Sweet wormwood (Artemisia annua L.) is an annual herb that produce artemisinin in its trichomes. Artemisinin is an endoperoxide sesquiterpene lactone which is used in malaria treatment. Globally, between 2000 and 2020, approximately 1.7 billion cases of malaria and 10.6 million malaria-related deaths were reported (WHO, 2021). Artemisinin-based combination therapies (ACTs) remain efficacious against Plasmodium species (malaria’s parasite). Despite the creation of a malaria vaccine, which the World Health Organization (WHO) recommended for use in preventing P. falciparum malaria among children residing in areas with moderate to high transmission rates, ACTs are the first- or second-line malaria treatment (WHO, 2021). Furthermore, artemisinin has activity against several type of cancers (Efferth, 2006) and Li et al. (2005) found that the plant extract has anti-viral activity against Severe Acute Respiratory Syndrome-associated coronavirus (SARS-CoV).

Besides, A. annua crops have the potential to provide low-cost raw materials to pharmaceutical industries and commercially support farmers engaged in large-scale production (Soni et al., 2022). Since, the only commercial source of artemisinin is the A. annua plant, the chemical synthesis has low yields and the process is long and relatively complex (Wang et al., 2014), artemisinin production is a key challenge in the artemisinin supply chain (Lee et al., 2023). Thus, raising the artemisinin content of A. annua is a viable strategy for lowering the cost of artemisinin while increasing the availability of ACTs.

The most significant and harmful form of stress to plant production worldwide is drought, which diminishes yields at various crucial stages throughout the crop cycle (Jacques et al., 2021; Agathokleous et al., 2022). In A. annua, artemisinin content decreased due to water stress (Marchese et al., 2010; Yadav et al., 2014; Soni and Abdin, 2017). Specifically, Yadav et al. (2014) found that mild and moderated water stress reduced artemisinin content, plant growth, development, and physiological status, as well as glandular trichomes density and size, in all the plant development stages (from the early vegetative stage to full bloom stage). However, in other reports different results were described. Soni and Abdin (2017) found slightly higher artemisinin amounts in severely stressed plants at the preflowering and flowering stages. Additionally, Marchese et al. (2010) quantified artemisinin content after 14, 38, 62 and 86 h of water deficit and obtained a significative increase (29%) with moderate water deficit (38 h). These studies highlight the need to further investigate how drought affects artemisinin production.

Drought stress also has detrimental effects on physiological, morphological and biochemical processes. These aspects were deeply discussed by Farooq et al. (2012) and Yang et al. (2021). The authors point out how growth parameters are negatively affected due to a decrease in cell expansion and energy supply, loss of turgor and impaired enzyme activity. To alleviate these effects, plants activate their stress response mechanisms, which includes the expression of functional gene products (such as aquaporins, proline, soluble sugars, etc.), regulatory gene products (such as mitogen-activated protein kinases) which can cause changes in plant morphology or physiology, and the synthesis of hormones (such as abscisic acid) (Yang et al., 2021). Most of these mechanisms serve to regulate plant water status through osmotic regulation and plant transpiration. These, as well as the antioxidant machinery and drought-induced proteins that help the plant to face drought stress are reviewed in depth by Yang et al. (2021).

Plant water relations are characterised by key features such as leaf water potential, osmotic potential, pressure potential, and relative water content. (Kirkham, 2005). In fact, the study of water stress commonly employs leaf water potential (Ψ_L) and relative water content (RWC) as these are significantly impacted by water deficiency (Farooq et al., 2012). Besides, drought can increase the production of reactive oxygen species (ROS), such as hydrogen peroxide (H₂O₂), and generate oxidative damages (Ahluwalia et al., 2021). These damages could affect different plant processes since it induces protein denaturation and loss of enzyme activity, lipid peroxidation (which affects membrane structure), and DNA strand breakage (Yang et al., 2021).

