Trichome Dynamics and Artemisinin Accumulation during Development and Senescence of Artemisia annua Leaves

• Abstract
• Introduction
• Materials and Methods
• Plant material and experimental design
• Processing and botanical analysis of leaves
• Processing, extraction and phytochemical analysis of leaves
• Statistical analysis
• Results
• Discussion
• Acknowledgements
• References

Abstract

Artemisinin is a sesquiterpene lactone endoperoxide and an important antimalarial drug produced in Artemisia annua. To unravel the diverse processes determining artemisinin yield in A. annua crops, artemisinin accumulation during the development of individual leaves was studied in two field experiments. During the life cycle of a leaf, artemisinin was always present. Quantities were low at leaf appearance and increased steadily. In leaves studied until after senescence, maximum quantities and concentrations were achieved after the leaf had turned brown. The total quantity of possible artemisinin precursors per leaf (dihydroartemisinic acid and other upstream precursors) was highest early in the leaf cycle when the leaf was still expanding. Dihydroartemisinic acid was more abundant than the other compounds and its quantity declined during leaf development whereas that of artemisinin increased. Dihydroartemisinic acid was not converted directly into artemisinin, because on a per leaf basis the decline in molar quantity of precursors in the earliest formed leaves was not compensated for by a simultaneous increase in artemisinin. Our results suggest that a (putative) intermediate such as dihydroartemisinic acid hydroperoxide temporarily may have accumulated in considerable quantities. The number of mature, capitate trichomes on the adaxial leaf side increased after leaf appearance until the end of leaf expansion, and then decreased, probably due to collapse of trichomes. Artemisinin production thus (also) occurred when trichomes were collapsing. Later formed leaves achieved higher concentrations of artemisinin than earlier formed leaves, because of a higher trichome density and a higher capacity per trichome.

Introduction

Artemisia annua L. (annual or sweet wormwood, Asteraceae) is an annual herb originating from Asia that produces the antimalarial compound artemisinin, a sesquiterpene lactone with an endoperoxide bridge. Artemisinin and some of its synthetic derivatives, e. g., dihydroartemisinin, artesunate, arteether and artemether [1], are highly effective against Plasmodium falciparum that causes the severe malaria tropica. Because chemical synthesis of artemisinin is complicated and expensive, the plant is the only feasible commercial source of artemisinin. Within the plant, highest concentrations of artemisinin are found in the leaves and inflorescences [2], [3], [4]. A. annua produces biseriate glandular trichomes on the leaves and inflorescences [5], [6], which are regarded to be the sites of artemisinin production or sequestration [6].

There is ongoing discussion about the biosynthetic pathway by which artemisinin is formed, and therefore about which sesquiterpenoids are precursors and which are not (Fig. [1]). The first step in artemisinin formation is the cyclisation of farnesyl diphosphate to amorpha-4,11-diene [7]. This sesquiterpene is hydroxylated to give artemisinic alcohol [8]. This alcohol is oxidised yielding artemisinic aldehyde after which the C11-C12 double bond is reduced to yield dihydroartemisinic aldehyde, which is subsequently oxidised to dihydroartemisinic acid [8]. Dihydroartemisinic acid can be converted non-enzymatically into artemisinin [9], [10], [11]. The ratios between these precursors and artemisinin vary in genotypes from different origins [12], [13], [14], which leads to the conclusion that different chemotypes of A. annua exist [12]. High artemisinin levels are often found together with high dihydroartemisinic acid levels and low artemisinic acid levels, e. g., in a Vietnamese genotype [12], [15]. Low artemisinin genotypes often have high artemisinic acid levels but low dihydroartemisinic acid levels. Wallaart et al. [12] suggested that the differences between chemotypes resulted from a rate-limiting step in the reduction of artemisinic acid to dihydroartemisinic acid in the genotypes with high artemisinic acid levels. The recent results of Bertea et al. [8], however, exclude artemisinic acid as a direct precursor for artemisinin in the Vietnamese genotype.

High artemisinin yield and high concentrations of artemisinin in the plant are important for high production and efficient extraction of artemisinin from the plant mass. To maximise artemisinin yield and concentrations under different production conditions, it is essential to understand how artemisinin production proceeds within a crop. Several different processes at the level of the cell, the individual leaf and the whole crop will govern the total quantities and concentrations of artemisinin produced.

