Antiplasmodial Properties of Aqueous and Ethanolic Extracts of Ten Herbal Traditional Recipes Used in Thailand against Plasmodium falciparum

This study evaluated the in vitro and in vivo antiplasmodial efficacy and toxicity of aqueous and ethanolic extracts from traditional recipes used in Thailand. The aqueous and ethanolic extracts of ten traditional recipes were tested for in vitro antiplasmodial activity (parasite lactate dehydrogenase assay), cytotoxicity (MTT assay), and hemolysis). Oxidant levels were measured using cell-permeable probe 5-(and-6)-chloromethyl-2′,7′-dichlorodihydrofluorescein diacetate fluorescent dye-based assays. The best candidate was chosen for testing in mouse models using 4-day suppressive and acute toxicity assays. An in vitro study showed that ethanolic extracts and three aqueous extracts exhibited antiplasmodial activity, with an IC50 in the range of 2.8–15.5 µg/mL. All extracts showed high CC50 values, except for ethanolic extracts from Benjakul, Benjalotiga, and Trikatuk in HepG2 and Benjalotiga and aqueous extract from Chan-tang-ha in a Vero cell. Based on the results of the in vitro antiplasmodial activity, an aqueous extract of Triphala was chosen for testing in mouse models. The aqueous extract of Triphala exhibited good antiplasmodial activity, was safe at an oral dose of 2 g/kg, and is a potential candidate as a new source for the development of antimalarial drugs.


Introduction
The malaria burden has an impact around the world and is the heaviest, according to the World Health Organization (WHO). The African region accounted for approximately 95% of all cases and 96% of all deaths in 2020 [1]. There are five Plasmodium species that cause devastation to human life-P. falciparum, P. vivax, P. malariae, P. ovale, and P. knowlesi. A key feature of these infections is their ability to invade erythrocytes [2]. Sequestered parasites can lead to impaired blood flow and induce negative effects on critical organs [3]. However, effective tools for malaria control and elimination rely on potent antimalarial agents. The WHO recommends six artemisinin combinations (ACTs) as first-line treatment for uncomplicated falciparum malaria: artemether-lumefantrine, artesunate-amodiaquine, artesunate-mefloquine, artesunate-sulfadoxine-pyrimethamine, dihydroartemisinin-piperaquine, and artesunate-pyronaridine [4]. Blood-stage malaria plays a crucial role in the development of symptoms and clinical complications and can be terminated by blood schizonticides. Unfortunately, artemisinin and partner drugs

Preparation of Crude Extract
The extraction process was performed as previously described [19]. The maceration method was performed to prepare ethanolic extracts. Sixty grams of the powdered recipe was extracted for 72 h in 600 mL of 80% ethanol at 25 • C. The decoction method was used to make aqueous extracts. Sixty grams of each herb powder was extracted three times by mixing with 600 mL of distilled water and allowed to boil for 30 min. Subsequently, the liquid portion was separated from the residue using filter paper (Whatman, Buckinghamshire, England). The marcs of ethanolic and aqueous extract were re-extracted twice, with 600 mL of solvents each time. The filtrates were combined and concentrated using a rotary evaporator (Rotavapor, Buchi, China) at 45 rpm and 45 • C. Further drying was performed in a freeze-drying machine at −89 • C (Martin Christ, Germany). The crude extracts were stored in a refrigerator until further use.

In Vitro Culture of Plasmodium Parasites
K1 chloroquine-resistant P. falciparum strain was used in this study. The cryopreserved parasite-infected blood was thawed using a 12% and 1.6% NaCl concentration gradient. The culture of the parasite was slightly modified from the original methods developed by Trager and Jensen [22]. It was cultured in a T-75 flask containing 2% human O+ erythrocytes, base media of RPMI-1640 medium (Gibco, Carlsbad, CA, USA) supplemented with 2 mg/mL NaHCO 3 , 4.8 mg/mL HEPES (Himedia, Mumbai, India), 10 µg/mL hypoxanthine (Sigma-Aldrich, New Delhi, India), 2.5 µg/mL gentamicin (Sigma-Aldrich, New Delhi, India), and 0.5% albumax II (Gibco MA, USA) in a saturated atmosphere of 5% CO 2 at 37 • C. In order to assess parasitemia and developmental stages, a thin blood smear was stained with Giemsa dye, and intracellular parasites were visualized under an oil immersion lens (100×) using a light microscope (Olympus CX31, Model CX31RBSFA, Tokyo, Japan).