Plants have two types of antioxidant protection systems, enzymatic and non-enzymatic, to maintain a moderate level of ROS which at low levels can act as secondary messengers (Kirova et al., 2021). The non-enzymatic antioxidant system consists of substances, such as phenols and flavonoids (Kirova et al., 2021). On the other hand, catalase (CAT), superoxide dismutase (SOD), ascorbate peroxidase (APX) and glutathione reductase (GR) are some of the main antioxidant enzymes (Yang et al., 2021). However, a largely unexplored aspect of the response to plant stress is the fatty acid profile changes. These are the main components of cell membranes as well as precursors of several signalling molecules, involved in moisture and heat loss prevention, and are associated with cell identification specificity and tissue immunity (Gu et al., 2020).

A well-known strategy to face different stresses and/or improve crop production is the use of biostimulants, since they promote plant growth and trigger natural defences (García-García et al., 2020). Chitosan and their derivatives have been widely studied due to their properties as biostimulants (García-García et al., 2020). It is also an abundant polymer that can be obtained from all crustaceans and insects shells, and in certain fungi, algae and yeast (Rinaudo, 2006). Chitosan oligosaccharides (COS) are chitosan degradation products obtained through physical, enzymatic, or chemical hydrolysis (Naveed et al., 2019; Yuan et al., 2019). Furthermore, COS have a lower molecular weight, a lower degree of polymerization, a higher degree of deacetylation, less viscosity, more solubility, biodegradability, biocompatibility, and exposed N-glucosamine units than chitosan and chitin (Benchamas et al., 2021). Besides, COS have been shown to have a variety of effects on plant cultivation, including inducing resistance to plant diseases, stimulating plant growth and development, and enhancing the yield and quality of plant products (Yuan et al., 2019).

Moreover, several studies showed that the use of COS is a promising option to increase the production of secondary metabolites with pharmacological significance. Thus, COS improved the pharmaceutical value of Catharanthus roseus (L.) G. Don by increasing the content of vindoline and catharanthine in the leaves, and by increasing the fresh weight of roots, stems and leaves (Tang et al., 2022). Moreover, greek oregano (Origanum vulgare ssp. hirtum (Link) Ietsw.) treated with 200 and 500 ppm of COS displayed superior plant height, while the use of 50 and 200 ppm COS doses significantly increased polyphenol content (Yin et al., 2012a). Moreover, COS increased the fresh and dry weight of mint, as well as secondary metabolites production (Ahmad et al., 2017). Jaleel et al. (2017) also found an enhancement in biomass parameters and an increased yield of essential oil and citral content in lemongrass (Cymbopogon flexuosus (Steud.) Wats) treated with COS. Foliar spray with COS improved shoot fresh weight, shoot dry weight and fruit fresh weight of chili (Capsicum frutescens L.) plants (Dzung et al., 2017). Promising results were also obtained with COS-treated wheat seedlings, where growth parameters improved (Zhang et al., 2016). Also, COS treatment afforded better yield and increased superoxide dismutase activity (SOD) and peroxidase (POD) activity in tea (Camellia sinensis (L.) O. Kuntze) plants (Ou et al., 2022). Besides, chitosan oligomers treatment increased growth, mineral uptake and yield in coffee (Coffea canephora Pierre ex A.Froehner) plants (Dzung et al., 2011). Furthermore, the authors proposed COS treatment to improve coffee drought resistance since it reduced the transpiration of leaves.

