Genomic and Metabolomic Investigation of a Novel Alaskan Soil Streptomycete

Modern medicine relies on effective antibiotics, but pathogens evolve resistances faster than new antimicrobials are commercialized. Most of the antibiotics currently in use are directly or indirectly derived from soil bacteria, specifically the actinomycetes, a bacterial phylum containing the prolific antibiotic-producing genus Streptomyces. Through the 40 years termed the "Golden Age" of antibiotic discovery, an average of 75 new antibiotics were discovered each year (1). Now the discovery rate has slowed to a standstill: only two antibacterial drugs for humans were approved by the FDA between 2008 and 2011 (2). Worse, only two new classes of antibiotics have been discovered since 1962 (3-5).;Screening bacteria isolated from environmental samples has fallen out of favor as a means for discovering new drugs, this is largely because the likelihood of finding a new compound among bioactive hits using the classic method of screening for activity against Gram-positive bacteria is believed to be only 0.2% (6). The bioinformatics-based alternative of using genomically identified targets in in vitro small molecule screens has also resulted in little progress (1, 3). Fruitful searches for new antibiotics will likely derive from combining traditional sources (environmental samples); whole-cell antimicrobial assays; and genetics, genomics, and biochemistry (1, 7).;We investigated the potential for antibiotic discovery with the Alaskan soil isolate, Streptomyces sp. 2AW, an organism that illustrates the unparalleled chemical virtuosity of the streptomycetes. Using bioassay-guided fractionation, highly sensitive mass spectrometry (MS), and nuclear magnetic resonance (NMR), we found that Streptomyces sp. 2AW simultaneously produces cycloheximide and two other biosynthetically unrelated antibiotics, neutramycin and hygromycin A.;The pathways for synthesis of neutramycin and cycloheximide are not known. This is surprising in the case of cycloheximide because of its 70-year history and widespread use. To identify the putative cycloheximide biosynthesis gene cluster, we sequenced the Streptomyces sp. 2AW genome, and based on genomic analysis, we propose a pathway for cycloheximide biosynthesis. We identified five open reading frames (ORFs) that encode a putative regulator and four putative biosynthesis enzymes, including a putative polyketide synthase (PKS) containing five modules. We propose that these enzymes catalyze the synthesis of a cyclohexenone intermediate through standard polyketide elongation starting with a non-traditional malonamic acid starter unit and featuring a non-standard Michael addition that results in glutarimide formation. Finally, we propose that the intermediate is reduced to cycloheximide by tailoring enzymes that are encoded by three clustered ORFs, which are not linked to the rest of the putative cycloheximide biosynthesis gene cluster.;We also identified the biosynthesis gene clusters likely responsible for the synthesis of neutramycin and hygromycin A. The pathway for neutramycin synthesis is not known, but a cluster responsible for hygromycin A synthesis was proposed in 2006 (8).;In addition to the clusters associated with cycloheximide, neutramycin, and hygromycin A, preliminary in silico analysis suggests that the Streptomcyes sp. 2AW genome contains other biosynthesis clusters. Cycloheximide, neutramycin, and hygromycin A are produced by other streptomycetes, who likely share these gene clusters. We determined through phylogenetic comparisons with 54 other streptomycetes that the production of shared metabolites does not correlate with the sequence similarity of six standard housekeeping genes. This suggests that horizontal gene transfer is the likely mechanism for the accumulation of these antibiotic gene clusters in Streptomyces sp. 2AW.;Our use of bioinformatics combined with structural chemistry to investigate the soil-borne Streptomyces sp. strain 2AW enabled the discovery of a strain that produces a constellation of three antibiotics that is not previously reported. The analysis also yielded two new proposed antibiotic biosynthesis pathways strongly supported by the bioinformatics, chemical analysis, and biochemical precedent. Future work will focus on the genetics of these biosynthesis clusters, which may yield tools for the synthesis of new antibiotics biochemically. Equally important will be to unlock the mysteries of biosynthesis regulation, which may enable approaches to induce the expression of new antibiotics in other environmental samples. We must utilize all the tools we can muster to address the humanitarian imperative of antibiotic discovery.