Abstract
There are differences in the pharmacological properties of phenylhistamines and histaprodifens between guinea pig histamine H1 receptor (gpH1R) and human histamine H1 receptor (hH1R). The aim of this study was to analyze species differences in more detail, focusing on histaprodifen derivatives and including the bovine histamine H1 receptor (bH1R) and rat histamine H1 receptor (rH1R). H1R species isoforms were coexpressed with the regulator of G protein signaling RGS4 in Sf9 insect cells. We performed [3H]mepyramine binding assays and steady-state GTPase assays. For a novel class of histaprodifens, the chiral histaprodifens, unique species differences between hH1R, bH1R, rH1R, and gpH1R were observed. The chiral histaprodifens 8R and 8S were both partial agonists at gpH1R, but only 8R was a partial agonist at the other H1R species isoforms. An additional phenyl group in chiral histaprodifens 10R and 10S, respectively, resulted in a switch from agonism at gpH1Rto antagonism at hH1R, bH1R, and rH1R. In general, histaprodifens showed the order of potency hH1R < bH1R < rH1R < gpH1R. An active-state model of gpH1R was generated with molecular dynamics simulations. Dimeric histaprodifen was docked into the binding pocket of gpH1R. Hydrogen bonds and electrostatic interactions were detected between dimeric histaprodifen and Asp-116, Ser-120, Lys-187, Glu-190, and Tyr-432. We conclude the following: 1) chiral histaprodifens interact differentially with H1R species isoforms; 2) gpH1R and rH1R, on one hand, and hH1R and bH1R, on the other hand, resemble each other structurally and pharmacologically; and 3) histaprodifens interact with H1R at multiple sites.
The histamine H1 receptor (H1R) is a prototypical G protein-coupled receptor (GPCR) that interacts with Gq proteins to activate phospholipase C (Hill et al., 1997; Jorgejan et al., 2005). H1R antagonists are clinically important for the treatment of allergic diseases, and H1R agonists are experimental tools to analyze H1R function. Numerous H1R agonists are known, including small agonists derived from histamine and bulkier agonists such as phenylhistamines (Leschke et al., 1995; Zingel et al., 1995), ergolines (Bakker et al., 2001; Pertz et al., 2006), and histaprodifens (Menghin et al., 2003). Suprahistaprodifen (Menghin et al., 2003) is a highly potent agonist at the guinea pig ileum (pEC50 = 8.26). Based on those data, Striegl (2006) synthesized a novel series of suprahistaprodifen derivatives.
The highly conserved Asp3.32 in transmembrane domain (TM) helix 3 (Ohta et al., 1994; Nonaka et al., 1998), Lys5.39 (Leurs et al., 1995; Bruysters et al., 2004; Jongejan and Leurs, 2005), Thr5.42 (Leurs et al., 1994; Ohta et al., 1994), Asn5.46 (Leurs et al., 1994; Ohta et al., 1994), and Phe6.55 (Bruysters et al., 2004) are involved in histamine binding. Ser3.36 and Asn7.45 are involved in hH1R activation (Jongejan and Leurs, 2005). Trp4.56, Lys5.39, Phe6.52, Phe6.55, and Lys5.39 interact with H1R antagonists (Wieland et al., 1999; Gillard et al., 2002).
Phenylhistamines and histaprodifens (Seifert et al., 2003) and ergolines (Pertz et al., 2006) were studied at the recombinant guinea pig H1 receptor (gpH1R) and human H1 receptor (hH1R) expressed in Sf9 insect cell membranes by [3H]mepyramine competition binding and Gq protein-catalyzed GTP hydrolysis. Those studies revealed substantial species differences between gpH1R and hH1 R. Asn2.61 acts as a selectivity switch between hH1R and gpH1R (Bruysters et al., 2005).
To date, the cDNAs of 12 mammalian H1R species isoforms are known. A phylogenetic tree of H1R species isoforms is shown in Fig. 1. The cDNAs of gpH1R (Traiffort et al., 1994), hH1R (Fukui et al., 1994), bovine H1R (bH1R) (Yamashita et al., 1991), and rat H1 receptor (rH1R) (Fujimoto et al., 1993) are available in our laboratory. Figure 2 shows a sequence comparison of hH1R, bH1R, rH1R, and gpH1R. Table 1 provides a summary on the percentage of identical amino acids for the N terminus, C terminus, each extra- and intracellular loop, each transmembrane domain, and for the complete sequence, respectively, among two H1R species isoforms. For the complete sequences, similarities range from 66.73% (bH1R versus gpH1R) to 81.87% (hH1R versus bH1R). In general, amino acids in the transmembrane domains show >80% identity. TM3 shows no species differences. The similarities of the N terminus, C terminus, and the loops are in a range from 100% (I1) to only 26% (N terminus). Collectively, there are only moderate similarities between H1R species isoforms. Thus, we hypothesized that there are significant differences in pharmacological properties not only between hH1R and gpH1R but also rH1R and bH1R. To test this hypothesis, we coexpressed hH1R, bH1R, rH1R, and gpH1R with the regulator of G protein signaling RGS4 in Sf9 insect cells, and we characterized known and novel histaprodifens (Fig. 3) in [3H]mepyramine competition binding and GTPase assays. Moreover, an active state model of gpH1R was generated.
