Abstract
The histamine H4 receptor (H4R) is involved in the chemotaxis of leukocytes and mast cells to sites of inflammation and is suggested to be a potential drug target for asthma and allergy. So far, selective H4R agonists have not been identified. In the present study, we therefore evaluated the human H4R (hH4R) for its interaction with various known histaminergic ligands. Almost all of the tested H1R and H2R antagonists, including several important therapeutics, displaced less than 30% of specific [3H]histamine binding to the hH4R at concentrations up to 10 μM. Most of the tested H2R agonists and imidazole-based H3R ligands show micromolar-to-nanomolar range hH4R affinity, and these ligands exert different intrinsic hH4R activities, ranging from full agonists to inverse agonists. Interestingly, we identified 4-methylhistamine as a high-affinity H4R ligand (Ki = 50 nM) that has a >100-fold selectivity for the hH4R over the other histamine receptor subtypes. Moreover, 4-methylhistamine potently activated the hH4R (pEC50 = 7.4 ± 0.1; α = 1), and this response was competitively antagonized by the selective H4R antagonist JNJ 7777120 [1-[(5-chloro-1H-indol-2-yl)-carbonyl]-4-methylpiperazine] (pA2 = 7.8). The identification of 4-methylhistamine as a potent H4R agonist is of major importance for future studies to unravel the physiological roles of the H4R.
Histamine exerts many (patho-)physiological effects through its interaction with four histamine receptor subtypes that all belong to the family of G protein-coupled receptors (Hough, 2001). The histamine H1 receptor (H1R) and H2 receptor (H2R) were pharmacologically identified long before their cDNAs were cloned (Gantz et al., 1991; Yamashita et al., 1991), and they have been successful blockbuster targets for more than two decades. The cDNA encoding the histamine H3R was cloned more recently (Lovenberg et al., 1999), and bioinformatic analysis of human genome databases resulted in identification of the gene encoding the human H4R (hH4R) based on its sequence homology to the H3R gene (37%) (Oda et al., 2000; Liu et al., 2001a; Morse et al., 2001; Nguyen et al., 2001; Zhu et al., 2001). Although the hH3R is mainly present in the nervous system, the hH4R is distributed mainly in hematopoietic cells (Oda et al., 2000; Liu et al., 2001a; Morse et al., 2001; Zhu et al., 2001). The H4R shows a different pharmacological profile compared with the closely related H3R, although many H3R ligands also interact with the H4R. Like the H3R, the H4R couples to pertussis toxin-sensitive Gi/o proteins and thereby inhibits forskolin-induced cAMP production (Oda et al., 2000; Liu et al., 2001a; Zhu et al., 2001). In addition, the H4R also activates mitogen-activated protein kinase (Morse et al., 2001) and mobilizes calcium in eosinophils and mast cells (Buckland et al., 2003; Hofstra et al., 2003).
The presence of the hH4R on leukocytes and mast cells suggests that this new histamine receptor plays an important role in the modulation of the immune system. This hypothesis is supported by the fact that IL-10 and IL-13 modulate hH4R expression (Morse et al., 2001) and that binding sites for cytokine-regulated transcription factors, such as interferon-stimulated response element, interferon regulatory factor-1, nuclear factor-κB, and nuclear factor-IL6, are present upstream of the hH4R gene (Cogé et al., 2002). Physiological roles of the hH4R include the control of IL-16 release by human CD8+ T cells (Gantner et al., 2002), chemotactic responses and cytoskeletal changes of human eosinophils (O'Reilly et al., 2002; Buckland et al., 2003; Ling et al., 2004), chemotaxis and intracellular calcium mobilization in mast cells (Hofstra et al., 2003), and control of leukotriene B4 production by mast cells that subsequently leads to neutrophil recruitment into peritoneum (Takeshita et al., 2003; Thurmond et al., 2004). These studies suggest that the hH4R is a potential drug target for immune system-related diseases.
Until recently, potent and selective H4R ligands were not available. In the early studies, the H3R antagonist thioperamide was identified as an equally effective H4R antagonist (Oda et al., 2000; Hough, 2001). High-throughput screening and subsequent medicinal chemistry efforts recently identified the indolylpiperazine JNJ 7777120 (Jablonowski et al., 2003; Thurmond et al., 2004) and the related benzimidazole analog VUF 6002 (Terzioglu et al., 2004) as selective and potent H4R antagonists. Studies directed toward selective H4R agonists have so far been less successful. Burimamide, clozapine, and clobenpropit are all known to act as H4R agonists, and clozapine and clobenpropit have been proven useful for initial pharmacological studies (Gantner et al., 2002; Buckland et al., 2003; Bell et al., 2004; Ling et al., 2004). Currently, the most selective H4R agonist is the imifuramine analog OUP-16, which displays a 15-fold selectivity for the H4R compared with its binding affinity for the H3R (Hashimoto et al., 2003). However, the lack of selectivity of currently known agonists for the H4R limits their use as H4R agonists.
In our search for selective H4R agonists, many known histaminergic ligands of different structural classes, including several important therapeutics, were evaluated for their interaction with the hH4R. Our studies resulted in the identification of 4-methylhistamine, a presumed moderately active and selective H2R agonist (Durant et al., 1975), as a high-affinity H4R agonist with a more than 100-fold selectivity over the H1R, H2R, and H3R.
