J Chem Ecol (2014) 40:100–117 DOI 10.1007/s10886-014-0384-6
REVIEW ARTICLE
Potential Ecological Roles of Artemisinin Produced by Artemisia annua L. Karina Knudsmark Jessing & Stephen O. Duke & Nina Cedergreeen
Received: 18 June 2013 / Revised: 16 October 2013 / Accepted: 21 January 2014 / Published online: 6 February 2014 # Springer Science+Business Media New York 2014
Abstract Artemisia annua L. (annual wormwood, Asteraceae) and its secondary metabolite artemisinin, a unique sesquiterpene lactone with an endoperoxide bridge, has gained much attention due to its antimalarial properties. Artemisinin has a complex structure that requires a significant amount of energy for the plant to synthesize. So, what are the benefits to A. annua of producing this unique compound, and what is the ecological role of artemisinin? This review addresses these questions, discussing evidence of the potential utility of artemisinin in protecting the plant from insects and other herbivores, as well as pathogens and competing plant species. Abiotic factors affecting the artemisinin production, as well as mechanisms of artemisinin release to the surroundings also are discussed, and new data are provided on the toxicity of artemisinin towards soil and aquatic organisms. The antifungal and antibacterial effects reported are not very pronounced. Several studies have reported that extracts of A. annua have insecticidal effects, though few studies have proven that artemisinin could be the single compound responsible for the observed effects. However, the pathogen(s) or insect(s) that may have provided the selection pressure for the evolution of artemisinin synthesis may not have been represented in the research thus far conducted. The relatively high level of phytotoxicity of artemisinin in soil indicates that
Electronic supplementary material The online version of this article (doi:10.1007/s10886-014-0384-6) contains supplementary material, which is available to authorized users. K. Knudsmark Jessing (*) : N. Cedergreeen Department of Plant and Environmental Sciences, University of Copenhagen, Thorvaldsensvej 40, 1871 Frederiksberg C, Denmark e-mail: [email protected] S. O. Duke United States Department of Agriculture, Natural Product Utilization Research, University of Mississippi, University, MS 38677-8048, USA
plant/plant allelopathy could be a beneficial function of artemisinin to the producing plant. The release routes of artemisinin (movement from roots and wash off from leaf surfaces) from A. annua to the soil support the rationale for allelopathy. Keywords Allelopathy . Antifeedant . Artemisinin . Ecological role . Fungicide . Sesquiterpene
Introduction Artemisia annua L. (annual wormwood, Asteraceae) synthesizes and accumulates artemisinin (Fig. 1, Klayman 1985), a unique sesquiterpene lactone with an endoperoxide bridge. No other plant species is known to produce this compound. Artemisia annua and artemisinin have gained much attention during the last four decades due to the antimalarial properties of artemisinin against chloroquine-resistant strains of Plasmodium falciparum (Klayman 1985). At present, total chemical synthesis (Abdin et al. 2003) or in vitro production (Arsenault et al. 2008) of the relatively complex structure artemisinin is not economically feasible, and cultivation of the plant is still the only cost effective source of artemisinin. Synthesis is possible from one of the precursors of artemisinin, dihydroartemisinic acid (Levesque and Seeberger 2012), making artemisinin medication more affordable, but the only economical source of artemisinic acid is extraction from field-grown A. annua. In addition to antimalarial activity, artemisinin and semisynthetic derivatives have activity against cancer cells, schistosomiasis, and certain viral diseases, all of great interest in human pharmacology and reviewed elsewhere (e.g., Efferth et al. 2008). Artemisia annua is of Asian origin, and the plant now is widely dispersed throughout worldwide temperate regions. It has become naturalized in many countries, including
J Chem Ecol (2014) 40:100–117
Fig. 1 Glandular trichome from Artemisia annua at the mature stage where the subcuticular space (indicated by arrow) is expanded and filled with secondary compounds, including artemisinin. Bar size=10 μm. From Ferreira and Janick (1995)
Australia, Argentina, Bulgaria, France, Hungary, Romania, Italy, Spain, and the United States (Dhingra et al. 2000b; Klayman 1993). For the production of artemisinin, A. annua is cropped on a large scale in China, Vietnam, Turkey, Iran, Afghanistan, and Australia (Bhakuni et al. 2001), and in Africa, where the area cultivated with A. annua was 5000 ha in 2009 (personal communication, Jessing et al. 2011). In Eastern Europe the plant also is cropped for its essential oils (Heemskerk et al. 2006). Artemisia annua is cultivated for experimental purposes in The Netherlands, Switzerland, Finland (Laughlin et al. 2002), and in Denmark. Artemisia annua is reported to grow well in all these regions, with varying artemisinin contents. Plant-produced endoperoxides are rare, with only a few known examples in addition to artemisinin. These are ascaridole from Chenopodium (Pollack et al. 1990) and yingzhaosu A from Artabotrys uncinatus (Szpilman et al. 2005). These are structurally complex compounds that require a significant amount of energy for plants to synthesize. So, what are the benefits to A. annua for producing this complex and unique molecule, and what is the ecological role of artemisinin? Most secondary compounds produced by plants are believed to be part of a defense mechanism towards herbivores, including insect pests, competing plants, and fungal or bacterial pathogens (Taiz and Zeiger 2002). Secondary metabolites also can have other ecological roles such as chelation of essential nutrients and regulation of soil biota in a way that affects soil fertility. In addition to interactions with other organisms in the ecosystem, secondary metabolite production can be influenced by abiotic factors (Inderjit et al. 2011). Knowledge of such abiotic interactions might be helpful in determination of ecological functions. To evaluate the potential ecological role of artemisinin, this review starts with information on the production of artemisinin in A. annua. Then, it provides an overview of abiotic factors known to affect artemisinin production, followed by a survey of the present knowledge on activities of artemisinin or A. annua extracts towards insects and a few other invertebrates, plants, fungi, and bacteria of ecological
101
interest. New data (see Supplemenatary material) on the toxicity of artemisinin towards soil and aquatic organisms and fungi are provided. A purpose of the review is to propose which organisms are the most likely targets of artemisinin, based on available literature and our new data. Another objective of the review is to emphasize gaps in knowledge that might point the way toward further experimentation that could provide a more complete understanding of the role of artemisinin in the chemical ecology of A. annua. We conclude the review with a discussion of the most likely ecological function(s) of artemisinin production for A. annua. Although the literature on the role of artemisinin as an anitmalarial drug is robust, evidence for its ecological role and benefit for the producing plant is limited. This review addresses this question, discussing existing and new information on the potential utility of artemisinin in protecting the plant from pathogens and insects, as well as competing plant species.
