Abstract

Antiplasmodial nortriterpenes with 3,4-seco-27-norlanostane skeletons, almost entirely obtained from fruiting bodies, represent the main evidential source for bioactive secondary metabolites derived from a relatively unexplored phytopathogenic fungus, Ganoderma boninense. Currently lacking is convincing evidence for antimicrobial secondary metabolites in this pathogen, excluding that obtained from commonly observed phytochemicals in the plants. Herein, we aimed to demonstrate an efficient analytical approach for the production of antibacterial secondary metabolites using the mycelial extract of G. boninense. Three experimental cultures were prepared from fruiting bodies (GBFB), mycelium cultured on potato dextrose agar (PDA) media (GBMA), and liquid broth (GBMB). Through solvent extraction, culture type-dependent phytochemical distributions were diversely exhibited. Water-extracted GBMB produced the highest yield (31.21 ± 0.61%, p < 0.05), but both GBFB and GBMA elicited remarkably higher yields than GBMB when polar-organic solvent extraction was employed. Greater quantities of phytochemicals were also obtained from GBFB and GBMA, in sharp contrast to those gleaned from GBMB. However, the highest antibacterial activity was observed in chloroform-extracted GBMA against all tested bacteria. From liquid-liquid extractions (LLE), it was seen that mycelia extraction with combined chloroform-methanol-water at a ratio of 1:1:1 was superior at detecting antibacterial activities with the most significant quantities of antibacterial compounds. The data demonstrate a novel means of assessing antibacterial compounds with mycelia by LLE which avoids the shortcomings of standardized
method
ologies. Additionally, the antibacterial extract from the mycelia demonstrate that previously unknown bioactive secondary metabolites of the less studied subsets of Ganoderma may serve as active and potent antimicrobial compounds.

• Keywords: deep learning;machine learning;microbiome;semi-supervised;supervised;unsupervised