Materials and Methods
Materials. Aminopotentidine, amthamine dihydrobromide, amselamine dihydrobromide, burimamide oxalate, and burimamide analogs (Vollinga et al., 1995); clobenpropit dihydrochloride, dimaprit dihydrobromide, histaprodifen dimaleate, homohistamine dihydrobromide, imbutamine dihydrobromide, imetit dihydrobromide, impentamine dihydrobromide, immepip dihydrobromide, immethridine dihydrobromide, iodophenpropit dihydrobromide, JNJ 7777120, 2-(3-bromophenyl)histamine dihydrobromide, methimmepip dihydrobromide, 2-pyridylethylamine (PEA) dihydrochloride, 2-(2-thiazolyl)ethylamine (TEA) dihydrochloride, thioperamide fumarate, and VUF 8328 were synthesized at the Department of Medicinal Chemistry (Vrije Universiteit Amsterdam, Amsterdam, The Netherlands). Famotidine, ketotifen fumarate, and 8R-lisuride were purchased from MP Biomedicals (Irvine, CA). Amoxapine, d-chlorpheniramine maleate, clozapine, cimetidine, N-desmethyl clozapine, diphenhydramine hydrochloride, doxepin hydrochloride, forskolin, histamine dihydrochloride, imipramine hydrochloride, loxapine, mepyramine (pyrilamine maleate), (R)-α-methylhistamine dihydrochloride, (S)-α-methylhistamine dihydrochloride, Nα-methylhistamine dihydrochloride, N-oxide clozapine, octoclothepin, pertussis toxin, polyethyleneimine, ranitidine hydrochloride, tripelennamine hydrochloride, and triprolidine hydrochloride were purchased from Sigma/RBI (Natick, MA). 2-Nitrophenol-β-d-pyranoside and G418 (geneticin) were from Duchefa (The Netherlands); promethazine was from VUMC Pharmacy Amsterdam (Amsterdam, The Netherlands); fexofenadine was from Ultrafine Chemicals (Manchester, UK); tiotidine was from Imperial Chemical Industries PLC (London, UK); and [3H]Nα-methylhistamine (85 Ci/mmol), [3H]histamine (12.4 Ci/mmol), and [3H]mepyramine (23 Ci/mmol) were from PerkinElmer Life and Analytical Sciences (Boston, MA). [125I]Iodoaminopotentidine and [125I]iodophenpropit were labeled at the Department Nuclear Medicine and PET Research (Vrije Universiteit Medical Centre, Amsterdam, The Netherlands) as described previously (Jansen et al., 1992), whereas [3H]JNJ 7777120 (84 Ci/mmol) was synthesized at Johnson & Johnson Pharmaceutical Research and Development, L.L.C. (La Jolla, CA) (Thurmond et al., 2004). Gifts of astemizole (Janssen Pharmaceuticals, Antwerp, Belgium); cyproheptadine hydrochloride (MSD, Haarlem, The Netherlands); cetirizine hydrochloride and hydroxyzine dihydrochloride (UCB Pharma, Brussels, Belgium); ebastine (Almirall Prodesfarma, Barcelona, Spain); loratidine (Schering Plough, Kenilworth, NJ); mianserin hydrochloride and ORG3770 (Organon NV, Oss, The Netherlands); mifentidine (Instituto De Angeli, Milan, Italy); mizolastine (Synthélabo Recherche, Bagneux, France); proxyfan dihydrochloride and iodoproxyfan dihydrochloride (Dr. J. A. M. Christiaans; Kovalainen et al., 1999); S(+)- and R(-)-sopromidine (Institute of Pharmacy, Free University Berlin, Berlin, Germany); and R-(+)- and S-(-)-terfenadine carboxylate (Sepracor, Marlborough, MA), 2-methylhistamine dihydrochloride, 4-methylhistamine dihydrochloride, and impromidine dihydrochloride (GlaxoSmithKline, Welwyn Garden City, Hertfordshire, UK) are greatly acknowledged.
Cell Culture. SK-N-MC cell lines, which stably express either the human H3R (SK-N-MC/hH3) or H4R (SK-N-MC/hH4) as well as a cAMP-responsive element (CRE)-driven β-galactosidase reporter gene SK-N-MC/hH3 or SK-N-MC/hH4 cells (Lovenberg et al., 1999; Liu et al., 2001a), were cultured in Eagle's minimum essential medium supplemented with 5% fetal calf serum, 0.1 mg/ml streptomycin, 100 U/ml penicillin, and 600 μg/ml G418 at 37°C in 5% CO2 and 95% humidity.
