Discovery of natural CdnP inhibitors through structure-based virtual screening and molecular dynamics simulations

Tuberculosis, caused by Mycobacterium tuberculosis, remains one of the most lethal infectious diseases both historically and in the post-coronavirus disease 2019 era. CdnP (Rv2837c) functions as a bifunctional oligoribonuclease and 3'(2')-phosphoadenosine 5'-phosphate-phosphatase that impedes host immune responses by recognizing Bacterial cyclic dinucleotides such as c-di-AMP, which serve as pathogen-associated molecular patterns. Despite its significance, strategies targeting CdnP remain limited. Through high-throughput virtual screening and enzymatic assays, we identified four natural product inhibitors: one coumarin derivative (macrosporusone A) and three flavonoid glucosides (ligustroflavone, rhoifolin, and neodiosmin). Surface plasmon resonance measurements confirmed direct binding of these compounds to CdnP with nanomolar to micromolar affinities. Molecular dynamics simulations elucidated a dual inhibitory mechanism wherein these compounds competitively occupy the product (AMP)-binding site while simultaneously constraining conformational plasticity of the substrate-binding domain. Evolutionary analysis demonstrated that these inhibitors exhibit broad-spectrum activity against Bacterial CdnP orthologs while showing minimal inhibition of host-derived 2',3'-cGAMP-specific phosphodiesterases, suggesting favorable selectivity. Notably, ligustroflavone exhibited superior inhibitory potency. In contrast, FDA-approved phosphodiesterase inhibitors showed poor activity against Bacterial orthologs. These findings provide a foundation for developing novel host-directed therapeutics against tuberculosis that could potentially enhance stimulator of interferon genes (STING)-mediated immune responses without exerting selective pressure for antimicrobial resistance.

Importance: Tuberculosis (TB) remains a leading cause of mortality worldwide, with drug resistance posing a significant challenge to global control efforts. This study represents a major contribution to the field by identifying novel natural product inhibitors targeting CdnP (Rv2837c), a c-di-AMP-specific phosphodiesterase critical for Mycobacterium tuberculosis pathogenesis. The significance of this work lies in its innovative approach to TB therapy by perturbing Bacterial nucleotide signaling pathways rather than directly inhibiting Bacterial growth. By selectively targeting Bacterial CdnP while avoiding host phosphodiesterases, these compounds-particularly ligustroflavone and Other flavonoid glucosides-offer a promising foundation for developing host-directed therapeutics with potentially reduced selective pressure for antimicrobial resistance. Furthermore, the detailed structural insights and inhibitory mechanisms elucidated through molecular dynamics simulations provide valuable knowledge for rational drug design. This research bridges natural product discovery with computational biology to address the urgent need for novel TB treatments, especially against drug-resistant strains, presenting a significant advancement toward more effective therapeutic interventions for this persistent global health threat.