Chemistry ETDs

Publication Date

Summer 7-11-2017


Discovery of new natural products (NPs) is critical both for diseases treatment and crops protection. Numerous NP biosynthetic gene clusters (BGCs) in sequenced microbial genomes allow identification of new NPs through genome mining. Developing an integrated bioinformatic/experimental approach for discovering novel type II polyketides (PK-IIs) facilitates investigation of this family of NPs in an efficient, systematic way. Here, we developed an approach to analyze ketosynthase α/β (KSα/β) gene sequences to predict PK-II core structures, allowing us to target novel PK-II BGCs either from isolated genomic DNA or genomes from the NCBI databank, and to isolate novel PK-IIs produced by these BGCs.

First, new degenerate primers were designed to amplify the region containing key “fingerprint residues” used as predictive indicators of the KSα/β gene product novelty. This work resulted in identification of several BGCs encoding potentially novel PK-IIs in the genomes of 54 Actinobacteria, including 38 unique environmental strains. Next, complete PK-II BGCs were obtained through whole genome sequencing of 5 strains of high priority. Through combined core structure prediction and bioinformatic analysis, Alloactinosynnema sp. L-07 was chosen for compound isolation. By optimizing fermentation conditions, we purified 3.8 mg of sample to elucidate the structure of this compound, a pentangular PK-II which we named alloactinomicin.

In another route, we bioinformatically identified over 500 PK-II BGCs from the NCBI databank, and selected 28 of these predicted to produce structurally novel PK-IIs for experimental characterization by a combined genetic/metabolic approach. As a proof of concept, the CRISPR/Cas9-based genome editing was utilized to achieve KSα gene inactivation in two Streptomyces PK-II BGCs. Comparison of wild-type and mutant metabolite profiles led to identification of new putative PK-IIs and correlation of genotype to chemotype for the PK-II BGC being characterized. Finally, we purified one of the metabolites identified and obtained uv-visible and mass spectral evidences consistent with an angucycline-type PK-II, which we named flavochromycin.

This work demonstrates two routes for discovery of novel PK-IIs using genomics-driven bioinformatic/experimental approaches. Results from both routes have laid the foundation for more targeted and efficient ways to discover novel NPs for drug and agrochemical development.




Natural products, Type II polyketides, Actinobacteria, Bioinformatics, Genome Mining, Ketosynthase α/β

Document Type


Degree Name


Level of Degree


Department Name

Department of Chemistry and Chemical Biology

First Committee Member (Chair)

Jeremy S. Edwards

Second Committee Member

Charles E. Melançon III

Third Committee Member

Lina Cui

Fourth Committee Member

Changjian Feng