Rational design and exploitation of the molecular diversity space of PRX1-derived SRX1 peptidic inhibitors containing proteinogenic and nonproteinogenic amino acids

Abstract

The peroxiredoxin-1 (PRX1) of redox peroxidase reduces reactive oxygen and nitrogen species levels in human cells and can be regulated by its interacting partner sulfiredoxin-1 (SRX1). A 9-mer core peptide segment (termed CTTc) located in the C-terminal tail of PRX1 was found previously to mediate the PRX1–SRX1 interaction, and proteinogenic amino acids were also examined for each residue of the CTTc peptide. However, only proteinogenic amino acids can guarantee sufficient exploitation of the structural diversity space of CTTc-derived peptidic inhibitors targeting SRX1. In this study, the structural diversity space was further expanded by introducing 34 nonproteinogenic amino acids plus 20 proteinogenic amino acids as the building blocks of nonapeptides, and a systematic single-position binding energy change (SSRBEC) profile of the nonapeptide against SRX1 was created computationally by performing machine learning scoring, molecular dynamics simulation, and binding energetics analysis. A reduced combinatorial peptide library consisting of over 500,000 nonproteinogenic amino acid-containing nonapeptide candidates was designed using a rational computational peptidology strategy based on the SSRBEC profile, from which a variety of top-scored hits were identified and their affinities to SRX1 were measured at the molecular level to substantiate the computational findings. Three nonproteinogenic amino acid-containing peptides, CTTc[ut8], CTTc[ut3], and CTTc[ut6] possessed high potency, and their affinities were improved considerably relative to the native CTTc peptide. Structural analysis revealed that the SRX1-binding peptide sequences can be divided into three sections, that is, amino-terminal section N, middle section M, and carboxy-terminal section C, in which section N and, secondarily, section C are primarily responsible for the peptide binding affinity and specificity to SRX1, while section M is exposed to solvent and does not directly interact with SRX1.