Sites of Artemisinin Accumulation Artemisinin is stored in glandular trichomes on the surface of the leaves and stems (Duke and Paul 1993) and on the corolla and receptacles of the florets (Ferreira and Janick 1995). The glandular trichomes consist of ten cells, of which the two most apical cells form a bilobed sac (Fig. 1) by the filling and expansion of the space between the external cell walls and cuticle with secretory products (Duke and Paul 1993; Duke et al. 1994). The subcuticular space of this sac stores secondary metabolites, including artemisinin. In addition to artemisinin, A. annua synthesizes and accumulates several other terpenoid secondary metabolites (Tellez et al. 1999). More than 600 secondary metabolites have been reported in A. annua. These cover several classes of bioactive compounds, including monoterpenes, sesquiterpenes, triterpenoids, flavonoids, and coumarins (Brown 2010). More than 15 of the terpenoids, including artemisinin, artemisitene, and several common monoterpenoids such as 1,8-cineole and pinocarone (Table 1), are localized almost entirely in the trichomes of the shoot (Duke et al. 1994; Duke et al. 2000; Tellez et al. 1999). Trichome localization of secondary compounds also has been observed in trichomes of tobacco (Wagner 1991) and cotton (Bell et al. 1987; Elzen et al. 1985), where trichomes were the exclusive or almost exclusive site of accumulation of certain terpenoids. The cuticles of A. annua trichomes burst as they mature, and artemisinin and the other secondary metabolites are leaked to the leaf surface (Duke and Paul 1993). From here, artemisinin can be washed off and reach the surrounding environment (Jessing et al. 2013). The more volatile monoterpenoid compounds found in the trichomes, such as 1,8-cineole, are probably lost to the atmosphere at this time.
102
J Chem Ecol (2014) 40:100–117
Table 1 Some of the constituents of essential oils of glanded and glandless biotypes of Artemisia annua L. (data taken from Tellez et al. 1999) Compounds
Glanded Glandless (Relative area to total peak area)
α-pinene camphene β-pinene dehydro-1,8-cineole yomogi alcohol 1,8-cineole artemesia ketone artemisia alcohol isophorone
26. 0.6 1.6 0.2 0.2 8.4 11.0 0.4 0.7
0 0 0 0 0 0 0 0 0
α-campholenal trans-pinocarveol pinocarvone borneol terpin-4-ol myrtenal myrtenol tridecene α-copaene β-caryophyllene α-humulene β-farnesene germacene-D bicyclogermacrene
0.5 9.0 15.8 0.5 0.6 0.7 0.4 0.2 0.4 2.6 0.1 0.7 6.1 0.3
0 0 0 0 0 0 0 0 1.9 25.1 1.3 5.4 49.8 3.5
The quantitative significance of root production of artemisinin by A. annua has been widely discussed. Most researchers have not detected artemisinin in roots (Fulzele et al. 1991; Tawfiq et al. 1989), but Nair et al. (1986) reported trace amounts, and Woerdenbag et al. (1991) found low levels of artemisinin and its precursors in secondary roots. Using silicone tubes to trap the compound in soil, Jessing et al. (2013) measured artemisinin at nanogram levels by in situ microextraction of artemisinin in pots with A. annua, where the roots were the only possible artemisinin source. Hence, roots do secrete artemisinin at relatively low levels, compared to the amount produced by trichomes. Under artificial conditions, hairy root cultures from A. annua can produce artemisinin in the mg/g root weight range (Liu et al. 2002; Wang and Tan 2002; Wang et al. 2001). It is possible that only root hairs produce the compounds, as root hairs are analogous to trichomes on shoots, and some plant anatomists now categorize root hairs as a specialized type of trichome, with both structures being under very similar genetic regulation (Dayan and Duke 2003; Kellogg 2001). If so, artemisinin might be secreted into the rhizosphere much like the potent phytotoxin sorgoleone, which is produced only by root hairs of Sorghum
spp. and is secreted as soon as it is formed, thereby avoiding autotoxicity (Czarnota et al. 2003; Dayan et al. 2010).
