Abstract
Leaves of holy basil (Ocimum sanctum L., Lamiaceae) are widely used in traditional Indian medicine, and preparations based on them enter the international market. Raw materials are sourced primarily from Asian countries with subtropical and tropical climates. Available data indicate variation in the chemical composition of O. sanctum leaf essential oil depending on ecological and geographical growing conditions, which subsequently affects the quality of the medicinal product and its efficacy. Due to the variability of the chemical composition, specialists face challenges in standardization. Therefore, the aim of this study was to investigate the variability of the chemical composition of O. sanctum leaf essential oil with subsequent in silico prediction of pharmacological action. Samples of O. sanctum leaves of Indian origin entering the Russian market were analyzed. Essential oil was isolated by steam distillation according to the method prescribed by the State Pharmacopoeia of the Russian Federation (XIV edition) and foreign pharmacopoeias. GC-MS profiling of the essential oil revealed pronounced heterogeneity (28–55 constituents per sample). Samples from northeastern India (Varanasi) were rich in β-elemene, caryophyllene, and eugenol, while those from southern India (Hassan) contained high levels of methyl eugenol, highlighting distinct geographical chemotypes. Phenylpropanoids and non-oxygenated sesquiterpenes were predominant; the former group was more prevalent in southern oils and the latter in northeastern ones, reflecting region-specific biosynthesis. Comprehensive two-dimensional gas chromatography (GC × GC), used for in-depth analysis, identified four novel minor constituents, p-cymene, longiverbenone, (E)-α-bergamotol, and longipinene epoxide, thereby expanding the known phytochemical repertoire of the species. Due to limited experimental pharmacological data, in silico bioactivity screening was performed using PASS software. Integrated profile analysis revealed potential complementary effects against inflammation, atherosclerosis, cancer, and related conditions, undetectable via single-compound analysis. Screening of 1957 types of biological activity identified 92 relevant effects and mechanisms, demonstrating the utility of in silico modeling for narrowing the experimental search space and prioritizing future research.