Design, Synthesis, and Biological Evaluation of Novel Hetero-Bivalent Ligands Targeting Kinases

Hetero-bivalent ligands consist of two copies of monomers that are connected through a linker. They show improved affinity towards their target through an avidity effect and can also enhance specificity, as part of the binding pocket may be absent in a similar target. This strategy was successful in yielding potent and specific kinase inhibitors by simultaneously targeting ATP and secondary peptide binding pockets. We innovatively apply this concept in two different kinase systems where we link low-affinity ATP binding pocket targeted small molecule PP2 with natural peptides that recognize secondary binding sites to produce potent and or specific hetero-bivalent ligands. In the first application, we targeted Extra-cellular Regulated Kinase5 (ERK5) that is known to play key roles in maintaining cancer stem cell signaling in a variety of cancers. The conventional ATP binding site inhibitors have not yet yielded expected levels of anti-cancer effects, due to complexities in converting ERK5 activation into Cancer Stem Cell (CSC) biological effects. Therefore, we hypothesized that designing a hetero-bivalent ligand, which simultaneously blocks a unique regulatory peptide interaction involved in upstream ERK5 kinase activation, and the conventional ATP binding pocket, produces stronger CSC biological effects. The active hetero-bivalent ligand ERK5.1 (i). inhibited ERK5 activation and kinase activity (IC50~8.5 µM) simultaneously, in two independent assay systems (ii). inhibited CSC activities, such as colony formation, migration and cell proliferation (IC50~6.5 µM) In the second application, we expanded the idea of using peptide sequences belonging to the same protein as the non-ATP binding site moiety in hetero-bivalent ligand design. Typically, these secondary peptide sequences are derived from substrate peptides from interacting proteins or structure based or combinatorial methods. The use of an already existing natural peptide sequence of the same kinase that uniquely interacts with a binding pocket within the same kinase as the non-ATP binding moiety in hetero-bivalent ligand design has never been investigated. We selected EphA3 kinase as a model system because of known natural peptide sequence in the SAM domain linker region, which turns back and binds to the bottom of EphA3. We connected this unique sequence to a PP2 analogue. Using a combination of structure and combinatorial approaches, we optimized the much longer linker (57 Å) to yield potent hetero-bivalent ligand EPHB2.3 (Kd~250 nM). We also report our effort to the convert the linker part of EPHB2.3 to gain additional binding affinity.