Radioligand Binding Assays. The SK-N-MC/hH3 cell homogenates were incubated for 40 min at 25°C with approximately 1 nM [3H]Nα-methylhistamine in 25 mM KPO4 buffer and 140 mM NaCl (pH 7.4 at 25°C), with or without competing ligands, whereas the SK-N-MC/hH4 cell homogenates were incubated 1 h at 37°C in 10 nM [3H]histamine and 50 mM Tris-HCl (pH 7.4 at 37°C), with or without competing ligands. Bound radioligands were collected on 0.3% polyethyleneimine-pretreated Whatman GF/C [and washed three times with 3 ml of ice-cold washing buffer (4°C) containing 25 mM Tris-HCl and 140 mM NaCl (pH 7.4 at 4°C) for the hH3R and 50 mM Tris-HCl (pH 7.4 at 4°C) for the hH4R]. Binding analysis of 10 nM [3H]JNJ 7777120 and 0.1 nM [125I]iodophenpropit to the hH4R was performed with the same conditions as described for [3H]histamine. In saturation binding analysis, the nonspecific binding of [3H]histamine or [3H]JNJ 7777120 was determined with 1 μM clobenpropit. The binding analysis of [3H]mepyramine and [125I]iodoaminopotentidine binding to human H1R and human H2R, respectively, was performed according to Bakker et al. (2004). The binding data were analyzed with Prism 4.0 (GraphPad Software Inc., San Diego, CA), and data are presented as mean ± S.E.M. Mouse and rat H4R radioligand binding assays were performed according to Liu et al. (2001b).
Colometric Cyclic AMP Assay. A reporter CRE-β-galactosidase reporter gene assay was used to determine (inverse) agonistic or antagonistic activity of either the hH3R or hH4R. Approximately 4 million cells/96-well plate of SK-N-MC/hH3 and SK-N-MC/hH4 cells were exposed for 6 h to histaminergic ligands in serum-free Eagle's minimum essential medium containing 1 μM forskolin. Thereafter, the medium was discarded, the cells were lysed in 100 μl of assay buffer (100 mM sodium phosphate buffer at pH 8.0, 4 mM 2-nitrophenol-β-d-pyranoside, 0.5% Triton X-100, 2 mM MgSO4, 0.1 mM MnCl2, and 40 mM β-mercaptoethanol), incubated overnight at room temperature, and the β-galactosidase activity was determined at 420 nm with a PowerwaveX340 plate reader (Bio-Tek Instruments Inc., Winooski, VT). The OD420 might differ between experiments due to intra-assay variability; therefore, intrinsic activity of agonists was determined relatively to activity of histamine.
Primary Cell Experiments. Cell culture of BALB/c mice-derived bone marrow mast cells (BMMCs) and in vitro BMMC chemotaxis assay was performed as described previously (Hofstra et al., 2003). Purification of human polymorphonuclear leukocytes and the human eosinophil shape change assay were performed as described previously (Ling et al., 2004). The mouse-derived BMMCs were obtained following approved protocols that follow National Institutes of Health/International Animal Care and Use Guidelines.
Results
Pharmacological Characterization of the hH4R Expressed in SK-N-MC Cells. Stable transfection of the hH4R cDNA in SK-N-MC cells resulted in the expression of functional hH4R proteins. The hH4R could be labeled with both agonist and antagonist radioligands. The H4R agonist [3H]histamine shows saturable binding to the expressed H4R with a minimal amount of nonspecific binding (Fig. 1A). Analysis of the [3H]histamine saturation binding yielded a Kd value of 11 ± 1.0 nM (n = 6) and a Bmax value of 1.8 ± 0.4 pmol/mg protein. Recently, JNJ 7777120 was described as a selective H4R antagonist (Jablonowski et al., 2003). In our hands, the nonimidazole JNJ 7777120 shows a 300-fold selectivity for the hH4R(pKi = 7.8 ± 0.1 against [3H]histamine) over the hH3R (pKi = 5.3 ± 0.1 against [3H]Nα-methylhistamine), allowing the use of [3H]JNJ 7777120 to label the H4R (Thurmond et al., 2004). The H4R antagonist [3H]JNJ 7777120 exhibits a somewhat higher level of nonspecific binding to hH4R expressing SK-N-MC cells, but it also binds saturably and shows an equipotent affinity (Kd = 11 ± 3.6 nM; n = 3) and results in a Bmax value of 1.7 ± 0.4 pmol/mg protein (Fig. 1B). The binding of either 10 nM [3H]histamine (Fig. 1C) or 10 nM [3H]JNJ 7777120 (data not shown) to the hH4R is fully displaced by histamine (pKi = 7.8 ± 0.1), the H3/4R antagonist thioperamide (pKi = 6.9 ± 0.1) and the H4R agonist/H3R antagonist clobenpropit (pKi = 8.1 ± 0.1), in a good agreement with the results reported previously (Oda et al., 2000; Liu et al., 2001a; Morse et al., 2001).
The SK-N-MC/hH4 cells used in this study coexpress a CRE-controlled β-galactosidase reporter gene and can therefore also be used for a functional analysis of H4R ligands. Stimulation of the hH4R with histamine resulted in the inhibition (58 ± 3%; n = 16) of the forskolin-stimulated (1 μM) cAMP-mediated reporter gene transcription with a pEC50 value of 7.7 ± 0.1 (n = 16) (Fig. 1D). Treatment of SK-N-MC/hH4 cells with the Gαi/o protein inhibitor pertussis toxin (100 ng/ml for 16 h) completely inhibited histamine induced responses, confirming the coupling of the H4R to Gαi/o proteins (Oda et al., 2000; Liu et al., 2001a; Zhu et al., 2001). In our hands, histamine exerted the maximally observed level of inhibition in this assay and is therefore referred to as a full agonist (intrinsic activity α = 1). As reported previously (Oda et al., 2000; Liu et al., 2001a; Morse et al., 2001), clobenpropit acts as a potent partial hH4R agonist with a pEC50 value of 7.7 ± 0.1 (n = 3) and an intrinsic activity of 0.8 (Fig. 1D).
