A Novel Phenylaminotetralin Radioligand Reveals a Subpopulation of Histamine H₁ Receptors

Materials and Methods

Chemicals.

(±)-trans-1-Phenyl-3-N,N-dimethylamino-1,2,3,4-tetrahydronaphthalene (H₂-PAT) was synthesized in a similar way to methods previously described (Wyrick et al., 1993). Briefly, the corresponding benzylstrylketone was cyclized to the tetralone intermediate and then reduced to the tetralol. This intermediate was tosylated, then converted to the free amine by reaction with sodium azide, and next followed by catalytic reduction to yield predominately the (±)-trans isomer that was purified as the HCl salt. The racemic mixture was resolved by derivatization with (−)-camphorsulfonic acid to afford the more active (−)-trans-H₂-PAT enantiomer, which was subsequently radiolabeled with tritiated methyliodide to yield [N-methyl-³H]-(−)-trans-H₂-PAT ([³H](−)-trans-H₂-PAT; specific activity = 85 Ci/mmol) (Wyrick et al., 1994). [³H]Mepyramine (30 Ci/mmol) was obtained from PerkinElmer Life Sciences (Boston, MA).d-Luciferin was purchased from Duchefa Biochemie BV (Haarlem, The Netherlands) and pNF-κB-Luc was from Stratagene (La Jolla, CA). Other compounds were obtained at the highest available purity from Sigma-Aldrich (St. Louis, MO) or Sigma/RBI (Natick, MA).

CHO Cell Culture.

Studies were conducted with CHOgpH₁ (Traiffort et al., 1994) or CHOhuH₁ (Smit et al., 1996). Cells were grown to confluency in 75-cm² flasks containing α-minimum essential medium (with 4500 g/l glucose), supplemented with 10% fetal bovine serum, 2 mM l-glutamine, and 0.1% penicillin/streptomycin (100 units/100 μg/ml), in a humidified atmosphere of air/CO₂ (95:5%) at 37°C.

Radioligand Binding Assays Using CHO Cell Membranes.

CHO cells were harvested by scraping from flasks with rinses (3 × 5 ml) of ice-cold phosphate-buffered saline (pH 7.5) and centrifuging at 1000g for 10 min. To prepare membranes, the resulting pellet was resuspended in 50 mM Na⁺-K⁺ phosphate buffer (pH 7.5; 25°C) at a volume of 1 ml/scraped flask. The suspension was homogenized using a Wheaton Teflon-glass homogenizer (Pittsburgh, PA) (10 strokes) and centrifuged at 50,000g (10 min, 4°C); the resulting pellet was resuspended in 50 mM Na⁺-K⁺ phosphate buffer (pH 7.5) at 1 ml/scraped flask (∼800 μg of protein/ml) and stored at −80°C.

For saturation isotherms using CHO cell membranes, 100 μl of stock cell membrane homogenate described above was incubated for 30 min at 25°C with 0.01 to 7.0 nM [³H]mepyramine or 0.01 to 1.0 nM [³H](−)-trans-H₂-PAT in a total assay volume of 400 μl (50 mM Na⁺-K⁺ phosphate buffer). Nonspecific binding for both radioligands was defined by addition of 10 μM triprolidine. Results were analyzed by nonlinear regression using the rectangular hyperbola curve-fitting algorithm in the microcomputer program Prism 3.0 (GraphPad, San Diego, CA) to determineK_D andB_max. Data were fit to one- and two-site models; however, no statistically significant (by F-test) improved fit was achieved using a two-site model. Each experimental condition was run in triplicate, and each experiment was repeated at least three times to determine S.E.M.

For competition binding assays using CHO cell membranes, 100 μl of the membrane preparation was incubated with 0.5 nM [³H]mepyramine or 0.1 nM [³H](−)-trans-H₂-PAT (about K_D) and 0.1 to 10,000 nM test ligand in a total assay volume of 400 μl (50 mM Na⁺-K⁺ phosphate buffer). Nonspecific binding was defined by the addition of 10 μM triprolidine for both radioligands. In assays where the effect of 1 mM GTP (Li₄⁺) on binding of ligands was measured, the buffer was 50 mM Tris HCl (pH 7.5) containing 4 mM MgCl₂.

