The Journal of Physical Chemistry B Article diffuses to the extracellular mouth of the binding cavity of DiR, changes, the cytoplasmic segment of TM6 shifts significantly the electrostatic attraction between the negatively charged out of the receptor core, and this shift in turn opens the region of the DiR and positively charged I-SPD assists in their docking pocket for g protein and leads to the activation of initial association( Figure 7, Rt1). The approach of extracellular DIR. However, in the antagonistic mechanism of I-SPD ends of TM3 and TM6 to each other is then bridged by D2Rand D3R, no hydrophobic stacking is formed betweer hydrogen bonding of 1-SPD with N6.55 and electrostatic 13.40, F6.44, and w6.48 in D2R and d3R to lead to an outward attraction with D3. 32. As a result, 13. 40, F6.44, and w6.48 are movement of the cytoplasmic segment of TM6 repositioned to form hydrophobic stacking, leading to a 7. 1 A Our finding of key residues of DiR, D2R, and D3R involved outward movement of the cytoplasmic segment of TM6 in 1-Spd binding could guide future experimental work on ( Figure 7, Rt2). At this stage, the receptor reaches its Rt2 state these receptors and has significance in the mechanism-based eady for G-protein binding, while not fully activated as design of functional drugs targeting dopamine receptors. protein binding are required to fully activate GPCRs. Finally, ASSOCIATED CONTENT TMge separation between the cytoplasmic ends of TM3 and S Supporting Information eadily binding of G-1 Sequence alignments of Dir with B2AR and of D2R with D3R, fully activated R* state. This stepwise activation process of DiR Ramachandran plots of all DR models, a brief summary of all is supported by the currently accepted postulate about the simulation systems, TM3-TM6 distances and ionic lock dynamics of GPCRs; that is, GPCRs are more like molecular formation and breakage in representative active and inactive consistent with the activation mechanism of P, AR. 24.25 are rheostats rather than simple two-state switches and are GPCR crystal structures, binding affinity data for 7-SPD analogues, and the definition of n-o distance as an indication of ionic lock formation and breakage. This material is available CONCLUSION g We report herein a more comprehensive molecula and molecular dynamics simulation study of the AUTHOR INFORMATION DIR, D2R, and D3R and 1-SPD-DIR, I-SPD-D2R Corresponding author D3R complexes, all simulated in a lipid bilayer mem Fax+8621-50807188Ph+86-21-50806600, e-mail kayu homology models of both DiR and D2R were mail. shcncaccn(KQ.Y Fax +86-21-51980010, Ph+86-21 usingthemorereliableactivestateB2ARandinactivestate51980010,e-mailwfu@fudan.edu.cn(WF. D3R as templates, respectively. Structural analyses of molecular Author contributions dynamics simulations were carried out followed by calculation of binding energy. SThese authors contributed equally We found that the separation in TM3 -TM6 for unliganded Notes DIR shortened gradually from ca 20 to ca. 12 A, whereas the The authors declare no competing financial interest. distance in TM3-TM6 was kept at ca. 1s A during the entire mulation of I-SPD-DIR complex, and this kind of separation ACKNOWLEDGMENTS in TM3 and TM6 is a typical feature of active state GPCrs This work was supported by the National Natural Science while the distance in TM3-TM6 in all D2R and D3R Foundation of China(No. 81172919)and grants simulations remained relatively stable below 14 A, which is National High Technology Research and deve typical feature of inactive state GPCRs. Although there was no Program of China(863 Program)(No. 2012AA02110 ionic lock formed when the unliganded DiR relaxed to its the State Key Program of Basic Research of Chir inactive state, the ionic lock R3.60 and E6.30 approaches (2009CB918502) SPD bound with DIR, D2R, and D3R correlated perfectly with REFERENCES ying that our in silico (1)Overington, ]. P Al-Lazikani, B. Hopkins, A L. How many drug simulations probably resembled the binding poses of 1-SPD targets are there? Nat. Rev. Drug Discovery 2006, 5, 993-6 to DIR, D2R, and D3R. A more in-depth analysis of the 2)Cherezov, V i et al. High-resolution crystal structure of ar interaction maps of 1-SPD with DIR, D2R, and D3R revealed ngineered human beta2-adrenergic G protein-coupled receptor. that hydrogen bonding with N6. 