(109d) Reductive loop swaps in polyketide synthases as a route to designer chemical products

Polyketide synthases (PKS) are a class of enzymes that behave similarly to fatty acid synthases. However, where fatty acid synthases fully reduce the b-carbonyl of the growing carbon chain iteratively, Type I modular PKSs generally vary the extent of reduction to create complex macromolecules using discrete, collinear enzymatic domains. In principle, this collinearity allows the engineer to produce “designer†chemicals in a predictable manner by combining various PKS domains. As biocatalysts, PKSs provide a route to commodity and specialty chemicals without a dependence on petrochemicals.

As b-carbonyl reduction is a fundamental component of PKS flexibility, we seek to explore the design principles in non-native reductive loop exchanges. Here, we introduced heterologous fully reductive domains with varied native acyltransferase (AT) selectivity and substrate size, from several PKS gene clusters, into the first extension module of lipomycin polyketide synthase (LipPks1). These engineered enzymes should programmatically produce a fully reduced short-chain branched fatty acid when fused with a heterologous thioesterase. In vivo screening in Streptomyces albus showed that substrate size compatibility, more than AT-selectivity, was critical to acid production, with engineered titers reaching 33.2 mg/L. Substrate size and AT-selectivity were further tested as critical design principles by identifying two more fully reductive loops with ClusterCAD, a computational platform for modular PKS design. These loops produced more fully reduced short-chain acids than reductive loops selected without these filters. This work strengthens the scientific literature regarding modular swapping in PKSs and brings us closer to a priori designed bioproducts.