Document Type


Publication Title

ACS Omega









First Page


Last Page


Publication Date



Structure-based computational peptide design methods have gained significant interest in recent years owing to the availability of structural insights into protein–protein interactions obtained from the crystal structures. The majority of these approaches design new peptide ligands by connecting the crucial amino acid residues from the protein interface and are generally not based on any predicted receptor–ligand interaction. In this work, a peptide design method based on the Knob–Socket model was used to identify the specific ligand residues packed into the receptor interface. This method enables peptide ligands to be designed rationally by predicting amino acid residues that will fit best at the binding site of the receptor protein. In this, specific peptide ligands were designed for the model receptor CD13, overexpression of which has been observed in several cancer types. From the initial library of designed peptides, three potential candidates were selected based on simulated energies in the CD13 binding site using the programs molecular operating environment and AutoDock Vina. In the CD13 enzymatic activity inhibition assay, the three identified peptides exhibited 2.7–7.4 times lower IC50 values (GYPAY, 227 μM; GFPAY, 463 μM; GYPAVYLF, 170 μM) as compared to the known peptide ligand CNGRC (C1–C5) (1260 μM). The apparent binding affinities of the peptides (GYPAY, Ki = 54.0 μM; GFPAY, Ki = 74.3 μM; GYPAVYLF, Ki = 38.8 μM) were 10–20 times higher than that of CNGRC (C1–C5) (Ki = 773 μM). The double reciprocal plots from the steady-state enzyme kinetic assays confirmed the binding of the peptides to the intended active site of CD13. The cell binding and confocal microscopy assays showed that the designed peptides selectively bind to the CD13 on the cell surface. Our study demonstrates the feasibility of a Knob–Socket-based rational design of novel peptide ligands in improving the identification of specific binding versus current more labor-intensive methods.

Creative Commons License

Creative Commons Attribution 4.0 International License
This work is licensed under a Creative Commons Attribution 4.0 International License.