Irreversible competitive antagonists are generally accepted to bind pharmacological receptors irreversibly, presumably by forming strong covalent bonds with the receptor. Thereby, they prevent other ligands from binding to the primary (orthosteric) binding site(s) on these receptors [the reader may consult Neubig et al. (2003) regarding specific terminology]. The classical irreversible α-adrenoceptor-blocking drug phenoxybenzamine (Dibenzyline; Wellspring Pharmaceutical Corporation, Neptune, NJ) has been used since the 1960s in the clinical setting to treat pheochromocytoma (Crago et al., 1967). Most irreversible competitive antagonists, however, found application in experimental pharmacology to investigate and eliminate spare receptors, implementing the Furchgott analysis (Furchgott, 1966) to estimate the relative intrinsic efficacy of agonists and the apparent equilibrium dissociation constants of agonist-receptor complexes (Herepath and Broadley, 1990; Adham et al., 1993; Agneter et al., 1993; Zhu, 1993; Tian et al., 1996; Kenakin, 1997; Morey et al., 1998; Koek et al., 2000). Irreversible antagonists have also been used to label and count receptor subtypes, to investigate drug and receptor specificity and receptor structure (Jenkinson, 2003), to investigate receptor trafficking (Taouis et al., 1987; McKernan et al., 1988), and to unravel drug action mechanisms (Timmermans et al., 1985). Selective inactivation of “unwanted” pharmacological receptor subtypes, although protecting the receptors of interest with a highly selective reversible competitive antagonist, has also been implemented to study and characterize receptors (Hieble et al., 1985; Eglen et al., 1994).

Phenoxybenzamine is known to irreversibly bind to α-adrenergic, H₁-histamine, and muscarinic acetylcholine (mACh) receptors (Timmermans et al., 1985; Eglen et al., 1994; Amobi and Smith, 1995; Giardinà et al., 1995, 2002; Van der Graaf and Danhof, 1997; Ruffolo and Hieble, 1999; Van der Graaf and Stam, 1999; Frang et al., 2001). Benextramine has been known to irreversibly bind to α₂-adrenoceptors (Melchiorre, 1981; Belleau et al., 1982b; Lew and Angus, 1984; Hieble et al., 1985; Timmermans et al., 1985; Taouis et al., 1987; McKernan et al., 1988; Brink et al., 2000; Umland et al., 2001), 5-HT_1A-serotonergic receptors (Stanton and Beer, 1997), H₂-histaminergic receptors (Belleau et al., 1982a), and neuropeptide Y receptors (Melchiorre et al., 1994). Present knowledge regarding the mechanism of action of benextramine is limited to the observations from predominantly radioligand binding studies, and it is generally assumed that benextramine irreversibly inactivates the ligand (primary or syntopic) binding sites at these receptors. 4-Diphenylacetoxy-N-(2-chloroethyl)piperidine (4-DAMP mustard) is known to bind to mACh receptors (Thomas et al., 1992; Eglen et al., 1994; Ehlert and Griffin, 1998; Sawyer and Ehlert, 1999; Ragheb et al., 2001; Umland et al., 2001) with moderate selectivity for M₃-mACh receptors. It does not, however, discriminate between M₁, M₂, or M₄-mACh receptors (Eglen et al., 1994).

The elimination of pharmacological receptors by an irreversible competitive antagonist depends on both the concentration used and the incubation time (Furchgott, 1966; Kenakin, 1997). Phenoxybenzamine has been typically used in vitro at concentrations of up to 10 μM for 30 min (Piascik et al., 1988), benextramine at concentrations of up to 100 μM for 120 min (Van der Graaf et al., 1996), and 4-DAMP mustard at concentrations of up to 40 nM for 4 h (Sawyer and Ehlert, 1999).

Reports in literature suggest that some irreversible competitive antagonists may also display irreversible allosteric interactions (noncompetitive antagonism) (Van der Graaf et al., 1996; Brink et al., 2000) by altering agonist binding affinity, modifying agonist-receptor binding kinetics (Van Ginneken, 1977), or even influencing signal transduction mechanisms. However, this has been difficult to prove (Jenkinson, 2003). The irreversible antagonism of benextramine at prostanoid TP receptors cannot be prevented by the selective prostanoid TP-receptor antagonist SQ 30,741 (Van der Graaf et al., 1996), suggesting that benextramine and SQ 30,741 bind to different binding sites (i.e., benextramine exhibits allotopic interactions at prostanoid TP receptors). The underlying mechanism, however, remains elusive.

In the present study, we examined three classical irreversible competitive antagonists, namely, benextramine at α_2A-adrenoceptors and 5-HT_2A receptors, phenoxybenzamine at α_2A-adrenoceptors, and 4-DAMP mustard at mACh receptors. For any observed noncompetitive interactions, the mechanisms were investigated, including possible allosteric interactions, where agonist affinity at the syntopic binding site of the GPCR is altered or signal transduction is modulated downstream. We have been able to show that benextramine displays irreversible noncompetitive antagonism by binding at a site affecting the α_2A-adrenoceptor-G_i protein coupling or G protein function, without affecting the GTP binding capacity of the G_i-related Gα_o protein. In addition, we have been able to show that benextramine also irreversibly and noncompetitively inhibits the signaling of α_2A-adrenoceptors through G_s proteins, as well as the signaling of mACh receptors and serotonin 5-HT_2A receptors through G_q proteins.

Materials and Methods

Cell Lines. In this study, cultured cell lines expressing α-adrenoceptors and mACh receptors, respectively, were used. Two Chinese hamster ovary (CHO-K1) cell lines, stably transfected to express the wild-type porcine α_2A-adrenoceptor at high numbers (cell line denoted α_2A-H) and low numbers (cell line denoted α_2A-L) respectively, were kindly provided by Dr. Richard Neubig (Department of Pharmacology, University of Michigan, Ann Arbor, MI). The pharmacological profiles and receptor expression characteristics of the α_2A-H and α_2A-L cell lines have been previously characterized. The determined α_2A-adrenoceptor concentrations were reported as 19 ± 2 pmol/mg membrane protein for α_2A-H and as about 1 pmol/mg membrane protein for α_2A-L (Brink et al., 2000). We confirmed the high receptor expression for α_2A-H in our laboratory, determined as 46 ± 5 pmol/mg membrane protein. These cell lines were used to investigate the mechanisms of antagonism of the irreversible α-adrenoceptor-blocking drugs phenoxybenzamine and benextramine.