In fact, it has been reported a reduction in plant transpiration with chitosan and COS treatment in several species (Lee et al., 1999; Bittelli et al., 2001; Iriti et al., 2009; Khokon et al., 2010; Dzung et al., 2011; Pongprayoon et al., 2013; Mirajkar et al., 2019). This is a common strategy to avoid water loss under drought stress, and therefore, chitosan and its derivatives have been proposed to increase drought tolerance, displaying excellent results in different species. Thus, COS addition to Buddhist Pine (Podocarpus macrophyllus (Thunb) Sweet) enhanced drought tolerance by causing starch allocation towards roots, where δ¹³C abundance (used to quantify newly formed photosynthates in trees) and antioxidant enzyme activity were upregulated (He et al., 2020). Muley et al. (2019) reported that 75 and 50 mg · L^-1 of chitosan and COS, respectively, were effective to improve drought tolerance in potato (Solanum tuberosum, L.). In cowpea (Vigna unguiculata (L) Walp), treatment with 250 mg · L^-1 of chitosan increased plant growth and yield under normal and drought stress conditions (Farouk and Amany, 2012). Besides, the treatment of apple (Malus sieversii (Ledeb.)M.Roem) trees with chitosan and COS has increased their drought tolerance (Yang et al., 2009), and similar effects have been observed in maize (Zea mays L.) (Rabêlo et al., 2019), sugarcane (Saccharum hybrid cultivar) (Mirajkar et al., 2019), barley (Hordeum vulgare L.) (Hafez et al., 2020) and in two basil species (Ocimum ciliatum Hornem and O. basilicum L.) (Pirbalouti et al., 2017). This superior resistance has been associated with an improvement in osmotic adjustment and antioxidant systems.

Furthermore, COS seems to be a promising option to enhance the production of secondary metabolites and biomass as well as drought tolerance. Specifically, after 48 h of treating A. annua with 100 mg · L^-1 chitosan, a significant increase in artemisinin production was observed (Lei et al., 2011). Also, in two species of basil, the highest total phenolic content was obtained by foliar spraying chitosan (0.4 g·L^-1) under low watering or mild drought stress (Pirbalouti et al., 2017). In summer savory (Satureja hortensis L.) the interaction between chitosan treatment and drought stress also increased total phenolic content (Alizadeh et al., 2020). Thus, our current study analyses the effect of different COS concentrations, drought stress and their interactions in growth, drought stress tolerance and artemisinin production of A. annua plants with the aim of artemisinin production optimisation.

Leaf water potential (Ψ_L) and relative water content (RWC) are two measurements widely used to indicate plant water status. In the present study, both were significantly affected by drought stress. However, under drought stress 100 and 200 mg · L^-1 doses improved the water potential in the plant but only 200 mg · L^-1 increased RWC compared to non-stressed plants. Mirajkar et al. (2019) obtained similar results in sugarcane treated with normal chitosan, while COS obtained by irradiating chitosan was more effective in improving RWC.

The decrease in RWC would primarily be associated with the reduction in protoplast volume, even though a sizable portion of the water in the leaf is apoplastic (i.e., contained within the cell walls or xylem) (Sack et al., 2018). Until extremely severe tissue dryness caused embolism and drained xylem conduits, this apoplastic water would be “locked” by surface tension in the cell wall pores or xylem conduits (Sack et al., 2018). Thus, RWC represents the physiological result of cellular water deficit and evaluates the quantity of water available for leaf transpiration (El-Hendawy et al., 2019). However, Boyer et al. (2008) showed that measuring RWC with fully turgid water content when leaf floated on water can produces artefacts in RWC since this measure assumed that the osmotic potential was constant. Osmotic potential changes are caused by osmotic adjustment in cells which also modify Ψ_L since it consists of the sum of osmotic potential (Ψ_S) and pressure potential (Ψ_P). This could explain the results in 100 mg · L^-1 which produce an osmotic adjustment that improved Ψ_L but produced an artefact that reduced RWC. In fact, in white clover (Trifolium repens) was observed that chitosan treated plants subjected to drought stress induced by polyethylene glycol (PEG) presented an increase in soluble sugar (Li et al., 2017). Soluble sugars serve as osmolytes to face osmotic stress (Afzal et al., 2021; Jiménez-Arias et al., 2021). Thus, the improvement in the hydric status of plants treated with higher doses of COS under drought stress could be due to an accumulation of soluble sugars elicited by COS.

