Structure-Activity Relationships of Small Phosphopeptides, Inhibitors of Grb2 SH2 Domain, and Their Prodrugs Wang-Qing Liu, † Michel Vidal, † Catherine Olszowy, Emmanuelle Million, Christine Lenoir, † He ´le `ne Dho ˆtel, and Christiane Garbay* ,† De ´ partement de Pharmacochimie Mole ´ culaire & Structurale, INSERM U266, CNRS FRE 2463, UFR des Sciences Pharmaceutiques et Biologiques, 4, Avenue de l’Observatoire, 75270 Paris Cedex 06, France Received August 12, 2003 To develop potential antitumor agents directed toward HER2/ErbB2 overexpression in cancer, we have designed inhibitors of the recognition between the phosphotyrosine of the receptor and the SH2 domain of the adaptor protein Grb2. In the first part of the paper, we report the synthesis of mimetics of the constrained (R-Me)phosphotyrosine residue such as (R-Me)-4- phosphonomethylphenylalanine (-CH 2 PO 3 H 2 ), (R-Me) 4-phosphonodifluoromethylphenylalanine (-CF 2 PO 3 H 2 ), and (R-Me)-4-phosphonophenylalanine (-PO 3 H 2 ). The incorporation of these residues in the mAZ-pTyr-Xaa-Asn-NH 2 series provided compounds with very high affinity for the Grb2 SH2 domain, in the 10 -8 -10 -9 range of K d values. These compounds behave as potent antagonists of the Grb2-Shc interaction. Our results highlight the importance of the doubly negative charge borne by the pY + 1 amino acid in accordance with the interactions observed in the complex crystallized between mAZ-pTyr-(RMe)pTyr-Asn-NH 2 and the Grb2 SH2 domain. mAZ-pTyr-(RMe)pTyr-Asn-NH 2 was derivatized as the S-acetyl thioester (SATE) of the phosphotyrosine residues, and its surrogates provided prodrugs with very potent antiprolif- erative activity on cells overexpressing HER2/ErbB2, with ED 50 values amounting to 0.1 µM. Finally a new prodrug is put forth under the form of a monobenzyl ester of phosphate group that is as active as and much easier to synthesize than SATE prodrugs. These compounds show promising activity for further testing on in vivo models. Introduction Cellular proliferation and differentiation are regu- lated by a variety of signaling mitogens such as growth factors that bind to the extracellular domain of their receptors. This process induces receptor dimerization and trans-phosphorylation of several intracellular ty- rosine residues in its C-terminal part and results in protein recruitment and transduction of the growth factor signal inside the cell. Deregulation of the Ras signaling pathway has been involved in a number of diseases that include leukemia and several cancers. 2 Along this pathway, the small adaptor protein Grb2 (growth factor receptor-bound protein 2) constitutes a connector between the receptor and Sos, the exchange factor of Ras. Grb2 is composed of a single SH2 (Src homology) domain flanked by two SH3 domains. 3 Grb2 SH2 domain binds numerous tyrosine phosphorylated proteins including activated RTKs such as the members of erbB family, 4 docking proteins such as Shc, 5 and cytoplasmic tyrosine kinases such as Bcr-Abl. 6 It was shown that direct binding of Grb2 through its SH2 domain with the Bcr-Abl is required for efficient induc- tion of chronic leukemia-like diseases in mice. 7 Grb2 forms a complex through its SH3 domains with Sos, which in turn activates Ras by exchanging its GDP binding form to the GTP binding one. Because of the Grb2 role in the Ras signaling pathway and its up- regulation in human breast cancer 8 and human bladder cancer 9 and in the early events of mice liver carcino- genesis, 10 inhibition of Grb2 constitutes an attractive strategy for developing new antitumor agents. 11,12 Since the Grb2 SH2 domain recognizes with high affinity and specificity the phosphotyrosyl consensus motif -pY-X-N- (pY, phosphotyrosine; X, any hydropho- bic amino acid; N, asparagine) on its targets, 13 the development of SH2 domain inhibitors was carried out following several directions. The first is the search for peptide inhibitors, encompassing phosphonate and car- boxylate-based pY mimetics that are resistant to intra- cellular phosphatases. 14 The second consists of modify- ing the structure of peptide inhibitors to circumvent the lack of cell permeability that is due to the presence of negatively charged groups. 15-18 The third consists of the design of peptidomimetics and even nonpeptidic com- pounds to inhibit Grb2-SH2 interactions. 19 These directions of research include (i) the optimiza- tion of the N- and C-terminal groups and of the modified hydrophobic residue X of the minimum pY-X-N pep- tide, 20-23 (ii) the design of phosphorylated as well as unphosphorylated cyclic peptides, 25,26 and (iii) the search for peptidomimetics retaining little or no peptidic character. 27-31 In a previous paper, 22 we reported the rational design and synthesis of derivatives in the mAZ-pTyr-Xaa-Asn- NH 2 series, which had high affinity for Grb2 especially when Xaa is an (R-Me)pTyr residue. The lower affinities of the peptides containing carboxylate mimetics of the (R-Me)pTyr residue confirmed that the doubly nega- * To whom correspondence should be addressed. Phone: 33-1-42- 86-40-80. Fax: 33-1-42-86-40-82. E-mail: christiane.garbay@ univ-paris5.fr. † Present address: Laboratoire de Pharmacochimie Mole ´culaire et Cellulaire, FRE CNRS 2718, INSERM U266, UFR Biome ´dicale, 45, Rue des Saints-Pe `res, 75270 Paris Cedex 06, France. 1223 J. Med. Chem. 2004, 47, 1223-1233 10.1021/jm031005k CCC: $27.50 © 2004 American Chemical Society Published on Web 02/03/2004