We also used human neuroblastoma cells (SH-SY5Y; American Type Culture Collection, Manassas, VA), that endogenously express predominantly M₃-mACh receptors (Slowiejko et al., 1996), with some evidence for M₁ and M₂-mACh receptors (Kukkonen et al., 1992). The K_D value of [³H]N-methyl scopolamine binding at the endogenous mACh receptors in intact cells was previously reported as 0.2 nM and the B_max as 100 to 150 fmol/mg membrane protein (Lambert et al., 1989). This cell line was used to investigate the mechanisms of antagonism of the irreversible mACh receptor-blocking drug 4-DAMP mustard. The fourth cell line used was SH-SY5Y cells stably transfected to express the human 5-hydroxytryptamine-2A (5-HT_2A) receptor, denoted 5-HT_2A-SH-SY5Y cells (see transfection protocol below). The human 5-HT_2A plasmid cDNA in the pIRES (Neo^r) mammalian expression vector was kindly provided by Dr. Bryan Roth (Department of Biochemistry, Case Western Reserve University, Cleveland, OH). Human neuroblastoma SH-SY5Y cells were transfected with this vector using DOTAP liposomal transfection reagent (Roche, Mannheim, Germany) according to the manufacturer's instructions. Since the plasmid encodes for G-418 resistance, the cells were subjected to G-418 (400 μg/ml; Invitrogen, Carlsbad, CA) treatment after 48 h, and the surviving colony was harvested and implemented in this study. Transfected cells were not cloned, but the transfection mix was used. Successful transfection was confirmed pharmacologically by dose-response curves with the 5-HT_2A receptor agonist serotonin (with control SH-SY5Y cells showing no serotonin-induced response, data not shown) and by the observed K_D value of [³H]spiperone at 5-HT_2A receptors (K_D = 8.3 ± 3.0 nM, B_max = 3166 ± 421 receptors/cell) as determined from saturation binding curves, using 10 μM ritanserin to define nonspecific binding. This K_D value corresponds with reported K_i values for [³H]spiperone at 5-HT_2A receptors against various radioligands, ranging between 0.12 nM (Roth et al., 1987) and 50.11 nM (Boess and Martin, 1994). The 5-HT_2A-SH-SY5Y cells were used to determine the effect of benextramine pretreatment on 5-HT_2A receptors by determining agonist-stimulated, G_q-mediated activation of phospholipase C (PLC), by measuring [³H]IP_x accumulation. These cells were also used for radioligand binding studies to investigate the effect of benextramine on 5-HT_2A-receptor numbers.

The α_2A-L and α_2A-H cells were maintained in a humidified environment and grown to 95% confluency in 150-cm² cell culture flasks with Ham's F-12 medium containing 10% fetal bovine serum, 100 units/ml penicillin, and 100 μg/ml streptomycin at 37°C in 5% CO₂. The SH-SY5Y cells were similarly maintained, but the growth medium used was a 1:1 ratio mixture of Ham's F-12 and Dulbecco's modified Eagle's medium (DMEM; Highveld Biological, Gauteng, South Africa) containing 10% bovine serum albumin. The 5-HT_2A-SH-SY5Y cells were maintained as the SH-SY5Y cells but with 400 μg/ml G-418.

Preparation and Pretreatment with Irreversible Competitive Antagonists. Cells were seeded in 24-well plates in preparation for the [³H]cAMP, [³H]IP_x, or ligand binding assays as described below and maintained for at least 18 h. The α_2A-H and α_2A-L cells were seeded at a density of approximately 3 × 10⁵ cells/ml, whereas SH-SY5Y and 5-HT_2A-SH-SY5Y cells were seeded at a density of 6 × 10⁵ cells/ml. When cell pretreatments were intended for membrane preparation, the pretreatments were performed directly in the 150-cm² culture flasks. Cells attached adequately to the well bottoms, allowing several aspirations and new additions of medium without significant cell loss (confirmed by microscopic observation).

When the experiments were performed, the pretreatments were initiated by incubating the cells with an appropriate concentration (0 M or a concentration >1000 × K_D value) of the reversible antagonist (allowing equilibrium of ligand-receptor binding), whereafter the cells were exposed to different concentrations of the appropriate irreversible antagonist plus the reversible antagonist. This was followed by several rinsing and incubation steps (washing procedure) with phosphate-buffered saline [PBS; containing (w/v) 0.8% NaCl, 0.02% KCl, 0.09% Na₂HPO₄, and 0.02% KH₂PO₄] and DMEM to remove all unbound and reversibly bound drugs. The following pretreatment steps were used.

Step 1. α_2A-L or α_2A-H cells were incubated with either 0 or 10 μM of the reversible competitive α₂-adrenoceptor antagonist (Becker et al., 1999) yohimbine hydrochloride (Sigma-Aldrich) in DMEM for 30 min at 37°C in 5% CO₂, to allow equilibrium of ligand-receptor binding. Likewise, SH-SY5Y cells were incubated with either 0 or 10 μM of the reversible competitive nonselective mACh receptor antagonist (Zwart and Vijverberg, 1997) atropine sulfate (Sigma-Aldrich; reported average K_i = 0.50 nM for M₃-mACh receptors by Hirose et al., 2001) in DMEM for 30 min at 37°C in 5% CO₂ to allow equilibrium of receptor binding. 5-HT_2A-SH-SY5Y cells were incubated with either 0 or 10 μM of the reversible competitive 5-HT₂-receptor antagonist ritanserin (Sigma-Aldrich; reported average K_i = 0.25 nM for 5-HT_2A receptors by Bonhaus et al., 1997) in DMEM for 30 min at 37°C and 5% CO₂.

Step 2. Thereafter, α_2A-L or α_2A-H cells were correspondingly incubated with either 0 or 10 μM yohimbine plus the indicated concentration of freshly prepared phenoxybenzamine hydrochloride (Sigma-Aldrich) or benextramine tetrahydrochloride (Sigma-Aldrich) (0, 1, 10, or 100 μM) for 20 min at 37°C in 5% CO₂. SH-SY5Y cells were correspondingly incubated with either 0 or 10 μM atropine plus the indicated concentration of 4-DAMP mustard (Sigma-Aldrich) (0, 10, or 100 nM) for 20 min at 37°C in 5% CO₂. Since 4-DAMP mustard is completely converted to its corresponding active aziridinium ion only after about 30 min in aqueous solution (Thomas et al., 1992), it was kept for at least 30 min at 37°C in DMEM before use in the assay. 5-HT_2A-SH-SY5Y cells were correspondingly incubated with either 0 or 10 μM ritanserin plus the indicated concentration of freshly prepared benextramine (0, 10, or 100 nM) for 20 min at 37°C in 5% CO₂.

Step 3. Cells were then rinsed twice with PBS and incubated twice with DMEM for 20 min at 37°C in 5% CO₂ to allow dissociation of any reversibly bound drugs, whereafter the cells were used for the functional or ligand binding assays as described below.

Measurement of Whole-Cell [³H]cAMP Accumulation. [³H]cAMP accumulation was determined in whole α_2A-L cells in 24-well plates as described previously (Wong, 1994; Wade et al., 1999). Briefly, cells were radiolabeled by adding 1 μCi per well [2-³H]adenine (19–23 Ci/mmol; Amersham Biosciences UK, Ltd., Little Chalfont, Buckinghamshire, UK), plus 100 ng/ml pertussis toxin (PTX; Sigma-Aldrich) when indicated, at least 18 h before the assay. After the washing procedure (as described above), the assay was initiated by adding DMEM with 1 mM 3-isobutyl-1-methylxanthine (IBMX) (Sigma-Aldrich) and 30 μM forskolin (Sigma-Aldrich) and the appropriate concentration of the full α_2A-adrenoceptor (Brink et al., 2000) agonist 5-bromo-N-(4,5-dihydro-1H-imidazol-2-yl)-6-quinoxalinamine (UK 14,304; Sigma-Aldrich) to construct appropriate semilogarithmic dose-response curves. After a 20-min incubation time at 37°C in 5% CO₂, the medium was aspirated and the reaction terminated with 1ml of ice-cold 5% (w/v) trichloroacetic acid (TCA; Sigma-Aldrich) containing 1 mM ATP (Sigma-Aldrich) and 1 mM cAMP (Sigma-Aldrich) and allowed to stand for 30 min at 4°C to lyse the cells. The acid soluble nucleotides were separated on Dowex and Alumina columns as described before (Salomon et al., 1974), and radioactivity was determined by liquid scintillation counting. The cAMP accumulation was normalized by dividing the [³H]cAMP counts by the total [³H]nucleotide counts. This value was then divided by the corresponding value obtained in the presence of IBMX and forskolin without agonist (to calculate the percentage of control).

