QSAR-based drug discovery of 2-((4-Imino-3,4-dihydroquinazolin-2-yl)thio-substituted analogs targeting Mycobacterium...

Dihydroquinazolin-4(3H)-imines have emerged as a promising scaffold for developing novel antimycobacterial agents. Here, we integrated QSAR modeling, in vitro screening, molecular dynamics (MD), binding free-energy calculations, and in vivo toxicity assessment to identify potent inhibitors against Mycobacterium tuberculosis (Mtb). Four QSAR models demonstrated strong internal and external reliability (R² > 0.68; Q²_LOO > 0.62; R²_extup to 0.82). Additional validation parameters met recommended thresholds (Q²-F1/F2/F3 ≥ 0.72; CCC_ext ≥ 0.88; r²_{m aver} ≥ 0.72; r²_{m delta}: 0.04), and Williams plots confirmed that all predictions fell within the applicability domain (h*: 0.286). Biological assays identified 11 active compounds, with MIC values ranging from 25 to 200 μM. Compound 2c (p-methylphenyl derivative) and 3d (m, p-dichlorophenyl analog) were the most potent, displaying MIC values of 50 and 25 μM, respectively. MD simulations revealed stable and specific interactions with Eis, an acetyltransferase linked to kanamycin resistance. Compound 2c exhibited a mean ΔG_bind of -52.19 ± 3.21 kcal/mol, while 3d showed a more favorable ΔG_bind of -73.15 ± 3.16 kcal/mol, consistent with its superior in vitro potency. Distinct interaction profiles-especially the engagement of Tyr¹²⁶ and hydrophobic clusters-help explain their differential affinities. Moreover, both leads demonstrated low to moderate in vitro cytotoxicity against HepG2 cells at the concentrations evaluated. In vivo acute toxicity in Zophobas morio indicated LD₅₀ values of 500 mg/kg for 2c and 100 mg/kg for 3d, with transient tremors and melanization observed only for 3d. Since compound 2c exhibited a safer in vivo toxicity profile, this compound was investigated for its association with antimicrobial drugs. These compounds were validated as promising anti-TB candidates, supported by robust QSAR predictivity, favorable binding energetics, and measurable in vitro and in vivo toxicity profiles, reinforcing their potentials for further optimization.