Phylogeny as selection tool for exploring CNS-activity in the Amaryllidaceae

To discover whether DNA-based phylogenies can be used predictively to identify species for bioprospecting, we analysed 37 taxa of the tribe Haemantheae (Amaryllidaceae). DNA sequences from the nuclear ribosomal internal transcribed spacer (ITS) and the plastid trnL-F regions were used. Maximum parsimony analyses divided the Haemantheae into two main clades, A corresponding to the genera Clivia and Cryptostephanus, and B corresponding to the genera Apodolirion-Gethyllis-Haemanthus-Scadoxus. Acetylcholinesterase inhibitory activity was detected in all investigated clades except the Apodolirion-Gethyllis clade. No alkaloids were detected by GC-MS in extracts of the Gethyllis species, which could explain the lack of acetylcholinesterase activity. Within the Haemantheae, dose-dependent affinity to the serotonin transporter was restricted to the genus Haemanthus. The GC-MS profiles indicated that Haemanthus species contain alkaloids of a different type than alkaloids in other members of the tribe. In conclusion, the phylogeny produced in this study can be used to predict in which clades compounds with CNS activity can, or cannot, be found – and thus where to search for candidate species with potential for drug discovery.