Measurement of Whole-Cell [³H]IP_x Accumulation. [³H]IP_x accumulation was determined in whole SH-SY5Y and 5HT_2A-SH-SY5Y cells in 24-well plates according to the principles described before (Godfrey, 1992; Casarosa et al., 2001) but with minor modifications. Briefly, cells were labeled by adding 1 μCi per well myo-[2-³H]inositol (17 Ci/mmol; Amersham Biosciences UK, Ltd.) in inositol-free medium (minimum essential medium, Earle's base + bovine serum albumin) at least 18 h before the assay. After the pretreatment and washing procedure (as described above), the [³H]IP_x assay was initiated by adding a mixture of DMEM, 20 mM LiCl (Sigma-Aldrich), 25 mM N-[2-hydroxyethyl]piperazine-N′-[2-ethanesulphonic acid] (HEPES) (Sigma-Aldrich), and the appropriate concentration of the full mACh receptor agonist (Olianas and Onali, 1991) methacholine chloride (Sigma-Aldrich) or the endogenous 5HT_2A receptor agonist serotonin creatinine sulfate (Sigma-Aldrich) to construct appropriate semilogarithmic dose-response curves. After a 60-min incubation at 37°C and 5% CO₂, the medium was aspirated, and the reaction was terminated with 1 ml of ice-cold 10 mM formic acid (Saarchem-Holpro Analytic, Krugersdorp, Gauteng, South Africa) and was let to stand for at least 90 min at 4°C to lyse the cells. The [³H]IP_x was separated on Dowex columns (250-μl Dowex 1 × 8–400, 200–400 mesh, 1-chloride form per 2 ml; Bio-Rad Poly-Prep column), and radioactivity was determined by liquid scintillation counting. The [³H]IP_x accumulation was expressed as the percentage of the control value measured in the absence of agonist.

Ligand Binding Assays. We determined the K_D value of [O-methyl-³H]yohimbine from radioligand saturation binding experiments in whole α_2A-H cells. Nonspecific binding was defined with 10 μM yohimbine. We report the K_D value of [O-methyl-³H]yohimbine at α_2A-adrenoceptors in α_2A-H cells as 2.67 nM.

To determine relative receptor numbers in α_2A-H, SH-SY5Y, and 5HT_2A-SH-SY5Y cells after drug pretreatments as described above, we performed single-dose saturation binding assays. Cells were plated and incubated for at least 18 h at 37°C in 5% CO₂ as before but without a radioligand. The cells were then rinsed once with reduced serum minimum essential medium (UltraMEM), whereafter the assay was initiated by adding UltraMEM with 0 or 10 μM reversible competitive antagonist (to define nonspecific binding) plus the appropriate concentration of radioligand. After a 30-min incubation at 37°C in 5% CO₂, the medium was aspirated and the cells rinsed twice with ice-cold PBS, and the reaction was terminated with 1-ml ice-cold 5% (w/v) TCA and was let to stand for at least 30 min at 4°C to lyse the cells. The TCA from each well was then transferred directly into scintillation vials, and the radioactivity was counted. The relative α_2A-adrenoceptor concentrations in α_2A-H cells (after the various pretreatments with the appropriate irreversible competitive antagonists and washing procedure as described above) were determined from specific binding of 5 nM [O-methyl-³H]yohimbine (83–92 Ci/mmol; Amersham Biosciences UK, Ltd.). Nonspecific binding was defined with 10 μM yohimbine. Likewise, the mACh receptor-relative concentrations in SH-SY5Y cells were determined from specific binding by 5 nM [N-methyl-³H]4-DAMP (80.5 Ci/mmol; PerkinElmer Life and Analytical Sciences, Boston, MA). Nonspecific binding was defined with 10 μM atropine. The 5-HT_2A-receptor relative concentrations in 5-HT_2A-SH-SY5Y cells were determined from specific binding by 5 nM [³H]ketanserin (63.3 Ci/mmol; PerkinElmer Life and Analytical Sciences). Nonspecific binding was defined with 10 μM ritanserin.

To determine the pK_i value of the α_2A-adrenoceptor agonist UK 14,304 at α_2A-adrenoceptors in α_2A-H cells after pretreatment with 10 μM yohimbine and 100 μM phenoxybenzamine or 100 μM benextramine as described above, we performed competition binding assays in whole α_2A-H cells against 5 nM [O-methyl-³H]yohimbine. In preparation for the competition binding experiments, the cells were plated and incubated as before but without a radioligand. Cells were then rinsed once with UltraMEM, whereafter the assay was initiated by adding UltraMEM with 5 nM [O-methyl-³H]yohimbine and different concentrations of UK 14,304. After a 30-min incubation at 37°C in 5% CO₂, the medium was aspirated, the washing procedure as described above was followed, and the reaction was terminated with 1 ml of 5% (w/v) TCA and let to stand for at least 30 min to allow the cells to lyse. The TCA from each well was then transferred directly into scintillation vials, and the radioactivity (bound [O-methyl-³H]yohimbine) was counted.

Preparing Membranes from α_2A-H Cells. After the appropriate pretreatment of whole α_2A-H cells (see above), the cells were washed twice with PBS, the cell monolayer was loosened with ethylenediaminetetraacetic acid (EDTA) in PBS (0.02% w/v), and the cells were scraped from the culture flask surface with a cell scraper. The cell suspension was centrifuged in a bench top centrifuge (5411g, 4°C, 15 min), the supernatant was discarded, and the pellet was washed twice with ice-cold PBS, whereafter the pellet was resuspended in 1 mM Tris buffer (pH 7.4). The cell suspension was tumbled for 15 min at 4°C, homogenized with a Teflon homogenizer, and centrifuged at 1000g in a Beckman ultracentrifuge at 4°C for 15 min. The supernatant was collected and kept on ice while the pellet was resuspended in the Tris buffer and the preceding procedure repeated to collect all protein. The resulting supernatants were centrifuged at 40,000g in a Beckman ultracentrifuge at 4°C for 60 min. The resulting pellet was resuspended and homogenized in TME buffer (50 mM Tris, 10 mM MgCl₂, and 1 mM EDTA, pH 7.4). Protein concentrations were determined with the Bradford method (Bradford, 1976), using bovine serum albumin as a standard and determining absorbance with a 96-well plate reader and a 560-nm filter (Labsystems Multiskan RC; Thermo Electron Corporation, Waltham, MA). Snap-frozen aliquots were stored at -86°C for up to 4 weeks.