Knowledge on A. annua crop behaviour so far is based mainly on time series studies of whole crops or crop fractions. During crop growth, artemisinin concentrations usually first increase and later decrease [2], [13], [16], [17], but highest concentrations were achieved during vegetative stages [4], [15], just before or around flowering [16] or during flowering [2], [17] and highest artemisinin yields in the vegetative stage [4], [15], at flowering [17] or at full flowering [13], [14]. Laughlin [14] reported that highest yields of artemisinic acid occurred earlier during flowering than for artemisinin. Woerdenbag et al. [15] found concentrations of artemisinic acid and arteannuin B to decrease with time, comparable to artemisinin, whereas maximum artemisitene concentrations were found later in the vegetative stage than maximum artemisinin concentrations. However, all these time studies are hampered by the fact that new leaves are continuously initiated during the vegetative crop stage, whereas older leaves may be lost. Comparisons are thus based on different leaf classes. Within a crop, artemisinin concentrations are often higher in the upper leaves compared to lower leaves when plants are still vegetative [4], [14], but this pattern may change after flowering [14]. By contrast, Ferreira and Janick [3] found no consistent differences over different leaf fractions in different cultivars. Higher concentrations of artemisinin in the upper leaves may result from loss of artemisinin in older leaves, as suggested by Duke et al. [6], by higher concentrations in leaves initiated later in the season, as shown by Gupta et al. [4] for part of the growing season and/or by further ”dilution” of artemisinin with leaf dry matter when the young, upper leaves develop further and accumulate dry matter.

The present paper addresses the changes in quantities and concentrations of artemisinin and important precursors during the life cycle of individual leaves within an A. annua crop. This should reveal how an A. annua leaf develops, when artemisinin and putative precursors are produced, whether rate-limiting steps in artemisinin formation occur, if and how phytochemical production is related to leaf or trichome development, and whether losses of artemisinin occur after its production.

Materials and Methods

Plant material and experimental design

In two field experiments, the development, growth and phytochemical composition of individual leaves in A. annua crops were studied. Experiment 1 was carried out from May 22, 2002 to October 16, 2002 in Wageningen (N51°59′31′′ E005°39′08′′), using seeds from a Vietnamese genotype, obtained from Artecef BV (Maarssen, The Netherlands) and described earlier [7], [12]. Experiment 2 was carried out from June 18, 2002 to October 15, 2002 in Achterberg (N51°59′24′′ E005°34′58′′), using the hybrid cultivar Anamed A3 (Anamed, Winnenden, Germany).

Plants were raised from seeds under natural day length in glasshouses from April 2, 2002 to May 14, 2002 (Experiment 1) and April 26, 2002 to June 7, 2002 (Experiment 2), hardened outside for 8 (Experiment 1) and 11 (Experiment 2) days and transplanted to the field at a planting density of 6.7 plants per m² (75 cm between rows, 20 cm within rows) in sandy soil with pH - KCl 5.0 (Experiment 1) and 4.5 (Experiment 2). Plants were fertilised with 74 kg ha^-1 N, 94 kg ha^-1 P₂O₅, and 62 kg ha^-1 K₂O (Experiment 1), and 95 kg ha^-1 N (Experiment 2). Both experiments contained five replicate blocks with three tagging dates randomised to rows within the block. The outer rows and outer 1 m of plants of a block were guard plants and were not analysed. Tagging was done to recognise leaves of the same day of leaf appearance in time. On a tagging date, a coloured thread was tied to the third visible leaf from the apex of the main stem, being the youngest leaf that could be tagged without damage. A fixed leaf (No. 1 - 3, depending on the tagging date) above the tagged leaf was harvested at different time intervals between 0 and 96 days after leaf appearance, starting 3 days after tagging. Because plants had grown above reach by the time of the third tagging in Experiment 1, a leaf from the 24^th primary branch located at the eastern side of the plants was tagged on that day. The exact day of leaf appearance of the harvested leaf was calculated from the relation between leaf number and temperature sum elapsed after field planting using main stem leaf appearance rates. Leaves appearing on Julian day 168, 197 and 240 (leaf No. 20 and No. 31 on the main stem and on the 24th primary branch) were studied in Experiment 1 and leaves appearing on Julian day 186, 228 and 267 (leaves No. 19, 32 and 48 on the main stem) were harvested with time in Experiment 2. From each plant only one leaf was harvested. On a harvest date, leaves were collected from 10 - 20 plants per block within 1 h from solar noon, stored in a cool container, and immediately processed after transport to the laboratory. The larger samples of 20 leaves were taken when leaves were still very small in order to obtain enough leaf material for analysis. Average temperature during the field period of the respective experiments was 15.9 and 15.9 °C, average photoperiod 14.4 and 14.0 h and average global radiation 1449 and 1369 J m^-2 day^-1. Plants did not flower during the experimental period.

Processing and botanical analysis of leaves

The leaf area (one-sided) of the complete leaf samples was measured by a LiCor 3100 leaf area meter with high resolution f = 105 mm lens. The colour of each leaf was determined using Munsell Color Charts for Plant Tissues [18]. In case the colour of the leaf was changing (e. g., from green to yellow), the leaf was classified to the most prominent colour. Thereafter leaves were further dissected and subsamples were taken for dry matter determination, determination of the trichome number, and phytochemical analysis. Dry matter was determined after drying in a fan-forced oven at 70 °C for 24 h. The number of glandular trichomes was determined at the adaxial side of 10 random pieces of leaf material per block with an accurately determined leaf area of approximately 5 mm² per piece. The abaxial side of these pieces was glued by self-adhesive tape to paper. Thereafter, pieces were incubated for 2 days in the dark at 6 °C before counting using a binocular microscope at 310 × magnification. The gluing and incubation facilitated trichome counting by levelling the leaf surface and slightly reducing the cell turgor.

