Index | Search | Home | Table of Contents
J.F.S. Ferreira and J. Janick. 1996. Distribution of artemisinin in Artemisia annua. p. 579-584. In: J. Janick (ed.), Progress in new crops. ASHS Press, Arlington, VA.

Distribution of Artemisinin in Artemisia annua*

Jorge F.S. Ferreira and Jules Janick

  1. BOTANY
  2. ARTEMISININ DISTRIBUTION
    1. In Vivo
    2. In Vitro
  3. GLANDULAR TRICHOMES AS SITES OF ARTEMISININ ACCUMULATION
  4. CONCLUSION
  5. REFERENCES
  6. Table 1
  7. Table 2
  8. Table 3
  9. Fig. 1
  10. Fig. 2
Artemisia (Artemisia annua L., Asteraceae), known in the United States as sweet Annie or annual wormwood, is an annual herb native to Asia, most probably China, where it is known as qinghao. The plant has become naturalized in many countries including Argentina, Bulgaria, France, Hungary, Romania, Italy, Spain, the United States, and the former Yugoslavia (Gray 1884; Bailey and Bailey 1976; Klayman 1989, 1993). Artemisia is mentioned in the Chinese Handbook of Prescriptions for Emergency Treatments of 340 AD for treatment of fevers. In 1971, extraction of aerial parts of A. annua with low-boiling solvents, such as diethylether, produced a compound mixture with antimalarial properties on infected mice and monkeys. The main active principle, artemisinin (formerly referred to as arteannuin and as qinghaosu in Chinese), was isolated and had its structure correctly defined in 1972 in China as a sesquiterpene lactone with an endoperoxide bridge. Artemisinin is now available commercially in China and Vietnam as an antimalarial drug efficacious against drug-resistant strains of Plasmodium, the malarial parasite. A semisynthetic drug based on artemisinin (artemether) has been recently registered in Africa as Paluther. Artemisinin also has phytotoxic activity, even on A. annua, and is a candidate as a natural herbicide (Duke et al. 1987; Chen et al. 1991).

Artemisinin production by A. annua is usually in the range of 0.01% to 0.4% but some clones produce over 1% (Delabays et al. 1993). Artemisinin can also be obtained from artemisinic acid which occurs at concentrations as much as 10-fold higher than artemisinin (Acton et al. 1985). Recently, Vonwiller et al. (1993) reported an extraction method which makes possible the extraction of both compounds from the same plant material, thus increasing the final production of artemisinin.

BOTANY

Artemisia annua, a vigorous weedy annual (Hall and Clements 1923), is a short day plant with a critical photoperiod of 13.5 hr (Ferreira et al. 1995a). The chromosome number is 2n = 36 (Benn et al. 1982). The plant is usually single-stemmed reaching about 2m in height with alternate branches and alternate, deeply dissected, aromatic leaves ranging from 2.5 to 5.0 cm in length. Tiny yellow nodding flowers (capitula) only 2 or 3 mm across are displayed in lose panicles containing numerous, greenish or yellowish, bisexual central (disc) florets containing little nectar and pistillate marginal (ray) florets (Fig. 1). The involucre is imbricated with several rows of bracts. The central flowers are perfect and can be either fertile or sterile. Ovaries are inferior and unilocular and each generates one achene, ca. 1 mm in length and faintly nerved. The pistillate marginal florets in the capitulum produce numerous achenes without pappus. The pollen is tricolpate and smooth, typical of anemophilous species, and has vestigial or no spines (Stix 1960). It has an internal, complex, columellae-tecta configuration in the exine, which is common to all taxa of the tribe Anthemideae and varies from two to three layers in A. annua (Skvarla and Larson 1965). The plant is naturally cross-pollinated by insects and wind action, which is unusual in the Asteraceae (McVaugh 1984).

Non-glandular T-shaped trichomes and 10-celled biseriate glandular trichomes occur on leaves, stems, and inflorescences. The morphology and origin of the glandular trichomes has been described for leaves (Duke and Paul 1993) and capitula (Ferreira and Janick 1995) using light and/or scanning electron microscopy. The essential oils of A. annua contains at least 40 volatile compounds and several nonvolatile sesquiterpenes, of which artemisinin and related compounds are the ones of most interest due to their antimalarial properties (Charles et al. 1991; Woerdenbag et al. 1994).

ARTEMISININ DISTRIBUTION

In Vivo

Artemisinin can be quantified by various analytical procedures including thin layer chromatography, gas chromatography, high-performance liquid chromatography with ultraviolet or electrochemical detection (see Ferreira et al. 1994), radioimmunoassay, and enzyme-linked immunosorbant assay (see Ferreira and Janick 1996a). Artemisinin has been detected from aerial parts of the plant, mostly in leaves and inflorescences with low levels in stems and none in pollen or roots (Table 1). The occurrence of artemisinin in the achene (seed) is due to the presence of floral remnants.

