Conserved features of the sequences of dopamine receptors and of homologous G-protein-coupled receptors point to regions, and amino acid residues within these regions, that contribute to their ligand binding sites. Differences in binding specificities among the catecholamine receptors, however, must stem from their nonconserved residues. Using the substituted-cysteine accessibility method, we have identified the residues that form the surface of the water-accessible binding-site crevice in the dopamine D2 receptor. Of approximately 80 membrane-spanning residues that differ between the D2 and D4 receptors, only 20 were found to be accessible, and 6 of these 20 are conservative aliphatic substitutions. In a D2 receptor background, we mutated the 14 accessible, nonconserved residues, individually or in combinations, to the aligned residues in the D4 receptor. We also made the reciprocal mutations in a D4 receptor background. The combined substitution of four to six of these residues was sufficient to switch the affinity of the receptors for several chemically distinct D4-selective antagonists by three orders of magnitude in both directions (D2- to D4-like and D4- to D2-like). The mutated residues are in the second, third, and seventh membrane-spanning segments (M2, M3, M7) and form a cluster in the binding-site crevice. Mutation of a single residue in this cluster in M2 was sufficient to increase the affinity for clozapine to D4-like levels. We can rationalize the data in terms of a set of chemical moieties in the ligands interacting with a divergent aromatic microdomain in M2-M3-M7 of the D2 and D4 receptors.

The binding site of the dopamine D2 receptor, like that of homologous G-protein-coupled receptors (GPCRs), is contained within a water-accessible crevice formed among its seven transmembrane segments (TMSs). Using the substituted-cysteine-accessibility method (SCAM), we are mapping the residues that contribute to the surface of this binding-site crevice. We have mutated to cysteine, one at a time, 21 consecutive residues in the fourth TMS (TM4). Eleven of these mutants reacted with charged sulfhydryl-specific reagents, and bound antagonist protected nine of these from reaction. For the mutants in which cysteine was substituted for residues in the cytoplasmic half of TM4, treatment with the reagents had no effect on binding, consistent with these residues being inaccessible and with the low-resolution structure of the homologous rhodopsin, in which TM3 and TM5 occlude the cytoplasmic half of TM4. Although hydrophobicity analysis positions the C-terminus of TM4 at 4.64, Pro-Pro and Pro-X-Pro motifs, which are known to disrupt alpha-helices, occur at position 4.59 in a number of homologous GPCRs. The SCAM data were consistent with a C-terminus at 4.58, but it is also possible that the alpha-helix extends one additional turn to 4.62 in the D2 receptor, which has a single Pro at 4.59. In homologous GPCRs, the high degree of sequence variation between 4.59 and 4.68 is more characteristic of a loop domain than a helical segment. This region is shown here to be very conserved within functionally related receptors, suggesting an important functional role for this putative nonhelical domain. This inference is supported by observed ligand-specific effects of mutations in this region and by the predicted spatial proximity of this segment to known ligand binding sites in other TMs.

The binding-site of the dopamine D2 receptor, like that of other homologous G protein-coupled receptors, is contained within a water-accessible crevice formed among its seven membrane-spanning segments. Using the substituted cysteine accessibility method (SCAM), we previously mapped the residues in the third, fifth, sixth, and seventh membrane-spanning segments that contribute to the surface of this binding-site crevice. We have now mutated to cysteine, one at a time, 22 consecutive residues in the second membrane-spanning segment (M2) and expressed the mutant receptors in HEK 293 cells. Eleven of these mutants reacted with charged, hydrophilic, lipophobic, sulfhydryl-specific reagents, added extracellularly, and 9 of these 11 were protected from reaction by a reversible dopamine antagonist, sulpiride. We infer that the side chains of the residues at the 11 reactive loci (D80, L81, V83, V87, P89, W90, V91, V92, L94, E95, V96) are on the water-accessible surface of the binding-site crevice and that 9 of these are occluded by bound antagonist. The pattern of accessibility suggests an alpha-helical conformation. A broadening of the angle of accessibility near the binding site is consistent with the presence of a kink at Pro89. On the basis of the enhanced rates of reaction of positively charged sulfhydryl reagents, we infer the presence of an electrostatic microdomain composed of three acidic residues in M2 and the adjacent M3 that could attract and orient cationic ligands. Furthermore, based on the enhanced reactivity of the hydrophobic cation-containing reagent, we infer the presence of an aromatic microdomain formed between M2, M3, and M7.