55 coupled with hydroph Science2007,3l8,1258-65 (3)Rosenbaum, D. M; et al. GPCR engineering yields high to mediate the agonist effect of I-SPD on DIR, whereas the Science 2007, 318, 1266-73 ts into beta2-adrenergic receptor function. absence of hydrophobic stacking between 13.40, F6.44, and 4)Rasmussen, S. G; et al. Structure of a nanobody-stabilized active W6.48 in D2R and D3R excludes receptor activation ate of the beta(2 Based on the observations from our structural modeling ar (S)Rosenbaum, D M. et al Structure and function of an irreversible nolecular dynamics simulation, the agonistic and antagonistic onist-beta(2)adrenoceptor complex. Nature 2011, 469, 236-40 dual-action mechanisms of 1-SPD were proposed. In the (6) Warne, T i et al. Structure of a betal-adrenergic G-protein agonistic mechanism of 1-SPD on DIR, structural changes in d receptor. Nature 2008, 454, 486-91 transmembrane helices are ignited by the formation of a Warne, T ; et al. The structural basis for agonist and partial hydrogen bond between the hydroxyl group of ring D of 1-SPD onist action on a beta(1)-adrenergic receptor. Nature 2011, 469, and N6.5 of TM6 which causes a clockwise rotation of the 41-4 extracellular segment of TM6. Such clockwise rotation in turn Jaakola, V P; et al. The 2.6 angstrom crystal structure of a human A2A adenosine receptor bound to an antagonist. Science 2008, 322, repositions F6.44 and w6.48 of TM6 to form a hydrophobic stacking 13.40 of TM3, causing inward movement of the 9)Xu, F et al. Structure of an Agonist-Bound Human A2A extracellular part of TM3. Accompanying these structural adenosine Receptor. Science 2011 8129 dx. dolora/o.021/p30492351 Phys. Chem. B2012116,8121-813diffuses to the extracellular mouth of the binding cavity of D1R, the electrostatic attraction between the negatively charged region of the D1R and positively charged l-SPD assists in their initial association (Figure 7, Rt1). The approach of extracellular ends of TM3 and TM6 to each other is then bridged by hydrogen bonding of l-SPD with N6.55 and electrostatic attraction with D3.32. As a result, I3.40, F6.44, and W6.48 are repositioned to form hydrophobic stacking, leading to a 7.1 Å outward movement of the cytoplasmic segment of TM6 (Figure 7, Rt2). At this stage, the receptor reaches its Rt2 state ready for G-protein binding, while not fully activated as experimental observation showed that both agonist and G￾protein binding are required to fully activate GPCRs.4,5 Finally, the large separation between the cytoplasmic ends of TM3 and TM6 allows readily binding of G-protein, leading D1R to its fully activated R* state. This stepwise activation process of D1R is supported by the currently accepted postulate about the dynamics of GPCRs; that is, GPCRs are more like molecular rheostats rather than simple two-state switches and are consistent with the activation mechanism of β2AR.5,24,25 ■ CONCLUSION We report herein a more comprehensive molecular modeling and molecular dynamics simulation study of the unliganded D1R, D2R, and D3R and l-SPD-D1R, l-SPD-D2R, and l-SPD￾D3R complexes, all simulated in a lipid bilayer membrane. New homology models of both D1R and D2R were constructed using the more reliable active state β2AR and inactive state D3R as templates, respectively. Structural analyses of molecular dynamics simulations were carried out followed by calculation of binding energy. We found that the separation in TM3−TM6 for unliganded D1R shortened gradually from ca. 20 to ca. 12 Å, whereas the distance in TM3−TM6 was kept at ca. 15 Å during the entire simulation of l-SPD−D1R complex, and this kind of separation in TM3 and TM6 is a typical feature of active state GPCRs, while the distance in TM3−TM6 in all D2R and D3R simulations remained relatively stable below 14 Å, which is a typical feature of inactive state GPCRs. Although there was no ionic lock formed when the unliganded D1R relaxed to its inactive state, the ionic lock R3.60 and E6.30 approaches gradually to each other. Our predicted binding energies of l￾SPD bound with D1R, D2R, and D3R correlated perfectly with experimental