Although some authors reported artemisinin being highest during preflowering stages (Acton et al. 1985; Liersch et al. 1986; Woerdenbag et al. 1991; El-Sohly 1990; Woerdenbag et al. 1994), others reported artemisinin reaching its peak during flowering (Singh et al. 1988; Pras et al. 1991; Morales et al. 1993; Ferreira et al. 1995a; Laughlin 1995). Artemisinin reached its peak during full flowering in a Chinese clone for both greenhouse and field conditions (Ferreira et al. 1995b).

Reports on the distribution of artemisinin throughout the plant have been inconsistent. Artemisinin has been reported to be higher at the top of the plant in some clones (Charles et al. 1990; Laughlin 1995) and equally distributed in others (Laughlin 1995). We analyzed six clones derived from Chinese material during both vegetative and flowering stages with samples taken from the bottom, middle, and top parts of the plant, and found a relatively even distribution of artemisinin along the main stem (Table 2). Five of six clones showed the same or higher levels of artemisinin at the flowering stage. Although plants from all clones were harvested at the same date, they were in different stages of development. Branches of one clone (clone 1) were collected sequentially from the bottom to the top of the plant. In this clone artemisinin showed a slight increase in artemisinin toward the top of the plant (Table 3).

In Vitro

Artemisinin is produced by differentiated (shoots + roots) shoot cultures (Martinez and Staba 1988; Fulzele et al. 1991; Whipkey et al. 1992; Ferreira and Janick 1996b) but occur only in trace levels, if at all, in shoots without roots (Martinez and Staba 1988; Jha et al. 1988; Fulzele et al. 1991; Woerdenbag et al. 1993; Paniego and Giuliette 1994). Brown (1994) reported 0.0038% of artemisinin being produced by callus cultures bearing shoots but did not specify whether or not they had roots. Most workers (Martinez and Staba 1988; Tawfiq et al. 1989; Fulzele et al. 1991; Kim et al. 1992) did not detect artemisinin in roots, although Nair et al. (1986) and Jha et al. (1988) reported trace amounts. Weathers et al. (1994) reported high levels (0.4%) of artemisinin in hairy root cultures of A. annua transformed with Agrobacterium rhyzogenes, but this was not confirmed by Jaziri et al. (1995). With the exception of Jha et al. (1988), most authors (He et al. 1983; Nair et al. 1986; Tawfiq et al. 1989; Fulzele et al. 1991; Kim et al. 1992; Woerdenbag et al. 1993; Brown 1994; Paniego and Giuliette 1994; Ferreira and Janick 1996b) either reported no artemisinin or only trace amounts produced from callus, cell, or the spent liquid media from these cultures. Ferreira and Janick (1996b) presented evidence that artemisinin production in shoots is enhanced by the presence of roots. The highest levels of artemisinin (0.287% DW) were obtained in hormone-free medium when the root production was maximized.

GLANDULAR TRICHOMES AS SITES OF ARTEMISININ ACCUMULATION

Although artemisinin immunolocalization has not been achieved, there is strong circumstantial evidence that the compound is sequestered in the glandular trichomes (Fig. 1D, E and 2A). Duke and Paul (1993) and Duke et al. (1994) described the development of such glands in leaves of A. annua and reported that neither artemisinin or artemisitene were detected from a glandless biotype and that virtually all artemisinin could be extracted by a 5-sec leaf dip in chloroform, without visible damage to other leaf epidermal cells, from the biotype with glands. Fig. 2B shows florets of the glandless biotype photographed under the scanning electron microscope.

Artemisinin content (% DW) was shown to be 4 to 11 times higher in the inflorescences as compared to leaves (Ferreira et al. 1995a) and the presence and development of glandular trichomes in the inflorescences was associated with artemisinin production based on extraction studies (Ferreira and Janick 1995). The glandular trichomes are more prominent in the corolla and receptacles florets than in leaves, stems, or bracts. Although these glands are present since the early stage of development on both leaves and inflorescences, artemisinin increases at anthesis, suggesting that it accumulates as the glands reach physiological maturity, a stage which coincides with the end of cell expansion in floret development. As glands approach maturity, there appears to be a cellular discharge into the subcuticular space around the apical cells and the contents are spread over the epidermis when the glands burst. After anthesis, artemisinin decreases and so does the number of intact glands. The association of artemisinin with glandular trichomes sequestration explains why artemisinin was not detected in parts of the plant that do not bear glands, such as pollen or roots (Ferreira et al. 1995a) or in a glandless biotype (Duke et al. 1994). Glandular trichomes are observed in leaves and stems of differentiated shoot cultures and artemisinin content of shoot cultures in vitro was similar to artemisinin content in vegetative clones grown in the greenhouse (Ferreira et al. 1995b).