Processing, extraction and phytochemical analysis of leaves

Leaf subsamples for phytochemical analysis were dried in a fan-forced oven in the dark at 30 °C during 24 h and were kept dark and dry until analysis. About 300 mg crushed dry leaf material were accurately weighed, ground with some sea sand in a mortar and extracted three times with a total of 5 mL dichloromethane containing 20 μg mL^-1 cis-nerolidol (C₁₅H₂₆O) as internal standard for quantification. Phytochemicals analysed were artemisinic alcohol, dihydroartemisinic alcohol, artemisinic aldehyde, dihydroartemisinic aldehyde, artemisinic acid, dihydroartemisinic acid and artemisinin. Molar response factors for all these compounds compared with cis-nerolidol were determined using authentic standards that have been described by Bertea et al. [8]. The reproducibility of the method was determined and found to be over 80 % (methodological standard deviation of maximum 20 %). Samples of 2 μL were analysed by GC-MS on a HP 5980 series II gas chromatograph and HP5972A mass selective detector (70 eV) equipped with an HP-5MS column (95 % dimethylpolysilane, 30 m × 0.25 mm i. d., film thickness 0.25 μm). The carrier gas was He (1 mL min^-1), the injection temperature 250 °C and detector temperature 290 °C. The oven was programmed at 2 min 80 °C, 80 to 235 °C at 5 °C per min, 235 to 280 °C at 25 °C per min, and a final time of 5 min.

Statistical analysis

Analysis of variance (ANOVA) was conducted using Genstat 6.1 (Lawes Agricultural Trust, Rothamsted) for time series data of each leaf separately. Relevant transformations were performed when necessary. Since ln or square root transformations of leaf area per leaf, dry weight per leaf and trichome number per leaf did not lead to substantially different conclusions, untransformed data are presented in the graphs. Means were compared using LSD-tests at α < 0.05.

Results

In both experiments, the area of the investigated leaves increased after leaf appearance to a maximum, remained at that maximum for some weeks, and finally decreased when leaves shrivelled (Figs. 2A and B). The earlier formed leaves increased faster in size and achieved higher maximum leaf areas than the later formed leaves (Figs. 2A and B).

Also the dry weight of the investigated leaves increased after leaf appearance up to a maximum, but usually the leaf started to loose dry weight soon after reaching the maximum and then gradually declined in weight (Figs. 2C and D). The earlier formed leaves increased faster in dry weight and achieved higher maximum levels, but the differences were less pronounced than for leaf area (Figs. 2C and D). Maximum dry weight was achieved earlier or at the same moment as maximum leaf area.

Leaf colour changed through different shades of lighter, darker and again lighter green (leaf colour codes 7.5 GY 7/8, 7/6, 6/8, 6/6, 5/6, 5/4, 4/6, 4/4, 3/4, 5 GY 6/6) to different shades of yellow (2.5 GY 8/8, 5 Y 8/10, 8/8), and finally to different shades of brown (7.5 YR 6/6, 4/4 and 4/2) [18]. To facilitate presentation, leaves were grouped into green, yellow and brown. Senescence of leaves, indicated by the first colour change of leaves from green to yellow and brown, started just after 6 weeks from leaf appearance (Figs. 2E and F). The yellow leaf colour usually was only apparent for a very short time before leaves turned brown (Figs. 2E and F). No abscission of dead leaves occurred. The peak in senescence shown in Fig. [2] F on day 46 for leaf 32 in Experiment 2 was probably due to sampling variation. To be able to relate leaf senescence easily to other parameters, the curves describing the proportion of yellow and brown leaves are plotted in all other graphs below.

The dry matter concentration in the leaves was high, usually between 20 and 25 % (Figs. 3A and B). After some fluctuations in a very young leaf stage it tended to decrease slightly (Figs. 3A and B), possibly following the decrease in dry matter weight of the leaf (Figs. 2C and D). When leaves started to senesce, the dry matter concentration increased considerably (Figs. 3A and B), showing that leaves desiccated during senescence.

The specific leaf area (SLA), the ratio between the leaf area and the leaf dry weight, did not reach a clear stable level during the life cycle of individual leaves (Figs. 3C and D). Generally, SLA increased after leaf appearance (Figs. 3C and D), first as a result of the greater increase in area per leaf than in dry weight per leaf, and later as a result of the more or less stable area and decreasing dry weight per leaf (Figs. 2A - D). A peak in SLA was reached at the onset of senescence. After that peak, SLA dropped (Figs. 3C and D), following the strong decrease in leaf area (Figs. 2A and B) due to the shrivelling of leaves.

