Ligand Binding Analysis for Human r51 Integrin: Strategies for Designing New r51 Integrin Antagonists Luciana Marinelli, Axel Meyer, Dominik Heckmann, Antonio Lavecchia, Ettore Novellino, and Horst Kessler* ,† Department Chemie, Technische Universita ¨ t Mu ¨ nchen, Lichtenbergstrasse 4, D-85747 Garching, Germany, and Dipartimento di Chimica Farmaceutica e Tossicologica, Universita ` di Napoli “Federico II”, Via D. Montesano, 49-80131 Napoli, Italy Received December 29, 2004 Abstract: We report a three-dimensional model of the R51 integrin headgroup bound to the most potent and selective ligand (SJ749) known to date. The model was built using the comparative protein modeling method, and it is consistent with experimental data. From this study, we identified two poten- tially important regions in the R51 receptor that are peculiar to this integrin and might be worth considering for drug targeting. Integrins are ubiquitous cell adhesion receptors that bind ligands on the surface of other cells and in the extracellular matrix and are involved in bidirectional signaling across the plasma membrane, regulating cell adhesion, differentiation, migration, grow, and sur- vival. 1 Evidence exists that such receptors are linked to pathological conditions including tumor progression, thrombosis, immune dysfunction, inflammation, and osteoporosis, so integrins have been attractive thera- peutic targets for several diseases. 2,3 Integrin R51 and especially Rv integrins have attracted attention as targets for antiangiogenic therapy. 4 Brooks et al. have reported that various low molecular weight ligands, which are recognized by Rv3 and Rv5 integrins, block angiogenesis in response to growth factors in tumors and suppress the cancer growth and metastasis. 5 On the basis of these findings, the Rv3 receptor has been the main integrin target in the search for new antican- cer drugs in the past 2 decades, and significant progress has been made in the identification of selective and potent Rv3 integrin antagonists. 6 As a result of our research, the cyclic peptide cyclo(-RGDf[NMe]V-), known as cilengitide, is now in phase II clinical trials for patients with glioblastoma. 7,14 Nonetheless, recent evidence that mice lacking of 3 or 5 integrins exhibit in some cases enhanced angio- genesis 8 led to a reevaluation of Rv3 and Rv5 integrin function in angiogenesis. 9 The different hypothesis made in an effort to explain the discrepancy between the genetic results and those obtained using low-molecular- weight ligands are not fully convincing, and the debate is still open. Reconsidering each integrin as a regulator of angiogenesis, an important issue came out. The R51 receptor is the only unambiguously proangiogenic in- tegrin; genetic ablation experiments and pharmacologi- cal results are consistent and strongly support its importance in neovascularization processes. 10 Hence, the R51 integrin is expected to move into the forefront of research for new effective anticancer drugs. The research of R51 antagonists is not as advanced as that of Rv3, and only few ligands are known to bind the R51 integrin. 11 This constitutes a considerable obstacle to the rational ligand-based drug design. Moreover, the 3D structure of membrane spanning receptors such as R51 is hard to obtain through X-ray crystallography or NMR methods. Consequently, few detailed structural information about ligand-receptor interactions have been obtained until now. However, it is well-know that integrins, which are heterodimers of different combinations of noncovalently bound R and chains, share extensive structural homology. It was also demonstrated that the ligand binding to Rv, R51, and RIIb3 integrins is mediated through the Arg-Gly-Asp (RGD) recognition motif. 12,13 A first glimpse at the binding between integrins and the RGD tripeptide sequence was provided by the crystal structure 13 of the extracellular domain of Rv3 integrin in complex with cyclo(-RGDf[NMe]V-). 14 Ex- periments using X-ray scattering and single-particle electron microscopy pointed out that the overall shape, the domain organization, and the way in which the R and subunits assemble are very similar between Rv3 and R51. 13,15 Fibronectin, which is the physiological ligand of R51, was found to bind its receptor through the RGD-containing module 10 (Fn10) in a manner similar to cyclo(-RGDf[NMe]V-) in the Rv3 crystal complex. 13 These findings, the high sequence similarity between R51 and Rv3 receptors (Rv:R5 53% identity; 3:1 55% identity in the integrin’s headgroup), and the need to get more information about the R51 selectivity requirement prompted us to build a 3D model of the R51 receptor. A multiple sequence alignment was performed utilizing evolutionary information of all R and subunits in different organisms. The crystal structure of Rv3 in the bound conformation was used as a template, obtaining 10 3D models of R51 integrin by means of comparative protein modeling methods. The models, which differ mostly in the side chain orientation, were used one at a time for ligand docking studies using the Autodock program (see Supporting Information). Compound 1 (SJ749) was chosen as ligand because of its potency, receptor selectivity, conformational restric- tions, and the recently published SARs (Table 1). 16 Docking results obtained for the 10 receptor models were carefully inspected to evaluate the agreement with the experimental data (mutagenesis, cross-linking, SARs) and the convergence and the binding free energy achieved for each simulation. SJ749 was found to fit in one R51 model preferentially over the others. Such a complex was energetically minimized using 3000 steps of steepest descent algorithm with the CVFF force field, permitting only the ligand and the side chain atoms of the protein within a radius of 5 Å around the ligand to relax. The stereochemical quality of the resulting model was checked with the program PROCHECK. The ma- jority of the residues occupied the most favored regions (81.7%) of the Ramachandran plot. Other residues * To whom correspondence should be addressed. Phone: +49-89- 289 13300. Fax: +49-89-289 13210. E-mail: kessler@ch.tum.de. Universita ` “Federico II” di Napoli. Technische Universita ¨t Mu ¨ nchen. 4204 J. Med. Chem. 2005, 48, 4204-4207 10.1021/jm040224i CCC: $30.25 © 2005 American Chemical Society Published on Web 06/01/2005