Resulting inhibition data were analyzed by nonlinear regression using the sigmoidal curve-fitting algorithms in Prism 3.0 to determine concentration of competing ligands to inhibit specific binding of [³H](−)-trans-H₂-PAT by 50% (IC₅₀) and Hill slopes (n_H). Data were fitted to both one- and two-site models; however, no statistically significant (by F-test) improved fit was achieved using the two-site model. In light of the still, as yet, incompletely characterized nature of ligand interaction with the site, ligand affinity is expressed as an approximation ofK_i values by converting IC₅₀ data toK_0.5 values using the equationK_0.5 = IC₅₀/1 +L/K_D, where L is the concentration of radioligand having affinityK_D (Cheng and Prusof, 1973). Each experimental condition was run in triplicate, and each experiment was performed a minimum of three times to determine S.E.M.

[³H]IP Formation in CHOgpH₁ Cells.

Accumulation of [³H]IP was measured in CHOgpH₁ cells preincubated with [³H]myo-inositol, a precursor of the PLC substrate phosphatidylinositol. On the 3rd day of culture, confluent monolayers of CHOgpH₁ cells were rinsed 3 times with 5 ml of phosphate-buffered saline (pH 7.0; without Ca²⁺ and Mg²⁺). Trypsin (5 ml, 0.25%) was added, and cells were incubated at 37°C for 5 min. The cell suspension was then placed into a 15-ml conical tube, filled with α-minimum essential medium (MEM), and centrifuged at 1000g for 10 min. The resulting pellet was resuspended in α-MEM to obtain a cell density of approximately 1.8 × 10⁵ cells/ml, as determined via bright-line hemocytometer. Aliquots (400 μl) of this suspension were added to each well of a polystyrene 12-well culture tray (approximately 7.0 × 10⁴ cells/well). Trays were incubated overnight at 37°C.

After 16 to 20 h of incubation, CHOgpH₁cells were confluent and adhered to wells. The trays were decanted, supplemented with 400 μl of α-MEM, and allowed to incubate at 37°C for 6 to 9 h. After this incubation period, medium was decanted, and each well was washed with 500 μl of inositol-free DMEM. After decanting of the inositol-free DMEM, a 400-μl suspension of [³H]myo-inositol in DMEM (inositol-free) was added to each well, yielding approximately 0.8 μCi of [³H]myo-inositol per well. After overnight incubation at 37°C, 100-μl aliquots of drug stocks containing 125 mM HEPES and 50 mM LiCl were added to triplicate wells (total well volume = 500 μl); trays were incubated in a 37°C water bath for 45 min. Medium was aspirated, trays placed on ice, and cell contents liberated upon addition of ice-cold 50 mM formic acid to each well. After 15 min on ice, formic acid was neutralized by the addition of 0.66 ml of 150 mM NH₄OH to each well. Well contents were added to individual AG1-X8 200-400 formate resin anion exchange columns (Bio-Rad Laboratories, Hercules, CA). Columns were washed with 10 ml of water, followed by 10 ml of 50 mM ammonium formate to displace any weakly attached anions. [³H]IP were eluted into scintillation vials upon addition of 5 ml of 1.2 M ammonium formate/0.1 M formic acid. Scintillation vials were counted for tritium by liquid scintillation spectroscopy using a Tri-carb 2100TR scintillation analyzer (Packard Instrument Co., Meriden, CT) at 65% counting efficiency; columns were regenerated upon addition of 5 ml of 2 M ammonium formate/0.1 M formic acid, followed by duplicate washes with 10 ml of water. Each experimental condition was run in triplicate. Data are mean percent basal control [³H]IP formation, and potency for histamine is expressed as the concentration required to produce 50% maximal [³H]IP formation (EC₅₀) ± S.E.M. (n ≥ 3), using the nonlinear regression analysis algorithm in Prism 3.0. Statistical analysis was carried out using the Student's ttest; p values < 0.05 were considered to indicate a significant difference.

COS-7 Cell Culture, Transfection, and Measurement of Inverse Agonism.

COS-7 cells were grown at 37°C in a humidified atmosphere with 5% CO₂ in either DMEM containing 2 mM l-glutamine, 50 IU/ml penicillin, 50 μg/ml streptomycin, and 5% (v/v) fetal calf serum or with Glutamax I containing 50 IU/ml penicillin, 50 μg/ml streptomycin, and 0.5% (v/v) dialyzed fetal calf serum. Using the DEAE-dextran method (Brakenhoff et al., 1994), COS-7 cells were transiently transfected with either pcDEF₃ or pcDEF₃ containing the gene for the wild-type human histamine H₁ receptor (pcDEF₃hH₁) to yield 2.5 μg of pcDEF₃hH₁ and 12.5 μg of pNFκB-Luc/1 × 10⁷ cells (COShuH₁). Using this protocol, H₁-transfected (versus mock) cells show a high-affinity binding site for [³H]mepyramine, an increase in basal [³H]IP formation (consistent with H₁ receptor constitutive activity), and a selective concentration-dependent reduction in [³H]IP formation by H₁receptor antagonists (consistent with H₁ receptor inverse agonist activity) (Bakker et al., 2000a,b, 2001).