Measuring [³⁵S]GTPγS Binding in α_2A-H Cell Membranes. The [³⁵S]GTPγS binding assay to α_2A-H membranes was based on the procedures previously described (Sternweis and Robishaw, 1984; Yang and Lanier, 1999) but adapted for the present study. Immediately before the [³⁵S]GTPγS binding assays, the membranes were thawed on ice, and the protein concentrations were adjusted to 1.4 μg/μl by adding the appropriate volume of TME dilution buffer. Freshly prepared assay buffer (0.5 nM [³⁵S]GTPγS, 1 μM GDP, 50 mM Tris, 5 mM MgCl₂, 1 mM EDTA, 100 mM NaCl, and 1 mM dithiothreitol (DTT), pH 7.4, at 4°C) was used to prepare a concentration range of UK 14,304. In each test tube, 90 μl of the assay buffer containing the appropriate concentration of UK 14,304 was added and heated for 5 min at 25°C in a water bath. Thereafter, 10-μl membrane was added and incubated for 40 min at 25°C in the water bath. After the incubation, 3 ml of ice-cold TMN washing buffer (20 mM Tris, 25 mM MgCl₂, and 100 mM NaCl, pH 7.4) was added to each membrane sample, and it was immediately filtered through Whatman GF/C filters (Kent, UK), using a Hoefer filtration rig under vacuum. Each sample was washed three times with ice-cold TMN buffer. Nonspecific binding of [³⁵S]GTPγS was defined by samples with assay buffer but no membrane protein. Filters were air-dried, whereafter radioactivity was determined by liquid scintillation counting. The specific UK 14,304-induced binding of [³⁵S]GTPγS to the membranes was determined by subtracting the nonspecific binding from the total binding.

Assessment of Binding of [³⁵S]GTPγS to Gα_o Protein. The [³⁵S]GTPγS binding assay to Gα_o protein was based on the procedures previously (Sternweis and Robishaw, 1984; Graber et al., 1992) but adapted to accommodate the pretreatment with benextramine. The Gα_o protein was diluted to a concentration of 8 ng/μl with a sample dilution buffer (10 mM HEPES, 1 mM EDTA, 1 mM DTT, and 0.1% w/v nonaethylene glycol monododecyl ether, pH 8.0) and kept on ice. Before measuring the binding of [³⁵S]GTPγS to Gα_o, the protein was pretreated with either 0 M or 100 μM benextramine for 2 h at 4°C or for 30 min at 25°C. Immediately after pretreatment, the benextramine pretreatment groups were divided into 10-μl samples in test tubes, and 10 μl of dilution buffer was added on ice. Thereafter, a 20-μl binding cocktail (0.8 nM [³⁵S]GTPγS, 2 μM GTPγS in 50 mM HEPES, 1 mM EDTA, 40 mM MgCl₂, 200 mM NaCl, and 1 mM DTT, pH 8.0) was added to each test tube at the indicated temperature (4°C or 25°C). After incubation, 3 ml of ice-cold TMN washing buffer (pH 8.0) was added to each sample, and the bound [³⁵S]GTPγS separated from the free fraction by rapid filtration through type HAWP nitrocellulose membrane filters (Millipore, Billerica, MA), placed on a Hoefer filtration rig under vacuum. Each sample was washed three times with ice-cold TMN buffer. Nonspecific binding of [³⁵S]GTPγS was defined by samples with assay buffer but no Gα_o protein. Filters were air-dried, whereafter radioactivity was determined by liquid scintillation counting.

Data Analysis. Data from all studies were obtained from triplicate observations from at least three separate, comparable experiments, and results are expressed as S.E.M. Semilogarithmic dose-response curves were constructed as least-square nonlinear fits, utilizing the computer software GraphPad Prism (version 3.03 for Microsoft Windows; GraphPad Software, San Diego, CA). Where data of dose-response curves are expressed as a percentage of control without drug, no statistical significant differences were found in the control values of second messenger accumulation among the different pretreatments with each irreversible antagonist. One-site competition binding curves were constructed as least-square nonlinear fits, and the K_i values were calculated from the IC₅₀ value, applying the K_D value of the radioligand into the built-in Cheng-Prusoff correction of the software. Student's two-tailed, unpaired t test was implemented to compare the E_max, pEC₅₀, and pK_i values. All reported p values are after the Bonferroni correction for multiple comparisons (when appropriate), and a value of p < 0.05 was taken as statistically significant.

Results

Specific Binding before and after Pretreatment with the Irreversible Antagonists, with or without Receptor Protection.Figure 1, A and B, depicts the specific binding of [O-methyl-³H]yohimbine to α_2A-adrenoceptors in α_2A-H cells before and after pretreatment with different concentrations of phenoxybenzamine for 20 min, either without protection of the α_2A-adrenoceptors (0 M yohimbine) or with protection of the α_2A-adrenoceptors (10 μM yohimbine). It can be seen in Fig. 1A that increasing concentrations of phenoxybenzamine progressively decreased specific binding (i.e., decreased α_2A-adrenoceptor concentration) (p < 0.001 for comparison of all bars) in the absence of receptor protection. This decrease in specific binding is not seen in Fig. 1B (p > 0.05) when the α_2A-adrenoceptors are adequately protected from phenoxybenzamine by yohimbine. A small reduction in the specific binding of [O-methyl-³H]yohimbine could be noted in bar b₁ (84 ± 1.5%) as compared with bar a₁ (100 ± 1.1%), which may represent some residual cold (nonradioactive) yohimbine after the washing steps and of which the negligible effect on receptor-mediated functional response is illustrated further below (Fig. 2).

Likewise, Fig. 1, C and D depicts the specific binding of [O-methyl-³H]yohimbine to α_2A-adrenoceptors in α_2A-H cells before and after pretreatment with different concentrations of benextramine for 20 min, either without protection of the α_2A-adrenoceptors (0 M yohimbine) or with protection of the α_2A-adrenoceptors (10 μM yohimbine). It can be seen in Fig. 1C that increasing concentrations of benextramine progressively decreased specific binding (p ≤ 0.01 for comparison of all bars, except for comparison of bars c₃ and c₄ where p = 0.07) in the absence of receptor protection. This prominent decrease in specific binding is not seen in Fig. 1D (p > 0.05 for comparison of bars when the α_2A-adrenoceptors are protected from benextramine by yohimbine, except for comparison of bars d₁ and d₃ or bars d₁ and d₄ where p = 0.02 or p = 0.01, respectively). As with phenoxybenzamine, a small reduction in the specific binding of [O-methyl-³H]yohimbine should be noted in bar d₁ (87 ± 3.4%) as compared with bar c₁ (100 ± 1.5%), which may represent some residual cold (nonradioactive) yohimbine after the washing steps and of which the negligible effect on receptor-mediated functional response is illustrated further below (Fig. 2). The reductions in the specific binding of [O-methyl-³H]yohimbine observed in bars d₃ (73 ± 2.9%) and d₄ (71 ± 2.8%) compared with bar d₁ (87 ± 3.4%) may be due to incomplete receptor protection by 10 μM yohimbine but are practically insignificantly small in terms of its expected effect on receptor-mediated response (illustrated further below, Fig. 2). From the bars in Fig. 1, B and D (when compared with the corresponding bars in Fig. 1, A and C), it follows that 10 μM yohimbine prevented the irreversible antagonist from reducing receptor binding at α_2A-adrenoceptors.

Similarly, Fig. 1, E and F depicts the specific binding of [N-methyl-³H]4-DAMP to mACh receptors in SH-SY5Y cells before and after pretreatment with different concentrations of 4-DAMP mustard for 20 min, either without protection of the mACh receptors (0 M atropine) or with protection of the mACh receptors (10 μM atropine). It can be seen in Fig. 1E that 4-DAMP mustard decreased specific binding (p < 0.01 for comparison of bars e₁ and e₂ or of bars e₁ and e₃) in the absence of receptor protection. The difference in specific binding between 10 and 100 nM 4-DAMP mustard was not statistically significant (p = 0.11 for comparison of bars e₂ and e₃). As with phenoxybenzamine and benextramine, a small reduction in the specific binding of the radioligand should be noted in bar f₁ (84 ± 2.5%) as compared with bar e₁ (100 ± 3.1%), which may represent some residual cold (nonradioactive) atropine after the washing steps and of which the negligible effect on receptor-mediated functional response is illustrated further below (Fig. 2). From the bars in Fig. 1F (when compared with the corresponding bars in Fig. 1E), it can be seen that increasing concentrations of 4-DAMP mustard with receptor protection by atropine did not cause a similar prominent decrease in receptor number (p > 0.05 for comparison of bars f₁ and f₂, bars f₁ and f₃, and bars f₂ and f₃), and it follows that 10 μM atropine prevented the irreversible antagonist from reducing receptor binding at mACh receptors.

