摘要：Abstract Fungal endophytes are a major source of anti-infective agents and other medically relevant compounds. However, their classical blinded-chemical investigation is a challenging process due to their highly complex chemical makeup. Thus, utilizing cheminformatics tools such as metabolomics and computer-aided modelling is of great help deal with such complexity and select the most probable bioactive candidates. In the present study, we have explored the fungal endophytes associated with the well-known antimalarial medicinal plant Artemisia annua for their production of further antimalarial agents. Based on the preliminary antimalarial screening of these endophytes and using LC-HRMS-based metabolomics and multivariate analyses, we suggested different potentially active metabolites (compounds 1–8 ). Further in silico investigation using the neural-network-based prediction software PASS led to the selection of a group of quinone derivatives (compounds 1–5) as the most possible active hits. Subsequent in vitro validation revealed emodin ( 1 ) and physcion ( 2 ) to be potent antimalarial candidates with IC 50 values of 0.9 and 1.9 µM, respectively. Our approach in the present investigation therefore can be applied as a preliminary evaluation step in the natural products drug discovery, which in turn can facilitate the isolation of selected metabolites notably the biologically active ones.
其他摘要：Abstract Fungal endophytes are a major source of anti-infective agents and other medically relevant compounds. However, their classical blinded-chemical investigation is a challenging process due to their highly complex chemical makeup. Thus, utilizing cheminformatics tools such as metabolomics and computer-aided modelling is of great help deal with such complexity and select the most probable bioactive candidates. In the present study, we have explored the fungal endophytes associated with the well-known antimalarial medicinal plant Artemisia annua for their production of further antimalarial agents. Based on the preliminary antimalarial screening of these endophytes and using LC-HRMS-based metabolomics and multivariate analyses, we suggested different potentially active metabolites (compounds 1–8 ). Further in silico investigation using the neural-network-based prediction software PASS led to the selection of a group of quinone derivatives (compounds 1–5) as the most possible active hits. Subsequent in vitro validation revealed emodin ( 1 ) and physcion ( 2 ) to be potent antimalarial candidates with IC 50 values of 0.9 and 1.9 µM, respectively. Our approach in the present investigation therefore can be applied as a preliminary evaluation step in the natural products drug discovery, which in turn can facilitate the isolation of selected metabolites notably the biologically active ones.