Treatment of SK-N-MC/hH4 cells with pertussis toxin (100 ng/ml for 16 h) resulted in an increase of 1 μM forskolin-stimulated CRE activity by 130 ± 3%, suggesting that the hH4R shows a detectable level of constitutive activity in the SK-N-MC/hH4 cells. In line with previous observations on inverse agonism by thioperamide (Liu et al., 2001a; Morse et al., 2001), 1 μM forskolin-stimulated CRE activity was increased by thioperamide with a pEC50 value of 7.0 ± 0.1 (Fig. 1D). The inhibition of the constitutive activity of the hH4Rby thioperamide was of the same magnitude as observed after treatment with pertussis toxin and thioperamide is referred to as a full inverse hH4R agonist (intrinsic activity α = -1).
In SK-N-MC/hH4 cells the cAMP-driven β-galactosidase reporter-gene transcription also can be activated by endogenously expressed Gαs protein-coupled β adrenergic receptors (Bahouth et al., 2001). The β2 adrenergic receptor agonist feneterol induced β-galactosidase activity to a similar extent to that of forskolin, with a pEC50 value of 6.9 ± 0.1 (n = 6). H4R activation by histamine inhibited the 100 nM feneterol-induced β-galactosidase activity for 39 ± 3% with a pEC50 value of 7.4 ± 0.1 (n = 7). However, inverse agonistic activity of thioperamide, at the hH4R, could not be easily demonstrated with a feneterol-based assay system (data not shown). The evaluation of the functional activity of all the various histaminergic ligands was therefore performed using forskolin (1 μM) stimulated SK-N-MC/hH4 cells.
All compounds were preliminarily tested as displacers of [3H]histamine binding to the hH4R expressed in SK-N-MC/hH4 cells at a concentration of 10 μM. Compounds inhibiting the specific binding of 10 nM [3H]histamine to the hH4R by ≤30% are expected to have a Ki > 10 μM based on the Cheng and Prusoff equation (Cheng and Prusoff, 1973): Ki = IC50/(1 + [radioligand]/Kd) and were excluded for further testing. Active compounds (displacement ≥30%) were tested more extensively in both [3H]histamine displacement studies and the CRE-β-galactosidase-based functional H4R assay.
Most H1R Ligands Are Devoid of H4R Activity. Histamine potently displaces [3H]histamine from the hH4R with a pKi value of 7.8 ± 0.1 (Table 1), whereas H1R agonists with substituents at the 2-position of the imidazole ring show significantly lower affinities. Substitution of the imidazole ring with either a small methyl or large 3-bromophenyl substituent is not tolerated and causes an almost 100-fold drop of affinity. Bulkier substituents at the 2-position (1,1-diphenylpropyl in histaprodifen) even result in a total loss of affinity for the hH4R (Table 1). Agonists, lacking the imidazole ring, such as TEA, PEA, or 8R-lisuride (Bakker et al., 2004), are also not active at the hH4R (Table 1).
After an initial report that the H4R can be labeled with [3H]mepyramine (Nguyen et al., 2001), a large number of H1R antagonists (Table 1), including many clinically relevant drugs, were evaluated for their hH4R affinity as well. Almost all tested H1R antagonists, including mepyramine, showed pKi values <5 (Table 1) and did not show functional activity at 1 and 10 μM at the hH4R (data not shown). Although structurally similar to some tricyclic H1R antagonists devoid of H4R affinity, clozapine binds with moderate potency to the hH4R (pKi = 6.7 ± 0.1) and exerts full agonistic activity at the hH4R with a pEC50 value of 6.8 ± 0.1 (n = 5). N-Desmethyl clozapine, a clozapine metabolite, showed a slightly decreased affinity (pKi = 6.5 ± 0.1), whereas N-oxide clozapine, another clozapine metabolite, is totally devoid of hH4R affinity. Furthermore, we evaluated clozapine analogs of therapeutic importance as well. Loxapine and amoxapine showed >10-fold lower affinity (pKi 5.4 ± 0.1 and 5.3 ± 0.1, respectively), whereas octoclothepin did not show binding for the hH4R at all.
Some H2R Ligands Act as hH4R Agonists. Within the series of known H2R agonists that we have tested, only some ligands retain H4R activity. Replacement of the imidazole ring of histamine in the selective H2R agonists amthamine and amselamine (Leurs et al., 1994) results in a total loss of hH4R activity at concentrations up to 10 μM. Dimaprit, a H2R agonist/H3R antagonist lacking an imidazole group, binds the H4R with moderate affinity, showing a pKi value of 6.5 ± 0.1, and exerts partial H4R agonistic activity (Table 1). Impromidine, which was reported to bind to both H2R and H3R, also binds potently to the hH4R with a pKi value of 7.6 ± 0.1 and acts as a partial H4R agonist (α = 0.5). Both the R- and S-enantiomers of the related sopromidine bind, respectively, >10 and >100 times less potently. In fact, the first reported H2R selective agonist 4-methylhistamine (Durant et al., 1975) is the only known H2R agonist that also acts as full agonist at the H4R (Table 1). 4-Methylhistamine binds two times less potently than histamine to the hH4R, exhibiting a pKi value of 7.3 ± 0.1 (n = 3).