The correlation model showed that DB and dry weight have low positive correlation but is a valid model to estimate dry weight from DB. However, DB is calculated from the area of fresh plants which also showed changes in water content. It is important to take into account that foliar water content is intimately related to leaf shape, angle and structure (Chaves et al., 2002). These affect the area calculated from the phenotyping platform. Huang et al. (2019) showed that leaf fresh weight had a stronger relationship with the area than leaf dry weight and highlighted the importance of foliar water content in leaf growth and form. This should be considered for future research.

Water scarcity has been widely reported to reduce growth (Bistgani et al., 2017; Pirbalouti et al., 2017; Alizadeh et al., 2020; Ahluwalia et al., 2021). Specifically, A. annua presented a growth reduction with the increase of drought stress (Yadav et al., 2014; Soni and Abdin, 2017). The present study showed a reduction in dry weight and RGR under drought stress where COS treatment did not significantly improve growth parameters. Without stress, chitosan (Ait Barka et al., 2004; Mondal et al., 2016; Harfoush et al., 2017) and degraded chitosan (such as COS) have been reported to improve growth in different species (Dzung et al., 2011; Rahman et al., 2013; Dzung et al., 2017; Muley et al., 2019). However, in A. annua plants under well-irrigated conditions, differences between treatments were not statistically significant. This is in accordance with results reported by Lei et al. (2011), where the foliar application of 100 mg · L^-1 of chitosan did not produce significant differences in dry weight and height over 16 days. Besides, Bittelli et al. (2001) found that chitosan foliar treatment reduced biomass and yield in pepper plants (Capsicum sp.) but the differences were not significantly different. Similar results were found in summer savory (Satureja hortensis L.), where the differences in dry weight were not significantly different (Alizadeh et al., 2020). It seems to indicate that in these species, triggered strategies to face drought stress were not enough to improve the biomass. Thus, in the present results, COS treatment at a higher dose improved Ψ_L and RWC under drought stress but did not significantly improve dry weight and RGR.

In plants, fatty acids are broadly dispersed on the cell surface and serve to prevent moisture and heat loss, and they are linked to cell recognition specificity and tissue immunity (Gu et al., 2020). Thus, alterations in the fatty acid profile could potentially serve as a membrane adaptation mechanism to cope with prolonged stress (Qureshi et al., 2013).

The SFA results show an interesting, oppositive response in well-watered plants and under drought stress, particularly at 200 mg · L^-1 COS dose. Thus, while the SFA decreased for WW plants, it increased for DS plants. This is due to changes in the levels of myristic acid (C14:0) and palmitic acid (C16:0). In particular, higher COS doses increased palmitic acid content under drought stress, probably due to inhibition of lipid desaturase activities as has been reported in Kentucky bluegrass (Poa pratensis L.) (Xu et al., 2011). Concerning the increase of myristic acid levels under drought stress, Qureshi et al. (2013) also found that its levels increased under other osmotic stress (salt stress). Authors suggested that protein myristoylation plays a role in the response to this stress since it is a crucial process in membrane targeting and signal transduction. In fact, Sainas et al. (2018) describe this reaction as a lipid modification that involves the addition of myristic acid to the N-terminal glycine of a subset of proteins and promote their binding to cell membranes for a range of biological functions.

On the other hand, the increase of UFAs, including PUFA, in drought-stressed plants treated with 200 mg · L^-1 COS dose is mainly influenced by linoleic acid (C18:2) and slightly by α-linolenic acid (C18:3). The present study showed a higher linoleic acid (C18:2) content in drought stressed plants. The increase of linoleic acid under salt stress in A. annua (Qureshi et al., 2013) and drought stress in Kentucky bluegrass (Xu et al., 2011) was related to changes of desaturases activity. Furthermore, Tang et al. (2012) found that higher levels of C18:2 and C18:3 were related to lower susceptibility to cold stress. Besides, under drought stress, it has been reported an increase in UFA in coconut (Cocos nucifera L.) (Repellin et al., 1997), grapevine (Vitis vinifera) (Toumi et al., 2008), cork oak (Sebastiana et al., 2018; Sebastiana et al., 2019) and in Arabidopsis thaliana (Gigon et al., 2004).

