Intracellular FRET-Based Assay for Redesigning the Specificity of Secreted Proteases

Proteases constitute 2% of the human genome and regulate many biological processes including cell growth and migration, blood coagulation, and programmed cell death.  A generally applicable, high-throughput strategy to engineer proteases to cleave a target substrate with high specificity and high catalytic efficiency would greatly expand the use of proteases for biotechnological and therapeutic applications.  We have developed a cell-based assay for redesigning protease selectivity by screening protease mutant libraries for cleavage of a fluorogenic peptide substrate exhibiting Förster resonance energy transfer (FRET).  Our screening method relies on using fluorescence activated cell sorting (FACS) for detecting cleavage of the FRET reporter substrate, which will allow screening of large protease mutant libraries (~10⁷members) to identify the activity of interest.  Here we report the application of this method to the protease human kallikrein 7 (hK7) to identify variants that selectively cleave the central hydrophobic core of the amyloid beta (Aβ) peptide, involved in Alzheimer’s disease pathology. 

Using flow cytometry, we optimized expression conditions to detect hK7 activity in yeast with a co-expressed intracellular Aβ8 (KLVF↓F↓AED) FRET substrate.  hK7 displays modest activity towards the target Aβ8 substrate but prefers tyrosine (Y) at the P1 position.  We hypothesized that amino acid substitutions within hK7 may yield variants that prefer the phenylalanine (F) at P1 of Aβ8 and exclude tyrosine, thereby narrowing the specificity towards Aβ8.  We randomly mutated the hK7 gene using error-prone PCR to generate a library with an average of 2 base substitutions per gene.  Screening this library for cleavage of the Aβ8 FRET reporter using fluorescence activated cell sorting yielded hK7 mutants with improved activity against the therapeutically relevant Aβ8 substrate.  These mutants were then recombined using the staggered extension process (StEP) and the resulting library was screened to produce mutants with additive improvements in activity and specificity for the Aβ8 substrate.  This methodology can be further applied to engineer other human proteases for highly specific degradation of proteins implicated in a disease state.