Most tested H2R antagonists, including cimetidine, mifentidine, aminopotentidine, ranitidine, famotidine, and tiotidine, only displaced <30% of 10 nM [3H]histamine binding to the hH4R. Only the H2/3R ligand burimamide shows a high affinity for the hH4R pKi = 7.4 ± 0.1 (Table 1). Moreover, burimamide acted as a potent, albeit partial H4R agonist (pEC50 = 7.7 ± 0.1; α = 0.7). Previously, we reported on various burimamide analogs as H3R antagonists (Vollinga et al., 1995). In our search for H4R selective ligands, various burimamide analogs were therefore investigated for their H4R activity. In this series of compounds, the presence of an isopropyl (VUF 4683 and VUF 4616) or cyclohexyl (VUF 4617) moiety adjacent to the thiourea group improved the affinity for the hH4R (Table 2). Interestingly, this series of closely related compounds exerts partial agonistic, neutral antagonistic, and inverse agonistic activities at the hH4R (Table 2; Fig. 2A). Substitution on the thiourea with aromatic substituents, like a benzyl group in VUF 4686, results in a reduced H4R agonistic activity. A total loss of agonistic H4R activity, but not affinity (pKi = 7.6 ± 0.1) is surprisingly observed for VUF 4614. As can be seen in Fig. 2B, VUF 4614 was able to competitively block the hH4R agonistic responses of histamine, resulting in a pA2 value of 6.8. Finally, within this series we identified VUF 4742 as an hH4R inverse agonist (Fig. 2A). This burimamide analog bound with moderate affinity to the H4R(pKi = 6.9 ± 0.1; n = 4) and acted as a full inverse agonist with a pEC50 value of 7.2 ± 0.1 (n = 5), in accordance with its binding affinity.
Evaluation of H3R Ligands at the hH4R. The H4R shares its highest sequence homology with the H3R, and it is therefore not surprising that in the initial studies some H3R ligands were identified as H4R ligands as well. We therefore characterized in this study a large set of known H3R ligands for their interaction with the H4R. The histamine analogs Nα-methylhistamine, (R)-α-methylhistamine, and (S)-α-methylhistamine show an almost 2 order of magnitude lower affinity for the hH4R than for the hH3R. However, the hH4R retains some level of stereoselectivity for (R)-α- (pKi = 6.6 ± 0.1) and (S)-α-methylhistamine (pKi = 5.4 ± 0.1) (Table 3). Increasing the spacer length between imidazole and amine group from two carbon atoms (histamine, pKi = 7.8 ± 0.1) to three carbon atoms (homohistamine, pKi = 7.5 ± 0.1) slightly decreases the affinity for the hH4R, whereas four carbon atoms (imbutamine, pKi = 8.0 ± 0.1) results in a slightly higher hH4R affinity. A further increase of the spacer length proved to be detrimental for hH4R affinity. The highly potent H3R agonist impentamine shows only moderate affinity at the H4R(pKi = 6.6 ± 0.1) (Table 3). Interestingly, besides the affinity, impentamine also looses intrinsic activity for the hH4R (α = 0). Previously identified H3R agonists, including immepip, imetit, and VUF 8328 (an imetit analog) (Wieland et al., 2001), also potently bind the hH4R with pKi values of 7.7 ± 0.1, 8.2 ± 0.1, and 8.0 ± 0.1, respectively. At the hH4R these ligands also act as agonists, but exert somewhat lower intrinsic activity (α values of 0.9, 0.9, and 0.6, respectively) (Table 3). As reported previously (Kitbunnadaj et al., 2004), the recently identified H3R agonist immethridine (pKi = 9.1 ± 0.1) binds much less potently to the hH4R (pKi = 6.6 ± 0.1) and is also not able to fully activate the H4R (Table 3). In agreement with our findings with Nα-methylhistamine, the methylated immepip analog methimepip shows a large selectivity for the hH3R (pKi = 9.0 ± 0.1) over the hH4R (pKi = 5.7 ± 0.1), as reported previously (Kitbunnadaj et al., 2005). Also various H3R antagonists bind to the hH4R (Oda et al., 2000; Liu et al., 2001a; Morse et al., 2001; Zhu et al., 2001). In our hands the isothiourea-based H3R antagonists clobenpropit and iodophenpropit both potently bind to the hH4R (Table 3). Yet, both ligands have very distinguishable intrinsic activities at the H4R. Clobenpropit, which acts as inverse agonist at the hH3R (Wieland et al., 2001) behaves as a potent partial agonist at the hH4R (Table 3; Fig. 3, A and B). In contrast, iodophenpropit, which also acts as an inverse agonist at the hH3R (Wieland et al., 2001), behaves as a neutral antagonist at the hH4R with a pKi value of 7.9 ± 0.1 (n = 6) (Table 3; Fig. 3, A and B). As expected, iodophenpropit competitively antagonized the action of histamine at the hH4R, yielding a linear Schild-plot and a pA2 value of 8.0 (Fig. 3, C and D), in accordance with its binding affinity. Besides clobenpropit, also the known H3R ligands proxyfan and the related iodoproxyfan show reasonable hH4R affinity with pKi values of 7.3 ± 0.1 and 7.9 ± 0.1, respectively. As observed for their action at the hH3R, both compounds act as partial agonist at the hH4R (Table 3).