Abiotic Factors Influencing Artemisinin Production The content of artemisinin is usually in the range of 0.01– 0.4 % of the plant dry weight (Janick and Ferreira 1996). Some new varieties can produce as high as 1.6 % artemisinin (Simonnet et al. 2011). The plant content of artemisinin is partly determined by genetic factors (Ferreira et al. 2005), but also by abiotic factors. Evaluating the role of artemisinin in the context of its abiotic environment should improve our understanding of artemisinin production and release. Production of artemisinin usually peaks with flowering. Light influences artemisinin production, as flowering is induced by a photoperiod shorter than 13.3 h (Ferreira et al. 1995). Production of some secondary metabolites can be induced by low soil or low plant tissue concentrations of nutrients (Inderjit et al. 2011). Increasing the nitrogen supply increased leaf biomass production of A. annua, while artemisinin production was unaffected (Magalhães et al. 1996). Reduced nitrogen, as well as phosphorus and lime deficiency, caused an increase in artemisinin production measured in g per 100 g biomass. Severe potassium deficiency caused a 75 % increase in artemisinin production (per biomass) (Ferreira 2007). Artemisia annua deficient in iron, manganese, zinc, and copper produced 25–30 % less artemisinin (Srivastava and Sharma 1990), whereas a mild deficiency of boron (0.5 and 1 mM) caused increased artemisinin production measured per biomass (Aftab et al. 2010). Both lead- and salt-induced stress causes increased artemisinin production as well (Qian et al. 2007; Qureshi et al. 2005). A possible explanation for the observed correlation between increased artemisinin production and nutrient deficiency or artemisinin production and salt-induced stress, is that abiotic stresses like nutrient deficiencies cause oxidative stress, which in A. annua might cause rapid conversion of the precursors artemisinic acid and dihydroartemisinic acid to artemisinin by activated oxygen species (Qureshi et al. 2005). Moderate water deficiency over a short period of time (38 h) also increased artemisinin production (Marchese et al., 2010), as did night frost (Wallaart et al. 2000). Short exposures (3 h) of in vitro-propagated A. annua plantlets to UV-B radiation increased artemisinin production, at least partly due to larger trichomes (Pandey and Pandey-Rai 2013), With the present knowledge on abiotic factors affecting artemisinin production, there do not seem to be specific correlations explaining the ecological role. The content of artemisinin might also potentially respond to biotic stress factors, like herbivores, competing plants, or pathogen attacks. Effects of artemisinin on biota to which A. annua might be exposed are reviewed in the following sections.
J Chem Ecol (2014) 40:100–117
Biological Activity of Artemisia annua Compounds and Artemisinin Effects on Insects and Other Invertebrates Problems with synthetic pesticides have encouraged researchers to look for natural plant protection compounds, such as botanical insecticides. Such products are assumed or proven to be effective towards the natural enemies of the producing plants. This has led to studies of the potential of extracts of A. annua leaves or essential oil derived from the leaves as potential insect management agents. Insecticidal effects of artemisinin or extracts of A. annua, where artemisinin could be the primary bioactive compound, are listed in Table 2. In Table 2, the estimated effective concentration or artemisinin is calculated on the basis of its average content in leaves of 0.010.4 % when the amount of leaf material used in the experiments have been given. The following section is sub-divided by method of exposure: Dry A. annua leaves, leaf extracts, essential oils, pure artemisinin, and field experiments with cultures of A. annua. Effects of Dry A. annua Leaves The idea that the bioactive constitutions in dry A. annua leaves can be used as an insecticidal agent with low mammal toxicity in feed and food products was tested by Brisibe et al. (2011). They compared the effect of dried A. annua leaves with the effect of a synthetic insecticide, pirimiphos-methyl dust (2 % a.i.) in controlling bruchids (Callosobruchus maculatus), also known as the cowpea weevil, by mixing them into cowpea seeds, an important worldwide protein source that suffers from bruchid attacks during storage. The leaves of A. annua were not as efficient as the commercial insecticide, but a significant increase in mortality was observed at the lowest concentration tested of 0.01 g leaves per g cowpea seeds. This corresponds to an artemisinin content of 0.1–4 μg/g. In addition, the number of bruchid eggs was reduced almost 7-fold at a concentration of 8 mg/g cowpea seeds (Brisibe et al. 2011). In a study like this, it is uncertain which of the compounds in the leaves are responsible for the observed effects. The observed mortality can be due both to direct toxicity after intake of the leaves or starvation because of feeding deterrence, whereas the reduction in eggs is more likely to be a direct effect after intake. Effects of Leaf Extracts Several studies have evaluated the effects of mainly methanolic A. annua leaf extracts. Insect development was prolonged in 4th instar larvae of the lesser mulberry pyralid (Glyphodes pyloalis Walker) by an A. annua leaf extract applied topically at a concentration equivalent to 0.33 g leaf/L methanol, suggesting hormone-like activity of A. annua extract. This concentration killed half of the population (LC50 concentration). In the same treatment, fecundity