To assess inverse agonism activity here, the transfected COShuH₁ cells were seeded (ca. 1 × 10⁵ cells/well) in 96-well blackplates (Costar, Cambridge, MA) in serum-free DMEM and incubated with the H₁ antagonist acrivastine or PATs (0.1–10,000 nM). After 48 h, cells were assayed for luminescence by aspiration of the medium followed by addition of 25 μl/well luciferase assay reagent (0.8 mM ATP, 0.8 mM d-luciferin, 19 mM MgCl₂, and 0.8 μM Na₂H₂P₂O₇) in 40 mM Tris (pH 7.8) buffer containing 0.4% (v/v) glycerol, 0.03% (v/v) Triton X-100, and 2.6 μM dithiothreitol. After 30 min, luminescence was measured for 3 s/well in a Victor² luminometer (Wallac-PerkinElmer, Brussels, Belgium). The basal luminescence of mock transfected cells was 13.7% of H₁-expressing cells. Data are expressed as mean percent basal control luminescence, and potency of PATs, mepyramine, acrivastine, and histamine is expressed as the concentration required to inhibit or stimulate basal control luminescence by 50% (IC₅₀ or EC₅₀) ± S.E.M. (n ≥ 3), using Prism.

H₁-Mediated Contraction of Guinea Pig Ileum.

These assays were conducted similarly to methods previously reported (Leurs et al., 1991). Briefly, intestinal smooth muscle strips were prepared from male guinea pig ileum and mounted at 0.4 g of tension on a Hugo Sachs Hebel-Messovorsatz TL-2/HF-modem (Hugo Sach Electronik, Hugstetten, Germany) in 20 ml of Krebs buffer (117.5 mM NaCl, 5.6 mM KCl, 1.18 mM MgSO₄, 2.5 mM CaCl₂, 1.28 mM NaH₂PO₄, 25 mM NaHCO₃, and 5.5 mM glucose), continuously gassed with 95% O₂/5% CO₂ at 37°C. After equilibration for at least 45 min (fresh Krebs buffer replaced every 10 min), cumulative dose-contractile response curves were recorded using half-log increments of histamine. The contractile response produced by histamine was reevaluated after the addition (including a 5-min equilibration period) of (−)-trans-H₂-PAT (3.0–300 nM). Data are the mean percentage of contractile response, and potency is expressed as the concentration of histamine required to produce 50% maximal contractile response (EC₅₀) ± S.E.M. (n ≥ 3) using Prism; antagonist potency of (−)-trans-H₂-PAT was quantified by Schild analysis.

Results

Binding Parameters of [³H](−)-trans-H₂-PAT and [³H]Mepyramine in CHO Cells.

Wild-type (control) CHO cell membranes showed no measurable specific binding for [³H](−)-trans-H₂-PAT or [³H]mepyramine. On the other hand, in CHOgpH₁ membranes, [³H](−)-trans-H₂-PAT binds to an apparent single population of sites (B_max = 44 ± 2 fmol/mg of protein) with high affinity (K_D = 0.076 ± 0.004 nM); a representative saturation isotherm is shown in Fig. 2A. For comparison, we also examined the binding of the standard H₁antagonist radioligand [³H]mepyramine, which also binds to an apparent single population of sites (B_max = 280 ± 10 fmol/mg of protein) with high affinity (K_D = 0.48 ± 0.03 nM) in CHOgpH₁ membranes (Fig.2B). Similar results are obtained using membranes prepared from CHOhuH₁ cells (Fig.3). [³H](−)-trans-H₂-PAT and [³H]mepyramine bind to an apparent single population of sites (B_max = 75 ± 8 and 578 ± 39 fmol/mg of protein, respectively) with high affinity (K_D = 0.23 ± 0.02 and 1.07 ± 0.04 nM, respectively) (Fig. 3, A and B, respectively). The difference observed in B_max values (ca. 6- to 8-fold) for specific binding of the two radioligands to CHOgpH₁ and CHOhuH₁ cell membranes are several times larger than the difference (ca. 2-fold) observed for guinea pig brain (Booth et al., 1999) and about the same as the difference (ca. 7-fold) observed in rat brain tissue (Choksi et al., 2000).

