Cytochrome P450 monooxygenases play a crucial role in the biosynthesis of many natural products and in the human metabolism of numerous pharmaceuticals. This has inspired synthetic organic and medicinal chemists to exploit them as catalysts in regio-and stereoselective CH-activating oxidation of structurally simple and complex organic compounds such as steroids. However, levels of regio-and stereoselectivity as well as activity are not routinely high enough for real applications. Protein engineering using rational design or directed evolution has helped in many respects, but simultaneous engineering of multiple catalytic traits such as activity, regioselectivity, and stereoselectivity, while overcoming tradeoffs and diminishing returns, remains a challenge. Here we show that the exploitation of information derived from mutability landscapes and molecular dynamics simulations for rationally designing iterative saturation mutagenesis constitutes a viable directed evolution strategy. This combined approach is illustrated by the evolution of P450 BM3 mutants which enable nearly perfect regio-and diastereoselective hydroxylation of five different steroids specifically at the C16-position with unusually high activity, while avoiding activity−selectivity trade-offs as well as keeping the screening effort relatively low. The C16 alcohols are of practical interest as components of biologically active glucocorticoids.

Multidimensional fitness landscapes provide insights into the molecular basis of laboratory and natural evolution. To date, such efforts usually focus on limited protein families and a single enzyme trait, with little concern about the relationship between protein epistasis and conformational dynamics. Here, we report a multiparametric fitness landscape for a cytochrome P450 monooxygenase that was engineered for the regio- and stereoselective hydroxylation of a steroid. We develop a computational program to automatically quantify non-additive effects among all possible mutational pathways, finding pervasive cooperative signs and magnitude epistasis on multiple catalytic traits. By using quantum mechanics and molecular dynamics simulations, we show that these effects are modulated by long-range interactions in loops, helices and β-strands that gate the substrate access channel allowing for optimal catalysis. Our work highlights the importance of conformational dynamics on epistasis in an enzyme involved in secondary metabolism and offers insights for engineering P450s.

By using structure-guided directed evolution, the substrate scope of the FeII and α-ketoglutarate dependent halogenase Wi-WelO15 from Westiella intricata HT-29-1 was engineered to enable chemo-, regio-, and diastereoselective chlorination of unactivated C­(sp3)–H bonds using NaCl as chlorine source. While FeII dependent enzymes are often oxygen sensitive, variants of this halogenase could be screened in lysates under aerobic conditions. The developed biocatalysts offer a route to mild, late-stage chlorination on milligram scale of non-natural hapalindoles containing a ketone instead of an isonitrile functionality, thereby unlocking them for preparative biocatalysis.

Efficient and economic methods in directed evolution at the protein, metabolic, and genome level are needed for biocatalyst development and the success of synthetic biology. In contrast to random strategies, semirational approaches such as saturation mutagenesis explore the sequence space in a focused manner. Although several combinatorial libraries based on saturation mutagenesis have been reported using solid-phase gene synthesis, direct comparison with traditional PCR-based methods is currently lacking. In this work, we compare combinatorial protein libraries created in-house via PCR versus those generated by commercial solid-phase gene synthesis. Using descriptive statistics and probabilistic distributions on amino acid occurrence frequencies, the quality of the libraries was assessed and compared, revealing that the outsourced libraries are characterized by less bias and outliers than the PCR-based ones. Afterward, we screened all libraries following a traditional algorithm for almost complete library coverage and compared this approach with an emergent statistical concept suggesting screening a lower portion of the protein sequence space. Upon analyzing the biocatalytic landscapes and best hits of all combinatorial libraries, we show that the screening effort could have been reduced in all cases by more than 50%, while still finding at least one of the best mutants.