Materials and Methods
Materials. The cDNAs from bH1R and rH1R were kindly provided by Dr. H. Fukui (Department of Pharmacology, University of Tokushima, Tokushima, Japan), cloned into the pEF-BOS plasmid (bH1R) and pUC-18 plasmid (rH1R), respectively. Phusion high-fidelity polymerase, all restriction enzymes, and T4 DNA ligase were from New England Biolabs (Beverly, MA). The anti-FLAG IgG (M1 monoclonal antibody) was from Sigma-Aldrich (St. Louis, MO), and the anti-RGS4 was from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA). [γ-32P]GTP (6000 Ci/mmol) and [3H]mepyramine (30.0 Ci/mmol) were from PerkinElmer Life and Analytical Sciences (Boston, MA). As liquid scintillation cocktail, we used Rotiszint ecoplus from Roth (Karlsruhe, Germany). Histaprodifens were synthesized as described by Elz et al. (2000) and Striegl (2006).
Construction of pGEMbH1R, pGEMrH1R, pVLbH1R, and pVLrH1R. bH1R and rH1R were cloned into the pGEM-3Z-SF-plasmid by overlap-extension PCR (PCR Ia, PCR Ib, and PCR II) with the pGEM-3Z-SF-gpH1R-plasmid and the pEF-BOS-bH1R (Yamashita et al., 1991) and pUC-18-rH1R (Fujimoto et al., 1993) as template. In PCR Ia, the DNA fragment with the signal peptide (S), the FLAG epitope (F), and the N terminus of the bH1R and rH1R, respectively, was amplified. In PCR Ib, the DNA fragments encoding bH1R and rH1R, respectively, and the His6 tag (CACCATCATCACCATCAC) were generated. In PCR II, the products of PCR Ia and PCR Ib were annealed in the SF encoding region, resulting in PCR fragments encoding the SF, bH1R and rH1R, respectively, the His6 tag, the stop codon and an XbaI site. For PCRs, the following primers were used: 5′-GCTCACTCATTAGGCACC-3′ (bH1R, forward, PCR Ia, PCR II), 5′-GGACAGGTCATGGCGTCATCATCGTCCTTG-3′ (bH1R, reverse, PCR Ia), 5′-GACGATGATGACGCCATGACCTGTCCCAACTCC-3′ (bH1R, forward, PCR Ib), and 5′-ATCCTCTAGATTAGTGATGGTGATGATGGTGGGAACGAATGTGCAGAATT-3′ (bH1R, reverse, PCR Ib, PCR II); 5′-GCTCACTCATTAGGCACC-3′ (rH1R, forward, PCR Ia, PCR II), 5′-GGCAAAGCTCATGGCGTCATCATCGTCCTTG-3′ (rH1R, reverse, PCR Ia), 5′-CAAGGACGATGATGACGCCATGAGCTTTGCCAATACC-3′ (rH1R, forward, PCR Ib), and 5′-ATCCTCTAGATTAGTGATGGTGATGATGGTGGGAACGAATGTGCAGAATC-3′ (rH1R, reverse PCR Ib, PCR II). The resulting PCR fragments with bH1R and rH1R were double-digested with HindIII and XbaI and cloned into the pGEM-3Z-plasmid using the pGEM-3Z-SF-gpH1R as template. The sequences of bH1R and rH1R, cloned into the pGEM-3Z plasmid were checked for their correctness by sequencing (Entelechon, Regensburg, Germany). Using pGEM-3Z-SF-bH1R and pGEM-3Z-SF-rH1Rastemplate, bH1R and rH1R were cloned into the pVL1392 baculovirus transfer vector using the restriction sites BssHII and XbaI for bH1R and SacI and XbaI for rH1R. Both species isoforms were N-terminally tagged with the signal-peptide and FLAG-epitope 5′-ATGAAGACGATCATCGCCCTGAGCTACATCTTCTGCCTGGTATTCGCCGACTACAAGGACGATGATGACGCC-3′ and C-terminally tagged with the His6-tag 5′-CACCATCATCACCATCAC-3′.
Cell Culture, Generation of Recombinant Baculoviruses, and Membrane Preparation. Sf9 insect cells were cultured using SF 900 II medium (Invitrogen, Carlsbad, CA), containing 5% (v/v) fetal calf serum (Sigma-Aldrich, Steinheim, Germany), and 0.1 mg/ml gentamicin (Lonza Walkersville, Walkersville, MD) in 250-ml Erlenmeyer flasks at 28°C and shaking at 125 rpm. Recombinant baculoviruses encoding H1R species isoforms were produced with Sf9 cells using the BaculoGOLD transfection kit (BD Biosciences PharMingen, San Diego, CA). After initial infection, two sequential virus amplifications followed to generate high-titer virus stocks. For final infection, yielding in membranes, Sf9 cells were sedimented by centrifugation and suspended in fresh medium, yielding in a final concentration of 3.0 × 106 cells/ml. Cells were cotransfected with a 1:100 dilution of higher titer baculovirus stock encoding one of the H1Rs and RGS4. After 48 h of incubation, membranes were prepared as described previously (Kelley et al., 2001; Houston et al., 2002).
Determination of Protein Concentration. Protein concentration was determined using the Bio-Rad DC protein assay kit (Bio-Rad, Hercules, CA).
SDS-Polyacrylamide Gel Electrophoresis and Immunoblot Analysis. Membrane proteins were separated on SDS polyacrylamide gels containing 12% (w/v) polyacrylamide. After transfer of the proteins onto Immobilon-P transfer membranes (Millipore Corporation, Bedford, MA), membranes were reacted with the M1 antibody or anti-RGS4 IgG (1:1000 each). Immunoreactive bands were visualized by sheep anti-mouse IgG (1:1000) (for the M1 antibody) or donkey anti-goat IgG (for the RGS4 antibody) as described previously (Seifert et al., 2003).
