Discovery of a Phosphonoacetic Acid Derived Natural Product by Pathway Refactoring.

The activation of silent natural product gene clusters is a synthetic biology problem of great interest. As the rate at which gene clusters are identified outpaces the discovery rate of new molecules, this unknown chemical space is rapidly growing, as too are the rewards for developing technologies to exploit it. One class of natural products that has been underrepresented is phosphonic acids, which have important medical and agricultural uses. Hundreds of phosphonic acid biosynthetic gene clusters have been identified encoding for unknown molecules. Although methods exist to elicit secondary metabolite gene clusters in native hosts, they require the strain to be amenable to genetic manipulation. One method to circumvent this is pathway refactoring, which we implemented in an effort to discover new phosphonic acids from a gene cluster from Streptomyces sp. strain NRRL F-525. By reengineering this cluster for expression in the production host Streptomyces lividans, utility of refactoring is demonstrated with the isolation of a novel phosphonic acid, O-phosphonoacetic acid serine, and the characterization of its biosynthesis. In addition, a new biosynthetic branch point is identified with a phosphonoacetaldehyde dehydrogenase, which was used to identify additional phosphonic acid gene clusters that share phosphonoacetic acid as an intermediate.

Combinatorial pathway engineering for optimized production of the anti-malarial FR900098.

As resistance to current anti-malarial therapeutics spreads, new compounds to treat malaria are increasingly needed. One promising compound is FR900098, a naturally occurring phosphonate. Due to limitations in both chemical synthesis and biosynthetic methods for FR900098 production, this potential therapeutic has yet to see widespread implementation. Here we applied a combinatorial pathway engineering strategy to improve the production of FR900098 in Escherichia coli by modulating each of the pathway's nine genes with four promoters of different strengths. Due to the large size of the library and the low screening throughput, it was necessary to develop a novel screening strategy that significantly reduced the sample size needed to find an optimal strain. This was done by using biased libraries that localize searching around top hits and home in on high-producing strains. By incorporating this strategy, a significantly improved strain was found after screening less than 3% of the entire library. When coupled with culturing optimization, a strain was found to produce 96 mg/L, a 16-fold improvement over the original strain. We believe the enriched library method developed here can be used on other large pathways that may be difficult to engineer by combinatorial methods due to low screening throughput.