https://orcid.org/0000-0003-2265-4972
Figures
Abstract
The use of teas made from locally cultivated Artemisia annua to fight malaria in remote areas where access to care is difficult is a matter of debate. This study aimed at document differences in the composition of A. annua teas cultivated in Benin to be sold as antimalarial teas, and in France, and their impact on antiplasmodial activity. A. annua teas were prepared with plants from one location in south France and from ten different plantations in Benin. Artemisinin was quantified in herbal teas with a liquid chromatography system coupled to mass spectrometry and multiple reaction monitoring detection methods. The herbal teas were tested against chloroquine-sensitive 3D7 strain of Plasmodium falciparum using isotopic microtest to determine IC50 values and calculate the concentration of artemisinin corresponding to the IC50 of the teas [ART(tea)_IC50]. Chemical profiles were determined by liquid chromatography coupled to high resolution mass spectrometry and a metabolomic analysis was performed to annotate compounds statistically linked to the antiplasmodial properties of the teas. Artemisinin content varied between 0.3 mg/L for tea with plants from France to 15.7 mg/L for teas made with plants from Benin with differences between locations. Artemisinin content was decreasing after a one-year storage of the plant for 3 localities in Benin with loss of 33%, 48% and 24% (P < 0.05). Artemisinin concentrations and antiplasmodial activity of teas were positively correlated although the comparison of ART(tea)_IC50s to IC50 of pure artemisinin suggested that other compounds present in the tea were involved in the activity, either enhancing or limiting it. Unknown alkaloids in A. annua teas correlated to antiplasmodial activity were also detected. These findings suggest that A. annua teas deserve further studies to identify other metabolites of interest and determine their role in antiplasmodial activity in relation to other molecules, particularly artemisinin.
Citation: Aubouy A, Olatoundé Amoussa AM, Lagnika L, Fabre N, Vitrai A, Claparols C, et al. (2025) Comparison of the composition and antiplasmodial activity of Artemisia annua teas using an untargeted metabolomic approach. PLoS One 20(8): e0330682. https://doi.org/10.1371/journal.pone.0330682
Editor: Shahin Tajeri,, Freie Universität Berlin: Freie Universitat Berlin, GERMANY
Received: April 17, 2025; Accepted: July 25, 2025; Published: August 22, 2025
Copyright: © 2025 Aubouy et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All the UPLC-HRMS chromatograms of the teas used for the metabolomic analysis, and their metadata are openly available in our institutional repository (DataSuds, IRD, France) under CC-BY license at https://doi.org/10.23708/HZ6LA1
Funding: The author(s) received no specific funding for this work.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Malaria remains a deadly disease with 263 million cases and 597 000 deaths estimated in 2023. Sub-Saharan Africa accounts for a very large proportion of malaria morbidity and mortality and 76% of malarial death in this Region are children under five [1]. Treatment of malaria caused by P. falciparum, the most widespread and most virulent species, is based on Artemisinin-based Combination Therapies (ACTs) containing a hemisynthetic derivative of artemisinin and a partner molecule. The emergence and spread of ACT-resistant P. falciparum isolates is a major concern worldwide given that no effective alternative treatment is currently available [2]. Artemisinin (ART), a highly effective antimalarial molecule, was first isolated and characterized from the plant Artemisia annua L. by the Chinese chemist Youyou Tu in 1972, as the plant was part of the traditional Chinese pharmacopeia for its antipyretic properties [3]. A. annua plays a key role in the fight against malaria, since its production is necessary for isolating ART and developing its hemisynthetic derivatives, such as artesunate and artemether, the best-known treatments of which are available.
The use of teas made from locally cultivated A. annua to fight malaria in endemic areas is a topic of debate. Such a strategy could constitute an alternative antimalarial treatment for people living in remote areas where access to healthcare is limited. A certain efficacy was previously proven, both to cure and to prevent malaria [4,5]. However, a high relapse rate was observed after 14 days of treatment, suggesting a risk of under-dosing of ART, conducive to the emergence of ART-resistant parasites [6]. The variability of the chemical composition of plants depending on their environment is a well-known phenomenon which can have a significant impact on the efficacy of bioactive plants [7,8]. Other parameters influence the therapeutic effectiveness of a plant material, including plant genetic factors, drying and storage conditions. For these reasons, the WHO does not recommend the use of non-pharmaceutical form of A. annua [9].
During the Covid-19 pandemia, A. annua herbal preparations gained popularity in different parts of the world. Our study carried out in 2022 in Benin highlighted that this fame has endured, and that a chain of production, sale and consumption of A. annua teas exists in Benin, independently of the network of traditional practitioners [10]. In Benin, A. annua teas have obtained market authorization as improved traditional medicine and are sold for the treatment of uncomplicated malaria, a fatal disease for non-immune subjects if treatment is ineffective or administered too late [11]. In this context, it is extremely important to document the variability in the composition of teas and its impact on their antiplasmodial efficacy. Thus, we studied the composition of A. annua teas prepared with plants grown in different plantations in southern Benin Republic and France, and evaluated their in vitro antiplasmodial activity. Additionally, we assessed the impact of compositional variability on activity and identified potential bioactive metabolites other than ART, using untargeted metabolomic analysis.
Materials and methods
Plant material
The plant names have been checked with https://www.worldfloraonline.org on the 6th. of March 2025. Aerial parts of A. annua were collected between September and November 2020, in ten plantations in Benin (Fig 1, Table S1 in S2 File). In each plantation, three samples were collected from different locations spaced at least 1 meter apart in the field and labelled sample_1, 2, 3. The plants were dried in a drying oven at 45°C for one week, ground, and stored at ambient temperature until further use. A voucher specimen from each plantation was botanically identified and deposited in the National Herbarium of Benin (University Abomey-Calavi, voucher numbers A003 A/HNB to A0012 A/HNB). Because we also needed Artemisia afra Jacq. ex Willd. for the quantification of ART, one sample of A. afra was collected from the Banigbé plantation, where both species A. afra and A. annua are cultivated. Post-harvest processing (drying and grinding) and tea preparation were carried out in the same way as for A. annua. A voucher specimen was deposited at the National Herbarium of Benin under the number A003 B/HNB.
The maps of West Africa and Benin used for this figure come from the geoBoundaries project: Runfola, D. et al. (2020) geoBoundaries: A global database of political administrative boundaries. PLoS ONE 15(4): e0231866. https://doi.org/10.1371/journal.pone.0231866.
Furthermore, two samples from the “conservatoire botanique de Soual” (Soual, France), were collected at two different time points “Soual_01” in January 2021, “Soual_02” in May 2021. The plant was dried at ambient temperature, ground, and stored at ambient temperature until further use. The samples were given to us by Dr Christel Fiorini, responsible for ‘Conservatoire Botanique Pierre Fabre’, Pierre Fabre Laboratories, Soual, France.
Molecular analysis of plant samples
Genomic DNAs from the various A. annua and A. afra samples were extracted using the DNeasy Plant Mini Kit (Qiagen) in accordance with the manufacturer’s instructions. Approximately 100 mg of each sample was ground with liquid nitrogen in a mortar with a pestle to a fine powder, followed by further crushing with the Biorad PRECESS24 homogenizer. Amplification of the ITS1 or ITS2 regions, was performed by PCR with GoTaq polymerase (Promega), with a hybridization temperature of 56°C, according to the supplier’s recommendations, and as described by Wang [12]. The PCR amplified regions of the expected size (707 bp and 337 bp for ITS1 and ITS2 markers, respectively) were visualized on 1% agarose gel and sent for sequencing to our provider (Genewiz).
Sequence comparisons between PCR amplified regions obtained in this study and the known ITS sequences of Artemisia sp., retrieved from the NCBI website (http://www.ncbi.nlm.nih.gov/), have been realized using the CLUSTALW software.
