capsaicin (6-iodo-nordihydrocapsaicin; Appendino et al., 2003) have also been reported to be antagonists of TRPV1.

Recently, several novel small molecule antagonists have also been reported that prevent activation of the native and heterologous TRPV1 channels by known activators (for review, see Szallasi and Appendino, 2004). These include N-(3-acyloxy-2-benzylpropyl)-N′-[4-(methylsulfonylamino)benzyl] thiourea analogs (Lee et al., 2003), BCTC (Valenzano et al., 2003), and SB-366791 (Gunthorpe et al., 2004). Efficacy in models of pain has only been investigated with a few of the TRPV1 antagonists, such as BCTC, I-RTX, and two of the thiourea analogs. BCTC produced antihyperalgesic effects in models of inflammatory and neuropathic pain (Pomonis et al., 2003), whereas I-RTX treatment inhibited capsaicin-induced nociceptive responses (Wahl et al., 2001). Two of the N-(3-acyloxy-2-benzylpropyl)-N′-[4-(methylsulfonylamino)-benzyl] thiourea analogs inhibited acetic acid-induced writhing in mice (Lee et al., 2003).

Here, we report the in vitro and in vivo characterization of a novel, potent, and competitive TRPV1 antagonist, AMG 9810 (Fig. 1). AMG 9810 inhibits capsaicin-, proton-, heat-, and endogenous ligand-induced uptake of ⁴⁵Ca²⁺ into TRPV1-expressing cells. Furthermore, AMG 9810 blocks capsaicin-induced calcitonin gene-related peptide (CGRP) release and capsaicin-evoked depolarization in rat dorsal root ganglia neurons. Finally, we demonstrate that AMG 9810 is the first cinnamide to prevent capsaicin-induced eye wiping and reverse hyperalgesia in an animal model of inflammatory pain.

Materials and Methods

In Vitro Pharmacology

Stable Transfections. CHO cells stably expressing different TRPV1 channels were generated by transfecting pcDNA3.1 expression vector containing rat, human, or rat/human (r/h) TRPV1 chimera cDNAs. Rat/human (r/h) chimeric TRPV1 was constructed by replacing the human TRPV1 coding region for 1 to 216 amino acids with rat TRPV1 to determine the role of N terminus in pharmacological differences between antagonists. CHO cells were maintained in Dulbecco's modified Eagle's medium supplemented with 10% dialyzed fetal bovine serum, penicillin, streptomycin, l-glutamine, and nonessential amino acids. A calcium phosphate method was used for stable transfection (5 μg of DNA per 2 × 10⁶ cells in 60-mm dishes) and 800 μg/ml Geneticin was used as a selection agent. After 2 weeks of selection, single colonies were picked and screened for expression of TRPV1 in the agonist-induced ⁴⁵Ca²⁺ uptake assay. Positive clones were expanded and used in studies described here.

Antagonist Assays with CHO Cells Expressing TRPV1. Antagonist assays used agonist-induced ⁴⁵Ca²⁺ uptake as a readout and hence represent pharmacology at the plasma membrane-expressed TRPV1. Two days before the assay, CHO cells expressing TRPV1 were seeded in Cytostar 96-well plates (Amersham Biosciences Inc., Piscataway, NJ) at a density of 20,000 cells/well. The activation of TRPV1 is followed as a function of cellular uptake of radioactive calcium (⁴⁵Ca²⁺; MP Biomedicals, Irvine, CA). All ⁴⁵Ca²⁺ uptake assays had a final ⁴⁵Ca²⁺ concentration of 10 μCi/ml. Stock solutions of agonists and antagonists were typically made in 100% DMSO and diluted to appropriate final concentrations in assay buffers, keeping the final DMSO concentration <0.5%. Antagonist assays using CHO-TRPV1 cells for activation by capsaicin, proton, and heat were conducted as described in Gavva et al. (2004).

Electrophysiology. Whole-cell membrane currents were recorded with the whole-cell patch-clamp technique (Hamill et al., 1981) using the PatchExpress 7000A (Axon Instruments, Union City, CA). Experimental conditions were optimized for the CHO cells expressing rat TRPV1, so that only cells that achieved a gigaohm seal formation were analyzed. The internal solution contained: 70 mM KCl, 70 mM KF, 10 mM HEPES, 5 mM N-hydroxy-EDTA, pH 7.3. The external solution contained 140 mM NaCl, 10 mM HEPES, 10 mM glucose, 5 mM MgCl₂, and 5 mM EDTA, pH 7.3. The capsaicin dose-response was performed in both ascending and descending concentration ranges where no significant differences were found (all of the data were averaged). Also, as a control for run-down of the capsaicin-evoked responses, repeated 10 μM capsaicin applications were given for 10 s every minute for 10 min. Less than 10% of the total current diminished throughout the time course of the experiment (data not shown), ruling out any potential effects of run-down. To normalize the current evoked in the presence of AMG 9810, a 10-s prepulse of 10 μM capsaicin was given followed by a 2-min preincubation of AMG 9810 alone.