Specific Binding and Second Messenger Formation before and after Pretreatment with the Reversible Antagonists.Figure 2A depicts the specific binding of [O-methyl-³H]yohimbine to α_2A-adrenoceptors in α_2A-H cells after pretreatment with 0 or 10 μM yohimbine, but no irreversible antagonist. Although there was a difference in specific binding of [O-methyl-³H]yohimbine of about 13.4% (p = 0.01) between the 0 or 10 μM yohimbine pretreatment groups, this difference appears to be functionally insignificant (i.e., negligible effect on receptor-mediated functional response) as shown by the corresponding functional data in Fig. 2B, depicting the dose-response curves of the α_2A-adrenoceptor full agonist UK 14,304 in α_2A-H cells after the pretreatments with 0 or 10 μM yohimbine. These dose-response curves were practically superimposed with no statistically significant difference between the pEC₅₀ values (p = 0.78) or E_max values (p = 0.37). This is predicted also by theory, where a reduction in receptor concentration of 13.4% in a system with significant spare receptors, should not significantly alter the EC₅₀ value of a dose-response curve.

Likewise, Fig. 2C depicts the specific binding of [N-methyl-³H]4-DAMP to mACh receptors in SH-SY5Y cells after pretreatment with 0 or 10 μM atropine, but no irreversible competitive antagonist. Although there was a difference in specific binding of [N-methyl-³H]4-DAMP of about 16% (p = 0.02) between the 0 or 10 μM atropine pretreatment groups, this difference appears to be functionally insignificant (i.e., negligible effect on receptor-mediated functional response) as shown by the corresponding functional data in Fig. 2D, depicting the dose-response curves of the mACh receptor full agonist methacholine in SH-SY5Y cells after pretreatments with 0 or 10 μM atropine. The dose-response curves were practically superimposed with no statistically significant difference between the pEC₅₀ values (p = 0.50) or E_max values (p = 0.76). Again, this is in line with theoretical predictions, as mentioned above.

A similar approach was intended for 5-HT_2A receptors, by measuring the specific binding of [³H]ketanserin to 5-HT_2A receptors in 5-HT_2A-SH-SY5Y cells after pretreatment with 0 or 10 μM ritanserin but no irreversible competitive antagonist. However, results from radioligand binding studies showed that, under the experimental conditions used, 5-HT_2A receptors could not be significantly protected against benextramine with ritanserin. Therefore, none of the dose-response curves in this cell line, as presented further below in this manuscript, were constructed in the presence of ritanserin and there was no need to include such curves in Fig. 2.

Agonist-Mediated Responses before and after Pretreatment with the Irreversible Antagonists with or without Receptor Protection.Figure 3A depicts dose-response curves of the α_2A-adrenoceptor full agonist UK 14,304 in α_2A-L cells after pretreatment with different concentrations of phenoxybenzamine for 20 min. As the concentration of phenoxybenzamine was increased, the dose-response curves of UK 14,304 shifted to the right (1 and 10 μM), and at the highest concentration (100 μM), the E_max value was also suppressed (E_max a₄/a₁ ratio = 0.37, p < 0.05). Similarly, Fig. 3C depicts dose-response curves of UK 14,304 in α_2A-L cells after pretreatment with different concentrations of benextramine for 20 min. As the concentration of the benextramine was increased, the dose-response curves of UK 14,304 initially shifted to the right (1 μM benextramine: EC₅₀ shift c₂/c₁ ratio = 23, p < 0.05), and at higher concentrations (10 and 100 μM), the E_max value was suppressed (E_max c₃/c₁ ratio = 0.56, p < 0.05; E_max c₄/c₁ ratio = 0.29, p < 0.05). However, when the α_2A-adrenoceptors were protected from binding to the irreversible competitive antagonist, the antagonism by the irreversible competitive antagonists were not abolished, as illustrated in Fig. 3, B and D. For phenoxybenzamine, the highest concentration (100 μM) showed a statistically significant, although small, rightward shift of the dose-response curve (EC₅₀ shift b₄/b₁ ratio = 3.1, p < 0.05). For benextramine, the highest concentration (100 μM) showed a relatively large rightward shift of the dose-response curve (EC₅₀ shift d₄/d₁ ratio = 138, p < 0.05), and a 10 μM concentration benextramine produced a small rightward shift of the dose-response curve (EC₅₀ shift d₃/d₁ ratio = 6.3, p < 0.05), but at 1 μM the rightward shift of the dose-response curve was not statistically significant (EC₅₀ shift d₂/d₁ ratio = 1.9; p > 0.05). The EC₅₀ values of all other curves differed statistically significantly (EC₅₀ shift d₃/d₂ ratio = 3.4, p < 0.05; EC₅₀ shift d₄/d₂ ratio = 74, p < 0.05; EC₅₀ shift d₄/d₃ ratio = 22, p < 0.05).

Figure 3E depicts dose-response curves of the mACh receptor full agonist methacholine in SH-SY5Y cells after pretreatment with different concentrations of 4-DAMP mustard for 20 min. The E_max value was statistically significantly suppressed at 100 nM 4-DAMP mustard for 20 min (E_max e₃/e₁ ratio = 0.75, p < 0.05) but not at 10 nM (E_max e₂/e₁ ratio = 0.89, p > 0.05). When the mACh receptors were protected from binding to the irreversible competitive antagonist, as presented in Fig. 3F, the dose-response curves were practically superimposed with no apparent changes in the antagonism.

Affinity of UK 14,304 for α_2A-Adrenoceptors before and after Pretreatment with Phenoxybenzamine or Benextramine. Since Fig. 3 indicated that irreversible antagonism by benextramine or phenoxybenzamine was not abolished by α_2A-adrenoceptor protection, it was investigated whether or not the irreversible noncompetitive antagonism by phenoxybenzamine or benextramine (i.e., observed rightward shift of dose-response curves of UK 14,304 after receptor protection) was due to an altered affinity of the α_2A-adrenoceptor for the full agonist UK 14,304. The pK_i value of UK 14,304 was determined from competition binding curves against 5 nM [O-methyl-³H]yohimbine in α_2A-H cells before and after pretreatment with 100 μM phenoxybenzamine (Fig. 4A) or benextramine (Fig. 4B) for 20 min, with α_2A-adrenoceptor protection by 10 μM yohimbine. Irreversible antagonist pretreatment did not alter the α_2A-adrenoceptor receptor affinity for UK 14,304 and apparently not the binding profile as well, since the competition binding curves before and after the irreversible antagonist were practically superimposed. The pK_i values were determined as pK_i = 6.64 ± 0.05 versus pK_i = 6.77 ± 0.06 after 0 or 100 μM phenoxybenzamine, respectively (p = 0.16), and pK_i = 6.91 ± 0.20 versus pK_i = 6.82 ± 0.13 after 0 or 100 μM benextramine, respectively (p = 0.73).

