Journal of Chemical Information and Modeling Article Proposed activation process of S-HTIAR by an FwO1-like agonist R represents the inactive state; R'depicts the recognition of an agonist electrostatic interaction; R"presents the agonist binding-induced major structural rearrangements(agonist-bound state, orange carto state, cyan cartoon); and R* refers to the fully activated S-HTIAR coupling with Gs protein (pink). (R* is a schematic diagram, and R was taken to represent R* since our simulation does not involve Gs protein. relative position of TM3, TMS, and TM6 of 5-HTIAR in an HTAR agonists were successfully identified. Three of them inactive state, is interrupted by the motion of PS. 50. As a result, reveal high potency of K, values less than 100 nM.These 13.40 and F6.44 are repositioned, and the switch of F6. 44 compounds deserve further considerations for the developmen causes a corresponding swing of TM6. Whereas unlike in BAR, of S-HTIAR"superagonists"through several rounds of W6.48 moves distinctly toward PS50 upon agonsit binding. So w6.48 might play a similar role as a rotamer toggle switch" in investigation of the binding mode of FwOI with 5-HTIAR provides molecular information for us to improve our F6.44 and 13.40 vice versa. Therefore, in the case of 5-HT1AR, pharmacophore model and perform more productive virtual Ser5.42 and w6.48 probably undergo synergy in the process of screening. Through MD simulations study, we uncovered promoting a large TM6 motion. Eventually, TM6 dramatically unique conformational changes of S-HT1AR induced by agonist Proposed Activation Mechanism of 5-HTAROn FwOl. Finally, a stepwise activation model of S-HTIAR induced basis of our MD simulations study and generally accepted by Fwol-like agonists was proposed. This model can help knowledge about the activation mechanism of other Family a better the understanding of the signal transduction process GPCRs members, a stepwise activation process model of 5- HTIAR was further proposed( Figure 7). Typically, unbound when activated by exogenous agonists and guide future itAR adopts the inactive or basal active state r through its structure-based agonist design. ramolecular interaction and traits between several energy minima conformations by molecular thermodynamic motion or ASSOCIATED CONTENT its inherent flexibility. When protonated FwOl diffuses to the extracullar mouth of avity of 5-HTIAR s Supporting Information electrostatic attraction between the negatively charged region of Figure S1. This material is available free of charge via the 5-HTIAR and positively charged FwOl motivates the ligand Internetathttp://pubs.acs.org ecognition and binding process(R). Then FwOl is stably cradled in the extracellar segments of TM3, TMS, TM6, and ■ AUTHOR INFORMATION TM7 through several anchoring interactions with D3. 32, SS. 42, and N7. 39. These residues, which in turn act like molecular Corresponding Authors *E-mail: hejiang@mail. shcncac cn. Corresponding author ad- middle part of the seven bundles: P5.50, 13. 40, F6.44, F6.48, dress: Shanghai Institute of Materia Medica, Chinese Academy and N7.45 are repositioned to form hydrophobic stacking and of Sciences, Shanghai 201203, China(H J) ultimately translate into a large-scale outward swing of TM6 *E-mail:Zhenxuechu(@suda.edu.cn.Corresponding the cytoplasmic end(R"). At this stage, the cytoplasmic ends of address: Department of Pharmacology, Soochow Ur TM3 and TM6 are largely separated, and 5-HT1aR is ready to College of Pharmaceutical Sciences, Suzhou 215123, bind Gs protein, leading itself to a fully activated state r". X.Z.) * Phone:+86-21-51980010.Fax:+86-21-51980010. E-mail: ■ CONCLUS|oN wfu@fudan.edu.cn.Correspondingauthoraddress:Department of Medicinal Chemistry, School of Pharmacy, Fudan University, The dynamic pharmacophore model for 5-HTIAR which 826 Zhangheng Road, Shanghai 201203, P R China(WF) corporates the receptor flexibility was built here by integrating Author Contributions a set of computational approaches including homology LIli Xu, Shanglin Zhou, and Kunqian Yu contributed equally to modeling, molecular dynamics simulation, GROMOS con pharmacophore virtual screening. Finally, through our dynamic Notes k formational cluster analysis, GRID mapping, and dynamical pharmacophore based virtual screening protocol, 10 new 5- The authors declare no competing financial interest. dx dolor/10.1021/c400481plJ Chem Inf Model. 