Rat E19 Embryonic Dorsal Root Ganglia Neuronal Cultures. Dorsal root ganglia were dissected under aseptic conditions from all spinal segments of embryonic 19 day-old (E19) uterus that were surgically removed from timed pregnant, terminally anesthetized Sprague-Dawley rats (Charles River Laboratories, Inc., Wilmington, MA). Dorsal root ganglia were collected in ice-cold L-15 medium (Invitrogen, Carlsbad, CA) containing 5% heat-inactivated horse serum (Invitrogen) and rinsed twice in Ca²⁺- and Mg²⁺-free Dulbecco's phosphate-buffered saline, pH 7.4 (Invitrogen). The dorsal root ganglia were then dissociated into single cell suspension using a papain dissociation system (Worthington Biochemicals, Freehold, NJ). The dissociated cells were pelleted at 200g for 5 min and resuspended in Earle's balanced salt solution containing 1 mg/ml ovomucoid inhibitor, 1 mg/ml ovalbumin, and 0.005% DNase. Cell suspension was centrifuged through a gradient solution containing 10 mg/ml ovomucoid inhibitor, 10 mg/ml ovalbumin at 200g for 6 min to remove cell debris and filtered through a 88-μm nylon mesh (Fisher Scientific Co., Pittsburgh, PA) to remove any clumps. Cells were seeded into poly-ornithine (100 μg/ml; Sigma-Aldrich, St. Louis, MO) and mouse laminin (1 μg/ml; Invitrogen)-coated 96-well plates at 10⁴ cells/well in complete medium. The complete medium consists of minimal essential medium and Ham's F-12 at a 1:1 ratio, 100 U/ml penicillin and 100 μg/ml streptomycin, 75 μM 5-fluoro-2′-deoxyuridine, 180 μM uridine, 10 ng/ml nerve growth factor, and 10% heat-inactivated horse serum (Invitrogen).

Capsaicin Antagonist Assay with E19 Rat Dorsal Root Ganglia Neurons. For measuring Ca²⁺ uptake, rat E19 embryonic dorsal root ganglia neurons of 3-day-old cultures were washed once with HBSS-HEPES buffer (0.1 mg/ml BSA in HBSS with 1 mM HEPES at pH 7.4) and incubated with serial concentrations of capsazepine or AMG 9810 in HBSS-HEPES buffer for 15 min at 37°C. Cells were then challenged with the TRPV1 agonist capsaicin (300 nM) in activation buffer containing 0.1 mg/ml BSA, 15 mM HEPES, pH 7.4, and 10 μCi/ml ⁴⁵Ca²⁺ (MP Biomedicals) in Ham's F-12 for 5 min at 37°C. Cells were washed three times with PBS containing 0.1 mg/ml BSA, incubated with Optiphase Supermix scintillation cocktail (PerkinElmer Wallac, Gaithersburg, MD) for 20 min, and radioactivity was measured using a MicroBeta Jet (PerkinElmer Life and Analytical Sciences, Boston, MA).

Adult Rat Dorsal Root Ganglia Neuronal Cultures. Primary sensory neuronal cultures were prepared from male Sprague-Dawley rats (150–200 g; Charles River Laboratories, Inc.) according to Lindsay (1988), with minor modifications. Briefly, C1-L6 dorsal root ganglia from adult rats terminally anesthetized with CO₂ were removed under aseptic conditions and collected in ice-cold L-15 medium containing 10% heat-inactivated horse serum. Ganglia were rinsed twice in Ca²⁺ and Mg²⁺-free Dulbecco's phosphate-buffered saline, incubated with collagenase at 37°C for 90 min, and digested with 20 U/ml papain and 0.005% DNase for 60 min. Ganglia were then triturated with fire-polished Pasteur pipettes, and the neuronal suspension was spun through 15% bovine serum albumin. Cells were resuspended into culture medium containing minimal essential medium and Ham's F-12 at 1:1 ratio, 100 U/ml penicillin and 100 μg/ml streptomycin, and 10% heat inactivated horse serum, supplemented with 1 mM l-glutamine and 10 ng/ml nerve growth factor and plated onto 100 μg/ml poly-dl-ornithine and 1 μg/ml mouse laminin-coated 96-well plates at a density of 1000 neurons/well. The cultures were kept in an incubator at 37°C with 5% CO₂ for 3 days before the assay.