In another set of saturation binding experiments using CHOhuH₁ membranes, the two radioligands were tested side-by-side using concentrations of up to about 30-timesK_D to reveal a possible additional binding site for [³H](−)-trans-H₂-PAT (Fig. 4). Data were fit to one- and two-site models; however, no statistically significant (by F-test) improved fit was achieved using a two-site model. Thus, an apparent single population of H₁ receptors is revealed by each radioligand, and the B_max for specific binding of [³H](−)-trans-H₂-PAT (160 ± 7.1 fmol/mg of protein) is about 14% of that observed for [³H]mepyramine (1180 ± 30 fmol/mg of protein).

Radioligand Competition Binding Studies.

Competition binding studies show that classical histamine H₁antagonists ([+]-chlorpheniramine, diphenhydramine, mepyramine, and triprolidine) have high affinity (K_0.5, <15 nM; Table1) for [³H](−)-trans-H₂-PAT labeled and [³H]mepyramine labeled sites in membranes prepared from CHOgpH₁ and CHOhuH₁ cells. Furthermore, the rank order of competition potency of characteristic H₁ ligands (i.e., doxepin > mepyramine ≥ triprolidine ≥ clozapine ≥ [+]-chlorpheniramine > diphenhydramine > [−]-chlorpheniramine > histamine; Table 1) is nearly identical when comparing [³H](−)-trans-H₂-PAT to [³H]mepyramine radiolabeling in CHOgpH₁ (r² = 0.90) and CHOhuH₁(r² = 0.96) cells. These results are consistent with the H₁ receptor binding profiles of both radioligands in tissue from guinea pig (Booth et al., 1999) and rat (Choksi et al., 2000) brain.

The affinity of PATs for H₁ receptors labeled by either [³H](−)-trans-H₂-PAT or [³H]mepyramine in both CHO cell lines shows stereochemical preference identical to that observed in guinea pig (Bucholtz et al., 1998) and rat (Choksi et al., 2000) brain. As expected,trans-(1R,3S)-(−)-H₂-PAT (K_0.5, ∼0.6–2 nM) shows the relatively highest affinity, whereas the correspondingtrans-(1S,3R)-(+)-H₂-PAT enantiomer (K_0.5, ∼12–40 nM) has about 20-fold lower affinity; racemic (±)-trans-H₂-PAT (K_0.5, ∼1.4–4.3 nM) is near the theoretically predicted half-potency of the more active (−)-enantiomer. Meanwhile, the affinity of thecis-H₂-PAT enantiomers are severalfold lower than their correspondingtrans-H₂-PAT diastereomers. Thecis-(1S,3S)-(−)-H₂-PAT isomer (K_0.5, ∼4–9 nM) has about 15-fold higher affinity thancis-(1R,3R)-(+)-H₂-PAT (K_0.5, ∼54–130 nM), and the affinity of (±)-cis-H₂-PAT (K_0.5, ∼8–15 nM) is about half the potency of the more active cis-(−)-enantiomer. Taken together, these results indicate that stereochemistry at the C1 phenyl group and especially at the C3 amino group (i.e., Sconfiguration shared by both [−]-cis- and [−]-trans-H₂-PAT; Fig. 1) is important for PAT affinity at histamine H₁receptors labeled with either [³H](−)-trans-H₂-PAT or [³H]mepyramine in CHOgpH₁ and CHOhuH₁ cells.

Using either [³H](−)-trans-H₂-PAT or [³H]mepyramine as the radioligand, competition binding curves for most tested ligands were sigmoidal shaped with Hill (n_H) coefficients close to unity (Table 1), expected from ligands that bind competitively (presumably as antagonists) to a single population of receptors. Competition with histamine, however, gave a shallow sloped (n_H, ∼0.7) concentration-response curve characteristic of agonist ligand binding at G protein-coupled receptors, according to the ternary complex model with limiting availability of G protein (De Lean et al., 1980). Virtually full displacement of either radioligand could be achieved by all tested ligands in either cell line. Radioligand displacement curves for representative ligands in Table 1 in competition for [³H]mepyramine labeled H₁receptors in CHOhuH₁ membranes are shown in Fig.5. Data were fit to one- and two-site models; however, no statistically significant (by F-test) improved fit was achieved using a two-site model.

