Cultured bacteria have been a very rewarding source of biologically active small molecules. However, the majority of the global microbiome remains recalcitrant to laboratory culture and so their vast repertoire of natural products has remained inaccessible. To address this challenge, we have spearheaded the development of culture-independent approaches for accessing environmentally encoded natural products. In our metagenomic discovery pipelines, the need to culture bacteria is circumvented by extracting DNA directly from environmental samples and cloning it into an easily cultured bacterium where it is interrogated for the potential to encode novel bioactive small molecules. While our metagenomic discovery methods can be applied to the identification of compounds to combat almost any disease, we have largely focused on the discovery of antibiotics due to a critical unmet medical need. In addition to using our existing metagenomic discovery pipelines to identify novel bioactive natural products, we also continue to develop improved tools for metagenomic discovery. This includes an interest in developing improved methods for cloning DNA from environmental samples, identifying clones of interest, retrieving target clones, and generating better heterologous hosts for natural product expression.
Existing methods for identifying bioactive natural products rely on biological processes to convert the genetic instructions contained in bacterial genomes into novel small molecules. Unfortunately, it is often impossible to coax laboratory grown bacteria into producing all of the different natural products they encode. We have therefore developed a “biology free” discovery approach where, instead of decoding genetic instructions using biological processes, bioinformatic algorithms are used to predict the chemical structures produced by bacteria followed by chemical synthesis of the predicted structures. We have called the resulting molecules Synthetic Bioinformatic Natural Products (syn-BNPs). As seen with traditional natural products, we have found that syn-BNPs are a very rewarding source of small molecules with novel or rarely observed modes of action and potent in vivo activity. We are currently working to expand the application of a syn-BNP approach to a more diverse group of biosynthetic gene clusters and bioactivities.
A natural extension of our exploration of environmental bacteria is the exploration of human commensal bacteria. Although the human microbiome is believed to play an important role in human health and disease, the mechanisms by which human associated bacteria affect host physiology remain poorly understood. In other ecosystems, bacteria are known to rely heavily on small molecules to interact with their environment and so it is likely that the human microbiota also relies heavily on small molecules to communicate with its human host. To date, only a limited number of small molecules that affect the human host through specific cell receptors have been characterized. Addressing this gap in our mechanistic-level characterization of host-microbial interactions is critical to understanding the role the microbiota plays in human health and disease. We are using direct functional screening against therapeutically relevant mammalian receptor families to identify small molecule ligands that are potentially involved in the host-microbial interactions. Animal models are then used to test the hypotheses suggested by these receptor ligand interactions.