Structure-Based Optimization of Conformationally Constrained Peptides to Target Esophageal Cancer TEAD Transcription Factor

Abstract

The transcriptional enhanced associate domain (TEAD) transcription factor is a core regulator of Hippo signaling pathway and has been recognized as a promising therapeutic target against gastrointestinal tumors. The intermolecular interaction of TEAD with its transcription coactivator vestigial-like protein (VGLL) is involved in the tumorigenesis of esophageal cancer, which possesses an additional double-stranded β-sheet (DSβS) motif in its TEAD-recognition domain as compared to the classic coactivators YAP and TAZ. Here, the DSβS motif is examined systematically based on the complex crystal structure of TEAD with VGL. It is found that the DSβS is one of the most important recognition sites at TEAD–VGLL complex interface, which can confer stability and specificity to the complex recognition and elicit the transcription activation event. Protein segments with different lengths are derived from the DSβS motif of VGLL to define a number of linear peptides; their binding modes and affinities to TEAD are computationally analyzed to reveal that these linear peptides cannot maintain in native double-strand conformation in free solution state, thus unfavorable for their rebinding to TEAD–VGLL interface. A head-to-tail cyclization strategy is described to constrain the conformation of linear peptides, which can considerably reduce the peptide flexibility and help the peptide folding to an approximately native conformation in solution. We also demonstrate that the binding affinity and biological activity of cyclic peptides are improved effectively relative to their linear counterparts at molecular and cellular levels.