Since α-mangostin in mangosteen fruits was reported to be the main compound able to provide natural antioxidants, the microwave-assisted extraction process to obtain high-quality α-mangostin from mangosteen pericarp (Garcinia mangostana L.) was optimized using a central composite design and response surface methodology. The parameters examined included extraction time, microwave power, and solvent percentage. The antioxidant and antimicrobial activity of optimized and non-optimized extracts was evaluated. Ethyl acetate as a green solvent exhibited the highest concentration of α-mangostin, followed by dichloromethane, ethanol, and water. The highest α-mangostin concentration in mangosteen pericarp of 121.01 mg/g dry matter (DM) was predicted at 3.16 min, 189.20 W, and 72.40% (v/v). The verification of experimental results under these optimized conditions showed that the α-mangostin value for the mangosteen pericarp was 120.68 mg/g DM. The predicted models were successfully developed to extract α-mangostin from the mangosteen pericarp. No significant differences were observed between the predicted and the experimental α-mangostin values, indicating that the developed models are accurate. The analysis of the extracts for secondary metabolites showed that the total phenolic content (TPC) and total flavonoid content (TFC) increased significantly in the optimized extracts (OE) compared to the non-optimized extracts (NOE). Additionally, trans-ferulic acid and catechin were abundant among the compounds identified. In addition, the optimized extract of mangosteen pericarp with its higher α-mangostin and secondary metabolite concentrations exhibited higher antioxidant activities with half maximal inhibitory concentration (IC50) values of 20.64 µg/mL compared to those of the NOE (28.50 µg/mL). The OE exhibited the highest antibacterial activity, particularly against Gram-positive bacteria. In this study, the microwave-assisted extraction process of α-mangostin from mangosteen pericarp was successfully optimized, indicating the accuracy of the models developed, which will be usable in a larger-scale extraction process.
* Title and MeSH Headings from MEDLINE®/PubMed®, a database of the U.S. National Library of Medicine.