[³⁵S]GTPγS Binding to Gα_i Proteins in α_2A-H Membranes after Benextramine Pretreatment, with or without α_2A-Adrenoceptor Protection. Since competition binding studies with UK 14,304 above revealed that benextramine does not affect agonist affinity, it follows that the noncompetitive irreversible antagonism should involve the inhibition of the α_2A-adrenoceptor signal-transduction. We therefore conducted G_i protein-[³⁵S]GTPγS binding studies to investigate whether the noncompetitive antagonism by benextramine is due to irreversible binding to a site at the receptor and/or G protein level, or whether the binding is located downstream to the G protein in the signal transduction system. As before, α_2A-H cells were pretreated with 0 or 100 μM benextramine for 20 min at 37°C, plus 0 M yohimbine, or 10 μM yohimbine to protect the α_2A-adrenoceptors from binding to benextramine. After washing procedures to remove all unbound and reversibly bound drugs (as described above), membranes were prepared. Figure 5 depicts dose-response curves of the full α_2A-adrenoceptor agonist UK 14,304, measuring [³⁵S]GTPγS binding to G proteins in these membranes. In Fig. 5A, a concentration-dependent response of UK 14,304 was observed in membranes pretreated with drug-free medium (i.e., control curve a₁, pretreated with 0 μM yohimbine plus 0 μM benextramine). As expected, when the membranes were pretreated with benextramine (i.e., curve a₂, pretreated with 0 μM yohimbine plus 100 μM benextramine), the response to UK 14,304 was totally abolished (from E_max, p < 0.01). When the membranes were pretreated with yohimbine alone (Fig. 5B, curve b₁, pretreated with 10 μM yohimbine plus 0 μM benextramine), a dose-response curve was obtained similar to curve a₁, confirming previous results that residual yohimbine after the washing procedure did not affect receptor function, as evidenced by Fig. 2. Importantly, however, when the α_2A-adrenoceptors were protected from binding to benextramine during the pretreatment (Fig. 5B, curve b₂, pretreated with 10 μM yohimbine plus 100 μM benextramine), the dose-response curve of UK 14,304 was completely suppressed as was found without receptor protection, as represented in curve a₂ (from E_max, p < 0.01). Results suggest that the protection of α_2A-adrenoceptors with yohimbine does not prevent benextramine from inhibiting the binding of [³⁵S]GTPγS to the membranes and, therefore, G_i-mediated signaling through the receptor.

Binding of [³⁵S]GTPγS to Gα_o before and after Incubation with Benextramine at Different Incubation Times and Temperatures. Gα_o is a relatively stable GTP binding protein of the G_i/o family and has been shown to constitutively (in the absence of receptor and agonist) bind guanine nucleotides (Sternweis and Robishaw, 1984). To investigate whether benextramine directly binds to the G protein nucleotide binding site to inhibit GTP binding, Gα_o was pretreated with 0 or 100 μM benextramine for 120 min at 4°C, or for 30 min at 25°C, whereafter the [³⁵S]GTPγS binding was measured, as presented in Fig. 6. After pretreatment with 0 or 100 μM benextramine for 120 min at 4°C, the specific binding obtained is depicted in Fig. 6A. As can be seen from Fig. 6A, benextramine has not significantly reduced the binding of [³⁵S]GTPγS to Gα_o (p = 0.80), and the amount of [³⁵S]GTPγS bound after pretreatment with 0 or 100 μM benextramine was measured as 17.2 ± 3.0 and 16 ± 3.1 fmol/ng membrane protein, respectively. Likewise, after pretreatment with 0 or 100 μM benextramine for 30 min at 25°C, the specific binding obtained is depicted in Fig. 6B. As can be seen from Fig. 6B, benextramine has not significantly reduced the binding of [³⁵S]GTPγS to Gα_o (p = 0.54), and the amount of [³⁵S]GTPγS bound after pretreatment with 0 or 100 μM benextramine was measured as 2.9 ± 0.6 and 3.7 ± 1.1 fmol/ng membrane protein, respectively. Although pretreatment with benextramine at both incubation times and temperatures did not significantly decrease the binding of [³⁵S]GTPγS to Gα_o, overall binding is significantly lower at the shorter incubation time but higher temperature. These results suggest that incubation of Gα_o at higher temperatures with shorter incubation times reduce the binding of [³⁵S]GTPγS to the nucleotide binding site of the G protein. According to Sternweis and Robishaw (1984), purified G proteins (such as Gα_o), could be stored at -80°C or even on ice (4°C) for several weeks with little or no loss of binding activity. Thus the observed overall significant reduction in binding of [³⁵S]GTPγS to Gα_o after pretreatment with benextramine for 30 min at 25°C may be due to reduced biological activity of the Gα_o protein, since it is well known that higher temperatures cause conformational changes that may modulate or destroy its biological activity.

G_s-Mediated [³H]cAMP Accumulation in α_2A-H Cells after Pretreatment with Benextramine, with or without α_2A-Adrenoceptor Protection. Since our previous observations investigated the noncompetitive antagonism by benextramine at α_2A-adrenoceptors via a G_i-mediated response only, it was important to investigate whether similar noncompetitive antagonism by benextramine can be observed via a G_s protein-mediated response but with the same receptors and effector (adenylyl cyclase). It has been shown before that after PTX treatment of α_2A-H cells, the α_2A-adrenoceptors couple to G_s proteins to stimulate adenylyl cyclase activity (Brink et al., 2000). Figure 7 displays dose-response curves of UK 14,304 in PTX-treated whole α_2A-H cells after pretreatment with 0 or 100 μM benextramine plus 0 M (Fig. 7A) or 10 μM yohimbine (Fig. 7B) for 20 min at 37°C. Without protection of the α_2A-adrenoceptors (Fig. 7A), 100 μM benextramine abolished the agonist-induced response (E_max for a₁ = 100 ± 13.1% and E_max for a₂ = -3.6 ± 7.5%, p < 0.01). When the α_2A-adrenoceptors were protected by 10 μM yohimbine, the G_s-mediated response was only partially inhibited (E_max of curve b₁ = 100 ± 18.9% and E_max of curve b₂ = 53.0 ± 5.8%, p < 0.05) and the EC₅₀ value remained unchanged.

Agonist-Induced, G_q-Mediated [³H]IP_x Accumulation in SH-SY5Y and 5-HT_2A-SH-SY5Y Cells after Pretreatment with Benextramine, with or without Receptor Protection. We investigated whether benextramine pretreatment would antagonize the response in a signal transduction system with a receptor type, G protein type and effector totally different from those in previous observations. In SH-SY5Y cells mACh receptors signal through G_q proteins to activate PLC. Likewise, in 5-HT_2A-SH-SY5Y cells, 5-HT_2A receptors signal through G_q proteins to activate PLC. Figure 8 depicts dose-response curves of methacholine in SH-SY5Y cells and of serotonin in 5-HT_2A-SH-SY5Y cells, measuring agonist-stimulated [³H]IP_x accumulation, after pretreatment of the cells with 0 or 100 μM benextramine for 20 min at 37°C.

In SH-SY5Y cells, benextramine partially suppressed the methacholine-mediated response (E_max values were 100 ± 17.2% for curve a₁ and 25.2 ± 8.6% for curve a₂, p < 0.05). The EC₅₀ value, however, remained unchanged. Likewise, in 5-HT_2A-SH-SY5Y cells benextramine partially suppressed the serotonin-mediated response (E_max values were 100 ± 23.2% for curve b₁ and 34.7 ± 8.3% for curve b₂, p < 0.05). Again, the EC₅₀ value remained unchanged.

