Target based discovery of next generation pyrazinamide

结核病是艾滋病患者死亡的一个主要原因。在有免疫能力的人群中，药物

Tuberculosis (TB) is a main cause of death for people living with AIDS. In immune-competent populations, drug therapy and immunity join forces to win the fight against Mycobacterium tuberculosis (Mtb). In HIV-positive individuals on the other hand, chemotherapy must fully sterilize all infection sites. There is an urgent medical need to develop more potent `sterilizing' drugs. The inclusion of pyrazinamide (PZA) into the TB regimen allowed reduction of treatment time to six months, while maintaining low relapse rates. The pharmacological basis for PZA's remarkable sterilizing activity in patients remains obscure, considering the drug's poor in vitro potency (MIC = 30-100 µg/mL). In the rabbit model of active TB, an animal model that recapitulates the various lung lesion types observed in human TB disease, we showed that PZA not only penetrates caseous necrotic lesions but also sterilizes them. In an ex vivo assay using caseum from infected rabbits, we showed that PZA kills non-growing, drug tolerant Mtb. These findings provide an explanation for the clinically observed treatment shortening effect of PZA: the drug reaches difficult-to-penetrate TB lesions and kills recalcitrant `persister' Mtb. However, consistent with the modest in vitro potency of PZA, onset of lesion sterilization is slow and concentrations required to kill Mtb in ex vivo caseum are high. Based on these findings, the logical way forward is to improve the potency of PZA while maintaining its unique lesion penetration and sterilization properties. To enable the rational optimization of PZA, we identified aspartate decarboxylase PanD, required for coenzymeA biosynthesis, as a the first genetically, biochemically and biophysically validated target of PZA. Consistent with poor whole cell activities of PZA, affinity of the drug (more specifically its bioactive component pyrazinoic acid, POA) for PanD was in the µM range, confirming room for improvement. Interestingly, mechanism of action studies revealed a novel antibacterial on-target mechanism whereby, rather than inhibiting PanD's catalytic activity, binding of POA to PanD appears to trigger degradation of the protein by the caseinolytic protease ClpC1. Here, we propose to build on our discoveries and 1. fully characterize the novel on-target mechanism by which the drug induces degradation of its target, 2. exploit in vitro pharmacological `sterilizing' models and PanD-based assays for the discovery of novel PZA analogs with improved potency and sterilizing activity, and 3. characterize lesion specific growth and replication status of PZA resistant panD mutant Mtb vs. wild type Mtb in the TB rabbit model with and without PZA treatment. In summary, we have identified key pharmacological properties responsible for the treatment shortening activity of PZA in clinical settings, as well as the molecular target of the drug, and we propose a `first-in-class to best-in- class' program to exploit these findings and rationally design the `next generation' PZA with improved sterilizing activity.