Tea preparation
Teas from A. annua and A. afra samples were prepared according to Mueller [4]. Two series of teas were prepared using A. annua samples collected from Benin between September and November 2020: one in February 2021 (Series 1) and one in February 2022 (Series 2). For the samples from Soual, each sample was extracted in triplicate. Briefly, 1 g (± 0.005 g) of plant was put in a tea bag and 200 mL of boiling distilled water were poured over. The beaker was covered and left at ambient temperature. The tea bag was stirred from time to time. After 15 minutes, the tea bag was squeezed and removed, and the tea was left to reach ambient temperature then filtered over cotton. Aliquots of 2 mL were stored at −20°C until use for the quantification of ART and LC-HRMS profiling for metabolomic analysis. Five of these were freeze-dried to evaluate the total concentration of the tea, and were used for the antiplasmodial assay. We verified that the composition of frozen and freeze-dried samples remained unchanged under these conditions by comparing the LC-HRMS profiles of nine samples after three years of storage at −20°C (Figs S10, S11 and S12 in S1 File).
Quantification of ART
A standard stock solution of ART (98% purity, ref 361593, Sigma Aldrich, Saint Quentin Fallavier, France) of 0.5 mg/mL was prepared in methanol. The A. afra tea was diluted 1000x in milliQ water. The standard stock solution of ART was diluted in the diluted A. afra tea to a concentration of 5 µg/mL (solution A). Working solutions were prepared by dilution of solution A in diluted A. afra tea to afford the concentration range 0.5, 3.125, 5, 6.25, 12.5, 25 and 50 ng/mL for the calibration standards. Before analysis, A. annua teas were diluted 100x or 1000x in milliQ water to be within the linearity range and above the limit of quantification (LOQ) value.
Precision was defined as the relative standard deviation (%RSD) of the means, while accuracy was defined as the relative error (%RE) between the nominal, and the calculated concentration of ART. The limit of detection (LOD) and the LOQ were the concentrations resulting in a signal-to-noise ratio (S/N) of 3 and 10 respectively, with acceptable accuracy and precision.
Quantification of ART was done with a UPLC system coupled to a triple quadrupole-linear ion trap tandem mass spectrometry with multiple reaction monitoring detection methods. The 2 transitions 283 → 219 and 283 → 229 were used to quantify ART. The LC separation was carried out as followed: sample aliquots (10 μL) were injected onto an Acquity UPLCTM BEH C18 column (1.7 µm, 130Å, 100 × 2.1 mm), fitted with a VanGuard™ C18 1.7 μm guard column (Waters, Milford, MA, USA). Column oven temperature was maintained at 40°C. Flow rate was 0.3 mL/min. The mobile phase consisted in milliQ water (solvent A) and CH3CN (solvent B), each containing 0.1% of formic acid. The column was equilibrated in 5% B and elution was carried out 0.5 min after injection with the following steps: 9.5 min, 95% B; 11.5 min, 95% B; 12 min, 5% B. The column was equilibrated for 2 min between injections. Detection was achieved with an Applied Biosystems Sciex QTRAP® 4500 hybrid linear ion trap triple quadrupole (LTQ) mass spectrometer (Foster City, USA) equipped with a Turbolon-Spray Interface. The instrument was operated in a positive ESI MRM mode (dwell time, 25 ms). The operating parameters were: capillary voltage, 5.5 kV; source temperature, 450°C; gas N2; curtain gas, 25; ion source gas 1, 50; ion source gas 2, 50; CAD medium; DP25; EP 10; CE 15; CXP 14. For MS spectra and chromatogram acquisition and exploitation, Analyst 1.6.1 software from Applied Biosystems Sciex (Foster City, USA) was used.
LC-HRMS profiling of teas and metabolomic analysis
A 2 mL aliquot of tea was centrifuged for 5 min at 2000 g and the supernatant was analyzed using a UPLC-HRMS composed of an Ultimate 3000 (Dionex) UPLC controlled by Chromeleon Xpress 6.8 (Dionex), and a mass spectrometer LTQ Orbitrap XL (Thermo Fisher Scientific, Waltham, MA, USA). An aliquot of 10 µL of supernatant tea was injected and the chromatographic parameters were identical to those described for the quantification of ART. Spectrum were recorded using the LCQ Xcalibur 3.0 software (Thermo Fisher Scientific) in ESI + mode. The instrument settings were as follows: mass scan range m/z 100–2000 Da, capillary temperature 300 °C, sheath and auxiliary gas flow rates at 55 and 10 arbitrary units, spray voltage at 4.2 kV, source current at 100 μA, capillary and tube lens voltages set at 50 and 120 V, respectively. The mass measurement was externally calibrated before starting the experiment. Each full MS scan was followed by data dependent MS/MS analysis on the three most intense peaks using stepped collision-induced dissociation (CID) (normalized collision energy 35%, isolation width 1 Da, activation Q = 0.250). Data visualization and analysis were performed with MZmine 2.53 software [13].
For metabolomic analysis, each sample was injected three times in random order. Blanks (distilled water) and quality control (QC) samples (a mixture of all samples) were regularly added to the sequence. A first batch of analysis consisted of the first series of extraction run for the Beninese samples (90 injections), and Soual samples (18 injections). Another batch of analysis carried out 3 weeks later consisted of the second series of extraction run for the Beninese samples. A preliminary inspection was performed to check the homogeneity of the chromatograms obtained per sample and of those of the QC samples.
Raw data were processed as previously described, with no logarithmic transformation, to run multivariate analysis [Principal Component Analysis (PCA) and orthogonal partial least squares analysis (OPLS) associated with antiplasmodial activity (IC50, and ART(tea)_IC50)] on the W4M Platform (Workflow4 Metabolomics, https://workflow4metabolomics) [14]
In vitro antiplasmodial evaluation
The activity against the chloroquine-sensitive P. falciparum strain 3D7 (from the Malaria Research and Reference Reagent Resource Center, MR4) of lyophilized herbal teas and pure ART was evaluated using the isotopic 3H-hypoxanthine incorporation assays, as previously described [15] with minor modifications. Briefly, P. falciparum parasites were maintained in O+ human erythrocytes (provided by Etablissement français du sang, EFS, Rungis, France) at 5% hematocrit and suspended in complete culture medium RPMI 1640 supplemented with 25 mM HEPES, 20 mM D-glucose, 25 mM sodium bicarbonate, 5 mM L-glutamine, and 10% AB human serum. All culture assays were conducted at 37°C, under a N2-enriched atmosphere (5% CO2, 10% O2, and 85% N2). At ring stage, P. falciparum cultures were highly synchronized by two consecutive treatments with 5% sorbitol (Sigma-Aldrich) in PBS (v/v) at 40 h intervals and diluted down to 0.3–0.5% parasitemia and 2% hematocrit. Freeze-dried aliquots of herbal teas were dissolved in water and then serially diluted in 96-well plates at concentrations ranging from 0.017 µg/mL to 20 µg/mL. Parasites were then dispensed and plates were incubated in the presence of 5% 3H-hypoxanthine (Perkin Elmer, Waltham, Massachusetts, USA) for 42 h. Scintillation counting was used to evaluate 3H-hypoxanthine uptake (Micro β2, Perkin Elmer, Waltham, Massachusetts, USA) and results were expressed as the inhibitory concentrations (IC50) defined as drug concentrations at which 50% of 3H-hypoxanthine incorporation was inhibited compared with drug-free controls. IC50 values were established by non-linear regression with ICEstimator software (http://www.antimalarial-icestimator.net/ 28 December 2021) [16]. The tests on 96-well plates were conducted in triplicate, meaning that three batches of parasites were prepared independently and distributed across three different plates.
Calculation of the concentration of ART corresponding to the IC50 value of the tea, called ART(tea)_IC50
To determine if the presence of other compounds in the tea will increase or limit the antiplasmodial effectiveness of ART in the teas as compared to pure ART, we calculated for each tea the concentration of ART that corresponds to the IC50 of the tea. This value was called ART(tea)_IC50 and calculated as follows: ART(tea)_IC50 = C_ART x (IC50/C_Tea)
ART(tea)_IC50 is expressed in ng/mL
C_ART: concentration of ART in the tea, expressed in mg/L
C_Tea: Total concentration of the tea, expressed in mg/mL
IC50: 50% inhibitory concentration of the tea, expressed in µg/mL
Statistical analysis
Statistical analysis was performed using GraphPad Prism software (version 9.5.1). For descriptive analysis of ART concentrations and IC50 values, quantitative variables were presented as the means ± standard deviations. Unpaired t-test was used to compare the groups, after the normality of the values was verified using the Shapiro-Wilk test, which is suitable for small sample sizes. Spearman correlation was used to test the nonlinear association between ART concentrations and IC50 values of the teas. Differences were considered significant at P < 0.05.