The number of glandular trichomes, assessed on the adaxial side of the leaf, increased after leaf appearance and reached peak values around the moment that maximum leaf areas were obtained (Figs. 4A and B). Maximum numbers of adaxial trichomes per leaf were around 300,000 for the Vietnamese genotype in Experiment 1 and 100,000 for cv. Anamed A3 in Experiment 2. In the first days after leaf appearance, assessment of the number of glandular trichomes was complicated by the presence of non-glandular hairs and not all leaf pieces in the subsamples could be assessed. Hence, the estimates on those dates are less reliable and are therefore presented in the figures by smaller markers and thinner lines. During leaf senescence, the number of visible adaxial trichomes per leaf decreased (Figs. 4A and B). The density of trichomes varied in time and between leaves and experiments (Figs. 4B and C), reflecting the changes in trichome number and leaf area per leaf, and achieving values between approximately 20 to 80 trichomes per mm². There was no stable level. Maximum trichome densities per mm² were lowest in the earliest formed leaves.

Artemisinin was detected in all stages of leaf development (Table [1]), but only very low quantities per leaf were present early after leaf appearance (Fig. [5]), especially when compared to some of its putative precursors. The quantity of artemisinin per leaf increased and continued to increase over almost the complete life cycle of the leaf until after the leaf became senescent (Fig. [5]). Also the concentration of artemisinin increased until well after the onset of senescence (Table [1]). Only for leaf 19 in Experiment 2, a significant reduction in the total artemisinin quantity per leaf (Fig. [5] D) and the artemisinin concentration (Table [1]) was found in the last harvest.

Artemisinic alcohol, dihydroartemisinic alcohol, artemisinic aldehyde, dihydroartemisinic aldehyde, artemisinic acid and dihydroartemisinic acid all were present in leaves of A. annua, although dihydroartemisinic aldehyde was only found in younger leaves, and artemisinic alcohol and dihydroartemisinic alcohol were not always detected. Of the possible precursors, dihydroartemisinic acid was the most abundant over the whole life cycle of the leaf, and in young leaves even accounted for over 75 % of all sesquiterpenoids analysed (cf. Table [1]). The abundance of dihydroartemisinic acid gradually decreased, whereas that of artemisinin increased (Fig. [5]). Of the other sesquiterpenes, artemisinic aldehyde was second most abundant, artemisinic acid third. The total quantity of putative artemisinin precursors per leaf increased rapidly until 7 - 14 days after leaf appearance in the first leaves investigated (Figs. 5A, B, D, E). This peak was achieved well before the leaf had attained its maximum dry weight or its maximum number of glandular trichomes, and was achieved in a stage that the leaf was still rapidly expanding (cf. Figs. 1A and B). After peak values were achieved, the total quantities of precursors and especially dihydroartemisinic aldehyde declined until a more or less stable level was achieved

Maximum concentrations of artemisinin were higher in later formed leaves (Table [1]). Concentrations were already high soon after leaf appearance. The same was found for most precursors (Table [1]). The changes with time in the different leaves were remarkably similar over the experiments.

Discussion

This paper describes for the first time artemisinin accumulation during development of individual leaves in an A. annua crop. Artemisinin was detected from leaf appearance until leaves had senesced (Table [1]), and quantities of artemisinin per leaf increased continuously after leaf appearance until well after the onset of senescence (Fig. [5]). Concentrations of artemisinin were usually highest when leaves were ageing (Table [1]), both because total quantities of artemisinin per leaf increased (Fig. [5]) and total dry weight per leaf usually decreased during ageing (Figs. 2C and D). For those leaves studied until after senescence, maximum quantities and concentrations were achieved always when leaves already had turned brown. Our results are very different from what was expected by Duke et al. [6], who anticipated that young leaves would contain most artemisinin before rupture of trichomes would cause artemisinin loss. Their idea, however, was based on the fact that younger leaves on a plant often have higher concentrations of artemisinin than older leaves [4], [14]. Our results show that comparing younger to older leaves at the same plant stage can lead to incorrect conclusions on processes occurring during leaf development because of differences between leaf classes.

A statistically significant decrease in artemisinin quantity per leaf was only observed 96 days after leaf appearance in one of the leaf classes studied (leaf 19, Experiment 2) (Fig. [5] D). It is likely that this was due to loss of artemisinin per se and not to loss of leaf material. Changes in the quantities of the other precursors were insignificant (Fig. [5] D) and leaf abscission did not occur during our experiments. How artemisinin was lost is not clear. Perhaps it was leached by rain from the leaves or was chemically degraded, or metabolised by micro-organisms.