Preparation of Compound Stock Solutions. Chemical structures of the analyzed compounds are given in Fig. 3. Compounds 1 and 18 (10 mM each) were dissolved in double-distilled water. Compounds 2 to 16 (5 mM each) were dissolved in a solvent containing 30% (v/v) DMSO, 30% (v/v) Tris-HCl, pH 7.4 (10 mM), and 40% (v/v) double-distilled water. Compound 17 (1 mM) was dissolved in 50% (v/v) DMSO and 50% (v/v) Tris-HCl, pH 7.4 (10 mM). The final DMSO concentration in all assays was adjusted to 3 (v/v) or 5% (v/v) as appropriate for a given ligand. These DMSO concentrations did not shift pKi and pEC50 values of histamine.
[3H]Mepyramine Saturation Binding Assay. Membranes expressing a given H1R isoform and RGS4 were thawed, suspended in binding buffer (12.5 mM MgCl2, 1 mM EDTA, and 75 mM Tris-HCl, pH 7.4), and sedimented by a 10-min centrifugation at 4°C and 13,000g. The membrane pellet was resuspended in binding buffer. Each reaction mixture contained 50 to 75 μg of membrane protein. All saturation binding assays were performed in a concentration range from 0.2 to 20 nM [3H]mepyramine. Nonspecific binding was determined in the presence of 10 μM diphenhydramine. Incubations were conducted for 90 min at room temperature and shaking at 250 rpm. Bound [3H]mepyramine was separated from free [3H]mepyramine by filtration through GF/C filters (Whatman, Maidstone, UK) followed by three steps of washing with 2 ml of binding buffer (4°C). The radioactivity remaining on filters was determined with liquid scintillation counting, after an equilibration phase of at least 4 h with scintillation cocktail.
[3H]Mepyramine Competition Binding Assay. Preparation of membrane was performed as described for the [3H]mepyramine saturation binding assay. All competition binding experiments were carried out in presence of 5 nM [3H]mepyramine and unlabeled ligands at final concentrations from 1 nM to 100 μM as appropriate to obtain saturated competition curves. Incubations were conducted for 90 min at room temperature and shaking at 250 rpm. Bound [3H]mepyramine was determined as described for the [3H]mepyramine saturation binding assay.
Steady-State GTPase Assay. Membranes expressing a given H1R isoform and RGS4 were thawed, suspended in 10 mM Tris-HCl, pH 7.4, and sedimented by a 10-min centrifugation at 4°C and 13,000g. The membrane pellet was resuspended in 10 mM Tris-HCl, pH 7.4. Each assay tube contained Sf9 cell membranes (10–20 μgof protein/tube), 1.0 mM MgCl2, 0.1 mM EDTA, 0.1 mM ATP, 100 nM GTP, 0.1 mM adenylyl imidodiphosphate, 5 mM creatine phosphate, 40 μg of creatine kinase, 0.2% (w/v) bovine serum albumin in 50 mM Tris-HCl, pH 7.4, and several ligands in a concentration range from 1 nM to 1 mM as appropriate to obtain saturated concentration/response curves. Reaction mixtures (80 μl) were incubated for 3 min at 25°C. Thereafter, 20 μlof[γ-32P]GTP (∼0.1 μCi/tube) were added to each tube. Reactions were conducted for 20 min at 25°C and terminated by the addition of 900 μl of slurry consisting of 5% (w/v) activated charcoal and 50 mM NaH2PO4, pH 2.0. Charcoal-quenched reaction mixtures were centrifuged for 15 min at room temperature at 15,000g. Six hundred microliters of the supernatant fluid of reaction mixtures was removed, and 32Pi was determined by liquid scintillation counting.
The steady-state GTPase assay was performed in the agonist and antagonist mode. In the agonist mode, the percentage of increase in GTP hydrolysis relative to histamine was determined with increasing concentrations of each ligand from 0.1 nM to 100 μM. In the antagonist mode, the percentage of GTP hydrolysis was determined as described in agonist experiments, but assay tubes additionally contained 1 μM histamine.
Data Analysis. All data were analyzed with the software Prism 4.02 (GraphPad Software Inc., San Diego, CA). KB and pKB values were calculated according to Cheng and Prusoff (1973). All data are the means ± S.E.M. of at least three independent experiments. To compare two pairs of data, the significance of the deviation of zero p was calculated using the t test.