Ethical considerations
The study was approved by the Research Ethics Committee of the Institute of Applied Biomedical Sciences (CER-ISBA) in Benin with authorization n°146 of 08/06/2022. The protocol was explained to the A. annua producers, who signed an informed consent allowing us to collect three samples from their field.
Results
Validation of A. annua identification by molecular approach
As the ITS2 marker was described as the most robust for differentiating Artemisia species, PCR amplifications of the ITS2 region has been first carried out on genomic DNA from all the samples listed in Fig 1. Except for one sample (Artemisia afra, collected in the Banigbé plantation), a fragment of the expected size (337 bp) was obtained after PCR amplification in all the samples. Sequence alignments between newly PCR amplified sequences and the known sequences of Artemisia sp. demonstrated unambiguously that all this samples are identical to known sequence of A. annua with 100% identity. By the same way, for the sample of A. afra, PCR amplification was realized for ITS1 marker, and sequence alignment also confirmed unambiguously that our A. afra sample had 100% identity to known sequence of A. afra. We conducted this verification to eliminate any possibility of confusion between the two species that can be mistaken for each other.
Validation of the method for ART quantification
A linearity range between 0.5 and 50 ng/mL was first determined with ART diluted in methanol for both transitions 283 → 219 and 283 → 229 (Figs S1 and S2 in S1 File). The same linearity range was also determined with ART diluted in diluted A. afra teas (Figs S3 and S4 in S1 File). A. afra is another medicinal plant used to treat malaria, which contains no ART, but also contains flavonoids, chlorogenic acids derivatives, and coumarins, some of which are shared with A. annua [17]. We showed that the use of these calibration curves avoided matrix effect. Indeed, it was not possible to quantify directly ART in the tea. Dilution by at least a factor 100 was required to obtain a linear response (Figs S5 and S6 in S1 File). The standard addition method (avoiding matrix effect) and the method using a calibration curve made with diluted A. afra tea were compared for three samples (Table S2 in S2 File). The difference between the concentrations obtained by the two methods was less than 8%. Therefore, we concluded that matrix effect could be avoided by using a calibration curve made with diluted A. afra tea. The obtention of the validation parameters for the quantification of ART in the tea using this method is showed in the Sup. Data (Figs S3 and S4 in S1 File, Tables S3 and S4 in S2 File). LOD were determined at 1.0 and 2.5 ng/mL for the 283 → 219 and 283 → 229 transitions respectively. LOQ were determined at 2.5 and 6.25 ng/mL for the 283 → 219 and 283 → 229 transitions respectively They are summarized in the Table 1.
The response was linear between 0.5 and 50 ng/mL. However, sensitivity was better for the 283 → 219 transition, as showed by the LOQ and LOD values, and the non-satisfying interday accuracy value of 14.67% for the standard at 0.5 ng/mL measured using the 283 → 229 transition. Results for both transitions are summarized in Tables S5 and S6 in S2 File. For some samples, quantification was repeated several times independently (2–4 times), and the calculated relative standard deviation was less than 14% for the 283 → 219 transition and less than 16% for the 283 → 229 transition. For all these reasons, we choose to work with values obtained using the 283 → 219 transition. Interestingly, Table S7 in S2 File shows that the values obtained using the two transitions are comparable (differences < 12%). With this method, ART concentration can be measured down to 2.5 ng/mL in a tea diluted at least 100 times, which correspond to a minimum concentration of 0.25 mg/L in the tea.
ART concentrations differed according to the geographic origin of the plants, the series of extraction, and the harvesting period
The results for ART concentration in the different teas are summarized in Table S8 in S2 File. ART concentrations ranged from 0.3 mg/L (herbal teas made with plants from Soual, France) to 15.7 mg/L for teas made with plants from Kétou (AK). For herbal teas made with plants from Benin, ART concentrations ranged from 2.6 (Sèmè-Kpodji, AS) to 15.7 mg/L. As shown on Figs 2A and 2B, for the first series of preparations, the 10 herbal teas made with plants from Benin exhibited higher ART concentrations than the two herbal teas made with plants from France (see symbol * on the Fig 2A). In addition, the herbal teas made with plants from Banigbé (AB) and Kétou (AK) had higher ART concentrations than most of the herbal teas made with plants from other localities (all localities except the herbal tea from Lokossa – AMLH). In this first series, many others differences emerged between localities with P < 0.05 or P < 0.005 (Fig 2B). For the second series of preparations (Fig 2A), significant differences in ART concentration were found between the tea from Banigbé (AB) and three others teas (Sémé-Kpodji-AS, Adjarra-AD and Ouidah-AODH) and only AS and AD teas had different ART concentrations to Kétou (AK) tea (Fig 2B). Interestingly, the two series had strikingly different profiles. This indicates that series 2 had a lower ART content than series 1, particularly in locations where the ART content was highest in series 1.
(A) Concentrations were compared for each series of preparation between locations. To avoid overloading the graph, comparisons with Soual_01 and Soual_02 are symbolized by * for series 1 (the highest P value was considered between comparison to Soual_01 or Soual_02); comparisons with Banigbé (AB) are symbolized by § for both series; (B) the table shows the P values of the other comparisons between locations. (C) Concentrations were compared for each location between the two series of preparations carried out one year apart (series 1 and 2 prepared in Feb. 2021 and Feb. 2022); for Soual in France, the symbol $ indicates the comparison between two harvest periods (January and May 2021). Comparisons were carried out using a t-test. One symbol means P < 0.05, two symbols P < 0.005, three symbols P < 0.0005.
We then compared ART concentrations between the two series of tea preparation (Fig 2C). For most herbal teas (7/10), ART concentrations were similar between preparation series 1 and 2. However, for herbal teas prepared with plants from Banigbé (AB), Kétou (AK) and Dogbo (ACDT), the concentrations in the second series were lower, showing losses of 33%, 48% and 24% respectively (P = 0.02, 0.04 and 0.03, respectively). Additionally, Fig 2C presents a comparison between ART concentrations obtained during two different times of plant harvesting at the Soual site in France (Soual_01 and Soual_02). The comparison was highly significant (P = 0.0001), suggesting that harvest time is also a determining factor in the amount of ART in A. annua herbal teas.
In order to better understand the variability between tea samples prepared with plants of different origins, we used unsupervised metabolomic analysis on the LC-HRMS chromatograms. Principal component analysis (PCA) was initially performed with French and Beninese samples, and then with Beninese samples alone (Fig 3). To enhance statistical power, the PCA incorporated the values from both series of herbal tea preparations and, as well as the two harvesting periods for the French samples. Fig 3 shows that French samples had a different composition than Beninese samples. They appeared as outliers in Fig 3, on the right of the PCA plot, characterized by the lowest ART content. In contrast, samples from Banigbé (AB), Kétou (AK), Lokossa (AMLH) and Dogbo (ACDT), clustered on the left side of the PCA, represented the samples with the highest ART content. Indeed, when looking at variable loading, the feature at m/z = 209.1536 and rt = 8.22 min, corresponding to the base peak of ART, had one of the highest negative loading on horizontal axis. However, this is not the only clustering factor: samples from the same origin cluster together, even if they have different ART content. This can be seen with series 1 and 2 of the AK, AB and AMLH samples, for example. Furthermore, the samples from Banigbé (AB) and Kétou (AK) displayed similar profiles, despite being prepared with plants from different plantations.
Samples from Benin are called AS, AD, AODH, ADJR, AC, ACO, ACDT, AMLH, AB and AK. For these samples, numbers 1 and 2 refer to two series of tea preparations (Feb. 2021 and Feb. 2022). Samples from France are called Soual_01 and Soual_02 and were harvested at two different periods (Jan. 2021 and May 2021). a. PCA with French and Beninese samples. b. PCA with Beninese samples only.