The ternary complex model also predicts the so-called “GTP-shift” (i.e., lower agonist ligand affinity is obtained in the presence of excess GTP as a result of virtually all receptors being converted to a G protein-uncoupled state). To investigate differences in functional binding that may occur between [³H]mepyramine and [³H](−)-trans-H₂-PAT at H₁ receptors, we examined the effect of excess GTP (1 mM) on the competitive binding of the H₁antagonist (+)-chlorpheniramine, (−)-trans-H₂-PAT, and histamine. The CHOgpH₁ cell line was used for these studies since these cells also were used to measure H₁-mediated functional effects on IP formation (vide infra). There was no significant difference (p > 0.05) in affinity observed for (+)-chlorpheniramine or (−)-trans-H₂-PAT, with or without excess GTP, using either radioligand. Meanwhile, the affinity of histamine was significantly (p < 0.02) reduced with excess GTP using either [³H]mepyramine (K_0.5, ∼18 versus 25 μM, with and without excess GTP, respectively; Table 1) or [³H](−)-trans-H₂-PAT (K_0.5, ∼10 versus 18 μM, with and without excess GTP, respectively; Table 1). The ∼1.5- to 2-fold increase in K_i values for histamine in the presence of GTP is similar to results reported using guinea pig brain homogenates (Chang and Snyder, 1980). As expected, the Hill coefficient (n_H) for histamine significantly (p < 0.05) increased from about 0.7 to 0.8 in the presence of excess GTP (Table 1), suggesting binding occurred to a single population of low-affinity G protein-uncoupled H₁ receptors.

[³H]IP Formation in CHOgpH₁ Cells.

Previously, it was reported that stimulation of H₁ receptors on CHOgpH₁(and CHOhuH₁) cells leads to activation of PLC and increased production of IP (Leurs et al., 1994; Smit et al., 1996). In the current studies, histamine produced a concentration-dependent stimulation of [³H]IP formation in CHOgpH₁ cells preincubated with [³H]myo-inositol (Fig.6), consistent with the literature. Maximal stimulation was observed to be about 900% of basal control [³H]IP formation at about 100 μM histamine (EC₅₀ = 2.6 ± 0.2 μM; n = 4).

Antagonism of Histamine-Induced Stimulation of [³H]IP Accumulation.

Triprolidine, (+)- and (−)-cis-H₂-PAT, and, (+)- and (−)-trans-H₂-PAT also were tested for effects on [³H]IP accumulation in CHOgpH₁ cells. At concentrations spanning 0.01 to 10 μM, none of these H₁ ligands increased [³H]IP accumulation over the basal level (data not shown). Thus, it was concluded that the PATs, like triprolidine, are not agonists at H₁ receptors coupled to IP formation. Test compounds subsequently were assessed for ability to antagonize the effect of histamine (3 μM, about EC₅₀) to stimulate [³H]IP accumulation. As summarized in Fig. 7, histamine-induced [³H]IP accumulation was essentially fully blocked by 1.0 μM of the H₁antagonist triprolidine (percentage of the control histamine response = 4.1 ± 0.3) and (−)-trans-H₂-PAT (percentage of the control histamine response = 2.8 ± 0.1). Comparatively, the histamine effect was incompletely antagonized by 1.0 μM (−)-cis-, (+)-trans-, and (+)-cis-H₂-PAT (percentage of the control histamine response = 16.6 ± 1.0, 26 ± 1.0, and 65 ± 1.3; respectively); at 10 μM, however, these PATs fully blocked the histamine effect (percentage of the control histamine response = 0.6 ± 1.1, 2.1 ± 1.4, and 2.1 ± 1.3; respectively). In comparison to triprolidine and (−)-trans-H₂-PAT, the higher concentration of (−)-cis-, (+)-trans-, and (+)-cis-H₂-PAT required to fully antagonize histamine-induced stimulation of [³H]IP accumulation is consistent with their lower affinity for H₁ receptors in CHOgpH₁ cells (Table 1).

Effect of PATs on Constitutive H₁ Receptor Activity (Inverse Agonism) in COS-7 Cells.

