Protein Engineering by Coupling Rational
Design and Combinatorial Screening

Marco A. Mena
Department of Chemical Engineering
UCSB Doctoral Candidate

Date: Tuesday, Dec. 5, 2006
Time: 4:00 p.m.
Place: Engineering II, Room 3361

ABSTRACT

Protein engineering is typically performed via either rational design or combinatorial library screening. We have sought to develop strategies which effectively harness the strengths of the two methodologies. A set of computational algorithms were developed which use structural information to guide targeted library design. One algorithm couples a Monte Carlo search strategy with an energetic scoring function to find amino acid combinations that are structurally compatible with a target backbone. A second algorithm then finds and scores degenerate codon libraries based on these combinations. User-selected libraries are then constructed using gene assembly mutagenesis and screened for the property of interest.

We used computational library design strategies to engineer variants of blue fluorescent protein (BFP), a mutant of Aequorea victoria green fluorescent protein (GFP) that is too dim and photobleaches too easily for routine use. Libraries were designed and constructed which targeted positions neighboring the chromophore. Screening via flow cytometry yielded several bright variants with previously unreported substitutions. Characterization of one of these variants, Azurite, revealed that it contained a considerably enhanced quantum yield (0.55 vs. 0.34) and a 40-fold improvement in photobleaching half-life. The variant is well-suited for whole-cell cytometry, microscopy, or other fluorescent assays and as a Förster resonance energy transfer (FRET) partner to other fluorophores such as GFP or red fluorescent variants. These results suggest that targeted protein libraries may in some cases overcome some of the limitations inherent in using either structural design or library screening approaches alone.