As we noticed a change in the composition for the Beninese samples between the two series of tea preparation, we performed an Orthogonal Partial Least Squares-Discriminant Analysis (OPLS-DA) to compare the two batches and identify variables that may change during the storage of A. annua powders. The outputs of the OPLS-DA analysis are presented in Fig S7 in S1 File and the variables with the highest VIP scores are presented in Table 2. Entry 7 corresponds to the base peak of ART (m/z = 209.1536 and rt = 8.22 min), whose signal is higher in the first series, confirming that the quantity of ART decreased during the conservation of the powder. Noteworthy, entry 1 represents the signal with the highest VIP value; which is higher in the second extraction series. Entries 3, 4 and 8 are linked to it, corresponding respectively to an isotope and in source fragments. The m/z value of the MH+ ion associated with this signal is consistent with the molecular formula C14H19NO2 for the parent molecule. Entry 2 represents a MH+ ion corresponding to deoxyartemisinin, which has been confirmed by a comparison with a standard. Entries 5 and 9 also correspond to signals increasing during the storage of A. annua powders. Both signals have the same fragment ion at m/z = 130.0, which could correspond to quinoline or isoquinoline ion, and are consistent with compounds whose molecular formula are C15H27NO4 and C14H25NO4 respectively suggesting that these two compounds may share a similar structure.
The variability in plantation location, plant storage time and harvesting period had an impact on the antiplasmodial activity of A. annua teas
Next, we compared in vitro antimalarial activities of A. annua teas by location, by tea preparation series, and by harvesting period for the French samples. The results are summarized in Table S8 in S2 File. For Beninese herbal teas, mean 50% inhibitory concentration (IC50) values ranged from 0.2 µg/mL for the herbal tea from Banigbé (AB) of the first series to 1.3 µg/mL for the tea from Sémé-Kpodji (AS) of the second series (Fig 4A, Table S8 in S2 File). In comparison, IC50 values obtained with A. annua teas from France were higher with 3.5 and 5.0 µg/mL for Soual_01 and Soual_02, respectively. As shown on Fig 4A, for the first series of tea preparation, the comparisons of IC50s of herbal teas from Benin with those from Soual in France were all highly significant (P < 0.001). For both preparation series, Banigbé (AB) and Lokossa (AMLH) herbal teas exhibited lower IC50 values than herbal teas from other localities in Benin (P < 0.05 to P < 0.0005), except for the AMLH/ACO comparison. Other significant comparisons between localities for the two series are detailed in the table (Fig 4B).
The y-axis on the left is for Soual_01 and Soual_02, the one on the right is for all the other localities in the two series. (A) IC50 values were compared for each series of preparation between locations. To avoid overloading the graph, comparisons with Soual_01 and Soual_02 are symbolized by * for series 1 (the highest P value was considered between comparison to Soual_01 or Soual_02); comparisons with Banigbé (AB) are symbolized by § for both series; (B) the table shows the P values of the other comparisons between locations. (C) IC50s were compared for each location between the two series of preparations carried out one year apart (series 1 and 2 prepared in Feb. 2021 and Feb. 2022); for Soual in France, the symbol $ indicates the comparison between two harvest periods (January and May 2021). Comparisons were carried out using a t-test. One symbol means P < 0.05, two symbols P < 0.005, three symbols P < 0.0005.
Fig 4C highlights the impact of the preparation series on the IC50 values of the two herbal teas from Sémé-Kpodji (AS) and Abomey Calavi (AC). Mean values rose from 0.74 to 1.13 µg/mL for AS and from 0.33 to 0.44 µg/mL for AC between series 1 and 2. Fig 4C, also shows that IC50 values obtained during two different times of plant harvesting at the Soual site in France (Soual_01 and Soual_02) were significantly different (P = 0.004). As for ART concentration in plant, this indicates that harvest time is a determining factor in antiplasmodial activity. These results suggest that ART concentration in plant and antimalarial efficacy are related.
The ART concentration of A. annua teas was not the only factor determining their antimalarial activity in vitro
To find out if ART concentration of A. annua teas was a determining factor in their antiplasmodial activity, we studied correlations between ART concentrations in teas and IC50 values. The relation between the two variables could be modelled as power curves for the two series. Equations were: y = 3.640x—0.785 and y = 3.934x—0.599 with R2 values of 0.75 and 0.51 for series 1 and 2 respectively, indicating that the two variables were highly correlated (Fig 5). Spearman correlation tests also resulted in r values of −0.77 and −0.67 for series 1 and 2 (P < 0.0001), showing the association between ART concentrations and IC50s.
The curve is modelled as a power curve (Excel software) for the first (A) and the second series (B). Spearman correlation test was used to test the association between values.
However, it is interesting to note that the increasing order of ART concentrations by locality presented in Fig 2 was not the same for the IC50 values presented in Fig 4. In particular, the herbal tea obtained from Kétou (AK) plants had the highest ART content following the first series of herbal tea preparations, unlike the herbal teas from Lokossa (AMLH), Abomey Calavi (AC) and Cotonou (ACO). However, the AK herbal tea of the first series had higher IC50 values than the AMLH, AC and ACO herbal teas, suggesting that the other molecules present in the A. annua teas are involved in the activity, either enhancing or limiting it.
To explore this issue, we first calculated the concentration of ART corresponding to the IC50 value of the tea, named ART(tea)_IC50 (Table S8 in S2 File). The difference between this value and the IC50 of pure ART indicates if other molecules in the tea contribute significantly to antiplasmodial activity, enhancing it if ART(tea)_IC50 < IC50 of pure ART, limiting it if ART(tea)_IC50 > IC50 of pure ART. Fig 6 highlights the variation of this value according, as before, to the series of preparation, the location of plantations and the harvest period for the French plants (Soual_01 and Soual_02). As shown on Fig 6A, ART(tea)_IC50 mean values varied between 1.67 ng/mL (5.77 nM) for tea with plants from Ouidah (AODH) to 5.46 ng/mL (19.34 nM) and 8.18 ng/mL (28.97 nM) for teas with plants from Dogbo (ACDT) and from Soual (Soual_01), respectively. The comparison for each locality by series 1 and 2 did not show any significant differences (Fig 6A). However, for French teas as for IC50 values, the harvesting period in May led to lower value for ART(tea)_IC50 than in January (P = 0.01) (Fig 6A). Since the values were similar between the two series of preparation, the values from both series were considered for the comparison by locality (Figs 6B and 6C). ART(tea)_IC50 of the different teas were first compared to IC50 value of pure ART. Teas from Ouidah (AODH), Abomey Calavi (AC), Cotonou (ACO), Lokossa (AMLH) and Banigbé (AB) obtained lower ART(tea)_IC50 values compared to those of pure ART (mean IC50 = 2.82 ng/mL or 9.99 nM), suggesting that the molecules, other than ART, contained in these herbal teas act synergistically or additively to enhance their in vitro antimalarial activity. In contrast, Soual, Adjarra (ADJR), Dogbo (ACDT) and Kétou (AK) A. annua teas had higher ART(tea)_IC50 values than pure ART, suggesting that their composition limits the activity of the ART contained in the herbal teas. Many other differences were found between plantation locations (Fig 6A), including the significant differences between ART(tea)_IC50 values of teas prepared with plants from Soual_01 (harvest period January) and all Beninese teas.
(A) ART(tea)_IC50s were compared for each location between the two series of preparations carried out one year apart (series 1 and 2 prepared in Feb. 2021 and Feb. 2022); for Soual in France, the symbol $ indicates the comparison between two harvest periods (January and May 2021). (B) ART(tea)_IC50s were compared for each series of preparation between locations. Since the values were similar between the two series of preparation, the values from both series were considered for this comparison. * shows the comparisons to the IC50 values of pure ART; (C) the table shows the P values of the other comparisons between locations. Comparisons were carried out using a t-test. One symbol means P < 0.05, two symbols P < 0.005, three symbols P < 0.0005.
