Small-Molecule Targeting of Heat Shock Protein 90 Chaperone Function: Rational Identification of a New Anticancer Lead Massimiliano Meli, ²,‡ Marzia Pennati, ‡ Maria Curto, ‡ Maria Grazia Daidone, ‡ Janet Plescia, § Sam Toba, | Dario C. Altieri, § Nadia Zaffaroni, ‡ and Giorgio Colombo* ,² Istituto di Chimica del Riconoscimento Molecolare, Consiglio Nazionale delle Ricerche, Via Mario Bianco 9, 20131 Milano, Italy, Istituto Nazionale Studio e Cura Tumori, Department of Experimental Oncology, Unit 10, Via Venezian 1, 20133 Milano, Italy, Department of Cancer Biology and the Cancer Center, UniVersity of Massachusetts Medical School, Worcester, Massachusetts 01605, and Accelrys, Rational Drug DiscoVery, San Diego, California 92121 ReceiVed July 17, 2006 Heat shock protein 90 (Hsp90) is a significant target in the development of rational cancer therapy due to its role at the crossroads of multiple signaling pathways associated with cell proliferation and cell viability. Here we present a combined structure- and dynamics-based computational design strategy, taking the flexibility of the receptor and of a lead peptidic antagonist into account explicitely, to identify the nonpeptidic small molecule 5-aminoimidazole-4-carboxamide-1--D-ribofuranoside (AICAR) as a structurally novel inhibitor of Hsp90. The compound is selected to bind the Hsp90 N-terminal domain, mimicking the chemical and conformational properties of the recently described peptidic antagonist of the survivin-Hsp90 complex, shepherdin [Plescia et al. Cancer Cell 2005, 7, 457-468]. Experimental tests show that AICAR binds the Hsp90 N-domain, destabilizes multiple Hsp90 client proteins in vivo, including survivin, and exhibits antiproliferative and proapoptotic activity in multiple tumor cell lines, while not affecting proliferation of normal human fibroblasts. We propose that AICAR represents a viable lead for further development of anticancer drugs with wide therapeutic opportunities. Introduction Cancer therapy now aims at disabling oncogenic pathways that are selectively operative in tumor cells, so as to spare normal tissues and limit side effects in humans. This “targeted therapy” relies on a better understanding of cancer genes, particularly those implicated in tumor cell proliferation and survival. 1 Accordingly, targeted inhibition of the Bcr-Abl kinase with small molecule antagonists has produced dramatic clinical responses in malignancies driven by this oncogene. 2-4 However, such approaches may not be immediately available for the majority of tumors where multiple molecular abnormalities and genetic instabilities may elude the identification of one single, disease- driving oncogene. 1 Conversely, pathways that intersect multiple essential functions of tumor cells may provide wider therapeutic opportunities. A prime target for this strategy is the heat shock protein 90 (Hsp90), a molecular chaperone that oversees the correct conformational development of polypeptides and protein refolding through sequential ATPase cycles and stepwise recruitment of cochaperones. This adaptive pathway contributes to the cellular stress response to environmental threats, including heat, heavy metal poisoning, hypoxia, and so on, and is extensively exploited in cancer, where Hsp90 ATPase activity is upregulated ∼100-fold. 5 The repertoire of Hsp90 client proteins is restricted mainly to growth-regulatory and signaling molecules, especially kinases and transcription factors, which may contribute to tumor cell maintenance. 5,6 Therefore, targeted suppression of Hsp90 ATPase activity with a small molecule inhibitor, the benzoquinone ansamycin antibiotic 17-allylamino- 17-demethoxygeldanamycin (17-AAG), has shown promising anticancer activity in preclinical models and has recently completed safety evaluation in humans. 7,8 One Hsp90 client protein with critical roles in tumor cell proliferation and cell viability is survivin, an inhibitor of apoptosis (IAP) protein selectively overexpressed in cancer. 9,10 Accordingly, targeting the survivin-Hsp90 complex may provide a strategy to simultaneously disable multiple signaling pathways in tumors, and a peptidomimetic antagonist of this interaction structurally different from 17-AAG, shepherdin, inhibited the chaperone activity and exhibited potent and selective anticancer activity in preclinical models. 11 In this study, we have used shepherdin (LFACGSSHK, all D-amino acids) as a scaffold to rationally identify low-molecular- weight compounds that may act as structurally novel Hsp90 antagonists (Figure 1). We built a three-dimensional pharma- cophore to screen a database of nonpeptidic structures, and we identified a novel antagonist of Hsp90 chaperone function with promising anticancer activity. The results presented here open the possibility to expand the molecular diversity space of Hsp90 antagonists. Materials and Methods Simulation Setup, Docking Experiments, and Molecular Dynamics (MD) Refinement of Hsp90/Shepherdin Complex. MD simulations of shepherdin in isolation in solution were described in ref 11. The docking procedure can be summarized as follows: the representative of the main conformational clustering for shepherdin structures (-hairpin) was subjected to blind docking experiments on the putative N-terminal Hsp90 receptor using the program AutoDock. 12 The crystal structure of the protein was taken from the protein data bank (pdb code 1YET). 13 The original X-ray structure contains the ligand GA, which was removed from the active site to yield the apo-open form of Hsp90. Mass-centered grid maps were generated with 0.25 Å spacing by the program Autogrid for the whole Hsp90 protein target. Lennard-Jones parameters 12-10 and 12-6 (the default param- * To whom correspondence should be addressed. Tel.: +39-02- 28500031. Fax: +39-02-28901239. E-Mail: g.colombo@icrm.cnr.it. ² Istituto di Chimica del Riconoscimento Molecolare. ‡ Istituto Nazionale Studio and Cura Tumori. § University of Massachusetts Medical School. | Accelrys. 7721 J. Med. Chem. 2006, 49, 7721-7730 10.1021/jm060836y CCC: $33.50 © 2006 American Chemical Society Published on Web 11/22/2006