Directed enzyme evolution: beyond the low-hanging fruit

The field of directed evolution has progressed to the point where it is feasible to engineer enzymes for unnatural substrates and reactions with catalytic efficiencies and regio-specificity or stereo-specificity that rival those of natural enzymes. Here, we describe the conceptual and methodological advances that have enabled this progress. We address methodologies based on small libraries enriched with improved variants and carrying compensatory stabilizing mutations. Such libraries can be combined with low-throughput screens that provide high accuracy and directly target the desired substrate and reaction conditions, and thereby provide highly improved variants.

► Directed evolution can readily achieve >104-fold increases in rate and selectivity.
► Compensating stability losses is essential for multiple rounds and large changes.
► Success depends primarily on library design and less on screening throughput.
► Smart libraries and well-designed low-throughput screens are often preferred.