Identification of putative antiplasmodial metabolites in A. annua teas
To delve deeper the issue of the chemical composition of A. annua teas and its impact on their in vitro antimalarial activity, we performed OPLS analysis using the IC50 values as Y input. The outputs of the OPLS-DA analysis are presented in Fig S8 in S1 File. The variables with the highest VIP score are presented in Table 3. Entries 2 and 4 indicated that Arteannuin B was negatively correlated to activity. The following signals showed positive correlation with antiplasmodial activity: Entries 9 and 16 showed up signals of ART. Entries 3 and 5 represent an unknown alkaloid and deoxyartemisin, which were also linked to the ageing of the plant samples (Table 3). Entry 15 corresponds to an unknown alkaloid as well. Aminoacids (Entries 1, 6 and 10) could also contribute to activity, as well as flavonoids (entry 13).
The same OPLS analysis was conducted using ART(tea)_IC50 as Y input. The outputs of the OPLS-DA analysis are presented in Fig S9 in S1 File, and the features with the highest VIP values are presented in Table 4. This analysis shows features corresponding to aminoacids (entries 1, 3 and 5), deoxyartemisinin (entry 10), and the unidentified alkaloids (entries 6 and 12) previously detected in Tables 2 and 3. Four features corresponding to chlorogenic acid and derivatives were also identified (entries 11, 4, 7 and 9). Among these, dicaffeoylquinic acids were identified by comparison of their signals with 3,4-; 3,5- and 4,5-dicaffeoylquinic acids standards. Due to the close retention times under our chromatographic conditions, the three isomers could not be fully discriminated, therefore the signal in entry 7 represents a mixture of these 3 isomers. Fig 7 highlights the different peaks corresponding to most of the features described in Tables 2–4 on the LC-HRMS chromatograms of one sample from Benin (AB_1) and one sample from France (Soual_02).
Main features described in Tables 2–4 are highlighted. a: proline, b: leucine/isoleucine, c: phenylalanine, d: tryptophan, e: chlorogenic acid, f: unknown alkaloid and 5-feruoylquinic acid, g: unknown alkaloid, h: dicaffeoylquinic acid (mixture of isomers), i: chrysosplenol D, j: arteannuin B, k: ART, l: deoxyartemisinin.
Discussion
This study, conducted using plants from Benin and France, aimed to investigate the variability in the chemical composition of A. annua teas based on the location of the plantations, the storage time of the plants, and the harvest period. The study also examined the impact of this variability on the in vitro antimalarial activity of the teas and identified putative active metabolites.
The LC-MS/MS quantification method for ART used in this study was adapted from that developed by Suberu et al. [18,19], which reported LOD and LOQ values of 0.13 and 0.41 ng/mL respectively. Others published methods based on LC-MS/MS described LOD and LOQ values of 1.5 and 5 ng/mL, similar to those obtained in this study [20,21]. The quantity of ART in the tea was variable, as already observed in previous studies. This variability depends on the variety of A. annua used, as some hybrids are developed to maximize ART production, which is typically low in wild type plants [22]. ART content is also influenced by geographical location and climatic condition [23,24]. The recipe used in this work, 5 g of plant for 1 L of water, was based on that described by Mueller et al. for clinical trials. Their teas were prepared with the Artemis hybrid known for its high content in ART, grown for this purpose in Democratic Republic of Congo or in Germany. The teas contained 12 mg/L and 47 mg/L of ART respectively [4,6]. Similarly, Silva et al. reported ART concentrations of 40–46 mg/L in teas prepared under comparable conditions using a high ART content A. annua variety developed by the UNICAMP University in Brazil [25]. Mouton et al. also compared the amount of ART in teas prepared from 16 samples of A. annua of diverse origins, and found concentration ranging from 8.36 to 117.2 mg/L in teas prepared with 9 g of plant for 1 L of water [26] (corresponding to 4.18 and 58.6 mg/L with 5 g of plant for 1 L of water, [6]). Therefore, the concentrations of ART measured in this work are in the lower range compared to those already published. Unfortunately, information about the varieties of A. annua cultivated in the Beninese plantations was unavailable. However, according to PCA (Fig 3), they are rather homogenous. Series 2 extractions performed one year after series 1 showed significant ART losses in teas prepared with plants from Dogbo (ACDT), Banigbé (AB) and Kétou (AK) (Fig 2). These findings align with WHO data from 2001 which reported ART losses of 12,87% over six months and 32.76% over 12 months storage [27]. Similarly, in this study losses in ART ranged from 24% to 48% for the affected samples. Again, in 2006, WHO warned against a storage under high humidity and temperature, which renders raw vegetal material unsuitable for ART extraction after one year [28]. Simonnet et al. observed the same phenomenon when monitoring ART content of dried A. annua powder. However, they demonstrated that entire dry leaves can be stored at 20 or 30°C for up to one year without any loss of ART [29]. It is important to point out here that our previous survey in Benin revealed that A. annua herbal teas are sold as powder in tea bags with an expiry date of three years from the date of packaging [10]. This is far too much in the light of our findings and WHO data. Table 2 highlights the most significant features obtained from an OPLS-DA comparing the global composition of series 1 and 2. These results confirmed ART degradation over time. Furthermore, we found that an alkaloid (molecular formula C14H19NO2) and deoxyartemisinin were the most discriminant features between the 2 groups, being higher in series 2. Therefore, ART could be converted in these two molecules during storage. Deoxyartemisinin has been characterized in A. annua for a long time, and is also one of the phase I metabolites of ART in vivo [30]. Czechowski et al. also observed an accumulation of deoxyartemisinin during A. annua drying (14 days at 40°C), without concurrent ART loss, attributing this to a post-harvest conversion of hydroperoxydated dihydroartemisinic acid (DHAAOOH) into deoxyartemisinin [31].
Regarding the in vitro antiplasmodial activity, it was clearly correlated with the concentration of ART in the teas, regardless of the preparation series of the teas. The curve of IC50 as a function of the concentration of ART could be modelled as a power curve (Fig 5) which is coherent with the fact that ART is the principal active component of the tea. One of the reasons for the semi-synthesis of ART derivatives is its low bioavailability [32,33]. Interestingly, Räth et al. demonstrated through a pharmacokinetic study that ART derived from the consumption of A. annua herbal tea was present in the plasma much more rapidly than when absorbed through capsules containing only ART. This faster bioavailability is attributed to the presence of plant constituents that enhance the solubility of ART [34]. It has been demonstrated that A. annua tea inhibits the activity of the hepatic cytochromes CYP2B6 and CYP3A4, thereby preventing the degradation of ART in the liver. A. annua tea could also prevent the transcription and translation of these cytochromes [35]. In addition, the chemical constituents of A. annua herbal tea may enhance the antimalarial effects of ART. In our study, some teas precisely displayed low IC50 values, i.e., high antiplasmodial activity, but had lower content in ART than other teas, such as the teas from Abomey Calavi (AC) and Cotonou (ACO). The fact that others compounds present in the tea may enhance the in vitro antiplasmodial activity of ART has been previously studied [18,25,26,36] and one tool used to evidence this is to calculate the concentration of ART corresponding to the IC50 value of the tea, called here ART(tea)_IC50, and compare it to the IC50 of pure ART. If ART(tea)_IC50 is lower than the IC50 of pure ART, other compounds may contribute to the activity of the tea. In this study, we observed that ART(tea)_IC50 was lower than IC50 of pure ART for Ouidah (AODH), Abomey Calavi (AC), Cotonou (ACO), Lokossa (AMLH) and Banigbé (AB) samples, while it was higher for the Soual (both harvest time), Kétou (AK), Dogbo (ACDT), and Adjarra (ADJR) samples, indicating that other components of these teas may have a detrimental effect on the in vitro antiplasmodial activity. Such detrimental effect has never been reported before. De Donno et al. reported ART(tea)_IC50 values three time lower than the IC50 value of ART for tea made from A. annua cultivated in Italy and tested against D10 and W2 P. falciparum strains. Similarly, with an A. annua tea made from an ART high yielding variety, Suberu et al. reported that their tea was 3–7 times more potent that ART alone on HB3 and Dd2 strains respectively. Mouton et al. showed no significant differences for teas made from a collection of 16 different samples of A. annua from different varieties and origins, when tested on 3D7 strains. The same result was observed by Silva et al. with 3 samples of an ART high content A. annua variety from Manaus, and tested on 3D7 and K1 strains.