Detection of Parasite Lactate Dehydrogenase (pLDH) Activity
A stock solution (20 mg/mL) of the aqueous extracts was prepared by dissolving the extract in phosphate-buffered saline (PBS), whereas the ethanolic extracts were dissolved in DMSO (dimethyl sulfoxide). A two-fold serial dilution was prepared using PBS (aqueous extracts) or DMSO (ethanolic extracts), covering a range of final concentration from 1.56 to 100 µg/mL. The antiplasmodial activity was determined by the concentration of the extracts that inhibited 50 percent of the parasite growth (IC 50 ) by the measurement of pLDH activity, as previously described [23]. The crude extract (1µL) was incubated with 1% parasitized red blood cells (pRBCs) and 2% hematocrit (199 µL) in a 96-well cell culture plate (SPL Life Sciences, Pocheon-si, Gyeonggi-do, Korea) in an atmosphere of 5% CO 2 at 37 • C for 72 h. The test was performed in triplicate for each concentration. Artesunate (Sigma, St Louis, MO USA) was used as the positive control (final concentration ranging from 1.56 to 100 ng/mL). PBS and DMSO were used as negative controls. Non-infected red blood cells served as blank controls. The suspension was frozen at −80 • C and thawed at 37 • C to lyse red cell pellets. The pLDH enzyme was detected by the reaction between 20 µL of their contents from the released red cells, 100 µL of malstate reagent, and 20 µL of nitroblue tetrazolium/phenazine ethosulfate solution (Calbiochem, Sigma-Aldrich, New Delhi, India) in a new 96-well plate flat bottom. The absorbance was measured at 650 nm (Biotek Eon, Winooski, VT, USA) after the reaction was incubated in the dark for an hour. The percent inhibition was calculated compared to the negative control after subtraction of the background using the following formula: % inhibition = 100 × [(OD negative well − OD sample well)/OD negative well] IC 50 was determined for each sample by plotting % growth as a function of concentration and estimating the concentration which caused 50% growth inhibition. Crude extracts were evaluated for toxicity at concentrations ranging from 12.5 to 800 µg/mL. Serial dilutions were prepared using a 160 mg/mL stock solution. Cell cultures at a density of 10 4 cells/well in 200 µL culture medium were seeded in a 96-well cell culture plate and then incubated at 37 • C for 24 h in an atmosphere of 5% CO 2 until they reached confluence. HepG2 and Vero cell lines (199 µL) were treated with different concentrations of the extract (1 µL), with doxorubicin (final concentration ranging from 0.3 to 20 µg/mL) (Sigma-Aldrich, New Delhi, India) as a positive control. DMSO and PBS served as negative controls for ethanolic and aqueous extracts. Treated cells were incubated in an incubator at 37 • C for 48 h. The assays were performed in triplicates. MTT reagent (5 mg/mL) was added to each well at the end of the exposure period and then incubated at 37 • C for 3 h. Subsequently, the reagent was removed and replaced with 100 µL of DMSO. The absorbance was read at a wavelength of 590 nm using a microplate reader. The percent cytotoxicity was calculated to determine the cytotoxicity as follows: The data in this assay are presented as the 50% cytotoxicity concentration (CC 50 ), as determined by regression analysis using GraphPad Prism 6.

Hemolysis Measurement
The toxicity of crude extracts in human erythrocytes was evaluated by monitoring the hemolysis of the red cell suspension. Venous blood was drawn from healthy donors into EDTA-blood collection tubes, and the red cell pellet was washed three times with PBS. The plasma and buffy coat were discarded after centrifugation at 3000 rpm for 5 min. Cell suspension at 2% hematocrit was incubated with 50 µg/mL of the extracts in a final volume of 0.2 mL in an incubator at 37 • C for 72 h. Triton X-100 (Sigma-Aldrich, New Delhi, India) was used as a positive control. DMSO and PBS were used as negative controls. The plate was centrifuged at 3000 rpm for 5 min, and the supernatant was transferred to a new 96-well plate. The release of hemoglobin was measured at 570 nm, and the percentage of hemolysis was calculated as follows: % hemolysis = [100 × (OD sample well − OD negative well)/(OD positive well − OD negative well)]