In well-watered plants treated with 200 mg · L^-1 dose, the UFA/SFA ratio is significantly higher than in control plants. This indicated lower SFA with respect to UFA content. Fatty acids are the primary components of cellular membranes, and changes in unsaturation levels and FA composition are directly related to plant resistance to various stresses, including drought (Zhang et al., 2005). Due to the non-linear nature of fatty chains containing double bonds, highly unsaturated fatty acids are less densely packed into a membrane, resulting in a less rigid membrane structure. (Lehninger, 1977; Cyril et al., 2002). Thus, the 200 mg · L^-1 COS dose made the membrane less rigid under well-irrigation condition. This suggests that maintaining higher fatty acid unsaturation levels could favour proper membrane fluidity, required for several membrane-dependent processes for plant stress adaptation (Xu et al., 2011; He and Ding, 2020). It has been reported that under conditions of drought stress, chitosan can stabilise membranes through an increase of antioxidant activity, as shown for apple tress (Malus sieversii), potato (Solanum tuberosum) and Thymus daenensis (Yang et al., 2009; Jiao et al., 2012; Bistgani et al., 2017). Antioxidant activity prevents the lipid peroxidation that causes membrane damage. In present study, the increase in the activity of the antioxidant enzymes activity in well-watered plants treated with COS may promote membrane stability. However, this improvement was not observed under drought stress. This likely results from damage caused by ROS, which react with the double bonds of fatty acids, since unbalanced ROS can easily oxidize PUFA and trigger lipid peroxidation (Vistoli et al., 2013; Alché, 2019). Nevertheless, the reduction of membrane fluidity limits free radicals diffusion and thus reduces lipid peroxidation in membranes (Mukhtar Ahmed et al., 2020).

There are not many works about the A. annua fatty acid profile and its changes under stress. Carvalho et al. (2011) determined the fatty acids profile in different species of Artemisia spp. and as in the present work, the major fatty acids identified were α-linolenic acid followed by linoleic acid and palmitic acid. Qureshi et al. (2013) obtained similar results at 100 and 130 days after sowing (DAS). At 160 DAS, authors found higher content of linoleic acid followed by α-linolenic acid and palmitic acid. These differences are likely due to different growth conditions and physiological stages.

Chitosan treatment induced the production and accumulation of H₂O₂ in several species (Lin et al., 2005; Li et al., 2009; Mejía-Teniente et al., 2013). However, in our results COS treatment did not increase H₂O₂ in well-watered plants. In fact, the 50 mg · L^-1 dose significantly decreased it while higher doses maintained similar levels than non-treated plants under well-watered conditions. These results could be due to the increase in CAT, APX and GR activity, since these enzymes play an important role in H₂O₂ scavenging.

Some authors have shown that hydrogen peroxide elicited by chitosan treatment has a key role in stomata closure signalling and detected accumulation of H₂O₂ in guard cells (Iriti et al., 2009; Li et al., 2009; Srivastava et al., 2009). Thus, the authors showed that chitosan has antitranspirant activity due to the H₂O₂ accumulation and it is this feature the main reason to propose it to face drought stress. However, Iriti et al. (2009) and Mirajkar et al. (2019) point out that these results are valid under optimum conditions, but under drought stress the response can be different. In fact, Mirajkar et al. (2019) reported that COS treatment decreased H₂O₂ levels and increased SOD activity under drought stress. In contrast, the present study found an increase in H₂O₂ content under drought stress with higher doses of COS. Also, in COS-treated plants under drought stress, CAT and SOD activity did not increase while APX and GR decreased. This decay in the antioxidant enzyme machinery could be caused by high stress levels that damaged it. Thus, it must be pointed out that the present study was carried out after 9 days of stress while the mentioned anti-transpirant experiments were carried out in the first stages of stress. Nevertheless, it could be possible that under severe drought conditions the observed increase in H₂O₂ induced abscisic acid (ABA) signalling. This plant hormone plays a main role in stomata closure and Zhang et al. (2001) showed that H₂O₂ is involved in ABA signalling and how exogenous catalase partially prevents stomatal closure. Since this is a mechanism to avoid water loss, H₂O₂ production could be one of the mechanisms involved in the improvement of RWC and Ψ_L with 200 mg · L^-1 dose.