The hypothesis that ART is more effective when taken as a tea of A. annua, that is, combined with the other molecules present in the tea, is a topic of debate. Our results, together with those already published, demonstrate that this hypothesis may be true for certain plants characterized by their variety, origin, storage time, and harvest period, but the opposite may also be true. There are seminal studies showing that casticin [37] and chrysosplenol D [38], two flavonoids present in the tea, enhance the in vitro antiplasmodial activity of ART. However, Czechowski et al. concluded that “flavonoids do not contribute to antimalarial activity in whole leaf extracts”. This team compared the in vitro antiplasmodial activity of a wild A. annua plant and a genetically modified plant without flavonoids but producing the same amount of ART, and found no difference [39]. Suberu et al. showed that compounds present in the tea like sesquiterpenes, flavonoids, phenolic acid, caffeoylquinic acids may have synergetic, additive or antagonistic effect, depending on their ratio with ART and the P. falciparum strain used (CQ resistant or sensitive) [18].
These previous studies were focused on the study of known and targeted metabolites, and their capacity to enhance activity of ART in vitro. We deemed it relevant to model LC-MS data with untargeted metabolomic tools like OPLS to determine the link between the variation of LC-MS signals and the variation of biological activity, and uncover yet unknown metabolites potentially linked to the antiplasmodial activity of the tea. To the best of our knowledges, untargeted OPLS analysis has been used once with in vivo antimalarial activity of different extracts of A. annua, and led to the identification of several polymethoxylated flavones together with sesquiterpenes, like 5α-hydroperoxy-eudesmane-4(15),11-diene and arteannuin B, linked with antimalarial activity [40]. Another study linking by hierarchical cluster analysis (HCA) in vitro antiplasmodial activity of several extracts of A. annua and their compositions determined by GC-MS, identified various volatile compounds and deoxyartemisinin as positively correlated with the activity [41]. Our results of the OPLS analysis with IC50 as the response are summarized in Table 3, where the signals with the highest VIP scores are presented. For the sesquiterpenes, signals of ART (entries 9 and 16), deoxyartemisinin (entry 5) and an unidentified compound, which is not artemisinic acid (entry 11), contributed positively to activity. Deoxyartemisinin itself has a low antiplasmodial reported activity (IC50 > 10 µM on 3D7 strain) [42], and its link with the antiplasmodial activity of A. annua extracts was reported once [41]. An unidentified isomer of ART (entry 8) and arteannuin B (entries 2, 4 and 12) were correlated with a lower antiplasmodial activity. Arteannuin B has a weak antiplasmodial activity compared to ART, but has been previously identified to contribute positively to in vitro and in vivo antimalarial activity [18,43]. Indeed, for biogenetic reasons, its quantity in A. annua is negatively correlated to the quantity of ART [22], and this could explain its negative correlation with the activity of the tea. It should be noted that the same phenomenon may also apply to compounds that are positively correlated with activity. This may be because their quantity is positively linked to the quantity of ART, rather than because they affect antiplasmodial activity. The possibility of such indirect correlations poses a limitation to the use of OPLS for the search for new bioactive compounds in A. annua. Entry 13 shows that chrysosplenol D was positively correlated with activity, which is coherent with previous findings (vide supra). Surprisingly, our study highlights that several aminoacids (proline, phenylalanine, tryptophan, entries 1, 6 and 10) may contribute to antiplasmodial activity, which has not been previously reported. However, it has already been reported that short peptides including phenylalanine residues, exhibit antimalarial activity, likely by interfering with the activity of parasitic enzymes such as plasmepsin II, which is involved in hemoglobin degradation, or aminoacyl tRNA synthetase [44,45]. Interestingly, some alkaloids not previously reported in A. annua are positively correlated with activity (entries 3 and 14). We also modeled the link between variation of composition and variation of the ART(tea)_IC50 (Table 4), and this highlighted the correlation of chlorogenic acid and derivatives (entries 4, 7, 9, 11), to low values of ART(tea)_IC50. Suberu et al. have already reported interaction between ART, chlorogenic acid and chlorogenic acid derivatives, which may be additive, synergetic or antagonistic, depending on their ratio with ART and the strain of P. falciparum. These compounds showed a tendency to be synergetic at high dose on CQ-sensitive strains and antagonistic at low dose on CQ-resistant strains. The authors attributed this antagonistic interaction to the antioxidant properties of these compounds [18]. This is coherent with our results on 3D7, a chloroquine sensitive strain.
Noteworthy, the three OPLS of this study highlighted the potential role of unreported alkaloids in A. annua. Feature 1 in Table 2 (C14H19NO2, rt = 4.34 min) appeared also in Table 3 (entry 3) and Table 4 (entry 12). Its molecular formula was the same as rupestine G and rupestine K, previously reported in Artemisia rupestris L. [46] although their structures were not compatible with the fragmentation pattern observed for the unknown compound. Feature 15 in Table 3 (C14H17N3O, rt = 4.00 min) appeared in Table 4 (entry 6). Two alkaloids, tentatively identified as quinoline or isoquinoline appeared in Table 2 (entries 5 and 9). Alkaloids are rarely mentioned in the studies on A. annua and their presence in the tea is yet unreported. Alkylamides can be found in Asteraceae (Echinacea genus), but have not been detected in this study. Interestingly, toxic pyrrolizidines alkaloids have been reported in Artemisia capillaris Thunb. [47] and Artemisia scoparia Waldst. & Kit. [48]. One sample of A. annua among 386 samples of others herbs used in TCM was tested for its content of these molecules, and the amount detected (11 µg/kg) was considered low and not a threat to human health [48]. A review dealing with alkaloids in the genius Artemisia reports only 8 alkaloids previously described in A. annua. Three are purine alkaloids, three are indole alkaloids, and the others two are dipeptides [49]. Alkaloids are known for their wide biological activities, as well as their potential toxicity. Isolation and identification of the alkaloids present in the teas of A. annua teas is therefore a major challenge.
Conclusion
In the context of A. annua production for local consumption of antimalarial herbal teas, our results reveal that the teas prepared with Beninese plants present variability in their composition, and especially in their ART content, depending on the location of the plantations. Furthermore, we confirm that ART content decreases with storage duration. These findings emphasize the need for a better identification of A. annua varieties available on the market. Packaging should include harvest date, and shorter shelf lives should be recommended.
Further studies are also necessary to define a number of quality criteria for A. annua herbal teas sold as anti-malarial treatments, particularly regarding the minimum ART content for proven anti-malarial activity. Although ART is not the sole factor responsible for the activity, it is currently the most reliable chemical marker available. Defining this minimum amount would require extremely costly phase 1 and 2 clinical trials. Considering the WHO standards for efficacy (less than 10% total failure rate), achieving official approval for the use of these teas for the treatment of uncomplicated malaria seems challenging (Mueller et al. reported 77% efficacy rate at Day 7 with teas containing 47 mg of ART per litre, more than 20 years ago).
Our study also confirmed others findings showing that the claim of a better antiplasmodial activity of ART taken in a A. annua tea, rather than alone, thanks to the synergy of other compounds cannot be generalized, at least in vitro. Additionally, our results also highlighted the presence of unknown alkaloids in A. annua teas, which are worth identifying and studying to determine their antiplasmodial activity as well as their toxicity. Furthermore, potential contribution of amino-acids and yet unknown sesquiterpenes to the antiplasmodial activity deserves further investigations.