2013, 53, 3202-3211relative position of TM3, TM5, and TM6 of 5-HT1AR in an inactive state, is interrupted by the motion of P5.50. As a result, I3.40 and F6.44 are repositioned, and the switch of F6.44 causes a corresponding swing of TM6. Whereas unlike in β2AR, W6.48 moves distinctly toward P5.50 upon agonsit binding. So W6.48 might play a similar role as a “rotamer toggle switch” in the Rhodopsin receptor53 that initiates the displacement of F6.44 and I3.40 vice versa. Therefore, in the case of 5-HT1AR, Ser5.42 and W6.48 probably undergo synergy in the process of promoting a large TM6 motion. Eventually, TM6 dramatically moves outward as much as 17 Å. Proposed Activation Mechanism of 5-HT1AR. On the basis of our MD simulations study and generally accepted knowledge about the activation mechanism of other Family A GPCRs members, a stepwise activation process model of 5- HT1AR was further proposed (Figure 7). Typically, unbounded 5-HT1AR adopts the inactive or basal active state R through its intramolecular interaction and trasits between several energy minima conformations by molecular thermodynamic motion or its inherent flexibility. When protonated FW01 diffuses to the extracullar mouth of the binding cavity of 5-HT1AR, the electrostatic attraction between the negatively charged region of 5-HT1AR and positively charged FW01 motivates the ligand recognition and binding process (R′). Then FW01 is stably cradled in the extracellar segments of TM3, TM5, TM6, and TM7 through several anchoring interactions with D3.32, S5.42, and N7.39. These residues, which in turn act like molecular triggers, propagate the conformational changes to the inner￾middle part of the seven bundles: P5.50, I3.40, F6.44, F6.48, and N7.45 are repositioned to form hydrophobic stacking and ultimately translate into a large-scale outward swing of TM6 in the cytoplasmic end (R″). At this stage, the cytoplasmic ends of TM3 and TM6 are largely separated, and 5-HT1AR is ready to bind Gs protein, leading itself to a fully activated state R*. ■ CONCLUSION The dynamic pharmacophore model for 5-HT1AR which incorporates the receptor flexibility was built here by integrating a set of computational approaches including homology modeling, molecular dynamics simulation, GROMOS con￾formational cluster analysis, GRID mapping, and dynamical pharmacophore virtual screening. Finally, through our dynamic pharmacophore based virtual screening protocol, 10 new 5- HT1AR agonists were successfully identified. Three of them reveal high potency of Ki values less than 100 nM. These compounds deserve further considerations for the development of 5-HT1AR “superagonists” through several rounds of structural optimization and QSAR studies. In addition, the investigation of the binding mode of FW01 with 5-HT1AR provides molecular information for us to improve our pharmacophore model and perform more productive virtual screening. Through MD simulations study, we uncovered unique conformational changes of 5-HT1AR induced by agonist FW01. Finally, a stepwise activation model of 5-HT1AR induced by FW01-like agonists was proposed. This model can help better the understanding of the signal transduction process from the extracellular side to the cytoplasmic end of 5-HT1AR when activated by exogenous agonists and guide future structure-based agonist design. ■ ASSOCIATED CONTENT *S Supporting Information Figure S1. This material is available free of charge via the Internet at http://pubs.acs.org. ■ AUTHOR INFORMATION Corresponding Authors *E-mail: hljiang@mail.shcnc.ac.cn. Corresponding author ad￾dress: Shanghai Institute of Materia Medica, Chinese Academy of Sciences, Shanghai 201203, China (H.J.). *E-mail: Zhenxuechu@suda.edu.cn. Corresponding author address: Department of Pharmacology, Soochow University College of Pharmaceutical Sciences, Suzhou 215123, China (X.Z.). *Phone: +86-21-51980010. Fax: +86-21-51980010. E-mail: wfu@fudan.edu.cn. Corresponding author address: Department of Medicinal Chemistry, School of Pharmacy, Fudan University, 826 Zhangheng Road, Shanghai 201203, P.R. China (W.F.). Author Contributions § Lili Xu, Shanglin Zhou, and Kunqian Yu contributed equally to this work. Notes The authors declare no competing financial interest. Figure 7. Proposed activation process of 5-HT1AR by an FW01-like agonist. R represents the inactive state; R′ depicts the recognition of an agonist driven by electrostatic interaction; R″ presents the agonist binding-induced major structural rearrangements (agonist-bound state, orange cartoon; ligand-free state, cyan cartoon); and R* refers to the fully activated 5-HT1AR coupling with Gs protein (pink). (R* is a schematic diagram, and R″ was taken to represent R* since our simulation does not involve Gs protein.) Journal of Chemical Information and Modeling Article 3209 dx.doi.org/10.1021/ci400481p | J. Chem. Inf. Model. 2013, 53, 3202−3211