PERRC: Protease Engineering with Reactant Residence Time Control

Abstract

Proteases with engineered specificity hold great potential for targeted therapeutics, protein circuit construction, and biotechnology applications. However, many proteases exhibit broad substrate specificity, limiting their use in such applications. Engineering protease specificity remains challenging because evolving a protease to recognize a new substrate, without counterselecting against its native substrate, often results in high residual activity on the original substrate. To address this, we developed Protease Engineering with Reactant Residence Time Control (PERRC), a platform that exploits the correlation between endoplasmic reticulum (ER) retention sequence strength and ER residence time. PERRC allows precise control over the stringency of protease evolution by adjusting counterselection to selection substrate ratios. Using PERRC, we evolved an orthogonal tobacco etch virus protease variant, TEVESNp, that selectively cleaves a substrate (ENLYFES) that differs by only one amino acid from its parent sequence (ENLYFQS). TEVESNp exhibits a remarkable 65-fold preference for the evolved substrate, marking the first example of an engineered orthogonal protease driven by such a slight difference in substrate recognition. Furthermore, TEVESNp functions as a competent protease for constructing orthogonal protein circuits in bacteria, and molecular dynamics simulations analysis reveals subtle yet functionally significant active site rearrangements. PERRC is a modular dual-substrate display system that facilitates precise engineering of protease specificity.

Keywords: high-throughput screening; protease engineering; synthetic biology; tobacco etch virus (TEV) protease; yeast surface display.

Figure 1.

Dual integration enables orthogonal substrate display. (A) In the previous version of YESS CS and SS substrates were within one polypeptide. (B) The new configuration allows us to independently display two or more substrates orthogonally. (C) Schematic of PERRC yeast display system which enables control over CS to SS ratios. (D) Flow cytometry histogram plots of the four-stain dual substrate display cassette.

Figure 2.

Establishing substrate residence time control in the ER. (A) WEHDEL-Stop cassette where the CS is paired with a strong ERS (WEHDEL), and the SS substrate is paired with no ERS (STOP). (B) Stop-Stop cassette where the CS and SS substrate are both paired with no ERS (STOP). (C) Time course at low temperature (20°C) with dual substrate strains. Flow cytometry bar graph showing the HA mean fluorescence of (A) in light pink and (B) in dark pink. (D) Time course at low temperature (20°C) with dual substrate strains. Flow cytometry bar graph showing the c-Myc mean fluorescence of (A) in light blue and (B) in dark blue. Reactions were run in triplicate. Statistical significance between populations was determined by multiple unpaired t-tests. *p ≤0.05, **p ≤0.01, ***p ≤0.001, ****p ≤0.0001.

Figure 3.

Resident time control in the ER empowers protease engineering under stringent kinetic competition. (A) Stop-Stop cassette where the CS, SS, and TEVEp are all paired with no ERS (STOP). (B) WEHDEL-Stop cassette where the CS is paired with a strong ERS (WEHDEL) and the SS and TEVEp are paired with no ERS (STOP). (C) Flow cytometry bar graph showing HA mean fluorescence (in pink) and c-Myc mean fluorescence (in blue) for both Stop-Stop and WEHDEL-Stop cassettes. Reactions were run in triplicate. Statistical significance between populations was determined by multiple unpaired t-tests. *p ≤0.05, **p ≤0.01, ***p ≤0.001, ****p ≤0.0001.

Figure 4.

Molecular dynamics simulations of the TEVEp and TEVESNp systems. (A) MELD-competitive simulations of TEVESNp with either a flexible or a rigid protocol for the enzyme. The plots track the population of the bound peptide at each replica, highlighting the clear preference for the ENLYFESG peptide in the TEVESNp enzyme if protein flexibility is taken into account. In the MELD-competitive simulations, multiple binding and unbinding events are expected, with both peptides unbound at the highest replicas. (B) Structural analysis of the centroids of the top clusters from MELD-Bracket simulations. Substrates are shown in magenta, and the enzymes are shown in blue. Important residues in the binding site are represented in stick models.

Figure 5.

TEVESNp exhibits strong orthogonality in a post-translationally controlled protein circuit. (A) Schematic of GFP degradation through protease cleavage. The blue rectangle is a substrate sequence (ENLYFQ or ENLYFE). The pink square is a N-terminal Y-degron (YLFVQ). The yellow Pac-Mac is the protease (TEVp, TEVEp, or TEVESNp). (B) Orthogonality matrix showing fluorescence fold reduction in cell cultures co-expressing TEVp, TEVEp, or TEVESNp with GFP reporters containing either the ENLYFQ or ENLYFE recognition sequence.