Construction of an Active gpH1R Model with Dimeric Histaprodifen in the Binding Pocket. A homology model of the inactive gpH1R was constructed as described previously (Strasser and Wittmann, 2007). In addition, essential internal water molecules were placed into the receptor model. Afterward, the whole model was embedded into an environment consisting of 104 1-palmitoyl-2-oleoyl-phosphatidylcholine molecules, 12,675 intracellular and extracellular water molecules, and 8 sodium and 25 chloride ions to achieve electroneutrality. This template was used to generate a model of the active gpH1R based on the distance restraints given by Niv et al. (2006). Therefore, we performed restrained MD simulations with the software package GROMACS 3.2 (van der Spoel et al., 2004). This procedure will be described in detail elsewhere. In brief, a positively charged dimeric histaprodifen 17 was manually docked into the proposed binding pocket of the active gpH1R, and one sodium ion was deleted to conserve electroneutrality. The whole simulation box was energetically minimized. Thereafter, MD simulations were performed. The whole equilibration phase was divided in 12 cycles with different simulation times and position constraints as described in the following. The constraints for the backbone atoms (bb), side chain atoms (sc), and ligand atoms (lig) are given in kiloJoules per milliliter per squared nanometer. Cycle 1: 250 ps, bb 5000, sc 5000, lig 5000; cycle 2: 100 ps, bb 4000, sc 4000, lig 4000; cycle 3: 100 ps, bb 4000, sc 3000, lig 3000; cycle 4: 100 ps, bb 3000, sc 3000, lig 3000; cycle 5: 100 ps, bb 2000, sc 1000, lig 1000; cycle 6: 100 ps, bb 1000, sc 800, lig 800; cycle 7: 100 ps, bb 800, sc 600, lig 600; cycle 8: 100 ps, bb 600, sc 400, lig 400; cycle 9: bb 400, sc 200, lig 200; cycle 10: 100 ps, bb 200, sc 100, lig 100; cycle 11: 100 ps, bb 100, sc 0, lig 0; cycle 12: 100 ps, bb 0, sc 0, lig 0. A 1-ns productive phase without position constraints followed. Simulations were carried out at 310 K and 1 atmosphere. For MD simulations Berendsen, temperature, and pressure coupling was used. For all calculations we used the software package GROMACS 3.2 (van der Spoel et al., 2004), with the ffG53A6 force field (Oostenbrink et al., 2004). The force field parameters for dimeric histaprodifen 17 were adopted from the ffG53A6 force field.
Results
Immunological Detection of the H1R Constructs. Based on the amino acid sequences of H1R isoforms, we calculated the molecular masses (without N-glycosylation) to be 55,784 Da for hH1R, 55,957 Da for bH1R, 55,693 Da for rH1R, and 55,620 Da for gpH1R. All H1R species isoforms were immunologically detected with the M1 antibody (Fig. 4). bH1R and rH1R exhibited strong bands at ∼60 kDa. The increase by 4 kDa relative to the theoretical mass is probably due to N-glycosylation. hH1R showed a strong band at ∼85 kDa, probably because of a higher degree of N-glycosylation. In addition, weak bands were visible in a range from 25 to 30 kDa, and at ∼50 kDa. gpH1R showed very different behavior in migration, i.e., a strong band was detected at 25 kDa, intermediate bands were visible at ∼30 and 36 kDa, and faint bands were detected at ∼50 and ∼100 kDa, respectively. The results with hH1R and gpH1R are in accordance with previous studies from our laboratory (Seifert et al., 2003).
Some of the differences in electrophoretic mobility between the four H1R species isoforms may be explained with different N-glycosylation states. At the N terminus, the four species isoforms exhibit different numbers of N-glycosylation sites: hH1R (two), bH1R (three), rH1R (two), and gpH1R (one). In E2 (last amino acid), there is an additional potential N-glycosylation site in bH1R and rH1R because the homology model shows that this amino acid should be positioned at the surface of the receptor (Fig. 2).
Analysis of H1R Species Isoforms in the [3H]Mepyramine Saturation Binding Assay. The KD and Bmax values determined in the [3H]mepyramine saturation binding assay are given in Table 2. The KD values at hH1R, bH1R and rH1R were approximately 2-fold higher (p < 0.05) than at gpH1R. The Bmax value for bH1R was approximately 2-fold higher (p < 0.05) than for hH1R, rH1R and gpH1R. Representative saturation binding isotherms for the four H1R species isoforms are shown in Fig. 5. The KD values determined in the Sf9 cell system are in good accordance with the KD values reported in the literature for native tissues, cultured cell lines and recombinant mammalian expression systems (Chang et al., 1979; Aceves et al., 1985; Nakahata et al., 1986; Yamashita et al., 1991; Fujimoto et al., 1993; Gillard et al., 2002; Bruysters et al., 2005). The nonspecific binding of [3H]mepyramine in Sf9 cell membranes expressing H1R species isoforms ranged between 10 and 30% of total [3H]mepyramine binding, providing an excellent signal-to-noise ratio for competition experiments even for the H1R species isoforms exhibiting lower expression levels, i.e., gpH1R.
Analysis of Histaprodifens at H1R Species Isoforms in the [3H]Mepyramine Competition Binding Assay. The results of the competition binding assays are summarized in Table 3. The pKi values of histamine 1 at the four H1R species isoforms did not significantly differ from each other. The pKi values of histaprodifen 2 were significantly higher than the pKi values of histamine 1 at all H1R species isoforms (p = 0.0006 for hH1R, p = 0.03 for bH1R, p = 0.0011 for rH1R, and p = 0.0267 for gpH1R). The introduction of chlorine 3 or fluorine 4 in one of the phenyl moieties of histaprodifen did not result in significant differences in the binding affinity compared with histaprodifen 2 at all four species isoforms. Nα-Methylhistaprodifen 5 had similar pKi values as histaprodifen 2 at all four H1R species isoforms. Compounds 6 and 7 differ from each other in the substitution pattern of the imidazolyl moiety. The pKi values of these compounds were in a range from 6.33 to 7.11 at all four H1R species isoforms. The pKi value of suprahistaprodifen 7 at gpH1R was significantly higher than at hH1R(p = 0.0012). Otherwise, there were no significant differences between these two compounds at H1R species isoforms.
