On the Value of Homology Models for Virtual Screening: Discovering hCXCR3 Antagonists by Pharmacophore-Based and Structure-Based Approaches Dane Huang, † Qiong Gu, † Hu, Ge, † Jiming Ye, ‡ Noeris K. Salam, ∥ Arnie Hagler, §,∥ Hongzhuan Chen, ⊥ and Jun Xu †, * † School of Pharmaceutical Sciences & Institute of Human Virology, Sun Yat-sen University, Guangzhou, China 510006 ‡ Health Innovations Research Institute and School of Health Sciences, Royal Melbourne Institute of Technology (RMIT) University, P.O. Box 71, Melbourne, VIC 3083, Australia § Department of Chemistry, University of Massachusetts, 701 Lederle Graduate Research Tower, 710 North Pleasant Street, Amherst, Massachusetts 01003-9336, United States ∥ Shifa Biomedical, 1 Great Valley Parkway, Suite 8, Malvern, Pennsylvania 19355, United States ⊥ School of Medicine, Institute of Medical Sciences, Shanghai Jiao Tong University, Shanghai, China 200025 * S Supporting Information ABSTRACT: Human chemokine receptor CXCR3 (hCXCR3) antagonists have potential therapeutic applications as antivirus, antitumor, and anti-inflammatory agents. A novel virtual screening protocol, which combines pharmacophore-based and structure-based approaches, was proposed. A three-dimensional QSAR pharmaco- phore model and a structure-based docking model were built to virtually screen for hCXCR3 antagonists. The hCXCR3 antagonist binding site was constructed by homology modeling and molecular dynamics (MD) simulation. By combining the structure-based and ligand-based screenings results, 95% of the compounds satisfied either pharmacophore or docking score criteria and would be chosen as hits if the union of the two searches was taken. The false negative rates were 15% for the pharmacophore model, 14% for the homology model, and 5% for the combined model. Therefore, the consistency of the pharmacophore model and the structural binding model is 219/273 = 80%. The hit rate for the virtual screening protocol is 273/286 = 95%. This work demonstrated that the quality of both the pharmacophore model and homology model can be measured by the consistency of the two models, and the false negatives in virtual screening can be reduced by combining two virtual screening approaches. ■ INTRODUCTION Chemokines are small (8−10 KDa) peptides responsible among other things for leukocytes migration to sites of infection or injury. Chemokines exert their chemotactic functions by binding to chemokine receptors. Chemokine receptors are G protein-coupled receptors (GPCRs) with seven transmembrane domains. Abnormal expressions of chemokines and their receptors are implicated in the pathogenesis of several human diseases, including autoimmune, chronic inflammatory diseases, immunodeficiency, cancer, and viral infection. 1−4 Human chemokine receptor CXCR3 (hCXCR3) is a Gαi protein-coupled receptor in the CXC chemokine receptor super family. Binding of CXC chemokines CXCL9 (MIG), CXCL10, and CXCL11 (IP-10, I-TAC) to hCXCR3 is necessary for CXCR3 signal transduction and activation of the expression of Th1, NK cell, pDC as well as mast cells. 5,6 Dysregulated expression of CXCR3 is involved in the development of chronic hepatitis B, bronchial asthma, transplantation rejection, auto- immune diseases, and dermatosis. 7 Many hCXCR3 antagonists have been discovered 8−15 in the past few years. The first QSAR study for hCXCR3 antagonists was published in 2009. 16 Another QSAR model for CXCR3 antagonists was reported in 2010. This model was based on multiple linear regressions (MLR) and least-squares support vector machine (LS-SVM) approaches. 17 The binding model of hCXCR3 and its antagonists is unknown because of the lack of an hCXCR3 crystal structure. In order to improve the performance of virtual screening campaign for hCXCR3 antagonists, we propose a new virtual screening protocol that combines pharmacophore-based and structure-based ap- proaches. First, an hCXCR3 homology model was built. Received: February 5, 2012 Published: April 30, 2012 Article pubs.acs.org/jcim © 2012 American Chemical Society 1356 dx.doi.org/10.1021/ci300067q | J. Chem. Inf. Model. 2012, 52, 1356−1366