Regarding the content of phenolic compounds, these natural products are abundantly found in the Plant Kingdom and play an active role in protection processes against biotic and abiotic stresses (Naikoo et al., 2019). As an antioxidant defence, the content of these secondary metabolites usually increases under drought stress, as was reported for different species (Quan et al., 2016; Mohamed and Latif, 2017; Sarker and Oba, 2018; Gharibi et al., 2019; Kirova et al., 2021). Those results are in agreement with these obtained in the present work, where total phenolics and flavonoids increased under drought stress.

There are many reports on the increase of phenols and flavonoids under chitosan elicitation (Khan et al., 2019; Hawrylak-Nowak et al., 2021; Kahromi and Khara, 2021; Bavi et al., 2022). Besides, in sage (Salvia officinalis L.) extracts growing under reduced irrigation conditions and sprayed with chitosan, the total content of phenolic and flavonoids increased (Vosoughi et al., 2018). However, in the well-watered plants of the present study, the increase in phenols was not significant compared to the control. In fact, higher doses decreased flavonoids. Furthermore, their content decreased in drought-stressed plants under COS treatment, particularly at higher doses. A possible explanation is found in some reports about the toxic effects of high chitosan doses on cell cultures (Cabrera et al., 2006; Wang et al., 2008; Singh, 2016). Thus, chitosan treatment elicited ROS production, which in turn can promote the synthesis of defence compounds such as phenols, flavonoids, enzymes, etc. (Gechev et al., 2002). However, ROS can also cause cell damage and cell death when the production is excessive (Lawton and Lamb, 1987). Cabrera et al. (2006) and Singh (2016) related a decrease in the level of metabolites at higher chitosan and COS doses due to an imbalance of oxidative stress molecules production. This excess of ROS, caused by a high elicitor signal, disturbs the synthesis of defensive compounds, and the cell can no longer repair the damage. This is probably happening in the present study when drought-stressed plant underwent COS treatment, namely, cell damage affected phenolic and flavonoid synthesis. On the other hand, in well-irrigated plants, oligochitosan treatment increased CAT, APX and GR activity allowing low hydrogen peroxide content. This suggests a COS activation of antioxidant enzymes, but not the non-enzymatic antioxidant defence.

Drought stress was the main factor that affected the content of metabolites, and increased artemisinin production. Soni and Abdin (2017) reported that during the plant vegetative stage, artemisinin content was lower for drought-stressed plants control. Yadav et al. (2014) also reported a decrease in the artemisinin content under drought stress. However, in these previous works, the stress level was constant over time while in the current work there was a progressive decrease in SWC (soil water content). In another report, Vashisth et al. (2018) found a decrease in the artemisinin content under drought stress but this was not statistically significant. Marchese et al. (2010) studied artemisinin content withholding watering until 86 h and only found a significant increase at 38 h. As shown in Table 1 , well-watered and stressed non-treated plants gave similar results, except for the 200 mg · L^-1 dose, which increased artemisinin under drought stress.

