Targeted Discovery of β-Branched Conjugated Polyketides from Bacteria Based on Genomic and Metabolomic Hallmarks

A logical discovery method based on metabologenomic targeting was designed to search for new β-branched conjugated polyketides. PCR screening of a bacterial DNA library (1,638 strains) targeting the key enzyme hydroxymethylglutaryl-CoA synthase (HCS) homolog-encoding gene as a genomic hallmark identified 67 hit strains. Phylogenetic analysis of HCS homolog-encoding gene amplicons prioritized strains potentially producing new β-branched conjugated polyketides. Vibronic fine structures in ultraviolet spectra as a metabolomic hallmark clearly identified β-branched conjugated polyketides in bacterial extracts, logically leading to the discovery of paenillaene (<b>1</b>), paeniformicins A-D (<b>2</b>-<b>5</b>) from two strains belonging to the relatively unexplored genus <i>Paenibacillus</i>, and pulvomycins E-F (<b>6</b>-<b>7</b>) from a <i>Streptomyces</i> strain. Combined analysis of spectroscopic data, chemical reactions, quantum mechanics-based calculations, and biosynthetic gene clusters led to the structure elucidations of <b>1</b>-<b>7</b> as new β-branched conjugated polyketides. Paenillaene (<b>1</b>) dissociated amyloid-β aggregates, paeniformicin D (<b>5</b>) showed significant anti-tubercular activity against <i>Mycobacterium tuberculosis</i>, and pulvomycin E (<b>6</b>) displayed remarkable anti-proliferative activity against several human cancer cells, including a docetaxel-resistant breast cancer cell line. This metabologenomic targeting method provides an efficient tool for systematic discovery of new bioactive β-branched conjugated polyketides.