Bacteria are prolific producers of a group of small molecules known as secondary metabolites that serve as a major source of drugs and drug candidates. Penicillin and erythromycin are but two examples of bacterial secondary metabolites that, since their discoveries, have been invaluable in a clinical setting. In addition to therapeutic applications, the complex structures of secondary metabolites serve as a source of inspiration for synthetic chemists, who attempt to recreate these molecules in the laboratory, and for biochemists, who investigate Nature’s biosynthetic strategies for concocting them. Thus, bacterially-produced small molecules provide a common template for discovering potential therapeutics, finding novel, (bio)synthetically challenging structural scaffolds, investigating their physiological roles, understanding Nature’s biosynthetic repertoire, as well as elucidating the corresponding enzymatic reaction mechanisms. We are engaged in examining and understanding these facets underlying bacterially-generated secondary metabolites. Below are examples of ongoing projects in our group.
HiTES: A New Method for Activating Silent Biosynthetic Gene Clusters
Secondary metabolites are assembled by sets of contiguous genes called biosynthetic gene clusters (BGCs). The enzymes that they encode generate these often complex molecules in a step-wise fashion from simple building blocks. The beta-lactam penicillin, for example, is synthesized from three amino acids that are condensed into a tripeptide and further tailored to give the final product. Recent genome sequencing efforts have revealed that the vast majority of BGCs that can be identified bioinformatically in bacterial genomes are not, or only weakly, expressed under typical growth conditions. These so-called ‘silent’ or ‘cryptic’ BGCs represent a large and hidden reservoir of new and potentially useful small molecules. Methods that reliably induce their production would have a profound impact on natural products discovery. To investigate the products of silent gene clusters, we recently implemented a method we refer to as HiTES (high-throughput elicitor screening). In this approach, a reporter gene is inserted into the silent gene cluster of interest, allowing a rapid read-out of its expression. Small molecule libraries are then screened to identify small molecule elicitors (Fig. 1). With elicitors identified, the product of the gene cluster and the regulatory pathways that induce it can be elucidated. Our application of HiTES in Gram-negative and Gram-positive bacteria has surprisingly shown that antibiotics are the most effective elicitors of silent BGCs, suggesting they play a role in modulating secondary metabolism in bacteria, and that old antibiotics can be used to find new, cryptic ones. We are currently investigating the regulatory circuits that control this phenomenon and expanding the scope of HiTES to new bacterial species.
Microbial Symbiotic Interactions
An alternative strategy for the discovery of secondary metabolites that we are currently pursuing relies not on targeting biosynthetic gene clusters, as elaborated above, but rather on the physiological roles of these molecules. Because bacteria communicate using secondary metabolites, ‘listening in’ on these conversations provides an attractive search strategy. We are investigating a number of microbial interactions and the small molecules that mediate them. In one case, a naturally-occurring and wide-spread algal-bacterial symbiosis, we have discovered a novel family of small molecules, the roseobacticides, and a biphasic mode of interaction involving a mutualistic and a parasitic phase (Fig. 2). Under mutualistic conditions, the bacteria and algae exchange molecules beneficial to the symbiotic partner. However, when the algae begin to senesce, the bacteria produce the algaecidal roseobacticides, which kills the algal host. We have recently found that two molecules in this symbiosis are synthesized largely by the same biosynthetic gene cluster, the first example one gene cluster generating two different metabolites. These metabolites, tropodithietic acid and roseobacticides have different structures, biological activities, and are produced in different phases of the symbiotic interaction. Production of roseobacticides requires an unusual biosynthetic strategy, which we are examining further. Moreover, much like in the HiTES project described above, we are also addressing the regulatory pathways that trigger roseobacticide biosynthesis in response to the algal metabolite pCA (Fig. 2). Lastly, we are still finding new molecules that mediate the algal-bacterial association. Even though the Roseobacter are weak secondary metabolite producers by bioinformatic standards (that is, they have few recognizable biosynthetic genes), we have continually found new natural products using our knowledge regarding their ecological interactions.
Novel Biosynthetic & Enzymatic Chemistries
The remarkable structures of secondary metabolites provide opportunities for examining their exotic biosynthetic pathways and discovering novel enzyme-catalyzed transformations. The production of secondary metabolites may be grossly divided into two phases. The first involves synthesis of the scaffold or backbone, and the second the installation of unique, pathway-specific alterations. While the canonical mechanisms for generating the peptide, polyketide, and terpene backbones of secondary metabolites have been largely elucidated, the tailoring enzymes that provide unique functionalities have received less attention. Among these, metalloenzymes introduce often unusual, functionally essential, and mechanistically puzzling modifications. We are interested in elucidating the detailed mechanisms of these unusual enzyme-catalyzed transformations. We recently reported one such example: structural elucidation of streptide revealed an unprecedented post-translational modification, a carbon-carbon crosslink at unactivated positions between the side-chains of Lys and Trp. We subsequently showed that a radical SAM enzyme, StrB, installs this modification in a single step (Fig. 3). Other novel modifications introduced by radical SAM enzymes are currently being investigated in the group. We have also begun to examine the aromatic crosslinks that are a structural hallmark of the glycopeptide antibiotics, notably the antibiotic-of-last-resort vancomycin. We recently reconstituted the in vitro activity of a cytochrome P450 enzyme, OxyA, which introduces the second aryl ether bond in vancomycin. Studies addressing the mechanism of this transformation as well as the installation of the third biaryl crosslink in vancomycin are currently underway.