PHYSIOLOGY AND BIOCHEMISTRY Evidence of artemisinin production from IPP stemming from both the mevalonate and the nonmevalonate pathways Melissa J. Towler Æ Pamela J. Weathers Received: 26 April 2007 / Revised: 29 June 2007 / Accepted: 19 July 2007 / Published online: 21 August 2007 Ó Springer-Verlag 2007 Abstract The potent antimalarial sesquiterpene lactone, artemisinin, is produced in low quantities by the plant Artemisia annua L. The source and regulation of the iso- pentenyl diphosphate (IPP) used in the biosynthesis of artemisinin has not been completely characterized. Terpe- noid biosynthesis occurs in plants via two IPP-generating pathways: the mevalonate pathway in the cytosol, and the non-mevalonate pathway in plastids. Using inhibitors spe- cific to each pathway, it is possible to resolve which supplies the IPP precursor to the end product. Here, we show the effects of inhibition on the two pathways leading to IPP for artemisinin production in plants. We grew young (7–14 days post cotyledon) plants in liquid culture, and added mevinolin to the medium to inhibit the mevalonate pathway, or fosmidomycin to inhibit the non-mevalonate pathway. Artemisinin levels were measured after 7– 14 days incubation, and production was significantly reduced by each inhibitor compared to controls, thus, it appears that IPP from both pathways is used in artemisinin production. Also when grown in miconazole, an inhibitor of sterol biosynthesis, there was a significant increase in artemisinin compared to controls suggesting that carbon was shifted from sterols into sesquiterpenes. Collectively these results indicate that artemisinin is probably biosyn- thesized from IPP pools from both the plastid and the cytosol, and that carbon from competing pathways can be channeled toward sesquiterpenes. This information will help advance our understanding of the regulation of in planta production of artemisinin. Keywords Artemisinin  Mevalonate pathway  Nonmevalonate pathway  DMSO Introduction Overexploitation of natural sources for high-valued chemicals has increased interest in finding alternative production platforms to facilitate preservation of endan- gered species and the environment (Wu et al. 2006). Low concentrations of products in plants and expensive extraction processes motivate a deeper understanding of their complex synthesis. Terpenoids constitute the largest family of natural plant products with over 30,000 members (Sacchettini and Poulter 1997; Dewick 2002). Until the discovery of an alternative non-mevalonate pathway (MEP) (Fig. 1), it was widely accepted that the formation of isopentenyl diphosphate (IPP), the 5-carbon isoprene precursor to terpenoids, occurred via the meval- onate pathway (MVA). Some of the earliest data in plants came from the study of [ 13 C]-labeled glucose incorporation into ginkgolides (diterpenes) in Ginkgo biloba (Schwarz 1994), which showed that the resulting labeling pattern was incompatible with the mevalonate route of synthesis. Prior to such labeling studies, Bach and Lichtenthaler (1983) used mevinolin, a known 3-hydroxy-3-methylglutaryl-CoA reductase (HMGR) inhibitor, to study plant growth, sterol formation, and pigment accumulation. They suggested that Communicated by P. Lakshmanan. M. J. Towler  P. J. Weathers Department of Biology/Biotechnology, Worcester Polytechnic Institute, Worcester, MA 01609, USA P. J. Weathers (&) Arkansas Bioscience Institute, Arkansas State University, State University, AR 72467, USA e-mail: pweathers@astate.edu 123 Plant Cell Rep (2007) 26:2129–2136 DOI 10.1007/s00299-007-0420-x