binding affinity, signifying that our in silico simulations probably resembled the binding poses of l-SPD to D1R, D2R, and D3R. A more in-depth analysis of the interaction maps of l-SPD with D1R, D2R, and D3R revealed that hydrogen bonding with N6.55 coupled with hydrophobic stacking between I3.40, F6.44, and W6.48 was the key feature to mediate the agonist effect of l-SPD on D1R, whereas the absence of hydrophobic stacking between I3.40, F6.44, and W6.48 in D2R and D3R excludes receptor activation. Based on the observations from our structural modeling and molecular dynamics simulation, the agonistic and antagonistic dual-action mechanisms of l-SPD were proposed. In the agonistic mechanism of l-SPD on D1R, structural changes in transmembrane helices are ignited by the formation of a hydrogen bond between the hydroxyl group of ring D of l-SPD and N6.55 of TM6 which causes a clockwise rotation of the extracellular segment of TM6. Such clockwise rotation in turn repositions F6.44 and W6.48 of TM6 to form a hydrophobic stacking I3.40 of TM3, causing inward movement of the extracellular part of TM3. Accompanying these structural changes, the cytoplasmic segment of TM6 shifts significantly out of the receptor core, and this shift in turn opens the docking pocket for G protein and leads to the activation of D1R. However, in the antagonistic mechanism of l-SPD on D2Rand D3R, no hydrophobic stacking is formed between I3.40, F6.44, and W6.48 in D2R and D3R to lead to an outward movement of the cytoplasmic segment of TM6. Our finding of key residues of D1R, D2R, and D3R involved in l-SPD binding could guide future experimental work on these receptors and has significance in the mechanism-based design of functional drugs targeting dopamine receptors. ■ ASSOCIATED CONTENT *S Supporting Information Sequence alignments of D1R with β2AR and of D2R with D3R, Ramachandran plots of all DR models, a brief summary of all simulation systems, TM3−TM6 distances and ionic lock formation and breakage in representative active and inactive GPCR crystal structures, binding affinity data for l-SPD analogues, and the definition of N−O distance as an indication of ionic lock formation and breakage. This material is available free of charge via the Internet at http://pubs.acs.org. ■ AUTHOR INFORMATION Corresponding Author *Fax +86-21-50807188, Ph +86-21-50806600, e-mail kqyu@ mail.shcnc.ac.cn (K.Q.Y.); Fax +86-21-51980010, Ph +86-21- 51980010, e-mail wfu@fudan.edu.cn (W.F.). Author Contributions § These authors contributed equally. Notes The authors declare no competing financial interest. ■ ACKNOWLEDGMENTS This work was supported by the National Natural Science Foundation of China (No. 81172919) and grants from the National High Technology Research and Development Program of China (863 Program) (No. 2012AA021102) and the State Key Program of Basic Research of China grant (2009CB918502). ■ REFERENCES (1) Overington, J. P.; Al-Lazikani, B.; Hopkins, A. L. How many drug targets are there? Nat. Rev. Drug Discovery 2006, 5, 993−6. (2) Cherezov, V.; et al. High-resolution crystal structure of an engineered human beta2-adrenergic G protein-coupled receptor. Science 2007, 318, 1258−65. (3) Rosenbaum, D. M.; et al. GPCR engineering yields high￾resolution structural insights into beta2-adrenergic receptor function. Science 2007, 318, 1266−73. (4) Rasmussen, S. G.; et al. Structure of a nanobody-stabilized active state of the beta(2) adrenoceptor. Nature 2011, 469, 175−80. (5) Rosenbaum, D. M.; et al. Structure and function of an irreversible agonist-beta(2) adrenoceptor complex. Nature 2011, 469, 236−40. (6) Warne, T.; et al. Structure of a beta1-adrenergic G-protein￾coupled receptor. Nature 2008, 454, 486−91. (7) Warne, T.; et al. The structural basis for agonist and partial agonist action on a beta(1)-adrenergic receptor. Nature 2011, 469, 241−4. (8) Jaakola, V. P.; et al. The 2.6 angstrom crystal structure of a human A2A adenosine receptor bound to an antagonist. Science 2008, 322, 1211−7. (9) Xu, F.; et al. Structure of an Agonist-Bound Human A2A Adenosine Receptor. Science 2011. The Journal of Physical Chemistry B Article 8129 dx.doi.org/10.1021/jp3049235 | J. Phys. Chem. B 2012, 116, 8121−8130