The additional methyl group in suprahistaprodifen in (R)- and (S)-configuration (compounds 8R and 8S) resulted in significant differences between species isoforms and the (R)- and (S)-configuration as well, as illustrated in Fig. 6. There was no significant difference in the pKi values between 7 and 8R at hH1R and gpH1R, but the pKi values of 8S were significant lower (p = 0.0026 for hH1R and p = 0.0006 for gpH1R) at both species isoforms compared with the (R)-configuration of 8. The decrease amounted to ∼0.8 log units at gpH1R and ∼0.7 log units at hH1R. In contrast, at bH1R and rH1R, the pKi values of 8R and 8S were significantly lower than of 7, but there was no significant difference between 8R and 8S. Thus, the largest difference in affinity between the (R)- and (S)-configuration was found at gpH1R, followed by hH1R. The additional phenyl moiety in 9 did not result in significant differences in pKi values relative to suprahistaprodifen 7 at all four H1R species isoforms. Again, the highest pKi value was found at gpH1R, which was significantly higher than at hH1R(p = 0.0082) and bH1R(p = 0.0057). The pKi values between the related phenyl-substituted 10R and 10S did not significantly differ from each other at H1R species isoforms. At gpH1R, the pKi value of 9 was significantly higher than for 10R or 10S (p = 0.004 for 10R and p = 0.0072 for 10S). The pKi values of 11 and 12 were similar at each H1R species isoform. Compounds 13 to 16, possessing a different substitution pattern of the terminal imidazolyl moiety compared with suprahistaprodifen 7 in combination with a varying length of the CH2-spacer, showed decreased affinity compared with histaprodifen 2 at all four H1R species isoforms. Except for 13 at bH1R, the decrease of approximately 0.4–1 log units was significant (p < 0.01). For the largest histaprodifen known so far, dimeric histaprodifen 17, we observed significant increases in pKi values compared with histaprodifen 2 for bH1R, rH1R and gpH1R. The pKi values for histamine 1, suprahistaprodifen 7, and dimeric histaprodifen 17 at hH1R and gpH1R are in good accordance with data published for mammalian expression systems (Gillard et al., 2002; Bruysters et al., 2005). Figure 7 provides an overview of the binding affinities of several histaprodifen derivatives compared with histaprodifen 2 itself. Affinities are given as ΔpKi values. There are significant species-differences between hH1R and gpH1R, whereas bH1R and rH1R exhibit an intermediate behavior. For some compounds, bH1R and rH1R are more similar to gpH1R, and for other compounds, bH1R and rH1R are more similar to hH1R.
Constitutive Activity and Maximum Stimulation with Histamine of hH1R, bH1R, rH1R, and gpH1Rinthe Steady-State GTPase Assay. The basal activity (Fig. 8) of hH1R, rH1R, and gpH1R in the steady-state GTPase assay ranged from 1.1 to 1.4 pmol/(mg min) without significant difference between these three species isoforms. However, the basal activity with bH1R [1.8 pmol/(mg min)] was significantly higher (p < 0.05) than with hH1R, rH1R, and gpH1R. In addition, the maximum stimulation with 100 μM histamine 1 relative to the basal activity (ΔHA) was significantly higher (p < 0.0001) at bH1R than at hH1R, rH1R and gpH1R (Table 4). The ratio ΔHA/Bmax was not significantly different between the four species isoforms, indicating that the large stimulatory effect of histamine at bH1R was a result of the high expression level of bH1R in Sf9 cell membranes. The inverse agonist mepyramine 18 (Fitzsimons et al., 2004) exhibited only small inhibitory effects on basal GTP hydrolysis, indicative for low constitutive activity of H1R species isoforms. The pEC50 values of 18 ranged from 7.66 to 8.96 (Table 4). Only the pEC50 values of bH1R and rH1R were significantly different from gpH1R(p < 0.0025).
Analysis of Histaprodifens at H1R Species Isoforms in the Steady-State GTPase Assay. The results of the GTPase assays are summarized in Tables 5 (agonist mode) and 6 (antagonist mode). There were no significant differences in the pEC50 values of the endogenous ligand histamine 1 among H1R species isoforms. Histaprodifen 2 was a partial agonist at all four species isoforms. At gpH1R and rH1R, the pEC50 of 2 increased approximately 0.8 log units relative to histamine 1. In contrast, there was no significant difference in the pEC50 values of histaprodifen 2 relative to histamine 1 at hH1R and bH1R. The pEC50 values of 2 between gpH1R and hH1R(p = 0.0005) or bH1R(p = 0.0302) were significantly different. However, there was no significant difference in the pEC50 values of 2 between gpH1R and rH1R, which were higher than at hH1R and bH1R. Thus, the additional diphenylpropyl moiety in histaprodifen 2 compared with histamine increased the pEC50 at gpH1R and rH1R but had no effect on hH1R and bH1R. The additional Cl 3 and F 4 substituent in the diphenylpropyl moiety of 3 had no significant influence on the pEC50 relative to histaprodifen 2. The pEC50 of 4 at gpH1R versus hH1R(p = 0.0102) and bH1R(p = 0.0121) and at rH1R versus hH1R(p = 0.0203) and bH1R (p = 0.0111) were significantly different. The pEC50 of 2 compared with the pEC50 of 5 at hH1R, bH1R, and rH1R were not significantly different. In contrast, the pEC50 of 5 relative to 2 was significantly different (p = 0.0006) at gpH1R. These data show that the pEC50 values of the small histaprodifens 2 to 5 were similar between gpH1R and rH1R and higher than at hH1R and bH1R. Compared with histamine 1, pEC50 values increased at gpH1R and rH1R, but not at hH1R and bH1R, except for 5 at bH1R.
