Biosynthesis, transport and combinatorial metabolic engineering of Tanacetum parthenium (feverfew) and Artemisia annua (sweet wormwood) sesquiterpene lactones

This thesis mainly focuses on sesquiterpene lactone biosynthesis, their (combinatorial) metabolic engineering and extracellular transport and accumulation. Terpenoids are the most diverse and largest class of natural products, many of which have important pharmaceutical, biological, nutraceutical and/or other beneficial properties. Chapter 1 of this thesis provides the background on the knowledge of biosynthesis of different terpenoid classes in plants at the onset of this thesis work. Also, a background is given of metabolic engineering approaches and different expression platforms (organisms) to overcome limitations of production of these attractive molecules. Finally the topic ‘terpenoid transport’ in plants is introduced, a research area which was largely underdeveloped at the start of my thesis.

There are different classes of sesquiterpene lactones from which particularly the biosynthesis of germacranolides has been studied, mostly in plant species belonging to the Asteraceae. In contrast, how members of the guaianolide class of sesquiterpene lactones are synthesized has been a mystery for a very long time. In chapter 2, I report the first characterization of an enzyme from feverfew responsible for the first committed step branching the guaianolide pathway off from the germacranolides. This enzyme, which converts costunolide into kauniolide, is very special as it performs several sequential reactions, from hydroxylation, water elimination and cyclisation to regio-selective deprotonation. The mechanism of action of this enzyme was elucidated by testing different putative substrates and intermediates (obtained through chemical synthesis) and through in-silico substrate docking studies.

Recent progress in metabolic engineering and pathway reconstruction in microbes and plants has resulted in the production of the bioactive costunolide and parthenolide from feverfew as well as dihydroartemisinic acid, the precursor of the antimalarial artemisinin, from Artemisia annua in heterologous expression platforms. In A. annua, a double bond reductase (AaDBR) enzyme catalyses a branch point in the pathway towards dihydroartemisinic acid. Because artemisinin biosynthesis in A. annua proceeds quite similar to costunolide biosynthesis in feverfew, we suspected that AaDBR may act on products from the feverfew pathway. This was confirmed in chapter 3 in a combinatorial metabolic engineering approach, in which I produced the novel dihydro-sesquiterpene lactones 3β-hydroxy-dihydroparthenolide and 3β-hydroxy-dihydrocostunolide.

In Chapter 4 I switch from biosynthesis of sesquiterpene lactones to their transport. In this chapter I describe the involvement of Lipid Transfer Proteins (LTPs) and Pleiotropic Drug Resistance (PDR) membrane transporters in extracellular accumulation of (dihydro)artemisinic acid [(DH)AA], using transient expression assays in the heterologous host Nicotiana benthamiana. We show that AaLTP3 and AaPDR2 enhance (DH)AA accumulation in the N. benthamiana leaf apoplast as well as the overall flux through the (DH)AA pathway. For functional analysis of the LTPs I developed two novel assays: an in planta substrate export assay and an in planta substrate exclusion assay. Using these assays, we show that AaLTP3 is more effective than AaPDR2 in preventing influx of (DH)AA from the apoplast into the cells. This chapter also describes the first documented case of artemisinin and arteannuin B production in N. benthamiana.

In chapter 5, I continue my research on sesquiterpene transport by functional characterization of Lipid Transfer Protein genes from feverfew trichomes. I demonstrate that some of these LTPs have acquired a specialized function in the transport of specific lipophilic products (sesquiterpene lactones) produced in feverfew trichomes. I show that TpLTP1 and TpLTP2 specifically improve costunolide export and that TpLTP3 highly specifically improves parthenolide export. Moreover, the substrate exclusion assays revealed TpLTP3 as the most effective in blocking influx of both costunolide and parthenolide. TpLTP3 has a GPI anchor domain and I show that this GPI-anchor domain is essential for the activity of TpLTP3, while addition of this domain to TpLTP1 resulted in loss of TpLTP1 activity.

In chapter 6 I discuss the biosynthesis and transport results obtained during my thesis research. I address remaining questions related to biosynthesis of sesquiterpene lactones in yeast and discuss opportunities for combinatorial metabolic engineering, e.g. what are putative applications of the enzyme-substrate promiscuity concept for the double bond reductase (DBR)? In this chapter I also discuss the many remaining questions related to transport of sesquiterpene lactones, as many details about transport over the plasma membrane and over the cell wall are still missing. Finally, I provide a wider perspective on the role of LTPs, based on clues from the exclusion assays and the chemical properties of other biologically active molecules in plants.

In summary, this thesis provides several novel discoveries in the field of sesquiterpene lactone metabolic engineering by identification of the biosynthetic branch point towards guaianolides, by combinatorial metabolic engineering to produce novel dihydro-sesquiterpene lactones and by revealing an important and very specific role of LTPs in extracellular transport of sesquiterpene lactones.