Recently, constitutive histamine H₁ receptor activity was shown in COS-7 cells transiently transfected with the human H₁receptor (COShuH₁) (Bakker et al., 2000a, 2001). Here, we evaluated PAT functional responses in COShuH₁ cells using the NF-κB reporter-gene assay that measures H₁ receptor-mediated bioluminescence (Bakker et al., 2000b, 2001). The H₁ antagonists mepyramine and acrivastine and the endogenous agonist histamine were used as reference compounds in this assay. The basal luminescence of mock transfected cells was 13.7% of H₁-expressing cells. Constitutive H₁ receptor activity in COShuH₁ cells, as measured by luminescence, is inhibited (versus basal control) by mepyramine (IC₅₀ = 20.7 ± 0.7 nM) and acrivastine (IC₅₀ = 60 ± 0.5 nM) (Fig.8). This activity of known H₁ antagonists has been interpreted as inverse agonism (Bakker et al., 2000a,b, 2001). Conversely, the endogenous agonist histamine stimulates (EC₅₀ = 250 ± 10 nM) H₁ receptor activity and luminescence. As shown in Fig. 8, all the PAT isomers also inhibited constitutive H₁ receptor activity in this assay with the following potency order: (−)-trans-H₂-PAT (IC₅₀ = 23 ± 0.4 nM) > (−)-cis-H₂-PAT (IC₅₀ = 170 ± 2 nM) > (+)-cis- H₂-PAT (IC₅₀ = 940 ± 30 nM) > (+)-trans-H₂-PAT (IC₅₀ = 1100 ± 21 nM). Consistent with its low H₁ binding potency (Table 1), (+)-cis-H₂-PAT did not fully inhibit basal luminescence even at 100 μM (about the limit of its solubility). These results indicate that the PATs behave functionally similarly to classical H₁ antagonists or inverse agonists (mepyramine and acrivastine) and not like the agonist histamine in this assay system.

We also determined the affinity of the PATs for [³H]mepyramine-labeled H₁receptors in membrane preparations from the COShuH₁ cell line. The rank order of affinity of PAT isomers is consistent with their rank order of functional potency in COShuH₁ cells (see above), i.e., (−)-trans-H₂-PAT (K_0.5 = 2.0 ± 0.01 nM) > (−)-cis-H₂-PAT (K_0.5 = 7.0 ± 0.02 nM) > (+)-trans-H₂-PAT (K_0.5 = 54 ± 0.3 nM) > (+)-cis- H₂-PAT (K_0.5 = 166 ± 1 nM). These results also are consistent with PAT rank order of affinity functional in CHOgpH₁ and CHOhuH₁cells (Table 1), rat brain (Choksi et al., 2000), and guinea pig brain (Bucholtz et al., 1998).

H₁-Mediated Contraction of Guinea Pig Ileum.

As shown in Fig. 9, histamine (in absence of competing ligand) produced a concentration-dependent contraction of guinea pig ileum intestinal smooth muscle, with maximal effect occurring at about 10 μM histamine (EC₅₀ = 0.1 μM). Previously, it has been reported that this effect of histamine is mediated by H₁ receptors coupled to PLC and formation of IP and is competitively inhibited by H₁ antagonists (Leurs et al., 1991). Accordingly, based on the results here that (−)-trans-H₂-PAT antagonizes histamine-induced accumulation of [³H]IP in CHOgpH₁ cells (vide supra), we assessed the ability of (−)-trans-H₂-PAT to act as an antagonist in the ileum contraction assay. At concentrations spanning 0.01 to 10 μM, (−)-trans-H₂-PAT had no effect on contraction of guinea pig ileum (data not shown). Meanwhile, (−)-trans-H₂-PAT competitively antagonized the stimulation produced by histamine, producing rightward shifts of the histamine concentration-response curve with increasing concentrations of (−)-trans-H₂-PAT (Fig. 9); Schild regression analysis slope = 0.90 ± 0.01; pA₂ = 9.2.

Discussion

Previously, (−)-trans-H₂-PAT was shown to activate stereospecifically H₁ receptors coupled to modulation of tyrosine hydroxylase activity in guinea pig and rat forebrain in vitro (Booth et al., 1999) and in vivo (Choksi et al., 2000). Meanwhile, in brain tissue homogenates from these same species, the novel radioligand [³H](−)-trans-H₂-PAT labels only a subpopulation of the total number of H₁ receptors labeled by the standard antagonist H₁ radioligand [³H]mepyramine (Booth et al., 1999; Choksi et al., 2000). In this article, we were able to more discretely examine the H₁ recognition features and associated functional activity of H₂-PAT by using cellular (CHO, COS) and tissue (guinea pig ileum strips) systems that are less complex than mammalian brain tissue.