Evaluation of the Potential Use of [125I]Iodophenpropit as H4R Radioligand. As reported in the previous section and in other studies, the hH4R can be labeled with either [3H]histamine or the H4R antagonist [3H]JNJ 7777120 (Oda et al., 2000; Liu et al., 2001a; Morse et al., 2001; Nguyen et al., 2001; Zhu et al., 2001; Thurmond et al., 2004). Previously, we described [125I]iodophenproprit as a suitable high affinity H3R radioligand (Jansen et al., 1992). Considering, the relatively high affinity of iodophenpropit at the hH4R and its high sensitivity, we investigated the potential of this radioligand to label the H4R. Due to the hH4R affinity of [125I]iodophenpropit, saturation binding experiments were not feasible. We therefore used homologous [125I]iodophenpropit displacement analysis to determine a Kd value of 34.4 ± 4.1 nM for [125I]iodophenpropit. The Bmax value obtained using [125I]iodophenpropit-binding displacement experiments (3.8 ± 0.4 pmol/mg protein) is approximately 2 times higher than those obtained with either [3H]histamine (1.8 ± 0.4 pmol/mg protein) and [3H]JNJ 7777120 (1.7 ± 0.4 pmol/mg protein). [125I]Iodophenpropit binding to membranes of SK-N-MC/hH4 cells was competitively displaced by a variety of H3/4R ligands (Fig. 4), despite a high level of nonspecific binding of approximately 60% as determined with 10 μM imetit. However, iodophenpropit and the related clobenpropit also displaced the nonspecific binding, resulting in a multiple site binding profile. The pKi values of compounds for the hH4R obtained using [125I]iodophenpropit displacement studies are consistent with their corresponding values obtained using displacement of either [3H]histamine or [3H]JNJ 7777120 binding to the hH4R; only the pKi value of thioperamide obtained using [125I]iodophenpropit displacement studies seems to deviate somewhat form the pKi values obtained using either [3H]histamine or [3H]JNJ 7777120 displacement studies (Fig. 4; Table 4).
4-Methylhistamine as a Selective H4R Agonist. After our initial observation of the relative high affinity of 4-methylhistamine for the hH4R, this histamine analog was evaluated in more detail. 4-Methylhistamine not only has high affinity for the hH4R(pKi = 7.3 ± 0.1; n = 3) but also exhibits considerable selectivity for the hH4R over the other three human histamine receptors (Fig. 5A). The human histamine H1, H2, and H3 receptors were tested for their interaction with 4-methylhistamine, using respectively 1 nM [3H]mepyramine (Kd = 1.6 nM), 0.5 nM [125I]iodoaminopotentidine (Kd = 0.5 nM) and 1 nM [3H]Nα-methylhistamine (Kd = 2.9 nM) binding to homogenates of transfected cells. As can be seen in Fig. 5A, 4-methylhistamine shows highest affinity for the hH4R and binds considerably less potently to the other histamine receptors, resulting in a >100-fold and >100,000-fold selectivity over the H3R and H2R, and H1R, respectively. 4-Methylhistamine not only binds to the hH4R but also has a high affinity, albeit reduced compared with the hH4R, for the mouse and rat H4R with Ki values of 73 and 55 nM, respectively (Fig. 5B). Moreover, 4-methylhistamine exerts full agonistic activity at the hH4R (Fig. 5C), resulting in a pEC50 value of 7.4 ± 0.1 (α = 1; n = 5). In contrast, 4-methylhistamine exhibits only moderate affinity for the hH2R (pKi = 5.1 ± 0.1; n = 3) and hH3R (pKi = 5.2 ± 0.1; n = 4), and partial agonistic hH3R activity (Fig. 5D). The hH4R agonistic effects of 4-methylhistamine can be antagonized by the selective H4R antagonist JNJ 7777120 (Fig. 5E). Schild-plot analysis of the JNJ 7777120 antagonism of the 4-methylhistamine-induced hH4R-mediated inhibition of forskolin-induced β-galactosidase activity yields a pA2 value of 7.8 (data not shown), which is in agreement with the hH4R affinity of JNJ 7777120 (Table 3) (Jablonowski et al., 2003; Thurmond et al., 2004). At the mouse and rat H4R, 4-methylhistamine also acts as a full H4 agonist, although with reduced pEC50 values of 5.8 ± 0.1 and 5.6 ± 0.1, respectively.
Previously, the H4R has been shown to be involved in the regulation of eosinophil and mast cell function (O'Reilly et al., 2002; Buckland et al., 2003; Hofstra et al., 2003; Takeshita et al., 2003; Ling et al., 2004; Thurmond et al., 2004). Indeed, 4-methylhistamine also acted as H4R agonist at human eosinophils and induced a rapid change in eosinophil cell shape, as measured by the gated autofluorescence forward scatter assay (Ling et al., 2004). The H4R agonist 4-methylhistamine acted as a full agonist and dose-dependently induced a change in forward scatter (Fig. 6A) with an EC50 value of 0.36 ± 0.09 μM, which was 3-fold less active compared with histamine. The effects of 4-methylhistamine on human eosinophils were not inhibited by H1R antagonist mepyramine or H2R antagonist ranitidine, but they could be antagonized by the H4R antagonist JNJ 7777120 (Fig. 6B). H3R antagonists were not included in this assay because Ling et al. (2004) have previously demonstrated that the H3R antagonist JNJ 637940 did not affect histamine-induced eosinophil shape change and that the H3R was not expressed in eosinophils. Finally, 4-methylhistamine was tested as an H4R agonist at mouse BMMCs as described previously (Hofstra et al., 2003). Like histamine, 4-methylhistamine dose-dependently induced migration of murine BMMCs with an EC50 value of 12 μM (Fig. 6C), again a somewhat lower potency compared with histamine (Hofstra et al., 2003). Furthermore, we found that the effect of 4-methylhistamine on murine BMMCs was completely inhibited by the selective H4R antagonist JNJ 7777120 in a dose-dependent manner (Fig. 6D).
