Discriminant analysis of Annona muricata and Rollinia mucosa extracts by multivariate curve resolution and partial least-squares regression of liquid chromatography-diode array data

Abstract

Species of the Annonaceae family are used in traditional medicine, and several antitumor acetogenins have been isolated from the leaves of Annona muricata and Rollinia mucosa. Leaves samples (self-shaded and sun-exposed) of these species were collected during winter and summer seasons. Extraction media were prepared from mixtures of five solvents according to a simplex centroid design. Plant extracts were analysed by injection into a high-performance liquid chromatography with diode array detection (HPLC-DAD) and evaluated after applying by multivariate curve resolution-alternating least-squares (MCR-ALS) and discriminant-unfolded partial least-squares (D-UPLS). The objective was to check whether the chemometric resolution of chromatographic profiles could help in the discrimination of samples according to taxonomic classification, sun exposition and/or harvest season. The results indicate that extraction media containing ethanol as main solvent achieved discrimination of self-shaded versus sun-exposed plants and harvest season of Annona muricata. Also, D-UPLS analysis allowed taxonomic discrimination between Annona muricata and Rollinia mucosa extract samples.