The Journal of Physical Chemistry B Article 24A25A 35A A Rt1 Figure 7. Proposed activation process of the dir by an 1-SPD like agonist. Rtl: electrostatically induced binding of an agonist. Rt2 agonist bi induced structural changes(cyan, agonist-bound state; orange, unliganded state; key residues are represented as spheres; red arrows indicate movement of TMs). R*: the fully activated DiR in complex with Ga(this model was constructed according to the complex of B2AR with G Protein).3.54 hydrogen bonding with Y7.35 and carbonyl group in backbone an ionic lock between R3. 50 and E6. 30; this shift in turn opens of 1184 and I183, respectively, through its two imidazole the docking pocket for g protein and leads to the activation of nitrogen atoms(Figure Sb, c). As the amide side chain of N6. 55 DiR(Figure 6b, c). Without binding of I-SPD, R3.50 and E6.30 is relatively small, the binding pocket 1 of DiR is sterically approach gradually to mimic an inactive-state Dir( Figure 6c) unhindered, making this binding pocket optimal for the binding To investigate the structural determinant of antagonist effects of d ring of I-SPD. On the contrary, the relative narrowness of of I-SPD on D2R, an unliganded D2R structure was binding pocket 1 of D2R and D3R caused by the bulkiness of superimposed to a typical 1-SPD stabilized structure. In contrast H6.55 made it only suitable for the binding of a ring of 1-SPD. to the relative transmembrane movements in 1-SPD-bound Furthermore, I-SPD formed polar contact with K2.61 by DIR, no significant rotation is found in transmembrane helices methoxyl group in A ring of DiR, whereas in D2R and D3R, in I-SPD-bound D2R( Figure 6d,e), except that the extracellular V2.61 exempted the formation of such polar contact. segment of TM3 moves away from the binding pocket as Mechanism for the Triple Action of /SPD on DiR, compared to unliganded D2R. In I-SPD-bound DIR, N6.55 D2R/D3R. It is now interesting to speculate how 1-SPD exerts moves toward the ligand binding site to form two hydrogen ts triple action on DIR, D2R, and D3R and how the bonds with 1-SPD Though in 1-SPD-D2R aromatic stacking was differences in binding mode are coupled with structural found between 1-SPD and F6.51, F6.52 and H6. 55, examination behavior in the cytoplasmic regions of TM3 and TM6. As of the downstream hydrophobic residues 13. 40, W6. 48, and an be seen from the binding pocket lining shown in Figure 5, F6.44 indicated that those residues were not induced to fit for when bound to DIR, I-SPD formed hydrogen bond with hydrophobic stacking(Figure 6d). The absence of such a agonist-sensitive residue in TM6(N6. 55) to mediate its agonist critical hydrophobic stacking exempted an outward movement activity, while bound to D2R and D3R, energetically favorable of the cytoplasmic segment of TM6( Figure 6e). In turn, as can edge-to-face aromatic stacking was found between I-SPD and be seen in Figure 6f, the ionic lock between R3. 50 and E6. 30 TM6, and instead of hydrogen bonding to I-SPD, H6.55 was maintained when D2R was bound with I-SPD formed two hydrogen bonds with Y7.35 and 1184 and 1183 of In the I-SPD-bound D3R model, prominent helical move ECL2. A plausible conformational link is shown in Figure 6. ment as compared to unliganded D3R is absent except that a Inactive and active DIR structures were superimposed to bulge located at $5.46 is tilted slightly outward from the monitor structural changes in ligand binding site. As shown in binding pocket to accommodate I-SPD. Though similar structures in the binding site are a 3.5 A movement of the F6. 52, and H6.55 in the 1-SPD-D3R complex, the absencegr aromatic stacking was found between I-SPD and F6.51 extracellular part of TM3 toward the core of the receptor, a hydrophobic stacking immediately downstream precluded an clockwise rotation of the extracellular segment of TM6 and a inward movement of the extracellular segment of TM6(Figure slight bulge of TMS at S5.46 toward TM3. Apparently, the 6h). As a result, a significant outward movement of the clockwise rotation of the extracellular segment of TM6 is the cytoplasmic segment of TM6 is prevented(Figure 6i,j), and the result of the hydrogen bonding between the D-OH of 1-SPD ionic lock between R3.50 and E6.30 is maintained( Figure 6) and N6.55(Figure 5a). Such clockwise rotation in turn Proposed Activation Process of D1R. Based on our positions F6.44 and w6.48 of TM6 to form a hydrophobic structural modeling, MD simulations, and the previously stacking with 13. 40 of TM3 reported signal transduction model of dir, a stepwise extracellular part of TM3, similar to the hydrophobic stacking activation process of DiR is proposed. Typically, the DiR bserved in the activation of B2Ar. Accompanying these adopts the inactive state through its intramolecular interactio tructural chat he cytoplasmic segment of TM6 shifts and transits between several energy minima conformations by ignificantly(7. 