Discussion
With the addition of the H4R to the histamine receptor family (Oda et al., 2000; Liu et al., 2001a; Morse et al., 2001; Nguyen et al., 2001; Zhu et al., 2001), this potential new drug target has created a lot of excitement in the field. The predominant expression of the histamine H4R on hematopoietic cells (Oda et al., 2000; Liu et al., 2001a; Morse et al., 2001; Zhu et al., 2001) and the H4R effects on, e.g., eosinophil and mast cell functions (Gantner et al., 2002; O'Reilly et al., 2002; Buckland et al., 2003; Hofstra et al., 2003; Takeshita et al., 2003; Ling et al., 2004; Thurmond et al., 2004) imply that this new histamine receptor subtype may play a role in various allergic and inflammatory conditions. So far, the search for selective H4R ligands has resulted in the discovery of potent neutral hH4R antagonists as JNJ 7777120 (Jablonowski et al., 2003; Thurmond et al., 2004) and VUF 6002 (Terzioglu et al., 2004), whereas potent and selective H4R agonists or inverse agonists have so far not been described. In search for new hH4R ligands, we therefore screened a library of known histaminergic ligands, using SK-N-MC cells stably expressing the hH4R. In this cell line the hH4R binds [3H]histamine and [3H]JNJ 7777120 with high affinity (Fig. 1, A and B) and functionally inhibits forskolin-induced CRE-mediated responses through pertussis toxin-sensitive Gi/o proteins (Fig. 1D). In these cells the hH4R also exhibits constitutive activity, which is blocked by pertussis toxin or the nonselective inverse agonist thioperamide (Fig. 1D).
Considering the H4R shares its highest sequence similarity with the H3R, it is not surprising that the H4R is targeted by various imidazole containing H3R ligands (Oda et al., 2000; Liu et al., 2001a; Morse et al., 2001; Zhu et al., 2001). The standard H3R inverse agonist thioperamide (Arrang et al., 1983) also acts as an inverse agonist at the hH4R. Moreover, in the present study we confirm that the presumed H3R agonists immepip, imetit, (R)-α-methylhistamine, and imbutamine also act as potent hH4R agonists. Furthermore, the H4R is activated by the H2R/H3R antagonist burimamide, the H3R antagonists clobenpropit, and the H3R agonist iodoproxyfan, indicating that for hH4R agonism considerable structural diversity (piperidine, isothiourea, thiourea, and ether) in the side chain of imidazole ring is allowed, including aromatic substitutions as indicated by the hH4R agonism displayed by clobenpropit. However, our detailed analysis of various H3R ligands indicates that hH4R efficacy can be modulated by differential hydrophobic substitution on the side chain. In the burimamide series we observed that differential substitution on the thiourea group gives rise to H4R (partial) agonists, a neutral antagonist (VUF 4614; pKi = 7.6) and a full inverse agonist (VUF 4742; pKi = 6.9). In addition, in the clobenpropit series, we observe that a slight change on the isothiourea substituent results in a modulation of H4R efficacy. The clobenpropit analog iodophenpropit (a phenylethyl substituent instead of a benzyl group) retains high H4R affinity (pKi = 7.9), but it has lost agonistic activity completely. In this study, we identified iodophenpropit as a high-affinity, neutral antagonist for the H4R. In view of the ∼15 nM affinity of iodophenpropit for the hH4R, we evaluated [125I]iodophenpropit as a potential new H4R radioligand. The hH4R can be labeled to the same extent with both the agonist [3H]histamine and the neutral antagonist [3H]JNJ 7777120. Surprisingly, the Bmax value determined with [125I]iodophenpropit was twice as much as that determined with either [3H]histamine or [3H]JNJ 7777120, suggesting that the radioligands might bind to different hH4R subpopulations, similarly to recent findings on the binding of two H1R radioligands to the H1R (Booth et al., 2002). Yet, the potential existence of different H4R subpopulations needs further investigation. The binding of the three radioligands to membranes of SK-N-MC/hH4 cells was displaced by a variety of H3/4R ligands, and the pKi values obtained from these displacement studies show a high correlation. Despite being shown as a potential hH4R radioligand, [125I]iodophenpropit has to be used with caution, as in our hands a high level of nonspecific binding limits its use.
From our screening of many H1R ligands, only the tricyclic clozapine shows reasonable H4R affinity, as reported previously (Oda et al., 2000; Liu et al., 2001a; Zhu et al., 2001). Despite their structural similarity to clozapine, other tested H1R antagonists do not show any appreciable affinity for the hH4R. We can therefore not confirm that mepyramine binds to the hH4R (Nguyen et al., 2001), either studied by displacement of [3H]histamine binding to the hH4R, by saturation [3H]mepyramine binding assays (data not shown), or by functional H4R assays. Clinically used H1R antagonists, such as cetirizine, ebastine, fexofenadine, and loratidine, demonstrate significant in vitro anti-inflammatory activity, which are not related to their H1R activity (Gelfand et al., 2004). The data from our study do not support the involvement of the hH4R in the anti-inflammatory effects of these H1R antagonists.
