Science 348, 1106–1112 (2015)

Credit: SCIENCE

Natural products derived from bacteria have served as a rich source for antibacterial lead compounds. For instance, griselimycin (GM) is a cyclic peptide isolated from Streptomyces strains that has antibacterial activity against Mycobacterium species, such as the tuberculosis-causing Mycobacterium tuberculosis (Mtb), including strains that are resistant to known drugs. Given the need to identify therapeutics for Mtb infection, Kling et al. focused on generating GM analogs with improved potency, metabolic stability and safety profile. They first defined a total synthesis route to GM and, in this process, identified cyclohexyl-griselimycin (CGM) as having similar activity against a range of Mtb strains in vitro, including strains that are mono-resistant to first- and second-line antituberculosis drugs. CGM had good ADME (drug-like) properties and was bactericidal in both acute and chronic tuberculosis (TB) mouse models of infection. Treatment with a combination of CGM, rifampicin and pyrazinamide was more effective in the acute model than the commonly used drug combination of rifampicin, pyrazinamide and isoniazid, suggesting a novel mechanism of action for CGM. To determine the target of CGM, the authors first explored a Streptomyces GM biosynthetic gene cluster and found a homolog of dnaN, which encodes the sliding clamp of DNA polymerase. Overexpressing dnaN increased resistance to GM but not to any other anti-TB drug. Likewise, GM-resistant strains isolated from Mycobacterim smegmatis and from an Mtb in vivo infection model had amplified DnaN expression. In vitro binding assays confirmed binding of GM and CGM to DnaN from M. smegmatis, Mtb and E. coli, and co-crystal structures helped define the binding sites. The authors also observed induction of the SOS response, indicative of DNA strand breaks in GM-exposed M. smegmatis, suggesting that clamp action in replication is inhibited by GM in cells. Even though resistance occurred with CGM, this was accompanied by considerable fitness costs, suggesting that targeting dnaN would be an effective TB treatment.