Our findings are consistent with the suggestion that dihydroartemisinic acid is one of the most direct precursors of artemisinin [8], [9], [10], [11]. Dihydroartemisinic acid was always more abundant than the other possible precursors (Table [1]). Quantities of dihydroartemisinic acid declined during leaf development when those of artemisinin increased (Fig. [5]). In young leaves, the conversion of dihydroartemisinic acid into artemisinin was likely the most important rate-limiting step in artemisinin production, because dihydroartemisinic acid accumulated initially in much higher quantities than artemisinin (Table [1]). The biosynthesis of dihydroartemisinic acid was postulated recently to take place from artemisinic alcohol through artemisinic aldehyde and dihydroartemisinic aldehyde, the most direct precursor of dihydroartemisinic acid [8] (Fig. [1]). The latter compound only accumulated in young leaves, when concentrations of dihydroartemisinic acid were still very high (Table [1]). This suggests that also the conversion of dihydroartemisinic aldehyde into dihydroartemisinic acid may have been limited in this stage, perhaps by a feed-back mechanism. When dihydroartemisinic aldehyde accumulated, also artemisinic aldehyde concentrations were high. Concentrations of artemisinic alcohol and artemisinic aldehyde, on the other hand, were not clearly correlated with the other compounds.

Dihydroartemisinic acid was not converted directly into artemisinin, because in the earliest formed leaves in both experiments, the decline in nmol of precursors on a per leaf basis was not compensated by a simultaneous increase in artemisinin (Figs. 5A and D). This implies that dihydroartemisinic acid was further converted into a compound that we did not analyse, or that it was lost. The former is most likely, because for the later increase in artemisinin quantity precursors are needed. Since artemisinin was also produced in senescent leaves which are not likely to have an active metabolism, precursors must have been present. Although the identity of these intermediate compound(s) is unknown, dihydroartemisinic acid hydroperoxide is a likely candidate. The first step in the conversion of dihydroartemisinic acid to artemisinin was postulated to be photooxidation to dihydroartemisinic acid hydroperoxide, that later would be converted further to artemisinin by air oxidation [9]. Shy and Brown [10] showed that dihydroartemisinic acid in vitro underwent slow spontaneous autooxidation to artemisinin and other compounds and confirmed that the first step could take place in vitro and required light whereas the second step took place through two intermediates in the dark. In our experiments, the peroxide could have been formed from dihydroartemisinic acid when leaves were still green and exposed to sun light, which is consistent with the drop in precursor quantities occurring well before leaf senescence (Figs. 5A and D). The further oxidation of the peroxide into artemisinin could have continued until well after leaves had senesced, turned brown, and were lower in the canopy where light penetration is poor. Further studies using labelled precursors should shed light on the importance of this intermediate.

The combination of relatively high levels of dihydroartemisinic acid and artemisinin and relatively low levels of artemisinic acid found in Experiment 1 in the Vietnamese genotype (Table [1], Fig. [5]) is consistent with what could be expected from a Vietnamese chemotype [12], [19]. The similar trend in our second experiment suggests that the genetic background of the hybrid cultivar Anamed A3 is also from this chemotype.

Two to three months after appearance of a leaf, artemisinin still comprised only about 50 % of the sesquiterpenoids analysed and dihydroartemisinic acid about 20 - 30 % (cf. Fig. [5]). This shows that theoretically, there still is room for a substantial increase in artemisinin production. The further conversion of dihydroartemisinic acid into artemisinin was postulated to be enhanced by night frost [12] and optimising post-harvest drying [20] or processing techniques.

It is generally accepted that within a leaf artemisinin is produced or sequestered only in glandular trichomes [6] on both leaf surfaces [5]. Bertea et al. [8] showed enzyme activity for production of most sesquiterpenoid precursors to occur in these glandular trichomes. Artemisinin is presumably sequestered between the cell wall of the upper six trichome cells and the cuticula. Glandular trichomes at various stages of development are already present on leaf primordia [5], but it is not clear until which leaf stage trichomes differentiate. Ascensão and Pais [21] working with Artemisia campestris, found differentiating glandular trichomes up to the 9th leaf from the top. Our assay evaluated the number of mature, capitate trichomes on the adaxial leaf side. The fact that the increase in the total number of adaxial trichomes per leaf was followed by a decrease (Figs. 4A and B) was unexpected, but very consistent over leaves. A similar pattern was reported in Mentha arvensis [22]. The decrease implies that trichomes collapsed during leaf senescence. Maximum numbers of mature trichomes per leaf (Figs. 4A and B) were observed around the time that the leaf area achieved its maximum (Figs. 1A and B). Maximum quantities of dihydroartemisinic acid and upstream artemisinin precursors were found earlier in the leaf life cycle, when the leaf was still expanding and increasing in dry weight (Fig. [5]). Further conversion and sequestration of intermediate compounds downstream of dihydroartemisinic acid then could have occurred while trichomes ”matured”. Maximum quantities of artemisinin were found well after maximum leaf dry weight and maximum trichome numbers, when leaves were senescing and trichome numbers were decreasing (Fig. [5]). The fact that artemisinin production continued when trichomes were already collapsing suggests that oxidative formation of artemisinin occurs also after rupture of the trichomes and thus (also) might take place ex planta. In addition, the results show clearly that during development in time, there is no positive correlation to be expected between the number of glandular trichomes and the quantity of artemisinin.