103
and egg hatchability also were decreased, and significant effects were measured on extracted α-amylase, protease, lipase, esterase, and gluthatione S-transferase activities of the insect (Khosravi et al. 2011). If the mortality observed in this study were due solely to the activity of artemisinin, the LC50 would correspond to 0.033–1.32 mg artemisinin/L, with the average artemisinin content of 0.01–0.4 %. In the three following studies of Zibaee and Bandani (2010), Shekari et al. (2008), and Hasheminia et al. (2011), the leaf extracts were prepared as 30 g dried leaves in 300 ml methanol. The evaporated residue of this extract was dissolved in 10 mlacetone or methanol and used as a stock solution. The estimated artemisinin concentration of this stock is 0.03–1.2 g/L. The insecticidal activities of the extract on the sunn pest (also known as corn bug) (Eurygaster intefriceps) were LC50 values of 32 and 17 % extract after 24 and 48 h of exposure, respectively (Zibaee and Bandani 2010). In this study, effects of A. annua extracts also could be measured on extracted enzyme activities. Extractable acetylcholinesterase activity was reduced with increasing extract concentration, and xenobiotic metabolism was affected, as changes could be measured in extractable esterases, glutathione-S-transferases, and phosphatase activities (Zibaee and Bandani 2010). Shekari et al. (2008) tested the extract against the elm leaf beetle (Xanthogaleruca luteola Mull.). The LC50 values in this study were 48 and 44 % extract after 24 and 48 h, respectively, and significant feeding deterrence was observed after 24 h of exposure at the lowest tested concentration of 0.625 % extract. At 10 % extract, growth regulatory disturbances, such as deformed wings and prolongation of the larval and pupae stages were observed. Fecundity and hatchability was reduced at 5 % extract, the only concentration where these parameters were tested (Shekari et al. 2008). In a similar setup, the 3rd instars of Pieris rapae had a LC50 value of 9.4 % A. annua extract, and 30 % feeding deterrence was observed at 0.625 % A. annua extract, which was the lowest tested concentration (Hasheminia et al. 2011). In the same study, the extract had no effect on developmental time of P. rapae, but changes in both enzymatic and non-enzymatic metabolic processes could be measured. In the studies by Hasheminia et al. (2011) and Shekari et al. (2008), the extract concentration causing feeding deterrence was much lower than the amount of extract having lethal effects. Several of the bioactive compounds reported in A. annua have distinct odors, which might cause feeding deterrence. As the A. annua leaves used in all these experiments were washed in water before they were dried and extracted, there is a risk that the trichomes were damaged and the content of bioactive compounds, including artemisinin, were reduced. This was not the case in the study by Maggi et al. (2005) who investigated the antifeedant activity of A. annua extracts obtained by Soxhlet extraction with ethanol against the leaf feeding coccinellidae Epilachna paenulata and the southern armyworm (Spodoptera
104
J Chem Ecol (2014) 40:100–117
Table 2 Insecticidal effects of artemisinin or extracts of Artemisia annua, where the effect could be due to artemisinin itself, some of the other bioactive compounds in A. annua or synergy between artemisinin and some of the other bioactive compounds in A. annua Insect/stage
Endpoint
Effect formulation or concentration
Cowpea brunchid Callosobruchus maculatus F Cowpea brunchid Callosobruchus maculatus F Lesser mulberry pyralid, Glyphodes pyloalis 4th instar larvae Sunn pest (Eurygaster Integriceps Puton)
Significant increased mortality
2.5 g leaves/250 g cowpea 0.1–4 μg/g seeds 20 g leaves/250 g cowpea 0.8–32 μg/g seeds 0.33 mg leaf/mL 0.033–1.32 mg/L
Elm leaf Beetle (Xanthogaleruca luteola Mull.) Elm leaf Beetle (Xanthogaleruca luteola Mull.) Elm leaf Beetle (Xanthogaleruca luteola Mull.) Small white Pieris rapae L. 3th instar larvae Small white Pieris rapae L. 3th instar larvae Epilachna paenulata,
LC50
7-fold decrease in eggs LC50 by topical application
Estimated or given artemisinin concentrations
Source
Brisibe et al. 2011 Brisibe et al. 2011 Khosravi et al. 2011
59 % feeding deterrent
32 % (24 h) and 17 % 9.6–380 μg/mL (24 h) and Zibaee and Bandani (48 h), methanolic 50–200 mg/L (48 h) 2010 extract 48 % (24 h) and 44 % 14.4–576 μg/mL (24 h) Shekari et al. 2008 (48 h), methanolic and 13.2–528 mg/L extract (48 h) 0.625 % methanolic extract 0.18–7.5 mg/L Shekari et al. 2008
Reproduction
10 % methanolic extract
3-120 m/L
Shekari et al. 2008
LC50
9.4 % extract
2.8–113 mg/L
30 % feeding deterrence
0.625 % extract
0.18–7.5 mg/L
1.5 mg/cm2
36 μg/cm2
Hasheminia et al. 2011 Hasheminia et al. 2011 Maggi et al. 2005
30 μg/cm2
Maggi et al. 2005
1.5 mg/cm2
36 μg/cm2
Maggi et al. 2005
1.5 mg/cm2 1.5 mg/cm2 1 g/L
36 μg/cm2 36 μg/cm2 0.2 mg/L
Maggi et al. 2005 Maggi et al. 2005 Durden et al. 2011
?
