Engineering Plant Cytochrome P450s for Enhanced Synthesis of Natural Products: Past Achievements and Future Perspectives

Abstract

Cytochrome P450s (P450s) are the most versatile catalysts and are widely used by plants to synthesize a vast array of structurally diverse specialized metabolites that not only play essential ecological roles but also constitute a valuable resource for the development of new drugs. To accelerate the metabolic engineering of these high-value metabolites, it is imperative to identify and characterize pathway P450s, and to further improve their activities through protein engineering. In this review, we focus on P450 engineering and summarize the major strategies for enhancing the stability and activity of P450s and successful cases of P450 engineering. Studies in which the functions of P450s were altered to create de novo metabolic pathways or novel compounds are discussed as well. We also overview emerging tools, specifically DNA synthesis, machine learning, and de novo protein design, as well as the evolutionary patterns of P450s unveiled from a massive number of DNA sequences that could be integrated to accelerate the engineering of these enzymes. These approaches would greatly aid in the exploitation of plant-specialized metabolites or derivatives for various uses including medical applications.

Keywords: amino acid co-evolution; plant cytochrome P450; plant natural product; protein design; protein engineering.

Figure 1

Engineering of P450s to Create Novel Reactions and their Potential Applications. The missing steps within the taxol metabolic pathway are indicated with a dotted line, and P450s are suggested to be involved in these steps. Two pathway P450s, T5αH and T10βH, which have substrates similar to those of the missing P450s (indicated with red backbone), are suitable for engineering to acquire new activities that can fill the pathway gap.

Figure 2

The Potential Applicaitons of P450 Engineering. Successful cases of engineering P450s with improved enzyme–host compatibility (A) and activities (B), and the ability to catalyze abiotic reactions creating de novo metabolites (C) or metabolic pathways (D). Cuol, cucurbitadienol; C, carbonyl; H, hydroxyl.

Figure 3

A Proposed Workflow for Future Plant P450 Engineering Using Integrated Tools. Metagenome data are explored using machine learning to identify co-evolving amino acids in P450s. This information is adopted by Rosetta structure prediction to generate highly accurate P450 models. Sequences of protein variants generated from direct evolution or rational design are further optimized by Rosetta to improve protein stability, and then directly synthesized for subsequent sensor-assisted screening. Iterative optimization may be needed to acquire desired properties for P450s.