High-Throughput Site-Directed Mutagenesis

Enzyme and Pathway Engineering for Improved Betanin Production in Saccharomyces cerevisiae

Betanin is a water-soluble red-violet pigment belonging to the betacyanins family. It has become more and more attractive for its natural food colorant properties and health benefits. However, the commercial production of betanin, typically extracted from red beetroot, faces economic and sustainability challenges. Microbial heterologous production therefore offers a promising alternative. Here, we performed combinatorial engineering of plant P450 enzymes and precursor metabolisms to improve the de novo production of betanin in Saccharomyces cerevisiae. Semirational design by computer simulation and molecular docking was used to improve the catalytic activity of CYP76AD. Alanine substitution and site-directed saturation mutants were screened, with a combination mutant showing an approximately 7-fold increase in betanin titer compared to the wild type. Subsequently, betanin production was improved by enhancing the l-tyrosine pathway flux and UDP-glucose supply. Finally, after optimization of the fermentation process, the engineered strain BEW10 produced 134.1 mg/L of betanin from sucrose, achieving the highest reported titer of betanin in a shake flask by microbes. This work shows the P450 enzyme and metabolic engineering strategies for the efficient microbial production of natural complex products.

Cluster learning-assisted directed evolution

Directed evolution, a strategy for protein engineering, optimizes protein properties (i.e., fitness) by expensive and time-consuming screening or selection of large mutational sequence space. Machine learning-assisted directed evolution (MLDE), which screens sequence properties in silico, can accelerate the optimization and reduce the experimental burden. This work introduces a MLDE framework, cluster learning-assisted directed evolution (CLADE), that combines hierarchical unsupervised clustering sampling and supervised learning to guide protein engineering. The clustering sampling selectively picks and screens variants in targeted subspaces, which guides the subsequent generation of diverse training sets. In the last stage, accurate predictions via supervised learning models improve final outcomes. By sequentially screening 480 sequences out of 160,000 in a four-site combinatorial library with five equal experimental batches, CLADE achieves the global maximal fitness hit rate up to 91.0% and 34.0% for GB1 and PhoQ datasets, respectively, improved from 18.6% and 7.2% obtained by random-sampling-based MLDE.