Binding Data for [³H]4-DAMP at mACh Receptors and [³H]Ketanserin at 5-HT_2A Receptors after Pretreatment with Benextramine, with or without Receptor Protection. Since benextramine inhibits the [³H]IP_x accumulation in both SH-SY5Y and 5-HT_2A-SH-SY5Y cells (Fig. 8), it was important to determine whether this inhibition could be ascribed to a reduction in receptor number, or whether noncompetitive antagonism is displayed by benextramine.

After pretreatment of SH-SY5Y whole cells with 0 M or 100 μM benextramine for 20 min at 37°C, plus 0 M or 10 μM atropine (>1000 × K_i value, to protect mACh receptors from binding to benextramine), radioligand binding assays were conducted to determine the relative number of receptors. Figure 9, A and B depict the specific binding of [³H]4-DAMP to mACh receptors in SH-SY5Y cells after pretreatment with 0 M or 100 μM benextramine. In the absence of receptor protection by atropine (Fig. 9A), benextramine pretreatment significantly reduced the specific binding of [³H]4-DAMP from 100 ± 4.4% (bar a₁) to 35.6 ± 3.3% (bar a₂) (p < 0.001). However, after pretreatment with 10 μM atropine to protect the mACh receptors from benextramine (Fig. 9B), the specific binding was also significantly decreased from 76.8 ± 3.2% (bar b₁) to 42.3 ± 2.2% (bar b₂) (p < 0.001).

Likewise, after pretreatment of 5-HT_2A-SH-SY5Y whole cells with 0 M or 100 μM benextramine for 20 min at 37°C, plus 0 M or 10 μM ritanserin (>1000 × K_i value, to protect 5-HT_2A receptors from binding to benextramine), radioligand binding assays were conducted to determine the relative number of receptors. Figure 9, C and D depicts the specific binding of [³H]ketanserin to 5-HT_2A receptors in 5-HT_2A-SH-SY5Y cells after pretreatment with 0 M or 100 μM benextramine. In the absence of receptor protection by ritanserin (Fig. 9C), benextramine pretreatment significantly reduced the specific binding of [³H]ketanserin from 100 ± 12.3% (bar c₁) to 31.1 ± 4.9% (bar c₂) (p < 0.01). However, after pretreatment with 10 μM ritanserin to protect the 5-HT_2A receptors from benextramine (Fig. 9D), the specific binding was also significantly decreased from 85.4 ± 19.3% (bar d₁) to 30.0 ± 6.0% (bar d₂) (p < 0.05).

Discussion

The Experimental Conditions and Pretreatments Are Suitable for the Evaluation of Noncompetitive Mechanisms by the Irreversible Competitive Antagonists. As expected, phenoxybenzamine (Fig. 1A) and benextramine (Fig. 1C) pretreatments decrease the α_2A-adrenoceptor number in α_2A-H cells and that 4-DAMP mustard (Fig. 1E) decreased the mACh receptor number in SH-SY5Y cells. The binding of phenoxybenzamine (Fig. 1B) and benextramine (Fig. 1D) to α_2A-adrenoceptors were effectively prevented by yohimbine, whereas atropine effectively prevented the binding of 4-DAMP mustard to mACh receptors (Fig. 1F). Therefore, the irreversible antagonists were used at sufficient concentrations and incubation times to eliminate a significant fraction of the operational receptors and the concentrations of the reversible antagonists used were sufficient to protect these receptors from significant binding by the irreversible antagonists.

In addition, the washing procedure after the pretreatments was sufficient to remove the reversibly bound antagonists. Results in Fig. 2A suggest that the small decrease in the specific binding of [O-methyl-³H]yohimbine to α_2A-adrenoceptors in α_2A-H cells after yohimbine pretreatment is functionally insignificant, since dose-response curves of UK 14,304 at α_2A-adrenoceptors in α_2A-L cells before and after yohimbine pretreatment were practically superimposed (Fig. 2B). Likewise, a small reduction in the specific binding of [N-methyl-³H]4-DAMP to mACh receptors in SH-SY5Y cells after atropine pretreatment (Fig. 2C), is functionally insignificant (Fig. 2D). The pretreatment procedures are therefore suitable for further investigation of any irreversible noncompetitive mechanisms of antagonism by the respective irreversible competitive antagonists.

Importantly, it was shown previously (Brink et al., 2000) and confirmed (data not shown) that in control Neo cells (control cell line, stably transfected with the empty vector without the α_2A-adrenoceptor) UK 14,304 does not elicit any change in [³H]cAMP accumulation, suggesting that all observed UK 14,304-mediated responses in α_2A-H and α_2A-L cells are mediated via α_2A-adrenoceptors.

Benextramine and Phenoxybenzamine, but Not 4-DAMP Mustard, Display Irreversible Noncompetitive Antagonism after 20-min Incubation Time. When receptor number in a system with spare receptors is progressively reduced by an irreversible antagonist, theory predicts that the dose-response curve of the agonist should progressively shift to the right until all spare receptors are eliminated, whereafter the maximal response is reduced (Furchgott, 1966; Kenakin, 1997). The results with phenoxybenzamine and benextramine in the current study followed this pattern (Fig. 3, A and C). In the absence of spare receptors, 4-DAMP mustard only suppressed the maximal response to methacholine (Fig. 3E). If the primary binding sites could be sufficiently protected by a reversible competitive antagonist, theory predicts that the dose-response curves of the agonist before and after antagonist pretreatment (plus sufficient washout so that any residual competitive antagonist does not influence receptor-mediated function) should superimpose. When dose-response curves do not superimpose under these conditions, it suggests noncompetitive antagonism by the irreversible antagonist. Results from the present study therefore suggest that benextramine, and to a lesser degree phenoxybenzamine, display irreversible noncompetitive antagonism at α_2A-adrenoceptors (Fig. 3, B and D). 4-DAMP mustard, however, does not display irreversible noncompetitive antagonism at mACh receptors (Fig. 3F) under the specific experimental conditions used.

The Noncompetitive Antagonism by Phenoxybenzamine and Benextramine Does Not Alter the Affinity of α_2A-Adrenoceptors for UK 14,304. The affinity (pK_i values) of UK 14,304 at α_2A-adrenoceptors in α_2A-H cells was not affected by the pretreatments with phenoxybenzamine or benextramine (Fig. 4). A logical alternative explanation for the observations in Fig. 3 could be that both drugs may modulate the signal transduction by exhibiting allotopic interactions at the α_2A-adrenoceptor macromolecule or by inhibiting G_i proteins or adenylyl cyclase downstream in the signal transduction pathway.

The Noncompetitive Irreversible Antagonism by Benextramine at α_2A-Adrenoceptors Can Be Explained by the Inhibition of Signaling at the Receptor-G_i Protein Level. Noncompetitive antagonism by benextramine at α_2A-adrenoceptors is also observed earlier in the signal transduction system at the G protein level, when measuring agonist-induced increase in [³⁵S]GTPγS binding in cell membranes (Fig. 5). Since GTPγS is a nonhydrolyzable GTP analog, it is not likely that altered GAP function (enhanced GTP hydrolysis to GDP with G protein inactivation) may explain this observation. α_2A-Adrenoceptors have previously been shown to couple with high efficiency to PTX-sensitive G_i proteins (Eason et al., 1992; Chabre et al., 1994), as is present in the CHO-K1 cells used (Eason et al., 1992), and with much lower efficiency to G_s and G_q proteins (Chabre et al., 1994; Wade et al., 1999; Brink et al., 2000). It is therefore reasonable to assume that the observed [³⁵S]GTPγS binding results predominantly from G_i protein activation, as confirmed by unpublished data from the laboratory of Dr. Richard Neubig (Department of Pharmacology, University of Michigan, Ann Arbor, MI; personal communication) that the α_2A-adrenoceptor-stimulated [³⁵S]GTPγS binding to α_2A-H cell membranes is completely abolished by PTX.