Measurement of Intracellular Oxidant in pRBCs
Oxidant generation was measured using the cell-permeable probe 5-(and-6)-chloromethyl-2 ,7 -dichlorodihydrofluorescein diacetate (CM-H2DCFDA) (Invitrogen, Carlsbad, CA, USA). In this study, parasite culture at 1% parasitemia and 2% hematocrit was incubated with the presence of proximity IC 50 concentration of drug or crude extracts in a 96-well plate for 72 h. Artesunate was used as the positive control. Negative controls were obtained using 0.5% DMSO and PBS. pRBCs were labeled with a fluorescence probe after disposal of the medium. The pellet was incubated in the dark with 100 µL CM-H2DCFDA solution at a final concentration of 10 mM for 30 min. Labeled cells were imaged using a Leica TCS SP5 confocal microscope (Leica, Mannheim, Germany) at an excitation/emission wavelength of 488/520 nm. In order to measure oxidant levels, five fields of each sample were acquired, and two hundred and twenty-five cells per sample were used for analysis. The integrated density, cell area, and background fluorescence were measured using the free software ImageJ Fiji. Corrected total cell fluorescence (CTCF), in terms of normalized values, was calculated as follows: CTCF = integrated density − (area of selected cell × mean fluorescence of background readings).
The relative change was obtained by the division of CTCF from negative controls.

In Vivo Antimalarial Activity Model
The antimalarial efficacy of the crude extract against early malarial infection was evaluated using Peters' 4-day suppressive test [24]. Male ICR mice (n = 25) aged 6-8 weeks were purchased from Nomura Siam International Co., Ltd., Bangkok, Thailand. They were randomly divided into five groups of five mice each. Group I was treated with PBS. Group II was treated with 25 mg/kg of chloroquine as the standard drug. Groups III, IV, and V received the aqueous extract of Triphala at doses of 200, 400, and 600 mg/kg, respectively. The animals were acclimatized for a week in the presence of food pellets and clean drinking water ad libitum before initiating the experiments. The Plasmodium berghei (P. berghei) ANKA strain (chloroquine-sensitive) was contributed by Thomas F. McCutchan and obtained from BEI Resources, NIAID, NIH. Mice infected with P. berghei were used as a donor, and then 25 mice were injected with 0.2 mL of 10 7 infected red cells per milliliter through the intraperitoneal route. Oral administration started at 3 h post-infection, followed by 24, 48, and 72 h daily. On day 4, blood samples from each mouse were collected to prepare a thin blood film and stained with Giemsa solution. The percentages of parasitemia and suppression were calculated using the following formulas: % parasitemia = (number of infected red blood cells/number of total red blood cells) × 100 % suppression = [(mean parasitemia negative control − mean parasitemia experimental group )/mean parasitemia negative control ] × 100

Acute Toxicity Test
The test was conducted according to the Organization for Economic Co-operation and Development (OECD) guidelines for testing chemicals, with a limit test at a dose of 2 g/kg (1). Ten male ICR mice were randomly divided into two groups: PBS solution and extract solution. A single oral dose was administered directly to the stomach through a feeding tube. Abnormalities were observed within the first 30 min and once daily for 14 days after administration. The parameters that were observed for signs of toxicity included general activity, body weight changes, tremors, convulsions, ataxia, diarrhea, urination, changes in skin fur, and death. On the last day, mice were anesthetized with 2% isoflurane and then euthanized by cardiac puncture after opening the thorax cavity. Blood samples were collected and placed into serum clot activator tubes for the analysis of liver and kidney enzymes, including aspartate aminotransferase (AST), alanine aminotransferase (ALT), alkaline phosphatase (ALP), blood urea nitrogen (BUN), and creatinine. The analysis was performed using an AU480 chemistry analyzer (Beckman Coulter, USA). In addition, histological alterations of the liver and kidneys were performed using hematoxylin and eosin staining as previously described [25]. The relative weights of the liver and kidneys were calculated using the following formula: Relative organ weight = (organ weight/body weight) × 100

The Liquid Chromatography-Quadrupole Time-of-Flight Mass Spectrometry (LC-QTOF-MS) Analysis
Product profiling of the crude extract was performed using an LC-QTOF-MS instrument (1290 Infinity II LC-6545 Quadrupole-TOF, Agilent Technologies, Santa Clara, CA, USA). Chromatographic separation was performed on a Zorbax Rapid Resolution HD Eclipse Plus C18 column (150 mm length × 2.1 mm inner-diameter, particle size 1.8 µm) from Agilent (Agilent, Waldbronn, Germany). The elution gradient was performed with 0.1% formic acid water (mobile phase A) and acetonitrile (mobile phase B) at a flow rate of 0.20 mL/min. The column was equilibrated (A: B; v/v) at 90:10 (1 min), and elution was carried out with the following steps; 80:20 (from 1 to 12 min), 75:25 (from 12 to 20 min), 70:30 (from 20 to 25 min), 65:35 (from 25 to 28 min), 60:40 (from 28 to 38 min), and 90:10 (from 38 to 45 min). The column temperature was set to 25 • C. The instrument was set to an MS range of 100-1200 m/z in both negative and positive modes. The injection volume was 2 µL. Data acquisition was controlled using the MassHunter WorkStation Qualitative Analysis Workflows V8 software (Agilent Technologies, Santa Clara, CA, USA). Compounds were identified by comparing retention times, mass data, and fragmentation patterns with a compound database in the library search of the Mass Hunter METLIN database (Agilent Technologies). The peak with similarity scores of 90% compared to the database was selected to confirm peak identification.

