Application of In Silico QSAR and Molecular Docking Studies to a Series of Xanthine-Based Analogues and Design, Synthesis and Pharmacological Evaluation of Identified New Potential Selective MAO-B Inhibitors

Background/Objectives: Methylxanthines, such as caffeine, exhibit neuroprotective properties in neurodegenerative conditions, partly linked to modulation of monoamine oxidase B (MAO-B) and oxidative stress pathways. This work aimed to design, synthesize and functionally characterize new caffeine-8-methylthioglycolic acid derivatives as selective MAO-B inhibitors with neuroprotective potential. Methods: A QSAR model was built on 94 studies of xanthine derivatives to guide the design of ten new semi- and thiosemicarbazides (Jas1–Jas10), followed by molecular docking to human MAO-B (PDB: 2V5Z) using Glide, GOLD and MM-GBSA binding free energy calculations. The target compounds were synthesized in relatively high yields, structurally confirmed by spectroscopic methods and tested in vitro for hMAO-A/B inhibition, as well as for neurotoxicity and neuroprotection in isolated mouse brain synaptosomes, mitochondria and microsomes under 6-hydroxydopamine (6-OHDA), tert-butyl hydroperoxide (t-BuOOH) and Fe/ascorbate (Fe2+/AA)-induced oxidative stress. Results: Docking and MM-GBSA identified Jas6 and Jas7 as the most stable MAO-B binders, with binding free energies approaching those of safinamide. All derivatives inhibited hMAO-A and hMAO-B in the submicromolar range, with Jas2 and Jas6 showing the highest MAO-B selectivity indices. At 100 µM, the series produced mild but significant pro-oxidant and cytotoxic effects when applied alone, yet under oxidative stress all compounds, especially Jas2 and Jas6, markedly preserved synaptosomal and mitochondrial viability, maintained glutathione levels, and reduced malondialdehyde production. Conclusions: The caffeine-based semi- and thiosemicarbazides, particularly Jas2 and Jas6, emerge as promising selective MAO-B inhibitors with pronounced antioxidant and neuroprotective activity, supporting their further optimization as multitarget candidates for neurodegenerative disorders such as Parkinson’s disease.