Supporting information
S1 File. Supporting figures
Fig S1 Calibration curve for 283 → 219 transition. Solutions of ART in methanol. Fig S2 Calibration curve for 283 → 229 transition. Solutions of ART in methanol. Fig S3 Calibration curve for 283 → 219 transition. Solutions of ART in A. afra tea diluted 1000 times. Fig S4 Calibration curve for 283 → 229 transition. Solutions of ART in A. afra tea diluted 1000 times. Fig S5 Calibration curves for 283 → 219 transition. Comparison between dilution in methanol, in A. Afra tea, in A. afra tea diluted 100 times, in A. afra tea diluted 1000 times. Fig S6 Calibration curves for 283 → 229 transition. Comparison between dilution in methanol, in A. Afra tea, in A. afra tea diluted 100 times, in A. afra tea diluted 1000 times. Fig S7 Result of the OPLS-DA analysis discriminating series 1 and series 2. Fig S8 Result of the OPLS analysis with IC50 value as Y input. Fig S9 Result of the OPLS analysis with ART(tea)_IC50 value as Y input. Fig S10 LC-HRMS chromatograms of lyophilized tea dissolved in water (in red) and frozen tea (in blue) for 3 samples of series 1 (AB2, AODH1, AS1). Fig S11 LC-HRMS chromatograms of lyophilized tea dissolved in water (in red) and frozen tea (in blue) for 3 samples of series 2 (AB1, AC2, ACO2). Fig S12 LC-HRMS chromatograms of lyophilized tea dissolved in water (in red) and frozen tea (in blue) for 3 samples of series 2 (ADJR1, AMLH2, AS2).
https://doi.org/10.1371/journal.pone.0330682.s001
(DOCX)
S2 File. Supporting tables.
Table S1 Location and date of sample collection, date of herbal tea preparation. Table S2 Comparison between results obtained with standard addition method, and with the use of a calibration curve made in A. afra tea diluted 1000 times, for three samples of A. annua tea (Soual_01_1, Soual_01_2, Soual_01_3) diluted 100 times. Table S3 Intraday variation for two standard solutions (5 and 12.5 ng/mL of ART in A. afra tea diluted 1000 times). Table S4 Interday variation for the standard solutions (0.5, 3.125, 5, 6.25, 12.5 ng/mL of ART in A. afra tea diluted 1000 times). Table S5 Results of determination of C_ART in A. annua teas using the 283 → 219 transition. Table S6 Results of determination of C_ART in A. annua teas using the 283 → 229 transition. Table S7 Comparison of the results for both transitions for the determination of C_ART in the teas expressed in mg/mL. Table S8: Summary of the results obtained for the 50% inhibitory concentration of the teas on 3D7 strain of P. falciparum (IC50), the concentration of ART in the teas (C_ART), the total concentration of the teas (C_Tea) and the concentration of ART corresponding to the IC50 of the teas [ART(tea)_IC50].
https://doi.org/10.1371/journal.pone.0330682.s002
(DOCX)
Acknowledgments
The authors are grateful to all the producers of A. annua who allowed plants to be collected from their fields for this study. The authors thank Dr Romain Duval (UMR 261 Merit, Paris, France) for the generous gift of chemical standards (3,4 dicaffeoylquinic acid, 3,5 dicaffeoylquinic acid, 4,5 dicaffeoylquinic acid, jaceidin and casticin), and the sharing of its knowledges on A. annua. The authors also thanks Dr. Christel Fiorini (Pierre Fabre Laboratories, Soual, France), responsible for ‘Conservatoire Botanique Pierre Fabre’, for the generous gift of the Soual_01 and Soual_02 samples of A. annua.
References
- 1.
WHO. World malaria report 2024. Geneva, Switzerland: World Health Organization. https://www.who.int/teams/global-malaria-programme/reports/world-malaria-report-2024
- 2. Ashley EA, Dhorda M, Fairhurst RM, Amaratunga C, Lim P, Suon S, et al. Spread of artemisinin resistance in Plasmodium falciparum malaria. N Engl J Med. 2014;371(5):411–23. pmid:25075834
- 3. Tu Y. The discovery of artemisinin (qinghaosu) and gifts from Chinese medicine. Nat Med. 2011;17(10):1217–20. pmid:21989013
- 4. Mueller MS, Karhagomba IB, Hirt HM, Wemakor E. The potential of Artemisia annua L. as a locally produced remedy for malaria in the tropics: agricultural, chemical and clinical aspects. J Ethnopharmacol. 2000;73(3):487–93. pmid:11091003
- 5. Ogwang P, Ogwal J, Kasasa S, Olila D, Ejobi F, Kabasa D, et al. Artemisia Annua L. Infusion Consumed Once a Week Reduces Risk of Multiple Episodes of Malaria: A Randomised Trial in a Ugandan Community. Trop J Pharm Res. 2012;11(3).
- 6. Mueller MS, Runyambo N, Wagner I, Borrmann S, Dietz K, Heide L. Randomized controlled trial of a traditional preparation of Artemisia annua L. (Annual Wormwood) in the treatment of malaria. Trans R Soc Trop Med Hyg. 2004;98(5):318–21. pmid:15109558
- 7. Camara A, Haddad M, Traore MS, Chapeland-Leclerc F, Ruprich-Robert G, Fourasté I, et al. Variation in chemical composition and antimalarial activities of two samples of Terminalia albida collected from separate sites in Guinea. BMC Complement Med Ther. 2021;21(1):64. pmid:33588819
- 8. Conforti F, Marcotullio MC, Menichini F, Statti GA, Vannutelli L, Burini G, et al. The influence of collection zone on glucosinolates, polyphenols and flavonoids contents and biological profiles of Capparis sicula ssp. sicula. Food Sci Technol Int. 2011;17(2):87–97. pmid:21436235
- 9.
WHO. The use of non-pharmaceutical forms of Artemisia. World Health Organization. 2019. https://www.who.int/publications/i/item/WHO-CDS-GMP-2019.14
- 10. Chassagne F, Akadiri S, Amoussa AMO, Agossa J, Yedomonhan H, Lagnika L, et al. Artemisia annua et Artemisia afra au Bénin: un engouement récent pour deux plantes non traditionnelles. Ethnopharmacologia. 2023;2(69):25–33.
- 11.
WHO. WHO guidelines for malaria. WHO. 2021. https://iris.who.int/handle/10665/343751
- 12. Wang X, Zheng S, Liu Y, Han J. ITS2, a Better DNA Barcode than ITS in Identification of Species in Artemisia L. Chinese Herbal Medicines. 2016;8(4):352–8.
- 13. Pluskal T, Castillo S, Villar-Briones A, Oresic M. MZmine 2: modular framework for processing, visualizing, and analyzing mass spectrometry-based molecular profile data. BMC Bioinformatics. 2010;11:395. pmid:20650010
- 14. Sorng S, Balayssac S, Danoun S, Assemat G, Mirre A, Cristofoli V, et al. Quality assessment of Curcuma dietary supplements: Complementary data from LC-MS and 1H NMR. J Pharm Biomed Anal. 2022;212:114631. pmid:35231794
- 15. Desjardins RE, Canfield CJ, Haynes JD, Chulay JD. Quantitative assessment of antimalarial activity in vitro by a semiautomated microdilution technique. Antimicrob Agents Chemother. 1979;16(6):710–8. pmid:394674
- 16. Le Nagard H, Vincent C, Mentré F, Le Bras J. Online analysis of in vitro resistance to antimalarial drugs through nonlinear regression. Comput Methods Programs Biomed. 2011;104(1):10–8. pmid:20828858
- 17. Maciuk A, Mazier D, Duval R. Future antimalarials from Artemisia? A rationale for natural product mining against drug-refractory Plasmodium stages. Nat Prod Rep. 2023;40(6):1130–44. pmid:37021639
- 18. Suberu JO, Gorka AP, Jacobs L, Roepe PD, Sullivan N, Barker GC, et al. Anti-plasmodial polyvalent interactions in Artemisia annua L. aqueous extract--possible synergistic and resistance mechanisms. PLoS One. 2013;8(11):e80790. pmid:24244716
- 19. Suberu J, Song L, Slade S, Sullivan N, Barker G, Lapkin AA. A rapid method for the determination of artemisinin and its biosynthetic precursors in Artemisia annua L. crude extracts. J Pharm Biomed Anal. 2013;84:269–77. pmid:23867088
- 20. Protti M, Mandrioli R, Mandrone M, Cappadone C, Farruggia G, Chiocchio I, et al. Analysis of Artemisia annua extracts and related products by high performance liquid chromatography-tandem mass spectrometry coupled to sample treatment miniaturisation. J Pharm Biomed Anal. 2019;174:81–8. pmid:31158609
- 21. Qiu F, Wu S, Lu X, Zhang C, Li J, Gong M, et al. Quality evaluation of the artemisinin-producing plant Artemisia annua L. based on simultaneous quantification of artemisinin and six synergistic components and hierarchical cluster analysis. Industrial Crops and Products. 2018;118:131–41.