The pharmacological profile of [³H](−)-trans-H₂-PAT labeled H₁ receptors in CHOgpH₁ and CHOhuH₁ cell membranes is similar to results obtained using the H₁ antagonist radioligand [³H]mepyramine. For instance, the rank order of stereoselective affinity of several known H₁ligands and H₂-PAT isomers for H₁ receptors labeled by [³H](−)-trans-H₂-PAT and [³H]mepyramine (Table 1) is nearly identical. However, the current studies show that the number of H₁ receptors labeled by [³H](−)-trans-H₂-PAT in CHOgpH₁ and CHOhuH₁cells is only about 15% of that labeled by [³H]mepyramine. These results are similar to those obtained using rat brain tissue (Choksi et al., 2000). Meanwhile, in guinea pig brain the B_max for [³H](−)-trans-H₂-PAT is about 50% the value for [³H]mepyramine (Booth et al., 1999)—the difference in rat versus guinea pig brain probably reflects the known species heterogeneity regarding the binding parameters of [³H]mepyramine (Chang et al., 1979). In any case, it remains apparent that [³H](−)-trans-H₂-PAT generally labels only a fraction (about 15 to 50%) of the total histamine H₁ receptor population labeled by [³H]mepyramine using either rodent brain tissue or clonal cell lines stably transfected with H₁cDNA.

Initially, we hypothesized that [³H](−)-trans-H₂-PAT may be an agonist-type radioligand that recognizes only a subpopulation of H₁ receptors in a high-affinity state (already coupled to G protein), whereas the H₁ antagonist radioligand [³H]mepyramine may recognize both high- and low-affinity (i.e., not already coupled to G protein) H₁ receptors. This hypothesis is consistent with the H₁-mediated functional effect of (−)-trans-H₂-PAT to activate tyrosine hydroxylase and catecholamine synthesis similarly to the endogenous agonist histamine in vitro (Marley and Robotis, 1998) and in vivo (Fleckenstein et al., 1993). However, results of functional assays conducted here, using CHOgpH₁ and COShuH₁cells, clearly indicate that the pharmacology of H₂-PAT is similar to H₁ antagonists or inverse agonists, such as triprolidine (Fig. 7), mepyramine (Fig. 8), and acrivastine (Fig. 8), rather than the endogenous H₁ agonist histamine (Figs. 7 and 8). Moreover, (−)-trans-H₂-PAT potently antagonizes histamine-induced H₁-mediated contractile effects in guinea pig ileum (Fig. 9). Finally, binding potency of (−)-trans-H₂-PAT was unaffected (similar to the H₁ antagonist [+]-chlorpheniramine) in CHOgpH₁ cell membranes where virtually all the H₁ receptors were presumed to be uncoupled to G protein as a result of excess GTP; in contrast, histamine showed the lower binding potency (∼50% decrease; Table 1) expected in systems where agonist ligand-receptor interaction is sensitive to the so-called GTP-shift. Results of the GTP-shift studies were the same regardless of whether [³H](−)-trans-H₂-PAT or [³H]mepyramine was used as the radioligand. Taken together, these results indicate that the fewer number of H₁ receptors labeled by [³H](−)-trans-H₂-PAT versus [³H]mepyramine in CHOgpH₁ and CHOhuH₁ cells probably is not due to differences in binding that result from differences in ligand-receptor functional interaction. Thus, we conclude that (−)-trans-H₂-PAT is not an agonist ligand (rather, it is an antagonist/inverse agonist) at H₁ receptors coupled to PLC/IP formation.