10 g/L
Durden et al. 2011
LC50
Complete feeding rejection with ethanol extract of A. annua Epilachna paenulata Complete feeding rejection, artemisinin Spodotera eridania, 87 % feeding inhibition with ethanol extract of A. annua Epilachna paenulata 100 % mortality in no-choice Spodotera eridania 50 % mortality in no-choice Codling moth, Cydia pomonella Feeding deterrent at P
REVIEW ARTICLE
Potential Ecological Roles of Artemisinin Produced by Artemisia annua L. Karina Knudsmark Jessing & Stephen O. Duke & Nina Cedergreeen
Received: 18 June 2013 / Revised: 16 October 2013 / Accepted: 21 January 2014 / Published online: 6 February 2014 # Springer Science+Business Media New York 2014
Abstract Artemisia annua L. (annual wormwood, Asteraceae) and its secondary metabolite artemisinin, a unique sesquiterpene lactone with an endoperoxide bridge, has gained much attention due to its antimalarial properties. Artemisinin has a complex structure that requires a significant amount of energy for the plant to synthesize. So, what are the benefits to A. annua of producing this unique compound, and what is the ecological role of artemisinin? This review addresses these questions, discussing evidence of the potential utility of artemisinin in protecting the plant from insects and other herbivores, as well as pathogens and competing plant species. Abiotic factors affecting the artemisinin production, as well as mechanisms of artemisinin release to the surroundings also are discussed, and new data are provided on the toxicity of artemisinin towards soil and aquatic organisms. The antifungal and antibacterial effects reported are not very pronounced. Several studies have reported that extracts of A. annua have insecticidal effects, though few studies have proven that artemisinin could be the single compound responsible for the observed effects. However, the pathogen(s) or insect(s) that may have provided the selection pressure for the evolution of artemisinin synthesis may not have been represented in the research thus far conducted. The relatively high level of phytotoxicity of artemisinin in soil indicates that
Electronic supplementary material The online version of this article (doi:10.1007/s10886-014-0384-6) contains supplementary material, which is available to authorized users. K. Knudsmark Jessing (*) : N. Cedergreeen Department of Plant and Environmental Sciences, University of Copenhagen, Thorvaldsensvej 40, 1871 Frederiksberg C, Denmark e-mail: [email protected] S. O. Duke United States Department of Agriculture, Natural Product Utilization Research, University of Mississippi, University, MS 38677-8048, USA
plant/plant allelopathy could be a beneficial function of artemisinin to the producing plant. The release routes of artemisinin (movement from roots and wash off from leaf surfaces) from A. annua to the soil support the rationale for allelopathy. Keywords Allelopathy . Antifeedant . Artemisinin . Ecological role . Fungicide . Sesquiterpene
Introduction Artemisia annua L. (annual wormwood, Asteraceae) synthesizes and accumulates artemisinin (Fig. 1, Klayman 1985), a unique sesquiterpene lactone with an endoperoxide bridge. No other plant species is known to produce this compound. Artemisia annua and artemisinin have gained much attention during the last four decades due to the antimalarial properties of artemisinin against chloroquine-resistant strains of Plasmodium falciparum (Klayman 1985). At present, total chemical synthesis (Abdin et al. 2003) or in vitro production (Arsenault et al. 2008) of the relatively complex structure artemisinin is not economically feasible, and cultivation of the plant is still the only cost effective source of artemisinin. Synthesis is possible from one of the precursors of artemisinin, dihydroartemisinic acid (Levesque and Seeberger 2012), making artemisinin medication more affordable, but the only economical source of artemisinic acid is extraction from field-grown A. annua. In addition to antimalarial activity, artemisinin and semisynthetic derivatives have activity against cancer cells, schistosomiasis, and certain viral diseases, all of great interest in human pharmacology and reviewed elsewhere (e.g., Efferth et al. 2008). Artemisia annua is of Asian origin, and the plant now is widely dispersed throughout worldwide temperate regions. It has become naturalized in many countries, including
J Chem Ecol (2014) 40:100–117
Fig. 1 Glandular trichome from Artemisia annua at the mature stage where the subcuticular space (indicated by arrow) is expanded and filled with secondary compounds, including artemisinin. Bar size=10 μm. From Ferreira and Janick (1995)
Australia, Argentina, Bulgaria, France, Hungary, Romania, Italy, Spain, and the United States (Dhingra et al. 2000b; Klayman 1993). For the production of artemisinin, A. annua is cropped on a large scale in China, Vietnam, Turkey, Iran, Afghanistan, and Australia (Bhakuni et al. 2001), and in Africa, where the area cultivated with A. annua was 5000 ha in 2009 (personal communication, Jessing et al. 2011). In Eastern Europe the plant also is cropped for its essential oils (Heemskerk et al. 2006). Artemisia annua is cultivated for experimental purposes in The Netherlands, Switzerland, Finland (Laughlin et al. 2002), and in Denmark. Artemisia annua is reported to grow well in all these regions, with varying artemisinin contents. Plant-produced endoperoxides are rare, with only a few known examples in addition to artemisinin. These are ascaridole from Chenopodium (Pollack et al. 1990) and yingzhaosu A from Artabotrys uncinatus (Szpilman et al. 2005). These are structurally complex compounds that require a significant amount of energy for plants to synthesize. So, what are the benefits to A. annua for producing this complex and unique molecule, and what is the ecological role of artemisinin? Most secondary compounds produced by plants are believed to be part of a defense mechanism towards herbivores, including insect pests, competing plants, and fungal or bacterial pathogens (Taiz and Zeiger 2002). Secondary metabolites also can have other ecological roles such as chelation of essential nutrients and regulation of soil biota in a way that affects soil fertility. In addition to interactions with other organisms in the ecosystem, secondary metabolite production can be influenced by abiotic factors (Inderjit et al. 2011). Knowledge of such abiotic interactions might be helpful in determination of ecological functions. To evaluate the potential ecological role of artemisinin, this review starts with information on the production of artemisinin in A. annua. Then, it provides an overview of abiotic factors known to affect artemisinin production, followed by a survey of the present knowledge on activities of artemisinin or A. annua extracts towards insects and a few other invertebrates, plants, fungi, and bacteria of ecological
101
interest. New data (see Supplemenatary material) on the toxicity of artemisinin towards soil and aquatic organisms and fungi are provided. A purpose of the review is to propose which organisms are the most likely targets of artemisinin, based on available literature and our new data. Another objective of the review is to emphasize gaps in knowledge that might point the way toward further experimentation that could provide a more complete understanding of the role of artemisinin in the chemical ecology of A. annua. We conclude the review with a discussion of the most likely ecological function(s) of artemisinin production for A. annua. Although the literature on the role of artemisinin as an anitmalarial drug is robust, evidence for its ecological role and benefit for the producing plant is limited. This review addresses this question, discussing existing and new information on the potential utility of artemisinin in protecting the plant from pathogens and insects, as well as competing plant species.