Noncompetitive Antagonism by Benextramine Does Not Involve Direct Inhibition of GTP Binding to the G Protein. From the results in Fig. 6, it follows that the pretreatment of purified Gα_o with benextramine at a relatively high concentration and at two extreme and distinct temperatures and incubation times does not inhibit the constitutive binding of [³⁵S]GTPγS to Gα_o. Gα_o and Gα_i belong to the same G protein family (i.e., G_i/o family) (Sternweis and Robishaw, 1984), are PTX-sensitive but cholera toxin-insensitive, and display similar agonist-mediated α_2A-adrenoceptor signaling properties as measured by [³⁵S]GTPγS binding (Yang and Lanier, 1999). Therefore, it can be reasonably assumed that benextramine will also not modulate [³⁵S]GTPγS binding to Gα_i, and we propose that the observed noncompetitive antagonism by benextramine at α_2A-adrenoceptors most likely results from an inhibition of receptor-Gα_i protein coupling. Importantly, however, the current data do not exclude the possibility that G proteins may be affected in other ways (e.g., altered GDP binding, modifications at the receptor, or effector coupling regions) or that different G protein types (i.e., Gα_i, Gα_s, or Gα_q proteins) may be affected differently. The data also do not exclude the possibility that other signaling molecules or downstream signaling entities, such as adenylyl cyclase or PLC, could also be affected by benextramine. Further clarification of the mechanisms of action of benextramine and related drugs may form the basis for future studies.

Benextramine Also Irreversibly Inhibits Signaling of α_2A-Adrenoceptors via G_s but to a Lesser Extent Than via G_i. α_2A-Adrenoceptors also signal through G_s proteins to activate adenylyl cyclase, although with 1000 times lower coupling efficiency than to G_i (Chabre et al., 1994). This is evident after PTX treatment in systems with a high α_2A-adrenoceptor concentration, measuring G_s-mediated increase in cAMP accumulation (Wade et al., 1999). The partial inhibition of the agonist-induced increase in [³H]cAMP accumulation after receptor protection in PTX-treated cells suggest that benextramine displays irreversible noncompetitive antagonism at α_2A-adrenoceptors also when measuring G_s-mediated responses (Fig. 7).

Taken together, it can be postulated that benextramine most likely exhibits allotopic interactions at the α_2A-adrenoceptor macromolecule or binds to a common GPCR coupling site on the G_i and G_s proteins. The noncompetitive antagonism by benextramine at α_2A-adrenoceptors is substantially less with G_s-than with G_i-mediated responses, suggesting that this action is dependent on the G protein type involved in the GPCR signaling. This complies with data suggesting distinct basic residues of the α_2A-AR that mediate G_i and G_s activation, respectively (Wade et al., 1999). Also, whereas classical irreversible antagonists such as the haloalkylamines are believed to form uncomplicated covalent bonds with the receptor, it has been proposed that benextramine inactivates receptors at distinct sites via a disulfide-thiol interchange reaction (Melchiorre et al., 1979; Brasili et al., 1980, 1986; Melchiorre, 1981; Melchiorre and Gallucci, 1983; Giardinà et al., 1996), thereby affecting the stereochemical properties of cysteine amino acid residues and probably disrupting the coupling of the α_2A-AR to G_i more than to G_s.

Benextramine Irreversibly Inhibits Signaling of Both Muscarinic Acetylcholine (mACh) Receptors and 5-HT_2A Receptors via G_q. Both mACh receptors (Sorensen et al., 1999) and 5-HT_2A receptors (Berridge, 1993) have been shown to signal through PTX-insensitive G_q proteins to activate PLC. We have also confirmed that the agonist-induced, mACh receptor-mediated PLC activation in SH-SY5Y cells is abolished by the PLC inhibitor U-73122 (Bleasdale and Fisher, 1993) and also that this mACh receptor-response is PTX-insensitive (data not shown). Data from the present study show that benextramine irreversibly inhibits the methacholine-induced G_q-mediated signaling of mACh receptors in SH-SY5Y cells (Fig. 8A) but that these receptors could not be protected by 10 μM atropine (i.e., >1000 × K_D value) (Fig. 9B). These results suggest that benextramine exhibits allotopic interactions at mACh receptors. In contrast, the antagonism by benextramine at mACh receptors in isolated guinea pig atrium and ileum was reversible when measuring smooth muscle contraction (Benfey et al., 1980). Species and tissue differences may explain this apparent contradiction.

Benextramine also irreversibly inhibits the serotonin-induced, G_q-mediated signaling of 5-HT_2A receptors (Fig. 8B), which cannot be prevented by ritanserin at a relatively high concentration of 10 μM (>1000 × K_D value) (Fig. 9D). These results suggest that benextramine exhibits allotopic interactions, displaying irreversible noncompetitive antagonism at 5-HT_2A receptors. Antagonism by benextramine at 5-HT_1A receptors has been shown before (Stanton and Beer, 1997), and it therefore does not seem to act selectively at a particular 5-HT receptor subtype.

These data suggest that benextramine binds to sites distinct from the primary binding sites (i.e., nonoverlapping) at mACh and 5-HT_2A receptors. These allotopic interactions may cause sterical changes in the receptor macromolecule, hindering binding of the ligand to the primary binding site of the receptor.

Final Conclusions and Implications of This Study. We provide here evidence of irreversible noncompetitive antagonism by phenoxybenzamine and benextramine at α_2A-adrenoceptors in addition to their known irreversible specific antagonism. Agonist affinity is not influenced, but signal transduction may be modulated via allotopic interactions. Irreversible noncompetitive antagonism by benextramine at α_2A-adrenoceptors is G protein-dependent. Benextramine inhibits agonist-induced responses at three different GPCR types, involving signaling via three of the four main families of G proteins (i.e., G_i/o, G_s, and G_q/11 families) in different pharmacological systems, presumably by binding to distinct binding sites at the various receptor types.

Under the conditions specified, 4-DAMP mustard does not display irreversible noncompetitive antagonism at mACh receptors, whereas benextramine exhibits allotopic interactions at mACh receptors and 5-HT_2A receptors, preventing ligand binding to the agonist binding site. However, the current data does not exclude the possibility that receptor-G protein coupling is also inhibited at mACh receptors and 5-HT_2A receptors, as was suggested for α_2A-adrenoceptors. Further investigation is needed to clarify the exact mechanisms of action of benextramine.

Benextramine may prove to be a useful experimental tool in investigating the signaling mechanisms of G protein-coupled receptors. Also, the mere confirmation of syntopic interactions (e.g., by receptor-radioligand binding studies) does not rule out irreversible noncompetitive antagonism. Irreversible competitive antagonists should be utilized with due care for the implementation of any procedure where irreversible competitive antagonism is desired or intended (e.g., the Furchgott analysis). In such cases, pure irreversible competitive antagonism should be verified under the experimental conditions applied. Finally, we propose that irreversible noncompetitive mechanisms of antagonism could potentially contribute to the therapeutic response of phenoxybenzamine in the treatment of pheochromocytoma, but the specificity, scope, or clinical significance of this mechanism needs further investigation.