Foliar application of 100 mg · L^-1 of chitosan increased artemisinin content 48 h after treatment (Lei et al., 2011). Since chitosan increased H₂O₂ and O₂ ^- content, authors related this result to a decrease in the artemisinin precursor, dihydroartemisinic acid. The hypothesis is that this precursor acted as ROS quencher, yielding artemisinin as end-product (Wallaart et al., 2000; Lei et al., 2011). Although dihydroartemisinic acid oxidation has been mainly related to O₂ ⁻, it has also been shown that H₂O₂ plays a major role in artemisinin synthesis (Kam and Yap, 2020). In this way, (Mannan et al., 2010) reported that dimethyl sulfoxide (DMSO) produced an increase in H₂O₂ and artemisinin levels. To verify that both events are related, authors added to in vitro cultures L-ascorbic acid, a natural scavenger of ROS, and observed that both H₂O₂ and artemisinin production were greatly reduced. Thus, an increase in ROS likely resulted in higher artemisinin content in DS plants treated with 200 mg · L^-1 COS.

Present study obtained non-significant differences in well-watered plants. This is in agreement with the studies by Yin et al. (2012b), where the increase in artemisinin content was similar in control and COS-treated plants. Therefore, the authors’ conclusion is that the application of elicitors may not be universally effective in increasing the yield of artemisinin from A. annua plants. Besides, the present study showed that CAT, APX and GR activity increased with oligochitosan treatment in well-irrigated plants allowing low ROS content. Thus, due to the crucial role of ROS in artemisinin production as mentioned before, it appears that well-watered plants utilise antioxidant enzymes instead of dihydroartemisinic acid to quench ROS, and artemisinin synthesis is not elicited.

However, chitosan and their derivatives characteristics such as degree of acetylation and polymerization, functional group availability or molecular weight can change their biological response (Harish Prashanth and Tharanathan, 2007; Mirajkar et al., 2019; Mukhtar Ahmed et al., 2020). This, as well as the studied specie, can affect the treatment response. Thus, further studies on the characteristics of chitosan and its biological effect are required since it showed to be not a universal elicitor. Ahmad et al. (2020) reviewed signalling in the production of plant secondary metabolites elicited by oligosaccharides and highlight the progress in this area. Thus, studies such as Zou et al. (2015); Zou et al. (2017); Mirajkar et al. (2019) and Muley et al. (2019) could allow a deep understanding of the mode of action and the relation between the structures and characteristics of chitosan and their derivative with their activity.

Artemisia annua has anti-malarial, anti-tuberculosis, anti-cancer and anti-microbial compounds with great commercial and medicinal relevance (Soni et al., 2022). A. annua plants are the best source of artemisinin production, but still low yield are obtained (0.5-0.8%), and thus an increase in production is a critical issue (Soni et al., 2022).

The effect of COS on artemisinin production in well-watered and drought-stressed plants was addressed in the present work, as well as A. annua growth, metabolism and water management. Thus, under drought stress, COS did not counteract the reduction in growth at any concentration, but higher doses improved plant water status, probably due to osmotic regulation and stomata closure mediated by ABA and H₂O₂ signalling. Furthermore, combined COS treatment and drought stress induced damage to antioxidant enzyme defence (especially APX and GR) and reduced the content of phenols and flavonoids. This boosted ROS production and led to an increase in artemisinin content in DS plants at a 200 mg·L-1 COS dose, probably due to dihydroartemisinic acid oxidation. In contrast, under well-watered conditions, COS did not enhance the growth of A. annua, and since the antioxidant enzyme machinery was not damaged, it prevented artemisinin elicitation.

These results highlight the significant role of ROS in artemisinin synthesis and suggest that COS application has the potential to increase artemisinin yield in crop production, even under drought conditions. However, a deeper understanding of COS characteristics and the elicitation process of artemisinin is necessary to achieve cost-effective production.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

DJ-A, JMdS, AB and AG-G designed the idea of the manuscript. AG-G performed the experiments with participation of AM, EF, and RCdC. All authors contributed to the article and approved the submitted version.