- 22. Czechowski T, Larson TR, Catania TM, Harvey D, Wei C, Essome M, et al. Detailed Phytochemical Analysis of High- and Low Artemisinin-Producing Chemotypes of Artemisia annua. Front Plant Sci. 2018;9:641. pmid:29868094
- 23. Fu C, Yu P, Wang M, Qiu F. Phytochemical analysis and geographic assessment of flavonoids, coumarins and sesquiterpenes in Artemisia annua L. based on HPLC-DAD quantification and LC-ESI-QTOF-MS/MS confirmation. Food Chem. 2020;312:126070. pmid:31911352
- 24. Li L, Josef BA, Liu B, Zheng S, Huang L, Chen S. Three-Dimensional Evaluation on Ecotypic Diversity of Traditional Chinese Medicine: A Case Study of Artemisia annua L. Front Plant Sci. 2017;8:1225. pmid:28744301
- 25. Silva LFR, Magalhães PM, Costa MRF, Alecrim MGC, Chaves FCM, Hidalgo AF, et al. In vitro susceptibility of Plasmodium falciparum Welch field isolates to infusions prepared from Artemisia annua L. cultivated in the Brazilian Amazon. Mem Inst Oswaldo Cruz. 2012;107(7):859–66. pmid:23147140
- 26. Mouton J, Jansen O, Frédérich M, van der Kooy F. Is artemisinin the only antiplasmodial compound in the Artemisia annua tea infusion? An in vitro study. Planta Med. 2013;79(6):468–70. pmid:23512495
- 27.
WHO. Meeting on antimalarial drug development, Shanghai, China, 16-17 November 2001: report. Regional Office for the Western Pacific. 2001. https://iris.who.int/handle/10665/207999
- 28.
WHO. WHO monograph on good agricultural and collection practices (GACP) for Artemisia annua L. World Health Organization. 2006. https://iris.who.int/handle/10665/43509
- 29.
Simonnet X, Grogg AF, Cutler M, Bathurst I. Artemisia annua L. Drying and Storage 2008 https://assets.publishing.service.gov.uk/media/57a08bc4ed915d3cfd000f1c/01_Artemisia_Guilin_2008.pdf].
- 30. Liu T, Du F, Wan Y, Zhu F, Xing J. Rapid identification of phase I and II metabolites of artemisinin antimalarials using LTQ-Orbitrap hybrid mass spectrometer in combination with online hydrogen/deuterium exchange technique. J Mass Spectrom. 2011;46(8):725–33. pmid:21766391
- 31. Czechowski T, Larson TR, Catania TM, Harvey D, Brown GD, Graham IA. Artemisia annua mutant impaired in artemisinin synthesis demonstrates importance of nonenzymatic conversion in terpenoid metabolism. Proc Natl Acad Sci U S A. 2016;113(52):15150–5. pmid:27930305
- 32. Ashton M, Nguyen DS, Nguyen VH, Gordi T, Trinh NH, Dinh XH, et al. Artemisinin kinetics and dynamics during oral and rectal treatment of uncomplicated malaria. Clin Pharmacol Ther. 1998;63(4):482–93. pmid:9585803
- 33.
Li Q, Pybus B. Pharmacokinetic and pharmacodynamic profiles of rapid- and slow-acting antimalarial drugs. In: Kasenga FH. Malaria. Rijeka: IntechOpen. 2019.
- 34. Räth K, Taxis K, Walz G, Gleiter CH, Li S-M, Heide L. Pharmacokinetic study of artemisinin after oral intake of a traditional preparation of artemisia annua L. (annual wormwood). Am J Trop Med Hyg. 2004;70(2):128–32.
- 35. Kane NF, Kiani BH, Desrosiers MR, Towler MJ, Weathers PJ. Artemisia extracts differ from artemisinin effects on human hepatic CYP450s 2B6 and 3A4 in vitro. Journal of Ethnopharmacology. 2022;298:115587.
- 36. De Donno A, Grassi T, Idolo A, Guido M, Papadia P, Caccioppola A, et al. First-time comparison of the in vitro antimalarial activity of Artemisia annua herbal tea and artemisinin. Transactions of the Royal Society of Tropical Medicine and Hygiene. 2012;106(11):696–700.
- 37. Elford BC, Roberts MF, Phillipson JD, Wilson RJ. Potentiation of the antimalarial activity of qinghaosu by methoxylated flavones. Trans R Soc Trop Med Hyg. 1987;81(3):434–6. pmid:3318019
- 38. Liu KC, Yang SL, Roberts MF, Elford BC, Phillipson JD. Antimalarial activity of Artemisia annua flavonoids from whole plants and cell cultures. Plant Cell Rep. 1992;11(12):637–40. pmid:24213368
- 39. Czechowski T, Rinaldi MA, Famodimu MT, Van Veelen M, Larson TR, Winzer T, et al. Flavonoid Versus Artemisinin Anti-malarial Activity in Artemisia annua Whole-Leaf Extracts. Front Plant Sci. 2019;10:984. pmid:31417596
- 40. Shi H, Wang Z, Xu F, Li J, Li J, Wang M. A Discovery-Based Metabolomic Approach Using UPLC-Q-TOF-MS/MS Reveals Potential Antimalarial Compounds Present in Artemisia annua L. Int J Mol Sci. 2022;23(23):14903. pmid:36499227
- 41. Sankhuan D, Niramolyanun G, Kangwanrangsan N, Nakano M, Supaibulwatana K. Variation in terpenoids in leaves of Artemisia annua grown under different LED spectra resulting in diverse antimalarial activities against Plasmodium falciparum. BMC Plant Biol. 2022;22(1):128. pmid:35313811
- 42. Antoine T, Fisher N, Amewu R, O’Neill PM, Ward SA, Biagini GA. Rapid kill of malaria parasites by artemisinin and semi-synthetic endoperoxides involves ROS-dependent depolarization of the membrane potential. J Antimicrob Chemother. 2014;69(4):1005–16. pmid:24335485
- 43. Cai T-Y, Ji J-B, Wang X, Xing J. Targeted screening of the synergistic components in Artemisia annua L. leading to enhanced antiplasmodial potency of artemisinin based on a “top down” PD-PK approach. Journal of Ethnopharmacology. 2024;322:117612.
- 44. Pérez-Picaso L, Velasco-Bejarano B, Aguilar-Guadarrama AB, Argotte-Ramos R, Rios MY. Antimalarial activity of ultra-short peptides. Molecules. 2009;14(12):5103–14. pmid:20032878
- 45. Xie SC, Griffin MDW, Winzeler EA, Ribas de Pouplana L, Tilley L. Targeting Aminoacyl tRNA Synthetases for Antimalarial Drug Development. Annu Rev Microbiol. 2023;77:111–29. pmid:37018842
- 46. He F, Nugroho AE, Wong CP, Hirasawa Y, Shirota O, Morita H, et al. Rupestines F-M, new guaipyridine sesquiterpene alkaloids from Artemisia rupestris. Chem Pharm Bull (Tokyo). 2012;60(2):213–8. pmid:22293480
- 47. Chen L-H, Wang J-C, Guo Q-L, Qiao Y, Wang H-J, Liao Y-H, et al. Simultaneous Determination and Risk Assessment of Pyrrolizidine Alkaloids in Artemisia capillaris Thunb. by UPLC-MS/MS Together with Chemometrics. Molecules. 2019;24(6):1077. pmid:30893797
- 48. Wang J, Zhang M, Chen L, Qiao Y, Ma S, Sun D, et al. Determination of Toxic Pyrrolizidine Alkaloids in Traditional Chinese Herbal Medicines by UPLC-MS/MS and Accompanying Risk Assessment for Human Health. Molecules. 2021;26(6):1648. pmid:33809536
- 49. Ur Rashid M, Alamzeb M, Ali S, Ullah Z, Shah ZA, Naz I, et al. The chemistry and pharmacology of alkaloids and allied nitrogen compounds from Artemisia species: A review. Phytother Res. 2019;33(10):2661–84. pmid:31453659