Capsaicin-Induced CGRP Release Assay. Cultured adult rat dorsal root ganglia neurons in 96-well plates were washed twice with release buffer (135 mM NaCl, 3.5 mM KCl, 2.5 mM CaCl₂, 3.3 mM glucose, 25 mM HEPES, and 2 μM phosphoramidon with 0.1% BSA, pH 7.4) to initiate the assay. CGRP release was induced by incubation of neurons with capsaicin for 10 min at room temperature. The EC₉₀ concentration of capsaicin was used to determine the antagonism of TRPV1 by capsazepine and AMG 9810. The cultures were preincubated with increasing concentrations of capsazepine or AMG 9810 for 15 min, followed by 300 nM capsaicin activation for 10 min at room temperature. The extracellular medium was collected and the CGRP content was determined using a commercially available enzyme-linked immunoassay kit using the double antibody sandwich technique (catalog no. A05482; SPI-BIO, Montigny-le-Bretonneux, France), and read by a 96-well microplate reader at 414 nm (SpectraMax 340 PC; Molecular Devices, Sunnyvale, CA). The concentration of CGRP released into the medium was determined by comparing the light absorption values of the samples to that of a CGRP standard curve. For each experiment, the basal CGRP release was measured by incubating the cells in release buffer containing no compounds.

In Vivo Pharmacology

Animals. In vivo experiments were conducted at Amgen Inc. (Thousand Oaks, CA) and the University of Arizona. All experiments were conducted under protocols approved either by Amgen Inc.'s Institutional Animal Care and Use Committee or by the Animal Care and Use Committee of the University of Arizona. Rats were housed on a regular 12-h light/dark cycle in a climate-controlled room with food and water ad libitum. For experiments conducted at Amgen Inc., male Sprague-Dawley rats (Charles River Laboratories, Inc.) with body weight between 250 to 350 g at the time for testing were used. For the capsaicin-induced eye wiping test conducted at the University of Arizona, male Sprague-Dawley rats (Harlan, Indianapolis, IN) weighing 200 to 250 g were used.

Capsaicin-Induced Eye Wipe Test. Rats were acclimated for 30 to 45 min in a 30 × 30 × 30-cm Plexiglas chambers before the intraperitoneal injection of either vehicle (DMSO) or AMG 9810. Injections were made over a 5-s period in the lower right ventral quadrant of the abdomen either 15, 30, or 60 min before intraocular application of capsaicin. Intraocular application of capsaicin (3 μg/20 μl in 10% ethanol/PBS) or vehicle (20 μl in 10% ethanol/PBS) was done with a pipette, and the number of front paw eye wipes was counted over a 5-min period in 1-min intervals. Each group consisted of 10 to 11 animals in this experiment.

Complete Freund's Adjuvant (CFA)-Induced Thermal and Mechanical Hyperalgesia. The effect of AMG 9810 against hyperalgesia associated with inflammation was investigated using the CFA model generated by an intradermal injection of 0.15 ml of CFA was made into the dorsal surface of the rat left hind paw. Thermal and mechanical hyperalgesia tests were conducted 5 to 7 days after CFA injection. Before drug treatment, baseline measurements were recorded in both procedures. The same cohort of rats was used for determining the effects of AMG 9810 on thermal and mechanical hyperalgesia in CFA-treated rats. Testing for thermal hyperalgesia preceded testing for mechanical hyperalgesia. Based on our experience, we have not found significant effects of this testing order (unpublished observations). After baseline or precompound testing, the animals were treated with vehicle (PEG400 or phosphate-buffered saline) or test compound (AMG 9810 or morphine, respectively). Next, animals were tested at each time point for thermal hyperalgesia immediately followed by testing for mechanical hyperalgesia. Dosing of animals was staggered such that the timing remained ±5 min with regard to each time point. The experiment was conducted over the course of 2 days.

Thermal Hyperalgesia Test. Thermal hyperalgesia was measured using a modified Hargreaves box (Hargreaves et al., 1988) designed at the University of Arizona Health Science Center (Tucson, AZ). Rats were placed on a glass surface through which a light beam was applied onto the plantar hind paws of the animals. The thermal nociceptive response was defined as the latency between light stimulus onset and paw withdrawal using a feedback-controlled shutdown unit. The intensity of the light stimulus was set such that naive rats responded with a latency of approximately 25 s. Each paw was tested two times. In the absence of a response within a predetermined maximum latency (40 s), the test was terminated to prevent tissue damage.

Mechanical Hyperalgesia Test. An analgesy-meter (Ugo-Basile, Varese, Italy) was used for testing. This apparatus administers a force on the paw that increases at a constant rate. For this test, the CFA-injected paw was held such that pressure was administered via the apparatus on the dorsal surface of the paw. The force at which the animal pulled the paw away was used for statistical analysis. Naive rats responded with a pressure latency of approximately 240 g.