Lower concentrations of artemisinin and other phytochemicals were found in earlier appearing leaves (Table [1]), because of two main reasons. The first was their lower trichome density, observed in Experiment 1 and for a major part of the life cycle in Experiment 2 (Figs. 4C and D). Trichome densities mainly varied over leaf classes because earlier appearing leaves had higher rates of leaf expansion and dry matter accumulation, resulting in larger areas and dry weight per leaf (Figs. 2A and B), whereas differences between leaves in the maximum number of trichomes per leaf (Figs. 4A and B) were smaller or even tended to be opposite. The second reason for the lower concentrations and in addition the lower quantities of sesquiterpenoids in earlier appearing leaves must be a lower production capacity per trichome and per unit area of weight of leaf: differences between leaves in total quantities of phytochemicals per leaf (Fig. [5]) and their concentrations (Table [1]) were much larger than the differences in maximum number of trichomes per leaf (Figs. 4A and B).

Acknowledgements

The authors would like to thank Dr. Ch. B. Lugt (Artecef BV, Maarssen, The Netherlands) for his gift of A. annua seeds, Prof. Dr. P. C. Struik and Dr. W. van der Werf (Wageningen University, Crop and Weed Ecology) for their comments on the manuscript and the teams of G. Versteeg and H. Masselink (PPW Wageningen) for their enthusiastic assistance in producing the transplants and growing the crops.

References

• 1 Dhingra V, Rao K V, Narasu M L. Current status of artemisinin and its derivatives as antimalarial drugs.  Life Sci. 2000;  66 279-300
  PubMedSearch in Google Scholar
• 2 Ferreira J FS, Simon J E, Janick J. Developmental studies of Artemisia annua: Flowering and artemisinin production under greenhouse and field conditions.  Planta Med. 1995;  61 167-70
  Thieme ConnectPubMedSearch in Google Scholar
• 3 Ferreira J FS, Janick J. Distribution of artemisinin in Artemisia annua .  In: Janick J, editor. Progress in new crops.  Arlington; ASHS Press; 1996: p. 579-84
  Search in Google Scholar
• 4 Gupta S K, Singh P, Bajpai P, Ram G, Singh D, Gupta M M. et al . Morphogenetic variation for artemisinin and volatile oil in Artemisia annua .  Ind Crop Prod. 2002;  16 217-24
  CrossrefPubMedSearch in Google Scholar
• 5 Duke S O, Paul R N. Development and fine structure of the glandular trichomes of Artemisia annua L.  Int J Plant Sci. 1993;  154 107-18
  PubMedSearch in Google Scholar
• 6 Duke M V, Paul R N, Elsohly H N, Sturtz G, Duke S O. Localization of artemisinin and artemisitene in foliar tissues of glanded and glandless biotypes of Artemisia annua L.  Int J Plant Sci. 1994;  155 365-72
  PubMedSearch in Google Scholar
• 7 Bouwmeester H J, Wallaart T E, Janssen M HA, van Loo B, Jansen B JM, Posthumus M A. et al . Amorpha-4,11-diene synthase catalyses the first probable step in artemisinin biosynthesis.  Phytochemistry. 1999;  52 843-54
  CrossrefPubMedSearch in Google Scholar
• 8 Bertea C M, Freije J R, van der Woude H, Verstappen F WA, Perk L, Marquez V. et al . Identification of intermediates and enzymes involved in the early steps of artemisinin biosynthesis in Artemisia annua L.  Planta Med. 2005;  71 40-7
  Thieme ConnectPubMedSearch in Google Scholar
