Virtual screening and biochemical evaluation to identify new inhibitors of mammalian target of rapamycin (mTOR) Hwangseo Park a,⇑ , Hyeonjeong Choe b , Sungwoo Hong b,⇑ a Department of Bioscience and Biotechnology, Sejong University, 98 Kunja-Dong, Kwangjin-Ku, Seoul 143-747, Republic of Korea b Department of Chemistry, Korea Advanced Institute of Science and Technology (KAIST), Daejeon 305-701, Republic of Korea article info Article history: Received 20 November 2013 Revised 17 December 2013 Accepted 19 December 2013 Available online 27 December 2013 Keywords: Virtual screening mTOR Docking Kinase inhibitor Anticancer agents abstract Mammalian target of rapamycin (mTOR) is a promising target for the development of anticancer medicines. Here, we report the first example for a successful application of the structure-based virtual screening to identify new mTOR inhibitors. Using the scoring function improved by implementing the ligand solvation effects on protein–ligand association, six novel mTOR inhibitors are found with IC 50 val- ues ranging from 8 to 60 lM. Because these new inhibitors are also computationally screened for having desirable physicochemical properties as a drug candidate, they deserve consideration for further devel- opment by structure–activity relationship studies to optimize the inhibitory and anticancer activities. Structural features relevant to the stabilization of the inhibitors in the ATP-binding site of mTOR are addressed in detail. Ó 2013 Elsevier Ltd. All rights reserved. Mammalian target of rapamycin (mTOR) belongs to the phos- phoinositide kinase-related kinase (PIKK) family, and occupies an important signaling node of PI3K/AKT/mTOR pathway. Although the normal activity of this pathway is indispensable for cell growth and proliferation, 1 its aberrant activation can be responsible for the pathogenesis of various human cancers. 2,3 Therefore, the inhibition of mTOR activity with small molecules has been considered a promising therapeutic strategy for the treatment of cancer. The clinical approval of its inhibitors such as rapamycin and related analogues (rapalogues) manifested the usefulness of mTOR as a target for the development of anticancer medicines. 4 However, the efficacies of rapalogues have been limited due to the only par- tial inhibition of mTOR activity and the existence of a negative feedback loop. 5,6 These prompted the discovery of new potent mTOR inhibitors with higher efficacy than rapalogues. Accordingly, a great deal of efforts have been devoted to the dis- covery of mTOR inhibitors with the aim to develop new anticancer medicine as recently reviewed in a comprehensive fashion. 7,8 These scientific endeavors led to the discovery of potent mTOR inhibitors including pyrimidoaminotropane, GDC-0349, N-methyl- ated imidazolo-pyrimidines, and CC214-2. 9–12 Because the PI3K/ AKT/mTOR pathway may be resistant to the inhibitors due to the feedback activation, 13 the discovery of dual PI3K/mTOR inhibitors has also been actively pursued. 14–16 Complementary to various experimental findings, computational studies have also been car- ried out based on 3D-QSAR and molecular docking to address the structural and energetic features relevant to the potencies of known mTOR inhibitors. 17 However, the lack of full-length struc- ture of mTOR has limited the applicability of computational meth- ods. Recently, three dimensional structure of mTOR was reported in complex with small-molecule inhibitors as well as in the resting form. 18 The presence of structural information about the nature of the interactions between mTOR and its potent inhibitors allows for a rational design of new lead compounds for anticancer medicines. In the present study, we aim to identify new classes of mTOR inhibitors by means of a structure-based drug design protocol involving the virtual screening with docking simulations and in vitro enzyme assay. Virtual screening with docking simulation has not always been successful due to the inaccuracy in the scoring function. In particular, the binding affinity of a molecule with many polar atoms has often been overestimated due to the neglect of ligand solvation effects in the scoring function. 19 Prior to the cal- culation of the binding free energies between mTOR and the puta- tive ligands, therefore, we improve the scoring function by the implementation of an accurate solvation model. Docking simula- tions with the modified binding free energy function are expected to be useful for enriching the chemical library with molecules that are likely to have the inhibitory activity against mTOR. We prepared the receptor model from the X-ray crystal struc- ture of mTOR in complex with the inhibitor Torin2 (PDB code: 4JSX) 18 to perform the virtual screening with docking simulations to identify new mTOR inhibitors from a large chemical database. 0960-894X/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.bmcl.2013.12.081 ⇑ Corresponding authors. Tel.: +82 2 3408 3766; fax: +82 2 3408 4334 (H.P.); tel.: +82 42 350 2811; fax: +82 42 350 2812 (S.H.). E-mail addresses: hspark@sejong.ac.kr (H. Park), hongorg@kaist.ac.kr (S. Hong). Bioorganic & Medicinal Chemistry Letters 24 (2014) 835–838 Contents lists available at ScienceDirect Bioorganic & Medicinal Chemistry Letters journal homepage: www.elsevier.com/locate/bmcl