Open-Field Locomotor Activity and Motor Coordination/Strength Tests. Location and ambulation were measured as interruptions of the infrared photo beams using open-field chambers (Hamilton-Kinder, San Diego, CA). The MotorMonitor software (Hamilton-Kinder) measures ambulations, rearing, distance traveled, and time spent in any predesignated zones within the open field. Rearing and total distance were measured over a 60-min period. The edge test is designed to measure motor coordination and strength, as reported in Bannon et al. (1998).

Statistical Analysis of in Vivo Data. GraphPad Prism (Graph-Pad Software Inc., San Diego, CA) software was used for all statistical analyses. For the thermal and mechanical hyperalgesia data, capsaicin-induced eye wiping data, as well as the open-field data (rearing and total distance), a two-way analysis of variance (ANOVA) was used with treatment as a between-subject factor and time as a within-subject factor. Follow-up analysis was conducted at each time point using a one-way ANOVA followed by Tukey's post hoc analysis if ANOVA results were p < 0.05. For the edge test data, a one-way ANOVA was conducted.

Materials. Reagents used in the present study were obtained from the following companies: capsaicin (catalog no. 211274; Calbiochem, San Diego, CA), capsazepine (catalog no. 0464; Tocris Cookson Inc., Ellisville, MO), SB-366791 (catalog no. S-0441, Sigma-Aldrich; catalog no. 1615, Tocris Cookson Inc.); iodo-RTX (catalog no. 1362; Tocris Cookson Inc.), ⁴⁵Ca²⁺ (catalog no. 62005; MP Biomedicals), morphine sulfate and CFA (Sigma-Aldrich), and AMG 9810 (G & J Research, Devon, UK). The cinnamide structure-activity relationship is described by Doherty et al. (2005); BCTC was synthesized, and nerve growth factor was recombinantly expressed and purified at Amgen Inc.

Results

Characterization of AMG 9810 as a TRPV1 Antagonist. AMG 9810 was identified as a TRPV1 antagonist in a capsaicin-induced ⁴⁵Ca²⁺ uptake assay of a series of cinnamide compounds. The ability of AMG 9810 to inhibit channel activation was investigated at human, rat, and r/h chimera TRPV1. Comparison of agonist and antagonist activity at r/h chimera versus rat or human TRPV1 indicated that the r/h chimera responded similarly to human TRPV1 (Table 1). The capsaicin dose response indicated that both r/h chimera and human TRPV1 were activated with similar EC₅₀ values (124 ± 40 and 115 ± 57 nM, respectively). Capsazepine blocked proton activation of both human TRPV1 and r/h chimera similarly (IC₅₀ values are 69 ± 10 and 93 ± 22 nM, respectively), but it had no effect on rat TRPV1. In addition, the rank orders of IC₅₀ values at capsaicin and proton activation by BCTC, I-RTX, and AMG 9810 were the same for both human TRPV1 and r/h chimera. These data demonstrate that r/h TRPV1 and human TRPV1 respond equivalently to agonists and antagonists (Table 1).

AMG 9810 potently and concentration-dependently inhibited capsaicin- and proton-induced ⁴⁵Ca²⁺ uptake into CHO cells expressing r/h chimera with IC₅₀ values of 19.1 ± 7.9 and 90.3 ± 53.1 nM, respectively. IC₅₀ values for blocking capsaicin activation of at rat and human TRPV1 were 85.6 ± 39.4 and 24.5 ± 15.7 nM, respectively (Fig. 2A). IC₅₀ values for blocking proton activation of rat and human TRPV1 were 294 ± 192 and 92.7 ± 72.8 nM, respectively (Fig. 2B). The ability of AMG 9810 to inhibit heat activation of TRPV1 was also demonstrated (Fig. 2C): IC₅₀ values for rat and human TRPV1 activation by heat (45°C) were 21 ± 17 and 15.8 ± 10.8 nM, respectively (Table 1). The ability of AMG 9810 to inhibit channel activation by proposed endogenous ligands was also studied. AMG 9810 was shown to be a potent antagonist of AEA, NADA, and OLDA in agonist-induced ⁴⁵Ca²⁺ uptake assays. IC₅₀ values for blocking human TRPV1 activation by AEA, NADA, and OLDA were 62 ± 27, 8.5 ± 4.6, and 264 ± 161 nM, respectively. AMG 9810 IC₅₀ values for blocking rat TRPV1 activation by AEA, NADA, and OLDA were 328 ± 210, 260 ± 74, and 440 ± 266 nM, respectively.

