Large‐scale Structure‐based Prediction and Identification of Novel Protease Substrates using Computational Protein Design

Characterizing the substrate specificity of protease enzymes is important for illuminating the molecular basis of their diverse and complex roles in key biological processes including blood clotting, apoptosis, angiogenesis, viral infections and drug resistance. Rapid and accurate prediction of their extended substrate specificity would also aid in the design of custom proteases capable of selectively and controllably cleaving biotechnologically or therapeutically relevant proteins. However, current in silico approaches for protease specificity prediction rely on and are therefore limited by, machine learning of sequence patterns in known experimental data. Here, we describe a general approach for predicting peptidase substrates de novo using protein structural modeling and biophysical evaluation of peptide‐peptidase complexes. We construct atomic resolution models of thousands of substrate‐enzyme complexes for each of six model proteases belonging to the four major protease mechanistic classes – serine‐, cysteine‐, aspartyl‐ and metallo‐proteases, and develop a discriminatory scoring function using enzyme design modules from Rosetta and Amber‐MMPBSA. We rank putative peptide substrates in order of calculated cleavage probability, which reflects the interaction energy of a given substrate peptide with a modeled near attack conformation of the enzyme active site. We show that the energetic patterns obtained from these simulations can be used to robustly rank and classify known cleaved and uncleaved peptides, and these structural‐energetic patterns have greater discriminatory power compared to purely sequence‐based statistical inference. Combining sequence and energetic patterns using machine learning algorithms further improves classification performance, and analysis of structural models affords physical insight into the structural basis for the observed specificities. We test the predictive capability of the approach in a blind test by designing and experimentally testing the cleavage of novel substrates for the Hepatitis C virus NS3/4 protease using a yeast‐based assay. The presented structure‐based approach complements existing experimental methods for specificity determination and should enable the identification of biological targets of functionally uncharacterized and/or drug resistant peptidases for which no experimental specificity data exist, as well as the design of peptidase enzymes targeting chosen substrates. Support or Funding Information StartUp Grant (Rutgers, Chemistry and Chemical Biology) 1An overview of the protocol used to generate the protease‐ peptide complexes and discriminative score function A. Models of the proteases built using crystal structures from the PDB B. High resolution crystal structures used to generate models. The discriminative score function is a weighted linear combination of C. Interfacial stability D. Electrostatic interaction energy E. Secondary Structure Propensity of the peptide sequence F. Catalytic constraint penalty2Receiver Operator Curves demonstrating discrimination efficiency3Comparison between Enrichment (black bars) and baseline(dotted bars) enrichment levels for all six proteases4Novel sequences identified using the discriminative score function ( using Rosetta). These sequences are tested using a Yeast based ER Sequestration Screen assay.