Mutational Analysis of the Antagonist-binding Site of the Histamine H1 Receptor

Abstract

Since the initial discovery of the role of histamine in allergic conditions (1) serious efforts have been made to develop drugs that inhibit the actions of histamine. Already in 1933, Fourneau and Bovet (2) reported the first "antihistamine" piperoxan. Following this finding many potent H 1 antagonists that can be considered as variations of diaryl-substituted ethylamines (e.g. diphenhydramine and mepyramine) have been developed (for review see Ref.3). These "first generation" H 1 antagonists are quite effective in humans in allergic rhinitis and urticaria, but because of central nervous system penetration and central H 1 receptor blockade their clinical use is hampered by sedative side effects (3-5). A "second generation" of nonsedative H 1 antagonists (e.g. astemizole, acrivastine, cetirizine, loratidine, and terfenadine) has recently been developed (for review see Ref.3). Their altered pharmacokinetics result in good clinical effectiveness combined with a strongly reduced sedative potential (3-5).The development of H 1 antagonists has so far been directed by traditional medicinal chemistry (3). With the availability of the genetic information of the histamine H 1 receptor (6), the rationalization of drug-protein interaction has become a major challenge for this therapeutically important class of drugs. Like all aminergic G-protein coupled receptors (GPCR), 1 the H 1 receptor contains an aspartate residue (Asp 116 ) in transmembrane domain (TM) III (6), that is involved in the binding of the protonated amine function found in both agonists and antagonists structures (7,8). Mutagenesis studies have furthermore shown that the imidazole ring of histamine is accommodated by Lys 200 and Asn 207 in TM V (9, 10). In view of the low sequence similarity between GPCRs and bacteriorhodopsin (BR) much controversy exists on the validity of models derived for GPCRs based on the homology with BR (11-13). Nevertheless, despite the speculative nature of BRderived GPCR models they have been quite helpful in understanding and predicting drug-receptor interactions for a variety of receptors (see e.g. Refs. 14 -16). Previously, we also developed a three-dimensional computer model of the histamine H 1 receptor based on the homology with BR, incorporating the results obtained from mutagenesis studies on the agonist binding site (17). In the present study this computer model of the H 1 receptor was combined with a pharmacophoric model of the H 1 antagonistic binding site (18). This ligand-based model for the H 1 antagonistic binding site is based upon an interaction of the protonated amine function of various first generation, semi-rigid H 1 antagonists with an aspartate residue (Asp 116 in the guinea pig H 1 receptor) (18) and precisely positions the cis-and trans-aromatic rings of the H 1 antagonists relative to the C ␣ and C ␤ carbon atoms of this aspartate residue. Combining the three-dimensional receptor model and the ligand-based pharmacophoric model of the H 1 antagonist binding site resulted in the prediction of interactions...

Structure-based discovery and binding site analysis of histamine receptor ligands

“…The other histamine binding residue Lys191(5.39) was also found far from either Asp107(3.32) or Asn198 (5.46) suggesting that the antagonist conformation shows significant differences from an agonist binding conformation. As proposed in previous modeling studies [8,10] Trp152(4.50) is facing the membrane and does not directly contribute to ligand binding. The 'ionic lock' feature that is found in the rhodopsin but missing from other GPCR structures is also missing from H1R: no interaction between Arg125(3.50) and either Glu410(6.30) or with Asp124(3.49) could be observed.…”

“…Timmermann and Leurs published the first atomistic models of the histamine H1 receptor (H1R) in 1995 [7] and 1999 [8]. In these first modeling attempts, the authors proposed binding modes for agonists [7] and antagonists [8] using a homology model of the guinea pig H1R based on the low-resolution bacteriorhodopsin structure.…”

“…In these first modeling attempts, the authors proposed binding modes for agonists [7] and antagonists [8] using a homology model of the guinea pig H1R based on the low-resolution bacteriorhodopsin structure. Despite the fact that bacteriorhodopsin is not a GPCR and it lacks some of the conserved GPCR sequence motifs the models provided useful insights.…”

“…Timmermann and Leurs published the first atomistic model of the histamine H1 receptor in 1995 [7] investigating the binding modes of histamine, 2-methylhistamine and 2-phenyl-histamine as agonists and cyproheptadine as antagonist on bacteriorhodopsin-based homology models. This model has been further refined in a subsequent study analyzing the relationship of the H1 antagonist pharmacophore model and the binding mode of H1 antagonists in a guinea pig H1R homology model [8].…”

Pharmacology of Antihistamines

“…In the case of the histamine H 1 -receptor, histamine cross-links sites on transmembrane domains III and V to stabilize the receptor in its active conformation, thus causing the equilibrium to swing to the on position[12] (Figure 1B). H 1 -antihistamines, which are not structurally related to histamine, do not antagonize the binding of histamine but bind to different sites on the receptor to produce the opposite effect.…”