• 9 Wallaart T E, van Uden W, Lubberink H GM, Woerdenbag H J, Pras N, Quax W J. Isolation and identification of dihydroartemisinic acid from Artemisia annua and its possible role in the biosynthesis of artemisinin.  J Nat Prod. 1999;  62 430-3
  CrossrefPubMedSearch in Google Scholar
• 10 Sy L K, Brown G D. The mechanism of the spontaneous autoxidation of dihydroartemisinic acid.  Tetrahedron. 2002;  58 897-908
  CrossrefPubMedSearch in Google Scholar
• 11 Brown G D, Sy L K. In vivo transformations of dihydroartemisinic acid in Artemisia annua plants.  Tetrahedron. 2004;  60 1139-59
  CrossrefPubMedSearch in Google Scholar
• 12 Wallaart T E, Pras N, Beekman A C, Quax W J. Seasonal variation of artemisinin and its biosynthetic precursors in plants of Artemisia annua of different geographical origin: Proof for the existence of chemotypes.  Planta Med. 2000;  66 57-62
  Thieme ConnectPubMedSearch in Google Scholar
• 13 Laughlin J C. Effect of agronomic practices on plant yield and anti-malarial constituents of Artemisia annua L.  Acta Hortic. 1993;  331 53-61
  PubMedSearch in Google Scholar
• 14 Laughlin J C. The influence of distribution of antimalarial constituents in Artemisia annua L. on time and method of harvest.  Acta Hortic. 1995;  390 67-73
  PubMedSearch in Google Scholar
• 15 Woerdenbag H J, Pras N, Nguyen Gia C han, Bui Thi B ang, Bos R, van Uden W. et al . Artemisinin, related sesquiterpenes, and essential oil in Artemisia annua during a vegetation period in Vietnam.  Planta Med. 1994;  60 272-5
  Thieme ConnectPubMedSearch in Google Scholar
• 16 Liersch R, Soicke H, Stehr C, Tüllner H U. Formation of artemisinin in Artemisia annua during one vegetation period.  Planta Med. 1986;  52 387-90
  Thieme ConnectPubMedSearch in Google Scholar
• 17 Morales M M, Charles D J, Simon J E. Seasonal accumulation of artemisinin in Artemisia annua L.  Acta Hortic. 1993;  344 416-20
  PubMedSearch in Google Scholar
• 18 Munsell . Munsell Color Charts for Plant Tissues,. 2nd edition. Baltimore; Munsell Color Division, Kollmorgen Corporation; 1972: p 19
  Search in Google Scholar
• 19 Wallaart T E, Pras N, Quax W J. Seasonal variations of artemisinin and its biosynthetic precursors in tetraploid Artemisia annua plants compared with the diploid wild-type.  Planta Med. 1999;  65 723-8
  Thieme ConnectPubMedSearch in Google Scholar
• 20 Laughlin J C. Post-harvest drying treatment effects on antimalarial constituents of Artemisia annua L.  Acta Hortic. 2002;  576 315-20
  PubMedSearch in Google Scholar
• 21 Ascensão L, Pais M SS. Glandular trichomes of Artemisia campestris (ssp. maritima) ontogeny and histochemistry of the secretory product.  Bot Gaz. 1987;  148 221-7
  CrossrefPubMedSearch in Google Scholar
• 22 Shanker S, Ajayakumar P V. Essential oil gland number and ultrastructure during Mentha arvensis leaf ontogeny.  Biol Plantarum. 1999;  42 379-87
  PubMedSearch in Google Scholar
• 23 Wallaart T E, Pras N, Quax W J. Isolation and identification of dihydroartemisinic acid hydroperoxide from Artemisia annua: a novel biosynthetic precursor of artemisinin.  J Nat Prod. 1999;  62 1160-2
  CrossrefPubMedSearch in Google Scholar
• 24 Woerdenbag H J, Lugt C B, Pras N. Artemisia annua L.: a source of novel antimalarial drugs.  Pharmaceutisch Weekblad, Sci Ed. 1990;  12 169-81
  PubMedSearch in Google Scholar