Sites of Artemisinin Accumulation Artemisinin is stored in glandular trichomes on the surface of the leaves and stems (Duke and Paul 1993) and on the corolla and receptacles of the florets (Ferreira and Janick 1995). The glandular trichomes consist of ten cells, of which the two most apical cells form a bilobed sac (Fig. 1) by the filling and expansion of the space between the external cell walls and cuticle with secretory products (Duke and Paul 1993; Duke et al. 1994). The subcuticular space of this sac stores secondary metabolites, including artemisinin. In addition to artemisinin, A. annua synthesizes and accumulates several other terpenoid secondary metabolites (Tellez et al. 1999). More than 600 secondary metabolites have been reported in A. annua. These cover several classes of bioactive compounds, including monoterpenes, sesquiterpenes, triterpenoids, flavonoids, and coumarins (Brown 2010). More than 15 of the terpenoids, including artemisinin, artemisitene, and several common monoterpenoids such as 1,8-cineole and pinocarone (Table 1), are localized almost entirely in the trichomes of the shoot (Duke et al. 1994; Duke et al. 2000; Tellez et al. 1999). Trichome localization of secondary compounds also has been observed in trichomes of tobacco (Wagner 1991) and cotton (Bell et al. 1987; Elzen et al. 1985), where trichomes were the exclusive or almost exclusive site of accumulation of certain terpenoids. The cuticles of A. annua trichomes burst as they mature, and artemisinin and the other secondary metabolites are leaked to the leaf surface (Duke and Paul 1993). From here, artemisinin can be washed off and reach the surrounding environment (Jessing et al. 2013). The more volatile monoterpenoid compounds found in the trichomes, such as 1,8-cineole, are probably lost to the atmosphere at this time.
102
J Chem Ecol (2014) 40:100–117
Table 1 Some of the constituents of essential oils of glanded and glandless biotypes of Artemisia annua L. (data taken from Tellez et al. 1999) Compounds
Glanded Glandless (Relative area to total peak area)
α-pinene camphene β-pinene dehydro-1,8-cineole yomogi alcohol 1,8-cineole artemesia ketone artemisia alcohol isophorone
26. 0.6 1.6 0.2 0.2 8.4 11.0 0.4 0.7
0 0 0 0 0 0 0 0 0
α-campholenal trans-pinocarveol pinocarvone borneol terpin-4-ol myrtenal myrtenol tridecene α-copaene β-caryophyllene α-humulene β-farnesene germacene-D bicyclogermacrene
0.5 9.0 15.8 0.5 0.6 0.7 0.4 0.2 0.4 2.6 0.1 0.7 6.1 0.3
0 0 0 0 0 0 0 0 1.9 25.1 1.3 5.4 49.8 3.5
The quantitative significance of root production of artemisinin by A. annua has been widely discussed. Most researchers have not detected artemisinin in roots (Fulzele et al. 1991; Tawfiq et al. 1989), but Nair et al. (1986) reported trace amounts, and Woerdenbag et al. (1991) found low levels of artemisinin and its precursors in secondary roots. Using silicone tubes to trap the compound in soil, Jessing et al. (2013) measured artemisinin at nanogram levels by in situ microextraction of artemisinin in pots with A. annua, where the roots were the only possible artemisinin source. Hence, roots do secrete artemisinin at relatively low levels, compared to the amount produced by trichomes. Under artificial conditions, hairy root cultures from A. annua can produce artemisinin in the mg/g root weight range (Liu et al. 2002; Wang and Tan 2002; Wang et al. 2001). It is possible that only root hairs produce the compounds, as root hairs are analogous to trichomes on shoots, and some plant anatomists now categorize root hairs as a specialized type of trichome, with both structures being under very similar genetic regulation (Dayan and Duke 2003; Kellogg 2001). If so, artemisinin might be secreted into the rhizosphere much like the potent phytotoxin sorgoleone, which is produced only by root hairs of Sorghum
spp. and is secreted as soon as it is formed, thereby avoiding autotoxicity (Czarnota et al. 2003; Dayan et al. 2010).