1 A)toward lipid, exempting the formation of molecular thermodynamic motion. As the protonated I-SPD 8128 dx. dolora/o.021/p30492351 Phys. Chem. B2012116,8121-813hydrogen bonding with Y7.35 and carbonyl group in backbone of I184 and I183, respectively, through its two imidazole nitrogen atoms (Figure 5b,c). As the amide side chain of N6.55 is relatively small, the binding pocket 1 of D1R is sterically unhindered, making this binding pocket optimal for the binding of D ring of l-SPD. On the contrary, the relative narrowness of binding pocket 1 of D2R and D3R caused by the bulkiness of H6.55 made it only suitable for the binding of A ring of l-SPD. Furthermore, l-SPD formed polar contact with K2.61 by methoxyl group in A ring of D1R, whereas in D2R and D3R, V2.61 exempted the formation of such polar contact. Mechanism for the Triple Action of l-SPD on D1R, D2R/D3R. It is now interesting to speculate how l-SPD exerts its triple action on D1R, D2R, and D3R and how the differences in binding mode are coupled with structural behavior in the cytoplasmic regions of TM3 and TM6. As can be seen from the binding pocket lining shown in Figure 5, when bound to D1R, l-SPD formed hydrogen bond with agonist-sensitive residue in TM6 (N6.55) to mediate its agonist activity, while bound to D2R and D3R, energetically favorable edge-to-face aromatic stacking was found between l-SPD and TM6, and instead of hydrogen bonding to l-SPD, H6.55 formed two hydrogen bonds with Y7.35 and I184 and I183 of ECL2. A plausible conformational link is shown in Figure 6. Inactive and active D1R structures were superimposed to monitor structural changes in ligand binding site. As shown in Figure 6a, the greatest differences between inactive and active structures in the binding site are a 3.5 Å movement of the extracellular part of TM3 toward the core of the receptor, a clockwise rotation of the extracellular segment of TM6 and a slight bulge of TM5 at S5.46 toward TM3. Apparently, the clockwise rotation of the extracellular segment of TM6 is the result of the hydrogen bonding between the D-OH of l-SPD and N6.55 (Figure 5a). Such clockwise rotation in turn repositions F6.44 and W6.48 of TM6 to form a hydrophobic stacking with I3.40 of TM3, causing inward movement of the extracellular part of TM3, similar to the hydrophobic stacking observed in the activation of β2AR.4 Accompanying these structural changes, the cytoplasmic segment of TM6 shifts significantly (7.1 Å) toward lipid, exempting the formation of an ionic lock between R3.50 and E6.30; this shift in turn opens the docking pocket for G protein and leads to the activation of D1R (Figure 6b,c).Without binding of l-SPD, R3.50 and E6.30 approach gradually to mimic an inactive-state D1R (Figure 6c). To investigate the structural determinant of antagonist effects of l-SPD on D2R, an unliganded D2R structure was superimposed to a typical l-SPD stabilized structure. In contrast to the relative transmembrane movements in l-SPD-bound D1R, no significant rotation is found in transmembrane helices in l-SPD-bound D2R (Figure 6d,e), except that the extracellular segment of TM3 moves away from the binding pocket as compared to unliganded D2R. In l-SPD-bound D1R, N6.55 moves toward the ligand binding site to form two hydrogen bonds with l-SPD. Though in l-SPD-D2R aromatic stacking was found between l-SPD and F6.51, F6.52 and H6.55, examination of the downstream hydrophobic residues I3.40, W6.48, and F6.44 indicated that those residues were not induced to fit for hydrophobic stacking (Figure 6d). The absence of such a critical hydrophobic stacking exempted an outward movement of the cytoplasmic segment of TM6 (Figure 6e). In turn, as can be seen in Figure 6f, the ionic lock between R3.50 and E6.30 was maintained when D2R was bound with l-SPD. In the l-SPD-bound D3R model, prominent helical move￾ment as compared to unliganded D3R is absent except that a bulge located at S5.46 is tilted slightly outward from the binding pocket to accommodate l-SPD. Though similar aromatic stacking was found between l-SPD and F6.51, F6.52, and H6.55 in the l-SPD-D3R complex, the absence of hydrophobic stacking immediately downstream precluded an inward movement of the extracellular segment of TM6 (Figure 6h). As a result, a significant outward movement of the cytoplasmic segment of TM6 is prevented (Figure 6i,j), and the ionic lock between R3.50 and E6.30 is maintained (Figure 6j). Proposed Activation Process of D1R. Based on our structural modeling, MD simulations, and the previously reported signal transduction model of D1R,20 a stepwise activation process of D1R is proposed. Typically, the D1R adopts the inactive state through its intramolecular interaction and transits between several energy minima conformations by molecular thermodynamic motion. As the protonated l-SPD Figure 7. Proposed activation process of the D1R by an l-SPD like agonist. Rt1: electrostatically induced binding of an agonist. Rt2: agonist binding￾induced structural changes (cyan, agonist-bound state; orange, unliganded state; key residues are represented as spheres; red arrows indicate relative movement of TMs). R*: the fully activated D1R in complex with Gα (this model was constructed according to the complex of β2AR with G￾protein).53,54 The Journal of Physical Chemistry B Article 8128 dx.doi.org/10.1021/jp3049235 | J. Phys. Chem. B 2012, 116, 8121−8130