Structural modeling of osteoarthritis ADAMTS4 complex with its cognate inhibitory protein TIMP3 and rational derivation of cyclic peptide inhibitors from the complex interface to target ADAMTS4

Abstract The ADAMTS4 (a disintegrin and metalloproteinase with thrombospondin motifs 4) enzyme is a matrix-associated zinc metalloendopeptidase that plays an essential role in the degradation of cartilage aggrecan in arthritic diseases and has been recognized as one of the most primary targets for therapeutic intervention in osteoarthritis (OA). Here, we reported computational modeling of the atomic-level complex structure of ADAMTS4 with its cognate inhibitory protein TIMP3 based on high-resolution crystal template. By systematically examining the modeled complex structure we successfully identified a short inhibitory loop ( 62 EASESLC 68 ) in TIMP3 N -terminal inhibitory domain (NID) that directly participates in blocking the enzyme’s active site, which, and its extended versions, were then broken from the full-length protein to serve as the peptide inhibitor candidates of ADAMTS4. Atomistic molecular dynamics simulation, binding energetic analysis, and fluorescence-based assay revealed that the TIMP3-derived linear peptides can only bind weakly to the enzyme ( K d  = 74 ± 8 μM), which would incur a considerable entropy penalty due to the high conformational flexibility and intrinsic disorder of these linear peptides. In this respect, we proposed a cyclization strategy to improve enzyme–peptide binding affinity by, instead of traditionally maximizing enthalpy contribution, minimizing entropy cost of the binding, where a disulfide bond was added across the two terminal residues of linear peptides, resulting in a number of TIMP3-derived cyclic peptides. Our studies confirmed that the cyclization, as might be expected, can promote peptide binding capability against ADAMTS4 substantially, with affinity increase by 3-fold, 9-fold and 7-fold for cyclic peptides , and , respectively.