To determine whether AMG 9810 competitively inhibits capsaicin binding to rat TRPV1, concentration-response curves for capsaicin were generated in the absence or presence of AMG 9810 (0.1, 0.3, and 1.0 μM). Preincubation of rat TRPV1-expressing CHO cells with AMG 9810 caused a rightward shift in the concentration-response to capsaicin, with no apparent change in maximum response in ⁴⁵Ca²⁺ uptake (Fig. 3A). A Schild analysis (Arunlakshana and Schild, 1959) indicated a pA₂ value of 7.46 and a slope factor of 1.506 ± 0.068 (Fig. 3B). Slope value of >1 indicates a cooperative nature of antagonism at capsaicin activation. Capsaicin antagonism by AMG 9810 was also studied by electrophysiology, using the whole-cell voltage-clamp configuration. Exposure to capsaicin produced concentration-dependent inward current responses in CHO cells expressing rat TRPV1 (Fig. 3C, left). However, after a 2-min preapplication of 0.1 μM AMG 9810 (which continued throughout the experiment), the capsaicin-evoked currents were smaller at all concentrations tested (Fig. 3C, right). AMG 9810 caused a rightward shift in the concentration-response to capsaicin-evoked currents (Fig. 3D), suggesting a competitive interaction with capsaicin, which is consistent with the ⁴⁵Ca²⁺ uptake assay results.

Since the reported IC₅₀ values for TRPV1 antagonists were generated either by fluorometric imaging plate reader assays measuring both calcium uptake and intracellular release or by electrophysiology, we have compared the antagonists in an agonist-induced ⁴⁵Ca²⁺ uptake assay, which represents pharmacology at the plasma membrane TRPV1. Among the studied antagonists, the rank order of potency at capsaicin for both rat and human TRPV1 was BCTC > I-RTX > AMG 9810 > SB-366791 > capsazepine (Table 1). The rank order of potency against capsaicin was the same for the proposed endogenous ligands AEA, NADA, and OLDA, as well as for heat. All the antagonists seemed to be more potent at blocking heat activation of TRPV1, compared with other modes of activation. Capsazepine was ineffective at inhibiting rat TRPV1 activation by protons (pH 5), consistent with the pharmacology reported by McIntyre et al. (2001). I-RTX was also ineffective at inhibiting proton (pH 5) activation of rat TRPV1. SB-366791 obtained from two different sources (Sigma-Aldrich and Tocris Cookson Inc.) was ineffective at proton (pH 5) activation of both rat and human TRPV1 expressed in CHO cells. AMG 9810 was found to be a more potent antagonist than capsazepine and SB-366791 under all modes of TRPV1 activation (Fig. 2; Table 1). The fact that it displayed a potent antagonism of proton (pH 5) activation made it a viable candidate for studying antihyperalgesic effects in rat model of inflammatory pain.

Pharmacological Selectivity of AMG 9810. AMG 9810 was found to be selective for TRPV1 among the recombinant TRP family members that we tested. IC₅₀ values were >4 μM for TRPV3 (heat-induced ⁴⁵Ca²⁺ uptake assay), TRPV4 (4-αPDD-induced increase in intracellular calcium in fluorometric imaging plate reader assay), TRPM8 (icilin-induced ⁴⁵Ca²⁺ uptake assay), and TRPA1 or ANKTM1 (cold-induced ⁴⁵Ca²⁺ uptake assay). In studies performed by Novascreen, AMG 9810 showed no significant inhibition (<35% inhibition at 10 μM) of ligand binding to a number of ion channels, G protein-coupled receptors, and transporter sites (IC₅₀ > 10 μM), including orphanin, adenosine transporter, adenosine (A₁), adrenergic (α1 ns, α2 ns), adrenergic β ns, cannabinoid CB₁, dopamine ns, GABA_A agonist site, benzodiazepine site, GABA_B, glutamate [α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid site, kainate site, NMDA agonist site, NMDA phencyclidine site, NMDA glycine (strychnine-insensitive site)], glycine strychnine-sensitive site, histamine H₁, H₃, imidazoline I₂ central, melatonin ns, muscarinic (ns central, ns peripheral, M₁, M₂, M₃), nicotinic (α-bungtox-insensitive), norepinephrine transporter, opiate (ns, κ, μ), purinergic, P₂Y, serotonin transporter, serotonin (ns, 5-HT_1A, 5-HT_2A, 5-HT_2C, 5-HT₃), σ (ns, 1, 2), estrogen, testosterone, L-type Ca²⁺ channel (benzothiazepine site), N-type Ca²⁺ channel, K⁺ channel (ATP-sensitive, Ca²⁺-activated VI, Ca²⁺-activated VS), Na⁺ site 1, inositol triphosphate IP₃, NOS (neuronal binding), GABA transport, leukotriene (LTB₄, LTD₄), thromboxane A₂, corticotropin releasing factor ns, oxytocin, platelet-activating factor, thyrotropin-releasing hormone, bradykinin, BK₂, cholecystokinin (CCK_B), endothelin (ET_A, ET_B), galanin ns, neurokinin (NK1, NK2), neuropeptide Y (ns, NPY1, NPY2), neurotensin, somatostatin ns, vasoactive intestinal peptide ns, vasopressin 1, acetylcholinesterase, choline acetyltransferase, glutamic acid decarboxylase, and monoamine oxidase (MAO-A peripheral, MAO-B peripheral). AMG 9810-displaced ligands for Ca²⁺ channel (L-type: DHP site), Na⁺ channel (site 2), dopamine transporter, and angiotensin II with IC₅₀ values of 2.35, 6.53, 1.08, and 5.31 μM, respectively.

