Open Collections

Abstract

Over 40 years of research have been spent investigating the camptothecin biosynthetic pathway in Camptotheca acuminata, yet it remains elusive. While there has been a large amount of work done in the last decade to generate new genomic, transcriptomic and metabolomic resources, there are still major gaps in the knowledge, such as the localization of the tissue responsible for the biosynthesis of camptothecin and the missing steps in the biosynthetic pathway. To fill this gap, isotopically labelled precursors were fed to six tissues from C. acuminata, which were then analyzed using metabolomics. It was found that stem tissue was best able to produce downstream products in the CPT biosynthetic pathway. To identify candidate genes encoding enzymes involved in the camptothecin biosynthetic pathway, a novel bioinformatics workflow was developed, which utilized a new machine learning algorithm CLEAN to generate new functional annotations of the C. acuminata genome. This annotation was then combined with orthology data generated using another machine learning algorithm, TOGA, from Nyssa sinensis and Davidia involucrata, two non-CPT producing plants closely related to C. acuminata in the family Nyssaceae to identify biosynthetic candidates highly expressed in stem tissues. To access cost-prohibitive and non-commercially available substrates, a workflow was developed to isolate Gram-scale quantities of secologanin and secologanic acid from Lonicera tatarica, and 20-(S)- deoxycamptothecin was produced semi-synthetically from camptothecin using a Marko-Lam deoxygenation. A new O-methyltransferase capable of methylating both 10 and 11-hydroxycamptothecin, as well as an array of flavonoids, was identified using correlation analysis of transcriptomic data. Finally, new activity of the previously discovered camptothecin-10-hydroxylase was identified when screened against 20-(S)-deoxycamptothecin putatively producing 10-hydroxy-20-(S)-deoxycamptothecin.