An important finding of this study is the discovery of 4-methylhistamine as a potent and selective hH4R agonist in both recombinant and endogenously expressing H4R systems. Whereas this compound is originally described as a relatively selective H2R agonist (Durant et al., 1975), our present data show that this histamine analog exhibits more than 100-fold selectivity over the recombinant H1R, H2R, and H3Rs. 4-Methylhistamine not only acts as a full agonist at the recombinant hH4R but also induces migration of mouse bone marrow derived mast cells and a shape change of human eosinophils. Both processes have recently been shown to be induced by histamine via interaction of the H4R (Hofstra et al., 2003; Ling et al., 2004). The relative potencies of histamine and 4-methylhistamine on human eosinophils are similar to those observed in recombinant systems. Similar to the observations with histamine (Hofstra et al., 2003; Ling et al., 2004), the potency of 4-methylhistamine on mouse mast cells is somewhat lower than in recombinant systems. Although the mouse H4R shows a lower affinity for both histamine and 4-methylhistamine compared with the hH4R, the low potency at BMMCs seems not merely an issue of species difference, but also it might be related to a low H4R expression level. In fact, the H4R in BMMCs is present at a very low density because it cannot be detected by radioligand binding studies (R. L. Thurmond, unpublished observations). Moreover, the cellular environment might dictate the potency of an agonist, such as composition of G proteins and accessories proteins in the cells (Kenakin, 2004).
In conclusion, from a large screening of many known histamine receptor ligands we have identified a variety of compounds with interesting H4R activities. The major significance of these findings is the reevaluation of numerous histaminergic ligands at the new histamine receptor subtypes. Based upon our data, many imidazol-containing H3R ligands, including various H3R reference compounds, show potent H4R activities and should be treated with caution. More recently developed H3R agonists, such as immethridine (Kitbunnadaj et al., 2004) or methimmepip (Kitbunnadaj et al., 2005), or nonimidazole H3R antagonists, such as JNJ 6379490 (Ling et al., 2004) or A-349821 (Esbenshade et al., 2004), hardly act at the H4R and will therefore provide good tools to selectively target the H3R. In the series of tested H3R ligands, we have identified iodophenpropit as potent neutral H4R antagonist and the burimamide analog VUF 4742 as the second identified H4R inverse agonist. From the screening of H2R ligands, we have identified 4-methylhistamine as the first high-affinity H4R agonist (Ki = 50 nM) that has a >100-fold selectivity for the hH4R over the other histamine receptor subtypes. The identification of 4-methylhistamine as a potent H4R agonist will be of major importance for future studies to unravel the physiological roles of the H4R.
Acknowledgments
We acknowledge the technical assistance on Wen Jiang, Steven Nguyen, and Pragnya Desai.
Footnotes
-
R.L. is a recipient of a PIONIER award of the Technology Foundation (Stichting voor de Technische Wetenschapppen) of the Netherlands Foundation of Scientific Research (Nederlandse Organisatie voor Wetenschappelijk Onderzoek).
-
Article, publication date, and citation information can be found at http://jpet.aspetjournals.org.
-
doi:10.1124/jpet.105.087965.
-
ABBREVIATIONS: HxR, histamine Hx receptor; hH4R, human histamine H4 receptor; IL, interleukin; JNJ 7777120, 1-[(5-chloro-1H-indol-2-yl)carbonyl]-4-methylpiperazine; VUF 6002, 1-[(5-chloro-1H-benzimidazol-2-yl)carbonyl]-4-methylpiperazine; VUF 4742, N-(4-chlorobenzyl)-N′-[5-(4(5)-imidazolyl)pentyl]thiourea; OUP-16, 2-cyano-1-methyl-1–3-{(2R,5R)-5-[1H-imidazol-4(5)-yl]tetrahydrofuran-2-yl}methylguanidine; PEA, 2-pyridylethylamine; TEA, 2-(2-thiazolyl)ethylamine; VUF 8328, S-[3-(4(5)-imidazolyl)propyl]isothiourea; ORG3770, 1,2,3,4,10,14b-hexahydro-2-methylpyrazino[2,1-a]pyrido[2,3-c][2]benzazepine; CRE, cAMP response element; BMMC, bone marrow mast cell; JNJ 637940, 7-methyl-2-[4-(3-piperidin-1-yl-propoxy)-phenyl]-imidazo[1,2-a]pyridine; A-349821, 4′-{3-[(R,R)2,5-dimethyl-pyrrolidin-1-yl]-propoxy)-biphenyl-4-yl}-morpholin-4-yl-methanone; VUF 4683, 1-[4-(imidazol-4-yl)-butyl)-3-isopropyl-thiourea; VUF 4616, 1-[5-(imidazol-4-yl)-pentyl]-3-isopropyl-thiourea; VUF 4617, 1-cyclohexyl-3-[5-(imidazol-4-yl)-pentyl]-thiourea.
-
↵1 These authors contributed equally.
- Received April 13, 2005.
- Accepted June 1, 2005.
- The American Society for Pharmacology and Experimental Therapeutics