Abiotic Factors Influencing Artemisinin Production The content of artemisinin is usually in the range of 0.01– 0.4 % of the plant dry weight (Janick and Ferreira 1996). Some new varieties can produce as high as 1.6 % artemisinin (Simonnet et al. 2011). The plant content of artemisinin is partly determined by genetic factors (Ferreira et al. 2005), but also by abiotic factors. Evaluating the role of artemisinin in the context of its abiotic environment should improve our understanding of artemisinin production and release. Production of artemisinin usually peaks with flowering. Light influences artemisinin production, as flowering is induced by a photoperiod shorter than 13.3 h (Ferreira et al. 1995). Production of some secondary metabolites can be induced by low soil or low plant tissue concentrations of nutrients (Inderjit et al. 2011). Increasing the nitrogen supply increased leaf biomass production of A. annua, while artemisinin production was unaffected (Magalhães et al. 1996). Reduced nitrogen, as well as phosphorus and lime deficiency, caused an increase in artemisinin production measured in g per 100 g biomass. Severe potassium deficiency caused a 75 % increase in artemisinin production (per biomass) (Ferreira 2007). Artemisia annua deficient in iron, manganese, zinc, and copper produced 25–30 % less artemisinin (Srivastava and Sharma 1990), whereas a mild deficiency of boron (0.5 and 1 mM) caused increased artemisinin production measured per biomass (Aftab et al. 2010). Both lead- and salt-induced stress causes increased artemisinin production as well (Qian et al. 2007; Qureshi et al. 2005). A possible explanation for the observed correlation between increased artemisinin production and nutrient deficiency or artemisinin production and salt-induced stress, is that abiotic stresses like nutrient deficiencies cause oxidative stress, which in A. annua might cause rapid conversion of the precursors artemisinic acid and dihydroartemisinic acid to artemisinin by activated oxygen species (Qureshi et al. 2005). Moderate water deficiency over a short period of time (38 h) also increased artemisinin production (Marchese et al., 2010), as did night frost (Wallaart et al. 2000). Short exposures (3 h) of in vitro-propagated A. annua plantlets to UV-B radiation increased artemisinin production, at least partly due to larger trichomes (Pandey and Pandey-Rai 2013), With the present knowledge on abiotic factors affecting artemisinin production, there do not seem to be specific correlations explaining the ecological role. The content of artemisinin might also potentially respond to biotic stress factors, like herbivores, competing plants, or pathogen attacks. Effects of artemisinin on biota to which A. annua might be exposed are reviewed in the following sections.
J Chem Ecol (2014) 40:100–117
Biological Activity of Artemisia annua Compounds and Artemisinin Effects on Insects and Other Invertebrates Problems with synthetic pesticides have encouraged researchers to look for natural plant protection compounds, such as botanical insecticides. Such products are assumed or proven to be effective towards the natural enemies of the producing plants. This has led to studies of the potential of extracts of A. annua leaves or essential oil derived from the leaves as potential insect management agents. Insecticidal effects of artemisinin or extracts of A. annua, where artemisinin could be the primary bioactive compound, are listed in Table 2. In Table 2, the estimated effective concentration or artemisinin is calculated on the basis of its average content in leaves of 0.010.4 % when the amount of leaf material used in the experiments have been given. The following section is sub-divided by method of exposure: Dry A. annua leaves, leaf extracts, essential oils, pure artemisinin, and field experiments with cultures of A. annua. Effects of Dry A. annua Leaves The idea that the bioactive constitutions in dry A. annua leaves can be used as an insecticidal agent with low mammal toxicity in feed and food products was tested by Brisibe et al. (2011). They compared the effect of dried A. annua leaves with the effect of a synthetic insecticide, pirimiphos-methyl dust (2 % a.i.) in controlling bruchids (Callosobruchus maculatus), also known as the cowpea weevil, by mixing them into cowpea seeds, an important worldwide protein source that suffers from bruchid attacks during storage. The leaves of A. annua were not as efficient as the commercial insecticide, but a significant increase in mortality was observed at the lowest concentration tested of 0.01 g leaves per g cowpea seeds. This corresponds to an artemisinin content of 0.1–4 μg/g. In addition, the number of bruchid eggs was reduced almost 7-fold at a concentration of 8 mg/g cowpea seeds (Brisibe et al. 2011). In a study like this, it is uncertain which of the compounds in the leaves are responsible for the observed effects. The observed mortality can be due both to direct toxicity after intake of the leaves or starvation because of feeding deterrence, whereas the reduction in eggs is more likely to be a direct effect after intake. Effects of Leaf Extracts Several studies have evaluated the effects of mainly methanolic A. annua leaf extracts. Insect development was prolonged in 4th instar larvae of the lesser mulberry pyralid (Glyphodes pyloalis Walker) by an A. annua leaf extract applied topically at a concentration equivalent to 0.33 g leaf/L methanol, suggesting hormone-like activity of A. annua extract. This concentration killed half of the population (LC50 concentration). In the same treatment, fecundity
