Abstract

Integrins V{beta}3 and 5{beta}1 play critical roles in tumor survival, invasion, metastasis, and angiogenesis and are validated targets for cancer therapy and molecular imaging. Increasing evidence suggests that targeting both integrins simultaneously with antagonists is more effective in cancer therapy because of concerns about resistance and paradoxical promotion of tumor growth with use of agents highly selective for a single integrin. Engineered Arg-Gly-Asp (RGD)-containing 3.5 kDa cysteine-knot proteins (knottins 2.5F and 2.5D) are attractive drug candidates due to their exceptional structural stability and high affinity binding to certain integrins. 2.5F binds both V{beta}3 and 5{beta}1, whereas 2.5D is V{beta}3-selective. To elucidate the structural basis of integrin selection, we determined the structures of 2.5F and 2.5D both as apo-proteins and in complex with V{beta}3. These data, combined with MD simulations and mutational studies, revealed a critical role of two V{beta}3-specific residues in the vicinity of the metal ion dependent adhesion site (MIDAS) in promoting an V{beta}3-induced fit of 2.5D. In contrast, conformational selection accounted for the specificity of 2.5F to both integrins. These data provide new insights into the structural basis of integrin-ligand binding specificity, and could help in development of integrin-targeted therapeutics.