Cysteine proteases of the Potyviridae family, specifically the nuclear inclusion A (NIa) proteases, represent a largely untapped source of sequence-specific endopeptidases. With more than 3,800 characterized family members, NIa proteases cleave a distinct seven-amino-acid substrate sequence. The potyviral family offers a diverse set of proteases for engineering programmable protein control systems, constructing orthogonal protease-based genetic circuits, and developing sequence-specific biotechnological tools1,2. Despite this natural diversity, only two NIa proteases, TEVp and tobacco vein mottling virus protease (TVMVp), have been thoroughly characterized and engineered for broad laboratory and biotechnological use3,4. This limited protease toolkit limits the design space for multicomponent circuits that require orthogonal protease activities5, thereby motivating efforts to further expand and characterize NIa proteases through directed evolution6,7.
Yeast surface display has emerged as one of the most powerful platforms for directed evolution, enabling high-throughput protein library screening via fluorescence-activated cell sorting (FACS)6,7. For example, the yeast endoplasmic reticulum sequestration screening (YESS) system couples protease activity to surface display through the secretory pathway of S. cerevisiae6,7. The YESS system has been applied to engineer novel protease-substrate specificities and has demonstrated utility for multi-round directed evolution campaigns. In this system, protease variants are co-expressed with a fusion protein containing a cleavage substrate cassette with epitope markers, and the Aga1p surface anchor is expressed in the EBY100 yeast strain. Differential labeling with fluorescent antibodies distinguishes cleaved from non-cleaved substrate cassettes displayed on the cell surface. This activity-dependent display mechanism is directly compatible with iterative FACS-based selection for libraries of 107 or more variants6,7.
A fundamental prerequisite for directed evolution is the ability to introduce variant libraries at a sufficient scale to achieve adequate coverage of the targeted sequence space. For site-saturation mutagenesis libraries spanning multiple amino acid positions, transformation yields of 108 to 109 unique transformants are needed to ensure that the diversity introduced into the library is well represented in the sorted population. Several methods exist for introducing exogenous DNA into S. cerevisiae, each with distinct throughput ceilings. The conventional chemical transformation method uses lithium acetate/single-stranded carrier DNA/polyethylene glycol (LiAc/ssDNA/PEG) due to its simplicity and accessibility. However, it typically achieves transformation efficiencies of only 104–106 colony-forming units (CFU) per microgram of DNA, falling short of the sequence variation needed for large library campaigns8. By contrast, the electroporation method applies a high-voltage electric field to transiently permeabilize the cell membrane and facilitate DNA uptake. Electroporation consistently achieves the highest transformation efficiencies among available transformation methods, with optimized protocols reporting yields of 108–109 CFU9. It has been observed that electroporation efficiency in S. cerevisiae is critically sensitive to cell growth phase, DNA mass, and buffer composition, providing a framework that has since been widely adapted for library-scale directed evolution9.
The protocol described here uses S. cerevisiae strain EBY100, a widely adopted host strain for yeast surface display. EBY100 was engineered by Boder and Wittrup by integrating an AGA1 expression cassette under the control of the galactose-inducible GAL1 promoter into the chromosomal AGA1 locus of the protease-deficient parent strain BJ546510. The strain is commercially available from the ATCC (catalog no. MYA-4941; depositor K.D. Wittrup), carrying the genotype MATa AGA1::GAL1-AGA1::URA3 ura3-52 trp1 leu2-Δ1 his3-Δ200 pep4::HIS3 prb1Δ1.6R can1 GAL (URA+, leu−, trp−). As a result, both Aga1p and the Aga2p-substrate cassette fusion are expressed from the same galactose-inducible promoter, ensuring that the two subunits are co-induced and assemble into the disulfide-linked Aga1p–Aga2p complex that anchors the displayed fusion to the cell wall.
Despite the widespread use of EBY100 in yeast display applications, no published protocol has been systematically optimized for electroporating this strain at library-scale yields. EBY100-specific parameters, such as the optimal growth phase at harvest, the medium composition during subculture, and the quantitative relationship between initial optical density at 600 nm (OD600) and transformation efficiency, remain to be rigorously characterized and optimized. Here, we report that transformation efficiency in EBY100 increases monotonically with the culture's OD600 at harvest, with cells in late log phase yielding substantially higher transformation rates than those harvested at early log phase. We evaluated critical protocol parameters, including initial OD600 values at inoculation, the use of 2 × Yeast Extract Peptone Dextrose (YPD) medium for subcultures, electroporation buffer formulations, and post-electroporation outgrowth conditions. The protocol presented here provides an optimized, step-by-step workflow (Figure 1A–C, Figure 2, Figure 3A–D, Figure 4A–E) for preparing electrocompetent EBY100 cells and performing high-efficiency DNA library electroporation, yielding transformation rates consistently above 109 CFU. To validate this protocol, we conducted a comprehensive directed evolution campaign to engineer TEVp with altered cleavage specificity across four rounds of FACS-based selection and enrichment. Sequencing of enriched protease variants confirmed substantial diversification at targeted residues within the protease's substrate-binding pocket, demonstrating that the protocol provides sufficient library coverage of sequence space to support productive evolution campaigns. This protocol will be of broad utility to researchers using EBY100 for yeast surface display-based directed evolution and FACS-compatible, high-throughput screening of proteases, antibodies9, kinases11, and receptors12, among other proteins.