The additional imidazolyl moieties in the histaprodifen derivatives 6, 7, and 13 differ in the position of substitution. Compounds 6, 7, and 13 exhibited only small differences in pEC50 values at H1R species isoforms compared with histaprodifen 2. Again, higher pEC50 values were found at gpH1R and rH1R for 6, 7, and 13. The substitution position of the imidazolyl moiety had only a small influence on the pEC50 values at a given species isoform. Compounds 11, 12, and 9 have additional moieties at the R3 position compared with suprahistaprodifen 7. The pEC50 values of 11 were only significantly different between gpH1R and bH1R (p = 0.0061). The additional ethyl group in 11 slightly decreased pEC50 values at rH1R, but it had a small increasing influence at gpH1R and hH1R with respect to suprahistaprodifen 7. There was no significant difference between the four H1R species isoforms with respect to the pEC50 values of the thienylmethyl-substituted suprahistaprodifen 12. Again, the additional substituent slightly decreased the pEC50 values relative to suprahistaprodifen 7. The pEC50 value of 9 at gpH1R was significantly higher (p < 0.05) than at hH1R, bH1R, and rH1R.
The chiral histaprodifens 8R and 8S showed partial agonism at gpH1R. The (R)-enantiomer 8R was significantly more potent than the (S)-enantiomer 8S (p = 0.0068). Only the (R)-enantiomer 8R showed partial agonism at hH1R, bH1R, and rH1R. The potencies and the efficacies were significantly lower than at gpH1R(p < 0.05). The (S)-enantiomer 8S showed antagonism at hH1R, bH1R, and rH1R, with similar pKB values (Table 6). Thus, the additional methyl group, introducing a center of chirality constitutes a unique agonism/antagonism-switch between the species [(S)-configuration 8S] and within the species [(R)/(S)-configuration 8R, 8S]. Figure 9 illustrates this switch for gpH1R and hH1R.
The enantiomeric compounds 10R and 10S were partial agonists with equal potency at gpH1R. The efficacy of the (R)-enantiomer 10R was significantly higher (p = 0.0062) than of the (S)-enantiomer 10S. Both enantiomers were antagonists at hH1R, bH1R and rH1R. There were no significant differences in the pKB values between the (R)-enantiomer 10R and the (S)-enantiomer 10S at hH1R, bH1R, and rH1R (Table 6). However, the pKB values for 10R and 10S were significantly higher at hH1R than at bH1R and at rH1R. At bH1R and rH1R, the pKB values are in the same range. Thus, this pair of enantiomers showed agonism/antagonism-switch between species.
In compounds 13 to 16, the length of the CH2-spacer was varied, resulting in differences in pEC50 values of approximately 0.7 log units at each species isoform between the compounds of this series. Compound 15, with a (CH2)4-spacer, showed the lowest potency at all four species isoforms. The highest potency in this series was evident for 13 or 14. For dimeric histaprodifen 17, species differences were detected for all pairs of H1R species isoforms (p < 0.05), except for the pairs gpH1R/rH1R and bH1R/rH1R. The potency increased in the order hH1R < bH1R < rH1R < gpH1R.
At gpH1R, all histaprodifens were partial agonists. The compounds showed Emax values from 0.23 to 0.92. Only 10S showed an Emax value <0.5. Emax values at gpH1R and rH1R were higher than at bH1R and hH1R in most cases. In addition, the Emax values between hH1R and bH1R, on one hand, and between rH1R and gpH1R, on the other hand, were similar.
Binding Mode of Dimeric Histaprodifen in gpH1R. MD simulations with docked dimeric histaprodifen 17 in gpH1R showed that the ligand interacts with a binding pocket straight through the entire receptor (Fig. 10A, only hydrogen bonds and electrostatic interactions between ligand and receptor are shown). The following amino acids participate in the binding of 17. Leu-39 (TM1, 1.35), Leu-43 (TM1, 1.39), Leu-97 (TM2, 1.65), Trp-112 (TM3, 3.28), Asp-116 (TM3, 3.32), Tyr-117 (TM3, 3.33), Ser-120 (TM3, 3.36), Ile-124 (TM3, 3.40), Trp-167 (TM4, 4.56), backbone carbonyl of Lys-187 (E2), Glu-190 (E2), Trp-429 (TM6, 6.48), Tyr-432 (TM6, 6.51), Trp-456 (TM7, 7.40), and Tyr-459 (TM7, 7.43). The quaternary amine moiety in the center of dimeric histaprodifen is electrostatically trapped between the negatively charged Asp-116 (TM3, 3.32) and Glu-190 (E2). One of the histamine moieties forms a stable hydrogen bond interaction with Ser-120 (TM3, 3.36) and Tyr-432 (TM6, 6.51). The second histamine moiety forms a stable hydrogen bond to the backbone carbonyl of Lys-187 (E2). Both diphenylpropyl moieties are embedded in hydrophobic pockets formed by amino acids Trp-429 (TM6, 6.48), Ile-124 (TM3, 3.40) and Trp-167 (TM4, 4.56) as well as Leu-39 (TM1, 1.35), Leu-43 (TM1, 1.39), Leu-97 (TM2, 1.65), Trp-112 (TM3, 3.28), and Trp-456 (TM7, 7.40), respectively.