AMG 9810 Is a Potent Antagonist of Activation of Endogenous TRPV1 by Capsaicin in Rat Dorsal Root Ganglia Neurons. The ability of AMG 9810 to antagonize capsaicin-induced ⁴⁵Ca²⁺ uptake in rat dorsal root ganglia neurons expressing endogenous TRPV1 was studied (Fig. 4). Capsaicin induced ⁴⁵Ca²⁺ uptake into dorsal root ganglia neurons in a dose-dependent manner, with an EC₉₀ value of 300 nM. It has been shown that capsaicin induces CGRP release in a Ca²⁺-dependent manner from a subset of dorsal root ganglia neurons through activation of TRPV1 and that the competitive antagonist capsazepine blocks the CGRP release (Sauer et al., 1999; Tognetto et al., 2001). Therefore, we examined the ability of AMG 9810 to block capsaicin-induced CGRP release in cultured rat dorsal root ganglia neurons. Capsaicin (300 nM) induced CGRP release from cultured neurons into the media by approximately 12-fold (350 ± 45 fmol/well/10 min) compared with the basal level (29 ± 10 fmol/well/10 min).

The competitive TRV1 antagonist capsazepine blocked both capsaicin-induced ⁴⁵Ca²⁺ uptake and CGRP release, with IC₅₀ values of 890 ± 360 and 1120 ± 580 nM, respectively. Compared with capsazepine, AMG 9810 was found to be more potent in blocking capsaicin-induced ⁴⁵Ca²⁺ uptake and CGRP release with IC₅₀ values of 9 ± 6 nM (Fig. 4A) and 6 ± 3 nM (Fig. 4B), respectively. Exposing neurons to either capsazepine or AMG 9810 alone up to 10 μM had no effect on basal ⁴⁵Ca²⁺ uptake or CGRP release in the neuronal cultures (data not shown).

AMG 9810 Blocks Capsaicin-Evoked Depolarization in Rat Dorsal Root Ganglia Neurons. The ability of AMG 9810 to affect capsaicin-evoked depolarization was studied in rat E19 embryonic dorsal root ganglia neurons. Capsaicin evoked a concentration-dependent membrane depolarization and elicited action potential firing at 1 μM (Fig. 5). Membrane potential was slowly hyperpolarized after capsaicin washout and recovered back to the normal level in about 2 min. Preapplication of 0.1 μM AMG 9810 for 1 min blocked membrane depolarization evoked by 0.1 μM capsaicin. However, 0.3 and 1 μM capsaicin evoked membrane depolarization in the presence of 0.1 μM AMG 9810 to a smaller magnitude compared with capsaicin alone. In addition, 1 μM capsaicin did not evoke action potential firing in the presence of 0.1 μM AMG 9810 (Fig. 5, middle). Capsaicin was able to evoke membrane depolarization (at 0.3 μM) and action potential firing (at 1 μM) after AMG 9810 washout for 2 min (Fig. 5, bottom). AMG 9810 alone failed to change membrane potential or cause action potential firing in neonatal rat DRG neurons (data not shown). In summary, AMG 9810 was able to decrease capsaicin-evoked depolarization in primary sensory neurons from prenatal rats and its effects were partially reversible in a 2-min washout.

AMG 9810 Blocks Capsaicin-Induced Eye Wiping. Intraocular application of capsaicin (3 μg/20 μl) resulted in a significant number of capsaicin-evoked front paw eye wipes in the rat (Fig. 6). For statistical analysis, data from the first-minute interval were used because capsaicin-evoked eye wiping behavior occurred only during the first minute of observation in vehicle-treated rats (data not shown). There was an overall effect of treatment [F_(3,36) = 12.33, p < 0.0001] and time [F_(2,36) = 4.413, p < 0.016], but not a significant interaction. A dose-dependent decrease of capsaicin-induced eye wipes was observed at 15 min with all AMG 9810 doses tested (3, 10, and 30 mg/kg intraperitoneal; Fig. 6). In animals treated with AMG 9810, 30 min before capsaicin treatment, statistically significant reductions of the number of eye wipes were observed with 10 and 30 mg/kg AMG 9810. In animals treated 60 min before capsaicin administration, significant reductions in eye wipes were only observed in animals treated with 30 mg/kg AMG 9810. AMG 9810 inhibition of capsaicin-evoked eye-wiping was performed side by side with vehicle-administered animals throughout all experiments to ensure positive capsaicin-evoked eye wiping response. Vehicle had no effect on capsaicin-evoked eye wipes.