103
and egg hatchability also were decreased, and significant effects were measured on extracted α-amylase, protease, lipase, esterase, and gluthatione S-transferase activities of the insect (Khosravi et al. 2011). If the mortality observed in this study were due solely to the activity of artemisinin, the LC50 would correspond to 0.033–1.32 mg artemisinin/L, with the average artemisinin content of 0.01–0.4 %. In the three following studies of Zibaee and Bandani (2010), Shekari et al. (2008), and Hasheminia et al. (2011), the leaf extracts were prepared as 30 g dried leaves in 300 ml methanol. The evaporated residue of this extract was dissolved in 10 mlacetone or methanol and used as a stock solution. The estimated artemisinin concentration of this stock is 0.03–1.2 g/L. The insecticidal activities of the extract on the sunn pest (also known as corn bug) (Eurygaster intefriceps) were LC50 values of 32 and 17 % extract after 24 and 48 h of exposure, respectively (Zibaee and Bandani 2010). In this study, effects of A. annua extracts also could be measured on extracted enzyme activities. Extractable acetylcholinesterase activity was reduced with increasing extract concentration, and xenobiotic metabolism was affected, as changes could be measured in extractable esterases, glutathione-S-transferases, and phosphatase activities (Zibaee and Bandani 2010). Shekari et al. (2008) tested the extract against the elm leaf beetle (Xanthogaleruca luteola Mull.). The LC50 values in this study were 48 and 44 % extract after 24 and 48 h, respectively, and significant feeding deterrence was observed after 24 h of exposure at the lowest tested concentration of 0.625 % extract. At 10 % extract, growth regulatory disturbances, such as deformed wings and prolongation of the larval and pupae stages were observed. Fecundity and hatchability was reduced at 5 % extract, the only concentration where these parameters were tested (Shekari et al. 2008). In a similar setup, the 3rd instars of Pieris rapae had a LC50 value of 9.4 % A. annua extract, and 30 % feeding deterrence was observed at 0.625 % A. annua extract, which was the lowest tested concentration (Hasheminia et al. 2011). In the same study, the extract had no effect on developmental time of P. rapae, but changes in both enzymatic and non-enzymatic metabolic processes could be measured. In the studies by Hasheminia et al. (2011) and Shekari et al. (2008), the extract concentration causing feeding deterrence was much lower than the amount of extract having lethal effects. Several of the bioactive compounds reported in A. annua have distinct odors, which might cause feeding deterrence. As the A. annua leaves used in all these experiments were washed in water before they were dried and extracted, there is a risk that the trichomes were damaged and the content of bioactive compounds, including artemisinin, were reduced. This was not the case in the study by Maggi et al. (2005) who investigated the antifeedant activity of A. annua extracts obtained by Soxhlet extraction with ethanol against the leaf feeding coccinellidae Epilachna paenulata and the southern armyworm (Spodoptera
104
J Chem Ecol (2014) 40:100–117
Table 2 Insecticidal effects of artemisinin or extracts of Artemisia annua, where the effect could be due to artemisinin itself, some of the other bioactive compounds in A. annua or synergy between artemisinin and some of the other bioactive compounds in A. annua Insect/stage
Endpoint
Effect formulation or concentration
Cowpea brunchid Callosobruchus maculatus F Cowpea brunchid Callosobruchus maculatus F Lesser mulberry pyralid, Glyphodes pyloalis 4th instar larvae Sunn pest (Eurygaster Integriceps Puton)
Significant increased mortality
2.5 g leaves/250 g cowpea 0.1–4 μg/g seeds 20 g leaves/250 g cowpea 0.8–32 μg/g seeds 0.33 mg leaf/mL 0.033–1.32 mg/L
Elm leaf Beetle (Xanthogaleruca luteola Mull.) Elm leaf Beetle (Xanthogaleruca luteola Mull.) Elm leaf Beetle (Xanthogaleruca luteola Mull.) Small white Pieris rapae L. 3th instar larvae Small white Pieris rapae L. 3th instar larvae Epilachna paenulata,
LC50
7-fold decrease in eggs LC50 by topical application
Estimated or given artemisinin concentrations
Source
Brisibe et al. 2011 Brisibe et al. 2011 Khosravi et al. 2011
59 % feeding deterrent
32 % (24 h) and 17 % 9.6–380 μg/mL (24 h) and Zibaee and Bandani (48 h), methanolic 50–200 mg/L (48 h) 2010 extract 48 % (24 h) and 44 % 14.4–576 μg/mL (24 h) Shekari et al. 2008 (48 h), methanolic and 13.2–528 mg/L extract (48 h) 0.625 % methanolic extract 0.18–7.5 mg/L Shekari et al. 2008
Reproduction
10 % methanolic extract
3-120 m/L
Shekari et al. 2008
LC50
9.4 % extract
2.8–113 mg/L
30 % feeding deterrence
0.625 % extract
0.18–7.5 mg/L
1.5 mg/cm2
36 μg/cm2
Hasheminia et al. 2011 Hasheminia et al. 2011 Maggi et al. 2005
30 μg/cm2
Maggi et al. 2005
1.5 mg/cm2
36 μg/cm2
Maggi et al. 2005
1.5 mg/cm2 1.5 mg/cm2 1 g/L
36 μg/cm2 36 μg/cm2 0.2 mg/L
Maggi et al. 2005 Maggi et al. 2005 Durden et al. 2011
?
10 g/L
Durden et al. 2011
LC50
Complete feeding rejection with ethanol extract of A. annua Epilachna paenulata Complete feeding rejection, artemisinin Spodotera eridania, 87 % feeding inhibition with ethanol extract of A. annua Epilachna paenulata 100 % mortality in no-choice Spodotera eridania 50 % mortality in no-choice Codling moth, Cydia pomonella Feeding deterrent at P
