Tuberculosis (TB) is a disease caused by bacteria of the Mycobacterium genus, with Mycobacterium tuberculosis (Mtb) being the predominant species responsible for this condition [1,2]. Although TB is typically curable with an appropriate medication regimen, it remains the leading cause of death globally attributable to a single infectious agent [3]. According to the World Health Organization (WHO), over 10 million TB cases and 1.25 million deaths associated with the disease were reported in 2023. This global burden, however, is not uniformly distributed, with certain regions experiencing a disproportionate impact. Regions with the highest burden of TB include Southeast Asia (45 %), Africa (24 %), and the Western Pacific (17 %) [3].

Despite the availability of several drugs to treat the disease [4,5], the emergence of drug-resistant strains has posed a significant public health challenge. The main causes of resistance are attributed to genetic mutations, lack of adherence to the prescribed treatment, and the misuse of prescribed medications to treat the disease [6]. The WHO has identified multidrug-resistant (MDR) and extensively drug-resistant (XDR) TB strains, which exhibit resistance to several first- and second-line drugs employed in current treatment regimens. To address this issue, it is essential to identify new drugs and therapeutic targets that enable the development of more effective treatments.

The mycobacterial eukaryotic-like serine/threonine kinase, protein kinase A (PknA), is one of the 11 serine/threonine kinases encoded by the Mycobacterium tuberculosis genome [7]. PknA plays essential roles in cell wall synthesis, bacterial growth, and morphological changes during cell division [[8], [9], [10]]. Structurally, PknA is a protein composed of 431 amino acids that contains the residues Glu96, Leu97, and Val98 located within the hinge region, which are crucial for proper substrate binding. Additionally, it includes Met95 as a gatekeeper residue and Lys42 and Asp159 within the phosphate-binding region, both of which are important for the catalytic activity of the protein [7,9,11].

The literature reports that PknA expression increases within 48 h post-infection of Mtb in the lungs, making it one of the most highly expressed proteins by this mycobacterium during the first 30 days of infection [12,13]. Furthermore, the overexpression of PknA has been observed to cause structural malformations of the bacterial cell [8]. PknA has also been confirmed to be essential for ensuring the pathogenicity and survival of Mtb within host cells [9]. A study by Wang et al. [14] demonstrated the antibacterial activity of the reference compound CJJ300 through its inhibition of Mtb PknA. This ligand binds to the ATP-binding region of the protein via hydrogen bond interactions with residues Glu96 and Val98 (Fig. 1).

Although experimental data on compounds inhibiting PknA have been reported in recent studies [[14], [15], [16]], these compounds lack significant activity, highlighting the need to identify new potential inhibitors of this protein.

The use of in silico techniques for drug discovery has proven efficient in reducing analysis time and costs. Among these techniques, pharmacophore-based screenings are notable for their low computational demands and their ability to ensure the presence of key interactions within screened compounds. This approach has proven effective in identifying novel inhibitors against other Mtb kinases, as previously reported for PknG, for which our research group has experimentally validated the protocol [17]. Consequently, in this study, a pharmacophore (Ph4) model was developed to conduct a large-scale virtual screening of 1 581 625 compounds sourced from the commercial ChEMBL21, SelleckChem, and LifeChemicals databases (Fig. 2). The screening yielded 70 potential PknA inhibitors. These were subsequently subjected to docking simulations to confirm their binding modes within the active site and to ensure the presence of key interactions. Molecular docking analysis identified 12 compounds with a binding energy of −8.02 ± 0.40 kcal/mol, closely matching that of the reference compound CJJ300. Additional filtering steps, based on consensus docking results from three different software platforms, reduced this number to three molecules with potential PknA inhibitory activity that consistently exhibited the best binding energies. These three candidates underwent further evaluation, including binding free energy calculations (linear interaction energy and MM-GBSA) and dynamic undocking analyses. Among these, CHEMBL552033 stood out, exhibiting superior binding affinity compared to the reference compound. Additionally, ADMET and selectivity analyses were conducted to assess the viability of this compound as a potential inhibitor targeting PknA.