AMG 9810 Reduces CFA-Induced Thermal and Mechanical Hyperalgesia. Intraplantar injection of 150 μl of CFA resulted in the development of thermal (paw withdrawal latency dropped to 15 from 25 s) and mechanical (paw withdrawal threshold dropped to 120 from 240 g) hyperalgesia. For thermal hyperalgesia (Fig. 7A), there was a significant effect of AMG 9810 treatment [F_(3,20) = 19.2; p < 0.0001], time [F_(4,20) = 19.45; p < 0.0001], and a significant treatment × time interaction [F_(12,20) = 2.861; p < 0.0025]. Treatment with 30 and 100 mg/kg (intraperitoneal) of AMG 9810 significantly reversed CFA-induced thermal hyperalgesia at 30, 60, and 90 min post-treatment (p < 0.05). The reversal of thermal hyperalgesia by these doses of AMG 9810 had diminished by 120 min. No statistically significant effects on thermal hyperalgesia were found with a 10-mg/kg (intraperitoneal) dose of AMG 9810 at any time measured. Limited in vivo pharmacokinetic analysis of AMG 9810 indicated a plasma half-life of approximately 1.8 h with T_max reached after 30 min (data not shown). This profile is consistent with the antihyperalgesic effects seen in vivo between 30 and 90 min after AMG 9810 administration. For mechanical hyperalgesia (Fig. 7B), there was a statistically significant effect of treatment [F_(3,20) = 6.498; p < 0.003], time [F_(4,20) = 11.62; p < 0.0001], and a significant treatment × time interaction [F_(12,20) = 2.973; p < 0.0018]. Treatment with 100 mg/kg (intraperitoneal) AMG 9810 significantly reversed CFA-induced mechanical hyperalgesia at 30 and 60 min post-treatment (p < 0.05). The effect had diminished to control levels by 90 min. No other significant effects were found with the lower doses of AMG 9810 on mechanical hyperalgesia.

Discussion

In the present article, we compare the in vitro pharmacology of AMG 9810, a novel cinnamide TRPV1 antagonist, with the reported antagonists BCTC, capsazepine, I-RTX, and SB-366791. We show that treatment with AMG 9810 prevents capsaicin-induced eye wiping behavior and significantly reverses both thermal and mechanical hyperalgesia associated with CFA-induced inflammation. We have demonstrated that AMG 9810 is a TRPV1 antagonist in agonistinduced ⁴⁵Ca²⁺ uptake assays and by electrophysiology in CHO cells stably expressing TRPV1 and in rat DRG neurons. AMG 9810 was found to be a potent antagonist of exogenous (capsaicin), proposed endogenous ligands (AEA, NADA, and OLDA) of TRPV1, and physical stimuli (protons and heat). AMG 9810 was more potent than capsazepine and SB-366791 in all antagonist assays tested. The ⁴⁵Ca²⁺ uptake assay IC₅₀ value, calculated from concentration-inhibition curves, was 85.6 ± 39.4 nM and is comparable with the potency measured by Schild analysis (pA₂ = 7.46). AMG 9810 interacts competitively at the capsaicin-binding pocket of rat TRPV1, as demonstrated by a concentration-dependent, parallel rightward shift in capsaicin dose-response curves.

Capsazepine, I-RTX, and SB-366791 were ineffective at blocking proton (pH 5) activation of rat TRPV1. In fact, both SB-366791 and I-RTX potentiated proton-induced ⁴⁵Ca²⁺ uptake (approximately 40% increase over pH 5 alone level) in a concentration-dependent manner, indicating that they are allosteric potentiators of TRPV1 activation by protons (N. R. Gavva, L. Klionsky, R. Tamir, K. Wild, and J. J. Treanor, unpublished data). Interestingly, electrophysiological studies did not show potentiation, perhaps owing to different readout in these assays. AMG 9810 was 4- to 5-fold less potent at rat TRPV1 compared with human TRPV1. However, AMG 9810 was equipotent at r/h chimera and human TRPV1, suggesting that N terminus of rat TRPV1 is not responsible for lower potency at rat TRPV1. Based on mutagenesis data, it was shown that capsazepine requires Leu⁵⁴⁷ for antagonism of TRPV1 activation by protons (Gavva et al., 2004; Phillips et al., 2004). SB-366791 was found to be inactive up to 40 μMat blocking the proton activation of TRPV1 that contains either Met⁵⁴⁷ (rat TRPV1) or Leu⁵⁴⁷ (human TRPV1) (Table 1). In contrast, AMG 9810 was equally effective at blocking proton activation of TRPV1 with either Met⁵⁴⁷ or Leu⁵⁴⁷ (Table 1) and was also equipotent at rabbit TRPV1 gain-of-function mutant (I550T; Gavva et al., 2004) with either Met⁵⁴⁷ or Leu⁵⁴⁷ (data not shown). Although the structure-activity relationships of the cinnamide series including the AMG 9810 are reported (Doherty et al., 2005), it is not clear yet what the specific molecular interactions occur between residues in TM3/4 region of TRPV1 and the antagonists. Although BCTC is the most potent TRPV1 antagonist reported, it also inhibits TRPM8 with an IC₅₀ value of ∼1 μM (Behrendt et al., 2004; N. R. Gavva, L. Arik, and L. Klionsky, unpublished data), whereas AMG 9810, capsazepine, and SB-366791 showed no effect on TRPM8 activation at concentrations up to 4 μM. Overall, these data demonstrate that AMG 9810 is a potent antagonist of all known modes of activation of TRPV1.

