Structure activity relationship of pyrazinoic acid analogs as potential antimycobacterial agents

Abstract

Tuberculosis (TB) remains a leading cause of infectious disease-related mortality and morbidity. Pyrazinamide (PZA) is a critical component of the first-line TB treatment regimen because of its sterilizing activity against non-replicating Mycobacterium tuberculosis (Mtb), but its mechanism of action has remained enigmatic. PZA is a prodrug converted by pyrazinamidase encoded by pncA within Mtb to the active moiety, pyrazinoic acid (POA) and PZA resistance is caused by loss-of-function mutations to pyrazinamidase. We have recently shown that POA induces targeted protein degradation of the enzyme PanD, a crucial component of the coenzyme A biosynthetic pathway essential in Mtb. Based on the newly identified mechanism of action of POA, along with the crystal structure of PanD bound to POA, we designed several POA analogs using structure for interpretation to improve potency and overcome PZA resistance. We prepared and tested ring and carboxylic acid bioisosteres as well as 3, 5, 6 substitutions on the ring to study the structure activity relationships of the POA scaffold. All the analogs were evaluated for their whole cell antimycobacterial activity, and a few representative molecules were evaluated for their binding affinity, towards PanD, through isothermal titration calorimetry. We report that analogs with ring and carboxylic acid bioisosteres did not significantly enhance the antimicrobial activity, whereas the alkylamino-group substitutions at the 3 and 5 position of POA were found to be up to 5 to 10-fold more potent than POA. Further development and mechanistic analysis of these analogs may lead to a next generation POA analog for treating TB.

Keywords: Pyrazinoic acid; Tuberculosis; pyrazinamide.

Figure. 1.

POA binds to PanD and activates a degradation tag, causing degradation of PanD by the ClpC1-ClpP caseinolytic protease complex.

Figure 2.

Proposed scaffold modifications to POA

Figure 3:

Sequential design of POA analogs.

Figure 4:

Cartoon model of PanD and POA. A) Cartoon Model showing the SAR points on POA-PanD ligand complex. B) Cartoon model revealing the tetrameric arrangement of Mtb PanD crystal structure (PDB 6P02). The zoomed subset highlights the probable entry of 6Cl-POA from the central tunnel lined by amino acids Q43 (spheres), while the overlay of C-terminal loop residues (highlighted in cyan spheres), from full length AlphaFold Mtb PanD homology model) show that C-terminal residues close the solvent accessibility to the ligand binding sites.

Figure 5.

Cytotoxicity of PZA, POA and analogs 23, 23, 54, 55 and 60. Vero cells grown to confluence were exposed to compounds at various concentrations (20 μM to 1.5 mM) and cell viability was monitored as an indicator of cytotoxicity. A) Cell viability at the highest concentration tested (1.5 mM) for the reported compounds. B) Percent cytotoxicity at varying drug concentrations.

Figure 6:

Correlation of antimicrobial activity with ITC binding

Scheme 1:

A and B) Synthesis of carboxylic acid bioisosteres; C) Synthetic scheme for aryl coupling

Scheme 2:

A) Synthetic scheme for amine-substituted analogs; B) Synthetic scheme for amide-substituted analogs; C) Synthetic scheme for 6-amido analogs.