CONCLUSION

Artemisinin is a sesquiterpene lactone which is produced both in vivo and differentiated in vitro cultures, by Artemisia annua and is equally distributed throughout the plant. Artemisinin appears to be sequestered in glandular trichomes which occur in stems, leaves, and inflorescences. The association of peak artemisinin with flowering is related to the abundance of glandular trichomes in the inflorescence, particularly florets and receptacle. In vitro studies indicate that the biosynthesis of artemisinin is enhanced by the presence of roots. Artemisia annua is unlikely to be produced economically by chemical synthesis or by in vitro production, thus A. annua is a potential new antimalarial crop for temperate areas.

REFERENCES

*We acknowledge the Southern Weed Science Laboratory (USDA/ARS), for the use of their electron microsopy facilities, and Rex N. Paul for taking the pictures.
Table 1. Artemisinin content of different organs and structures of greenhouse- and field- grown Artemisia annua, determined by HPLC-EC (Source: Ferreira et al. 1995a).

Artemisinin (% DW x 1000)
Organ/
structure		 Greenhouse Field
Leaves 3-30 6-60
Main stems 0-3 0.4-7
Side stems 0 0.4-14
Roots 0 0
Flowers 12-42 104-264
Pollen 0 NDz
Seed husks nd 116
Seedsy 36 81
zNot determined.
yContaining floral debris.

Table 2. Artemisin content of six clones of Artemisia annua by location on the plant and date.

Artemisinin (% DW)
Clone Datez Base Middle Top Mean
1 Sept 7 0.022 0.014 0.019 0.018
Oct 8 0.044 0.055 0.054 0.051
2 Sept 7 0.020 0.023 0.019 0.021
Oct 8 0.079 0.070 0.094 0.081
3 Sept 7 0.043 0.043 0.054 0.046
Oct 8 0.034 0.026 0.023 0.027
4 Sept 7 0.080 0.078 0.080 0.079
Oct 8 0.081 0.146 0.143 0.123
5 Sept 7 0.100 0.127 0.107 0.111
Oct 8 0.086 0.123 0.127 0.112
6 Sept 7 0.163 0.153 0.146 0.154
Oct 8 0.245 0.181 0.234 0.220
7 Sept 7 0.071 0.073 0.071 0.072
Oct 8 0.095 0.100 0.112 0.102
Mean Sept 7 0.071 ay 0.073 a 0.071 a 0.072a
Oct 8 0.095 b 0.100 b 0.112 b 0.102b
zAll plants vegetative Sept 7; all plants flowering Oct 8.
yMean separation in rows and columns by Duncan's Multiple Range Test (5% level).

Table 3. Artemisinin content of Artemisia annua branches (base to apex) of clone 1 harvested Oct. 8, 1994.

Positionz Artemisinin content
(% DW)
1 (apex) 0.040
2 0.046
3 0.043
4 0.041
5 0.033
6 0.037
7 (base) 0.023
zFive branches were composited at each position.

Fig. 1. Floral morphology of Artemisia annua. A Nodding capitulum.

B. Expanded capitulum showing calyx with imbricated bracts (b), receptacle (r), marginal pistillate floret (p), and central hermaphroditic (h) florets. Glandular trichomes are found abundantly on the receptacle, bracts, and florets of the capitulum.

C. Cross-section of the involucre showing imbrication of bracts. D. Unexpanded floret showing orientation of glandular trichomes. E. Fully developed, turgid, glandular trichome, based on SEM.

F. Details of hermaphroditic floret with lobed anthers attached to basal portion of the corolla (c), pistil with bifid stigma (s), style (st), and ovary (o). Note that in a hermaphroditic floret, the stigma reaches this state of development only after pollen shed.

G. Tricolpate pollen grain with vestigial spines, characteristic of wind-pollinated species, and germination pores (gp) bulging from the furrows. H. Pollen cross section based on light microscopy shows details of bulging germination pores. (Source: Ferreira and Janick 1995).

Fig. 2. Florets of Artemisia annua pictured by JEOL JSM840 SEM, at 5KV. Top, hermaphroditic florets of normal biotype showing glandular trichomes. Bottom, hermaphroditic florets of glandless biotype. Bar size = 100 µm.
Last update August 24, 1997 aw