Dr. W. J. M. Lommen

Crop and Weed Ecology Group

Wageningen University

Haarweg 333

6709 RZ Wageningen

The Netherlands

Fax: +31-317-485-572

Email: Willemien.Lommen@wur.nl

References

• 1 Dhingra V, Rao K V, Narasu M L. Current status of artemisinin and its derivatives as antimalarial drugs.  Life Sci. 2000;  66 279-300
  PubMedSearch in Google Scholar
• 2 Ferreira J FS, Simon J E, Janick J. Developmental studies of Artemisia annua: Flowering and artemisinin production under greenhouse and field conditions.  Planta Med. 1995;  61 167-70
  Thieme ConnectPubMedSearch in Google Scholar
• 3 Ferreira J FS, Janick J. Distribution of artemisinin in Artemisia annua .  In: Janick J, editor. Progress in new crops.  Arlington; ASHS Press; 1996: p. 579-84
  Search in Google Scholar
• 4 Gupta S K, Singh P, Bajpai P, Ram G, Singh D, Gupta M M. et al . Morphogenetic variation for artemisinin and volatile oil in Artemisia annua .  Ind Crop Prod. 2002;  16 217-24
  CrossrefPubMedSearch in Google Scholar
• 5 Duke S O, Paul R N. Development and fine structure of the glandular trichomes of Artemisia annua L.  Int J Plant Sci. 1993;  154 107-18
  PubMedSearch in Google Scholar
• 6 Duke M V, Paul R N, Elsohly H N, Sturtz G, Duke S O. Localization of artemisinin and artemisitene in foliar tissues of glanded and glandless biotypes of Artemisia annua L.  Int J Plant Sci. 1994;  155 365-72
  PubMedSearch in Google Scholar
• 7 Bouwmeester H J, Wallaart T E, Janssen M HA, van Loo B, Jansen B JM, Posthumus M A. et al . Amorpha-4,11-diene synthase catalyses the first probable step in artemisinin biosynthesis.  Phytochemistry. 1999;  52 843-54
  CrossrefPubMedSearch in Google Scholar
• 8 Bertea C M, Freije J R, van der Woude H, Verstappen F WA, Perk L, Marquez V. et al . Identification of intermediates and enzymes involved in the early steps of artemisinin biosynthesis in Artemisia annua L.  Planta Med. 2005;  71 40-7
  Thieme ConnectPubMedSearch in Google Scholar
• 9 Wallaart T E, van Uden W, Lubberink H GM, Woerdenbag H J, Pras N, Quax W J. Isolation and identification of dihydroartemisinic acid from Artemisia annua and its possible role in the biosynthesis of artemisinin.  J Nat Prod. 1999;  62 430-3
  CrossrefPubMedSearch in Google Scholar
• 10 Sy L K, Brown G D. The mechanism of the spontaneous autoxidation of dihydroartemisinic acid.  Tetrahedron. 2002;  58 897-908
  CrossrefPubMedSearch in Google Scholar
• 11 Brown G D, Sy L K. In vivo transformations of dihydroartemisinic acid in Artemisia annua plants.  Tetrahedron. 2004;  60 1139-59
  CrossrefPubMedSearch in Google Scholar
• 12 Wallaart T E, Pras N, Beekman A C, Quax W J. Seasonal variation of artemisinin and its biosynthetic precursors in plants of Artemisia annua of different geographical origin: Proof for the existence of chemotypes.  Planta Med. 2000;  66 57-62
  Thieme ConnectPubMedSearch in Google Scholar
• 13 Laughlin J C. Effect of agronomic practices on plant yield and anti-malarial constituents of Artemisia annua L.  Acta Hortic. 1993;  331 53-61
  PubMedSearch in Google Scholar
• 14 Laughlin J C. The influence of distribution of antimalarial constituents in Artemisia annua L. on time and method of harvest.  Acta Hortic. 1995;  390 67-73
  PubMedSearch in Google Scholar
• 15 Woerdenbag H J, Pras N, Nguyen Gia C han, Bui Thi B ang, Bos R, van Uden W. et al . Artemisinin, related sesquiterpenes, and essential oil in Artemisia annua during a vegetation period in Vietnam.  Planta Med. 1994;  60 272-5
  Thieme ConnectPubMedSearch in Google Scholar
• 16 Liersch R, Soicke H, Stehr C, Tüllner H U. Formation of artemisinin in Artemisia annua during one vegetation period.  Planta Med. 1986;  52 387-90
  Thieme ConnectPubMedSearch in Google Scholar
• 17 Morales M M, Charles D J, Simon J E. Seasonal accumulation of artemisinin in Artemisia annua L.  Acta Hortic. 1993;  344 416-20
  PubMedSearch in Google Scholar
• 18 Munsell . Munsell Color Charts for Plant Tissues,. 2nd edition. Baltimore; Munsell Color Division, Kollmorgen Corporation; 1972: p 19
  Search in Google Scholar
• 19 Wallaart T E, Pras N, Quax W J. Seasonal variations of artemisinin and its biosynthetic precursors in tetraploid Artemisia annua plants compared with the diploid wild-type.  Planta Med. 1999;  65 723-8
  Thieme ConnectPubMedSearch in Google Scholar
• 20 Laughlin J C. Post-harvest drying treatment effects on antimalarial constituents of Artemisia annua L.  Acta Hortic. 2002;  576 315-20
  PubMedSearch in Google Scholar
• 21 Ascensão L, Pais M SS. Glandular trichomes of Artemisia campestris (ssp. maritima) ontogeny and histochemistry of the secretory product.  Bot Gaz. 1987;  148 221-7
  CrossrefPubMedSearch in Google Scholar
• 22 Shanker S, Ajayakumar P V. Essential oil gland number and ultrastructure during Mentha arvensis leaf ontogeny.  Biol Plantarum. 1999;  42 379-87
  PubMedSearch in Google Scholar
• 23 Wallaart T E, Pras N, Quax W J. Isolation and identification of dihydroartemisinic acid hydroperoxide from Artemisia annua: a novel biosynthetic precursor of artemisinin.  J Nat Prod. 1999;  62 1160-2
  CrossrefPubMedSearch in Google Scholar
• 24 Woerdenbag H J, Lugt C B, Pras N. Artemisia annua L.: a source of novel antimalarial drugs.  Pharmaceutisch Weekblad, Sci Ed. 1990;  12 169-81
  PubMedSearch in Google Scholar

Dr. W. J. M. Lommen

Crop and Weed Ecology Group

Wageningen University

Haarweg 333

6709 RZ Wageningen

The Netherlands

Fax: +31-317-485-572

Email: Willemien.Lommen@wur.nl