Meanwhile, H₁ receptors also can mediate histamine-induced stimulation of cAMP formation in mammalian brain (Palacios et al., 1978) and adrenal cells (Marley et al., 1991), and it has long been known that cAMP-dependent protein kinase A can activate tyrosine hydroxylase (Morgenroth et al., 1975). As the potency and efficacy of histamine to stimulate cAMP formation and tyrosine hydroxylase activity in bovine adrenal cells is similar, it is suggested H₁ receptors may modulate catecholamine synthesis via this pathway. Multifunctional signaling is apparent for many GPCR systems (Milligan, 1993), and this phenomenon has been described as “receptor promiscuity”—an unfaithfulness of a receptor to any one G protein (Kenakin, 1995a). Implicit in the receptor promiscuity hypothesis is the concept that a ligand that acts as an agonist at a receptor coupled to one particular signal transduction pathway may be an antagonist at the same receptor coupled to another signaling pathway. This phenomenon was termed “functional selectivity” to describe the effects of dopamine D₂ receptor ligands that are agonists at postsynaptic D₂ receptors but antagonists at presynaptic D₂ receptors (Ghosh et al., 1996). Receptor promiscuity and functional selectivity merge with the phenomenon of “precoupling of receptor-G protein complexes” (Leff and Scaramellini, 1998). Such spontaneous receptor-G protein precoupling explains observed GPCR constitutive activity now abundantly documented, including for H₁ receptors (Bakker et al., 2000a,b). A critical assumption of these theories is that a heterogeneity of active receptor conformations exists and that agonists differ in their ability to induce, stabilize, or select among receptor conformations, as described in the “agonist trafficking” hypothesis (Kenakin, 1995b). Thus, a compelling body of theoretical and experimental evidence exists to suggest the hypothesis that (−)-trans-H₂-PAT could behave as an H₁ agonist or antagonist, depending on the associated signal transduction pathway, as influenced by ligand stabilization of particular H₁-G protein coupling. We note that this phenomenon may involve differences between pre- and postsynaptically expressed H₁ receptors (presynaptic neuronal H₁ receptors seem to be involved in modulation of brain catecholamine synthesis); thus, our future studies will include adrenal cells to measure postsynaptic H₁-mediated effects on tyrosine hydroxylase activity as we further test the proposed H₁functional selectivity of (−)-trans-H₂-PAT.

Our finding that (−)-trans-H₂-PAT can fully displace [³H]mepyramine binding to H₁ receptors in CHOhuH₁membranes (Fig. 5) (and vice versa) seems to be at odds with results indicating [³H](−)-trans-H₂-PAT labels only a fraction (about 15–50%, depending on the species) of the total H₁ receptors labeled by [³H]mepyramine. This situation, however, is not unique among GPCRs. For example, theB_max for the dopamine D₂ receptor radioligand [³H]spiperone has been known for some time to be severalfold lower than that for the D₂radioligand [³H]nemonapride; however, spiperone fully displaces [³H]nemonapride, and nemonapride fully displaces [³H]spiperone (Seeman et al., 1992). Subsequently, it was determined that the D₂ photoaffinity probe [¹²⁵I]-azidophenethyl-spiperone labels only D₂ monomers, whereas the D₂photoaffinity probe [¹²⁵I]-azido-nemonapride labels both D₂ monomers and oligomers (Zawarynski et al., 1998). Apparently, radioreceptor experiments using reversible ligands with similar apparent K_Dvalues do not distinguish subtle kinetic differences in GPCR monomer versus oligomer populations. Some other GPCR neurotransmitter systems for which single reversible radioligands did not predict monomer versus oligomer subpopulations but are now known to oligomerize include α₂ (Gouldson et al., 1997) and β₂ (Hebert et al., 1996) adrenergic, H₂ histamine (Fukushima et al., 1997), M₃ muscarinic (Maggio et al., 1999), and δ- and κ-opioid (Cvejic and Devi, 1997).

We speculate that H₁ receptors also may be expressed as monomers and oligomers. Previous studies using guinea pig brain membranes showed that a photoaffinity analog of mepyramine, [¹²⁵I]-iodoazidophenpyramine, labeled proteins of molecular weight 47, 56, 92, and 350 to 400 kDa (Ruat et al., 1988). Labeling of these proteins was prevented by an H₁antagonist (the band at 92 kDa was only partially inhibited), suggesting these proteins were H₁-like. However, labeling of the 47-kDa protein also was diminished in the presence of protease inhibitors, suggesting it probably represented a proteolysis product. Meanwhile, labeling of the 350- to 400-kDa proteins greatly increased in the absence of 2-mercaptoethanol, suggesting these proteins to be higher molecular weight complexes linked by disulfide bridges. At the time, the 350- to 400-kDa proteins were interpreted as representing a 56-kDa H₁ receptor linked to one or more other (nondefined) peptides or as artifactual disulfide linked peptides formed during membrane preparation (Ruat et al., 1988). In light of our results with [³H](−)-trans-H₂-PAT and recent reports documenting a variety of GPCRs capable of oligomerization, we suggest that the proteins labeled by [¹²⁵I]-iodoazidophenpyramine in earlier studies may have been a combination of H₁ receptor monomers (i.e., 56 kDa) and oligomers (i.e., 350–400 kDa). In this regard, we believe (−)-trans-H₂-PAT represents a promising lead toward developing a (photo)affinity probe to differentiate H₁ receptor monomers from hypothesized H₁ oligomers and to determine whether GPCR oligomerization influences GPCR functional heterogeneity and vice versa.