Computational Modeling of Stapled Peptides toward a Treatment Strategy for CML and Broader Implications in the Design of Lengthy Peptide Therapeutics

The oncogenic gene product Bcr-Abl is the principal cause of chronic myeloid leukemia, and although several therapies exist to curb the aberrant kinase activity of Bcr-Abl through targeting of the Abl kinase domain, these therapies are rendered ineffective by frequent mutations in the corresponding gene. It has been demonstrated that a designed protein, known as CCmut3, is able to produce a dominant negative inactivating effect on Bcr-Abl kinase by preferentially oligomerizing with the N-terminal coiled-coil oligomerization domain of Bcr-Abl (Bcr-CC) to effectively reduce the oncogenic potential of Bcr-Abl. However, the sheer length of the CCmut3 peptide introduces a high degree of conformational variability and opportunity for targeting by intracellular proteolytic mechanisms. Here, we have examined the effects of introducing one or two molecular staples, or cross-links, spanning i, i + 7 backbone residues of the CCmut3 construct, which have been suggested to reinforce α-helical conformation, enhance cellular internalization, and increase resistance to proteolytic degradation, leading to enhanced pharmacokinetic properties. The importance of optimizing staple location along a highly tuned biological construct such as CCmut3 has been widely emphasized and, as such, we have employed in silico techniques to swiftly build, relax, and characterize a large number of candidates. This approach effectively allowed exploring each and every possible staple location along the peptide backbone so that every possible candidate is considered. Although many of the stapled candidate peptides displayed enhanced binding characteristics for Bcr-CC and improved conformational stability in the (Bcr-CC) bound form, simulations of the stapled peptides in the unbound form revealed widespread conformational variability among stapled candidates dependent on staple type and location, implicating the molecular replacement of helix-stabilizing residues with staple-containing residues in disrupting the native α-helical conformation of CCmut3, further highlighting a need for careful optimization of the CCmut3 construct. A candidate set has been assembled, which retains the native backbone α-helical integrity in both the bound and unbound forms while providing enhanced binding affinity for the Bcr-CC target, as research disseminated in this manuscript is intended to guide the development of a next-generation CCmut3 inhibitor peptide in an experimental setting.