For suprahistaprodifen 7 and the chiral histaprodifens 8R and 8S, we found two possible orientations in the binding pocket. In orientation I, 7 interacts electrostatically with Asp-116 (TM3, 3.32), Ser-120 (TM3, 3.36), Lys-187 (E2, interaction with backbone carbonyl), Glu-190 (E2), and Tyr-432 (TM6, 6.51) (Fig. 10B). The diphenyl propyl moiety is embedded in a pocket formed by Ile-124 (TM3, 3.40) and Trp-167 (TM4, 4.56). In orientation II, 7 also interacts electrostatically with Asp-116 (TM3, 3.32), Ser-120 (TM3, 3.36), Lys-187 (E2, interaction with backbone carbonyl), Glu-190 (E2), and Tyr-432 (TM6, 6.51) (Fig. 10C). However, in this case the diphenyl propyl moiety is embedded in a hydrophobic pocket formed by Leu-39 (TM1, 1.35), Leu-43 (TM1, 1.39), Leu-97 (TM2, 1.65), Trp-112 (TM3, 3.28), and Trp-456 (TM7, 7.40). The additional methyl group in 8R and 8S is embedded in a small individual pocket facing toward TM7 (Fig. 10D). In orientation I, both enantiomers 8R and 8S fit well into the binding pocket. However, in orientation II only 8R fits well because of a steric clash between the additional methyl group in 8S and Ile-455 (TM7, 7.39). As consequence, a slight movement of the Ile-455 (TM7, 7.39) toward the binding pocket was observed (see blue circle in Fig. 10D).
Discussion
Pharmacological Differences between hH1R, bH1R, rH1R, and gpH1R Concerning the Interaction with Chiral Histaprodifens. The results of the steady-state GTPase assay with the novel chiral histaprodifens 8R, 8S, 10R, and 10S showed unexpected and unique pharmacological differences between gpH1R versus hH1R, bH1R, and rH1R. At gpH1R, all four chiral histaprodifens were partial agonists. However, only 8R was a partial agonist at hH1R, bH1R, and rH1R, whereas 8S, 10R, and 10S were antagonists at those H1R species isoforms. For GPCR activation, a conformational change in Trp6.48/Phe6.52, referred to as aromatic switch, is important (Crocker et al., 2006). Trp6.48 (Trp-429 in gpH1R) is important for direct interaction of histaprodifens with H1R species isoforms (Fig. 2). Thus, it is possible that the additional methyl group in 8S, 10R, and 10S occupies a part of the binding pocket in hH1R, bH1R and rH1R like a molecular wedge and prevents reorientation of the aromatic amino acids Phe6.52 and Trp6.48, with the consequence that the receptor cannot be activated.
Chiral histaprodifens 8R, 8S, 10R, and 10S were also analyzed at the guinea pig ileum (Striegl, 2006). All four compounds acted as partial agonists, and the (R)-enantiomers 8R (pEC50 = 7.67) and 10R (pEC50 = 7.36) were approximately 1 log unit more potent than the corresponding (S)-enantiomers (8S) (pEC50 = 6.81) and 10S (pEC50 = 6.44). These results are in good agreement with the results obtained in the recombinant system (Table 5; Fig. 9). Similar results were obtained for some ergoline derivatives at H1R. Specifically, 8S-lisuride acted as potent partial agonist at the guinea pig ileum, whereas the 8R-lisuride showed only weak antagonism (Pertz et al., 2006). An analogous agonism/antagonism-switch as for the enantiomeric pairs 8R/8S and 10R/10S was observed for 1-allyl-8S-lisuride in the GTPase assay. This compound is a partial agonist at gpH1R, but an antagonist at hH1R (Pertz et al., 2006). Species-dependent stereoselectivity of receptor ligands is well documented (Soudjin et al., 2003) and extends beyond GPCRs. As an example, for chiral C-cyclopropylalkylamides, the (+)-enantiomer preferentially activates the human pregnane X receptor, whereas the (–)-enantiomer preferentially activates the mouse receptor (Mu et al., 2005).
H1R Binding Pocket for Histaprodifens. The MD simulations showed that dimeric histaprodifen 17 is docked into a binding pocket straight through the entire gpH1R (Fig. 10A). Because of the symmetrical properties of dimeric histaprodifen 17, all histaprodifen derivatives with a dimeric histamine moiety (7-12) could bind in two different orientations. Suprahistaprodifen 7 and the chiral histaprodifens 8R and 8S are predicted to interact with Asp-116 (TM3, 3.32), Ser-120 (TM3, 3.36), Lys-187 (E2), Glu-190 (E2), and Tyr-432 (TM6, 6.51) in both orientations. The diphenylpropyl moiety is predicted to interact with Ile-124 (TM3, 3.40), Trp-167 (TM4, 4.56), and Trp-429 (TM6, 6.48) in orientation I (Fig. 10B) and with Leu-39 (TM1, 1.35), Leu-43 (TM1, 1.39), Leu-97 (TM2, 1.65), Trp-112 (TM3, 3.28), and Trp-456 (TM7, 7.40) in orientation II (Fig. 10C).
