LAUSR

products with antimicrobial properties has been hampered by the fact that many bacteria cannot be cultured in the laboratory, and those that can do not produce all of their natural products in this setting. Hover et al. have devised a culture-free method for the discovery of bacterial natural products and used it to identify malacidins (metagenomic acidic lipopeptide antibiotic-cidins) as a distinct class of antibiotics. Calcium-dependent antibiotics contain a conserved Asp-X-Asp-Gly calcium-binding motif and are biosynthesized by non-ribosomal peptide synthetases (NRPSs). To search for additional antibiotics containing calcium-binding motifs, the authors screened environmental DNA (eDNA) from >2,000 soil samples by PCR with primers targeting NRPS adenylation domains, followed by next-generation sequencing. 75% of sequenced soils had natural product sequence tags (NPSTs) that mapped to an adenylation domain from a known calcium-dependent antibiotic biosynthetic gene cluster (BGC). As only 13% of these NPSTs had a high nucleotide identity to adenylation domains in characterized calcium-dependent antibiotics, the majority of adenylation domaincontaining lipopeptides encoded by the global soil metagenome might be uncharacterized. By analysing their next-generation sequencing reads using a web-based tool called eSNaPD (environmental Surveyor of Natural Product Diversity), the authors created a phylogenetic tree derived from NRPS adenylation domain-associated NPSTs. One eDNA-specific clade that was not associated with known BGCs was found in 19% of metagenomes, and the authors proposed that the BGCs associated with these NPSTs represented an uncharacterized family of antibiotics, which they called malacidins. Next, the authors cloned DNA from a desert soil in the malacidin branch of their phylogenetic tree and recovered a complete malacidin BGC, which they expressed in Streptomyces albus J1074. Extracts from S. albus cultures were antibacterial towards multidrug-resistant Staphylococcus aureus (MRSA) and they contained clone-specific metabolites. Analysis of two of these metabolites by mass spectrometry and NMR spectroscopy revealed that malacidins are ten-membered cyclic lipopeptides that do not contain the Asp-X-Asp-Gly calcium-binding motif. Despite this, the antibacterial activity of malacidins towards MRSA was dependent on calcium. Indeed, the authors showed that malacidins kill MRSA by binding, in a calcium-dependent manner, to a downstream intermediate of a cell wall precursor, causing this cell wall precursor to accumulate. Further characterization of malacidins revealed that the topical administration of one of these compounds, malacidin A, eliminated MRSA from infected wounds on the skin of rats after 24 h. Furthermore, no malicidin A-resistant S. aureus clones were identified after 20 days in culture with sublethal levels of this antibiotic, suggesting that there is a high barrier to the emergence of antibiotic resistance, and malacidins were not toxic to mammalian cells in culture. In short, malacidins are a promising new class of calcium-dependent antibiotics. Their discovery validates the use of BGC mining for discovering natural products.