Ligand binding in the GPCR tertiary structure

The high-resolution structure of rhodopsin unraveled at atomic resolution the seven helical-bundle characterizing GPCRs (Palczewski et al. 2000; Teller et al. 2001) (Fig. 3.1). The seven helices within the bundle are organized in a distinctive counterclockwise fashion (as seen from the extracellular side) and contain a large numbers of kinks, bends, and twists (Palczewski et al. 2000; Teller et al. 2001). Helices 1, 4, 6, and 7 are characterized by the presence of kinks at conserved proline residues while helix 3 is the longest and mostly tilted as compared to the plane of the membrane (Palczewski et al. 2000; Teller et al. 2001). In full agreement with the predictions from earlier mutagenesis studies, the chromophore 11-cis-retinal is contained within a deep binding pocket formed mainly by helices 3, 4, 5, 6, and 7 (Palczewski et al. 2000; Teller et al. 2001) (Fig. 3.1). Through formation of a Schiffbase, 11-cis-retinal is covalently coupled to a lysine in transmembrane segment (TM) 7 (Lys2967.43)1 and the protonated Schiffbase is paired with a glutamate (Glu1133.28) in the outer part of TM3 (Palczewski et al. 2000; Teller et al. 2001). An interesting feature revealed by the crystal structure is that the second extracellular loop 2 (ECL2) connecting TM4 and 5 forms a lid in the binding pocket by diving down into the transmembrane region (Fig. 3.1) (Palczewski et al. 2000; Teller et al. 2001). Structurally, the loop contains two stretches of ^-strands of which one is directly above retinal and contains residues forming contacts with retinal (Palczewski et al. 2000; Teller et al. 2001). Thus, retinal is completely surrounded by protein within its binding pocket (Palczewski et al. 2000; Teller et al. 2001). A key question is to what extent other GPCRs structurally resemble rhodopsin. Recently, a thorough comparison of the rhodopsin structure with data obtained from systematic application of the substituted cysteine accessibility method (SCAM) to the dopamine D2 receptor was carried out (Ballesteros et al. 2001 b). In the SCAM studies of the D2 receptor, performed by Javitch et al. (Javitch et al. 2002) each residue in all seven TMs was one by one substituted with cysteine. The comparison indicated a remarkable structural similarity between the D2 receptor and rhodopsin in the transmembrane domains (Ballesteros et al. 2001b). The amino acids residues, inferred based on the SCAM analysis to form the water accessible binding crevice in between the transmembrane helices, were almost completely consistent with the predictions from the rhodopsin structure (Ballesteros etal. 2001b).

At present it is more difficult to answer whether ECL2 is forming a plug in the binding crevice in other GPCRs as it does in rhodopsin (Fig. 3.1). If it does, it will have a major implication for our understanding of the receptor activation mechanism, hence, a key event in receptor activation would necessarily involve significant conformational changes ofECL2 allowing for rapid access of the ligand to the binding crevice. The ECL2 sequence is rather poorly conserved among the different receptors; however it does contain one of the two highly conserved cysteines known to form a disulfide bond between the top of TM3 and ECL2 (Gether 2000). It is also remarkable that several studies have suggested that the loop region maybe functionally important. In the a1B adrenergic receptor, for example, it was found that residues in the loop are critical for the pharmacological specificity of some adrenergic ligands consistent with the notion that other small molecule ligands than 11-cis-retinal might form contacts with residues in this loop (Zhao et al. 1996). Furthermore, previous studies of the p2 adrenergic receptor using a fluorescent ligand have demonstrated that the ligand binding

1 The positions of residues are throughout the chapter indicated by their generic number followed by their number according to the Ballesteros-Weinstein nomenclature in superscript (Ballesteros and Weinstein 1995)

(see legend to Fig. 3.2 for details).

site is completely inaccessible to aqueous quenchers (Tota and Strader 1990); an observation consistent with the presence of a 'plug' in the binding crevice.

Although GPCR structure might be evolutionarily highly conserved it must be emphasized that an extensive amount of evidence suggests that there is no common 'lock' for all agonist 'keys' (Schwartz and Rosenkilde 1996; Gether 2000). Accordingly, it is not required for receptor activation that all agonist molecules form a specific set of contact points in the receptor structure to promote receptor activation. Even agonists for the same receptor may not necessarily have to share the same binding site and it does not seem to be a specific requirement for receptor activation that the agonist ligand is docked in the transmembrane binding crevice (for reviews see Ji et al. 1998; Gether 2000). In other words, the mode of interaction of a given agonist depends on its individual chemical structure and whether it is an agonist or antagonist is determined by the mode of interaction and whether this mode preferentially stabilizes an inactive or an active receptor state (Ji etal. 1998; Gether 2000). In case of, for example, peptide hormones and neuropeptides, the most critical points of interactions are, in contrast to the small molecule ligands (Family A) found in the extracellular loops and in the amino terminus (Ji et al. 1998; Gether 2000). A special case is the Family 3 receptors, including among others the metabotropic glutamate and GABA receptors, which are characterized by containing their 'small molecule' ligand binding site in their large amino terminus (Ji et al. 1998; Gether 2000). Nonetheless, despite the almost extreme diversity in how structurally distinct ligands interact with their receptor, it is still highly likely that an underlying fundamental activation mechanism involving the transmembrane regions has been conserved during evolution given the common ability of the receptors to activate the same intracellular signalling pathways though the same classes of G proteins.

0 0

Post a comment