The chiral histaprodifen 8R fits well into the binding pocket in both orientations. However, 8S only fits well in orientation I due to a steric clash between the additional methyl group and Ile-455 (TM7, 7.39) (Fig. 10D). At gpH1R the affinities of suprahistaprodifen 7 and the chiral histaprodifen 8R are similar, whereas a decrease in affinity is observed for 8S. Based on these results two hypotheses can be proposed. First, suprahistaprodifen 7 and the chiral histaprodifens 8R and 8S may only bind in orientation II into the binding pocket. As a consequence of the sterical clash, the affinity of 8S is decreased. Second, suprahistaprodifen 7 and the chiral histaprodifen 8R may bind in orientations I and II, but 8S may only bind in orientation I. Thus, the experimentally observed Ki value for 7 and 8R may result from summation of the Ki values for orientations I and II, whereas the Ki value for 8S may only be derived from orientation I. With our present experimental data, we cannot distinguish which of the two hypotheses is correct. By analogy to our data, Dezi et al. (2007) discuss two possible orientations for some butyrophenone derivatives at the serotonin 5-hydroxytryptamine2A receptor.
Molecular Basis for Pharmacological Differences between hH1R, bH1R, rH1R, and gpH1R. The alignment of the amino acid sequences is given in Fig. 2. Therein, the amino acids that interact directly with dimeric histaprodifen 17 in gpH1R are highlighted. The same amino acids are found in hH1R, bH1R and rH1R, too. The only difference occurs in the E2-loop at position 187 (gpH1R), with a lysine in gpH1R but an aspartate in hH1R, bH1R, and rH1R. However, our simulations showed that dimeric histaprodifen 17 forms a stable hydrogen bond only to the backbone of Lys-187 in gpH1R, but not to the side chain, which is pointing away from the binding pocket. Thus, the differences in this position should not be responsible for the observed species differences in pharmacology.
The alignment shows a difference between gpH1R versus hH1R, bH1R and rH1R in TM2 in position 2.61. In hH1R, bH1R, and rH1R, there is an asparagine, which is changed to a serine in the gpH1R. Bruysters et al. (2005) reported that this mutation is responsible for species differences between gpH1R and hH1R concerning suprahistaprodifen 7 and dimeric histaprodifen 17. Our studies showed an analogous species difference in pKi values of suprahistaprodifen 7 and dimeric histaprodifen 17 between hH1R and gpH1R. If the amino acid exchange in position 2.61 alone had been exclusively responsible for the observed species differences between hH1R and gpH1R, the pKi values of dimeric histaprodifen 17 at bH1R and rH1R would have been expected to be identical to that at hH1R. However, we noticed a trend that the pKi values and pEC50 values of suprahistaprodifen 7 and dimeric histaprodifen 17 exhibited the order gpH1R > rH1R > bH1R > hH1R. Thus, it can be concluded that the amino acid exchange in position 2.61 is responsible for part of the species differences between hH1R and gpH1R, but other differences in amino acid sequences also influence pEC50 and pKi values.
In the competition binding assay, we observed that, in general, histaprodifens exhibited higher pKi values at gpH1R and lower pKi values at hH1R and bH1R. The pKi values at rH1R were intermediate between those at gpH1R and hH1R/bH1R. In addition, in the steady-state GTPase assay, histaprodifens were, in general, more potent and efficacious at gpH1R, followed by rH1R, bH1R, and hH1R. However, as pointed out above, there are no species differences in the amino acid side chains that directly interact with dimeric histaprodifen in the binding pocket of gpH1R. Therefore, it can be concluded, that amino acids that are different between the four species isoforms and not in direct contact with the binding pocket, have an indirect influence onto the binding pocket and to the binding mode of ligands.
Conclusions
Our studies document substantial pharmacological differences between H1R species isoforms. With respect to their pharmacological profile, bH1R and rH1R can be classified intermediate between hH1R and gpH1R, bH1R being more similar to hH1R and rH1R being more similar to gpH1R. This trend is approximately reflected in Table 1, where the level of similarity of the amino acid sequences between hH1R, bH1R, rH1R, and gpH1R is given. Our studies support the usefulness of GPCR species isoforms as tools to analyze the binding mode of ligands. However, our data also clearly show that in some cases, it is difficult to explain species differences in pharmacology with a single defined difference in amino acid sequence. Finally, the chiral histaprodifens are novel unique ligands to dissect subtle differences in the activation mechanism of the H1R in various species.
Acknowledgments
We thank Dr. H. Fukui for providing cDNAs for bH1R and rH1R; Dr. S. Elz (Department of Pharmaceutical and Medicinal Chemistry I, University of Regensburg, Germany) for helpful discussions and support; K. Wohlfart, A. Seefeld, A.-C. Groß, and C. Huber for performing the GTPase and binding assays; and G. Wilberg for competent help with the cell culture and immunoblot analysis. We also thank the reviewers for constructive critique.
Footnotes
-
This work was supported by the Research Training Program (Graduierten-kolleg) GRK760 “Medicinal Chemistry: Molecular Recognition–Ligand-Receptor Interactions” of the Deutsche Forschungsgemeinschaft.
-
Article, publication date, and citation information can be found at http://jpet.aspetjournals.org.
-
doi:10.1124/jpet.107.129601.
-
ABBREVIATIONS: H1R, histamine H1 receptor; GPCR, G protein-coupled receptor; TM transmembrane domain; gp, guinea pig; h, human; b, bovine; r, rat; RGS, regulator of G protein signaling; PCR, polymerase chain reaction; S, signal peptide; F, FLAG; DMSO, dimethyl sulfoxide; MD, molecular dynamics; bb, backbone; sc, side chain; lig, ligand; E, extracellular loop; HA, histamine; cpd., compound; I, intracellular loop, MEP, mepyramine; ps, picosecond.
- Received August 1, 2007.
- Accepted October 9, 2007.
- The American Society for Pharmacology and Experimental Therapeutics