Agonist-induced ⁴⁵Ca²⁺ uptake studies using rat dorsal root ganglia primary cultures indicate that AMG 9810 is a potent antagonist of endogenous TRPV1. Electrophysiological studies showed that capsaicin evokes depolarization in dissociated rat dorsal root ganglia primary neurons (Blair and Bean, 2003). We show that AMG 9810 inhibits capsaicinevoked membrane depolarization in dorsal root ganglia neurons. This observation underpins the physiological basis for dose-dependent blockade of capsaicin-induced eye wiping behavior in rats and possibly for its inhibitory effects on CFA-induced thermal and mechanical hyperalgesia.

Based on the mutagenesis studies, it was reported that residues in the TM3/4 region of TRPV1 are responsible for vanilloid activation (Jordt and Julius, 2002; Gavva et al., 2004), whereas the residues in prepore loop region are responsible for proton activation (Jordt et al., 2000). BCTC, capsazepine, and AMG 9810 all seem to require critical determinants present in TM3/4 region of TRPV1 for antagonism of capsaicin activation (Gavva et al., 2004; N. R. Gavva, R. Tamir, L. Klionsky, and J. J. Treanor, unpublished data). This infers that agonists such as capsaicin and the antagonists BCTC, capsazepine, and AMG 9810 all share a common binding pocket. We propose different mechanisms of action for AMG 9810 at TRPV1 for blocking capsaicin and proton activation. AMG 9810 is a competitive antagonist at the capsaicin binding pocket, whereas it is an allosteric inhibitor for proton activation. Presumably, AMG 9810 is able to hold the TRPV1 conformation in a nonconducting state or to force the channel toward its closed state, even when the channel is activated through allosteric sites such as protons from extracellular sites.

The TRPV1 antagonist capsazepine was shown to inhibit capsaicin activation, but it was ineffective in blocking proton activation of rat TRPV1 and was similarly ineffective in reversing pain behavior associated with inflammation in rats (McIntyre et al., 2001; Walker et al., 2003). Capsazepine was effective at inhibiting both proton and capsaicin activation of TRPV1 in guinea pigs and reduces inflammatory and nerve injury-related hyperalgesia in guinea pigs (Walker et al., 2003). These species' differences indicate that in vivo efficacy correlates with capsazepine potency for blocking proton activation of TRPV1. BCTC was also effective at inhibiting both proton and capsaicin activation of rat TRPV1 and was shown to reduce inflammatory and nerve injury-related hyperalgesia in rats (Pomonis et al., 2003). In agreement with this, we show that AMG 9810 inhibits both capsaicin and proton (pH 5) activation of rat TRPV1 and reverses CFA-induced thermal and mechanical hyperalgesia in rats, a model of inflammatory pain. An antagonist capable of blocking all modes of TRPV1 activation may represent an ideal candidate for studies of in vivo efficacy in models of chronic pain.

Although differences in carrageenan-induced mechanical hyperalgesia were not reported in TRPV1 knockout mice, all three antagonists—BCTC, capsazepine, and AMG 9810 — showed reversal of mechanical hyperalgesia, albeit with different magnitudes, in models of inflammatory pain. These results suggest a role of TRPV1 for mediating responses to mechanical stimuli under inflammatory conditions. The observations demonstrating AMG 9810 antihyperalgesia is unlikely due to sedative effects, because no significant effects of AMG 9810 were observed on rearing behavior, total distance traveled, and motor coordination/strength (data not shown). Furthermore, the effects of AMG 9810 are not mediated through opioid receptors because AMG 9810 neither binds opioid receptors nor did naloxone treatment reverse or block the antihyperalgesic effects of AMG 9810 (data not shown).

In summary, there is a growing body of research suggesting that TRPV1 may contribute to inflammatory pain. AMG 9810 was shown to be a potent and selective antagonist of TRPV1 that can significantly reverse thermal and mechanical hyperalgesia in an animal model of inflammatory pain. Further study of AMG 9810 and related agents will help further define the role of TRPV1 as target for the generation of pain therapeutics.