The ␤ 2 -adrenergic receptor (␤ 2 AR) 1 has been extensively studied and, along with rhodopsin, has served as a prototype for our understanding of the structure and function of related G-protein-coupled receptors (1, 2). In these receptors, the binding site is formed among the seven-transmembrane segments (TMs) (2) in a water-accessible binding-site crevice. A number of residues inferred to contact bound agonist in the ␤ 2 AR have been identified. Asp113 3.32 (see "Experimental Procedures" for a description of the indexing of residues) in the third TM (TM3) is thought to interact with the protonated amine of catecholamines (3) and is completely conserved in all amine receptors. A cluster of aromatic residues in TM6 that is thought to interact with the aromatic ring of catecholamines is highly conserved as well (2). A classical study by Strader et al. (4) in the ␤ 2 AR demonstrated that two serines in TM5 directly interact with the catechol hydroxyls (OHs) of agonists. Specifically, 5.43 interacts with the meta-hydroxyl (mOH) and Ser-207 5.46 with the para-hydroxyl (pOH) (4). These interactions were shown to contribute to the affinity, potency, and efficacy of various catecholamine agonists. A potential role of 5.42 in binding and activation, however, was obscured by the lack of expression of the mutant ␤ 2 AR in which 5.42 was mutated to Ala (4). Nonetheless, because the catechol ring contains two hydroxyls and because two serines were inferred to hydrogen bond to these two hydroxyls, Ser-203 5.42 has been generally assumed not to play a significant role in agonist binding to and activation of the wild-type ␤ 2 AR.Curiously, ␤ 2 AR Ser-203 5.42 is completely conserved in all catecholamine receptors (Fig. 1), and this residue has been shown to play a role in ligand binding and receptor activation in the ␣ 1A , ␣ 2A , and ␣ 1B adrenergic (5-8) ) interacts with the mOH of catecholamines, and it is this H bond that is critical for ligand binding and receptor activation (5).Using the substituted cysteine accessibility method (14, 15), we found that in the dopamine D 2 receptor each of the three aligned serines ) are accessible in the binding-site crevice (16). This is consistent with experimental results in the D 2 receptor implicating each of the three serines in the binding of various ligands, although different serines were found to be more or less critical with different ligands (11,12). More recently we have applied the substituted cysteine accessibility method to the aligned portion of TM5 in the human ␤ 2 AR, 2 and we were surprised to observe that substitution of by Cys, unlike the published mutation of this Ser to Ala in the hamster ␤ 2 AR (4), did not impair expression of the receptor but instead lowered its affinity for