COMPUTATIONAL DESIGN OF RECEPTOR SELECTIVE TRAIL VARIANTS

Tumor necrosis factor-related apoptosis-inducing ligand (TRAIL) is a potentialanticancer drug that selectively induces apoptosis in a variety of cancer cells byinteracting with death receptors DR4 and DR5. TRAIL can also bind to decoy receptors (DcR1, DcR2, and osteoprotegerin receptor) that cannot induce apoptosis. Different tumor types respond either to DR4 or to DR5 activation and chemotherapeutic drugs can increase the expression of DR4 or DR5 in cancer cells. Thus, DR4 or DR5 receptor-specific TRAIL variants have enormous potential as new tumorselectivetherapies. Therefore, we designed receptor specific TRAIL variants that bind selectively to either DR4 (2) or DR5 (1) by using the computational protein design algorithm FoldX. Various in vitro receptor binding assays and cell based biological activity assays demonstrated that the designed TRAIL variants selectively bound to and induced apoptosis via either DR4 or DR5. DR5 selective TRAIL variants induced apoptosis more potently than wild-type TRAIL. Moreover, it was shown that the kinetics of apoptosis induction were increased by an order of magnitude (Szegezdi et al., submitted). This increase in kinetics could however not solelybe explained by the ∼3-fold improvement in association rate constant for DR5 binding.Computer simulation of TRAIL-receptor interaction revealed that preventionof ligand-induced receptor heteromerization combined with the increased affinitytowards DR5 produced these enhanced receptor-activation kinetics. In conclusion, computational protein design was effective in the development of receptor-specific TRAIL variants. Moreover, the designed receptor selectivity also enhanced the kinetics of receptor activation. We believe that this method could be generally applicable to design faster acting and more potent variants of other promiscuous cytokines as well