To avoid entering into a post-antibiotic era, we need to develop antibiotics that function through novel strategies. One such strategy is to target enzymes involved in the production of public goods – secondary metabolites produced by individual bacterium that increase the collective fitness of the entire colony. One such public good are the siderophores, peptide-like small molecules that facilitate acquisition of iron from the host. Herein, we report efforts to optimize 5’-O-[N-(salicyl)sulfamoyl]adenosine), salicyl-AMS, a biochemical inhibitor of the M. tuberculosis adenylate-forming enzyme MbtA, which inhibits the production of the mycobactin siderophores. Salicyl-AMS has nanomolar biochemical inhibitory activity, sub-micromolar antimicrobial activity, and modest in vivo efficacy. Nevertheless, it suffers from rapid clearance, low oral bioavailability, and modest mycobacterial wall permeability. We determined the observed dose-limiting toxicity was caused by a trace quantity of AMS as a synthetic impurity. We employed structure-toxicity relationship studies to elucidate the mechanism of toxicity, and de-risked a series of second-generation salicyl-AMS analogues. Furthermore, we developed novel salicyl-AMS analogues with superior physicochemical properties designed to pharmacologically optimize members of this inhibitor class.