Statistical Analysis
The data are expressed as the mean ± standard error of the mean. The relative CTCF was tested for normality before statistically assessing differences with an independent t-test; p < 0.05 was considered significant using SPSS for Microsoft Windows (version 17.0; IBM, Armonk, NY, USA). Statistical analysis of in vivo studies was performed using one-way analysis of variance at p < 0.05 after the values showed a normal distribution using SPSS for Microsoft Windows version 17.0.

Percentage Yield of Crude Plant Material
The percentage yield of each extract obtained from the dried crude extract divided by the initial amount of powdered plant part is presented in Table 2. The maximum extract yield (40.12%) was obtained from the aqueous extract of the Jatu-phala-tiga recipe, whereas the minimum yield (2.07%) was recorded for the ethanolic extract of the Benjatian recipe. The yield of aqueous extracts from Benjalotiga and Chan-tang-ha was lower than that of the ethanolic extract; however, other extracts exhibited a higher percentage yield than that of the ethanolic extract.

Phytochemical Profile
Qualitative analysis of the ten traditional recipes (aqueous/ethanolic extracts) revealed different phytocomponents, which are displayed in Table 3 and Figure 1. All extracts, except the ethanolic extracts from Trikatuk, Benjalotiga, Benjakul, and Chan-tang-ha, contained tannins. A majority of terpenoids were found in all extracts, excluding the aqueous extract from Benjagot. Anthraquinones and cardiac glycosides were not detected. Saponins were mainly found in the aqueous extracts. Table 3. Phytochemical results of the ten traditional recipes (aqueous/ethanolic extracts).

In Vitro Antiplasmodial Activity
The antiplasmodial activities of the ten traditional recipes are shown in Table 4. According to these criteria of in vitro antiplasmodial assessment of products derived from plants [26], extracts with IC50 < 5 µg/mL are considered as exhibiting potent antiplasmodial activity, IC50 between 5-15 µg/mL and 15-50 µg/mL as good and moderate activity, respectively, and IC50 >100 µg/mL as inactive. Crude ethanolic extracts from Gaysorn-tangha, Trikatuk, and Triphala exhibited potent activity against the P. falciparum K1 strain with IC50 values of 2.81, 4.37, and 4.39 µg/mL, respectively. The ethanolic extract of Benjatian revealed moderate antiplasmodial activity (IC50 = 15.5 µg/mL), and all other extracts showed good activity with IC50 ranging from 6.8-8.5 µg/mL. On the other hand, most of the aqueous extracts showed no activity (IC50 > 5 µg/mL). Interestingly, aqueous extract from Jatu-phalatiga had a potent activity with an IC50of 5.0 µg/mL, and extracts from Triphala and Trisamo

In Vitro Antiplasmodial Activity
The antiplasmodial activities of the ten traditional recipes are shown in Table 4. According to these criteria of in vitro antiplasmodial assessment of products derived from plants [26], extracts with IC 50 < 5 µg/mL are considered as exhibiting potent antiplasmodial activity, IC 50 between 5-15 µg/mL and 15-50 µg/mL as good and moderate activity, respectively, and IC 50 > 100 µg/mL as inactive. Crude ethanolic extracts from Gaysorn-tang-ha, Trikatuk, and Triphala exhibited potent activity against the P. falciparum K1 strain with IC 50 values of 2.81, 4.37, and 4.39 µg/mL, respectively. The ethanolic extract of Benjatian revealed moderate antiplasmodial activity (IC 50 = 15.5 µg/mL), and all other extracts showed good activity with IC 50 ranging from 6.8-8.5 µg/mL. On the other hand, most of the aqueous extracts showed no activity (IC 50 > 5 µg/mL). Interestingly, aqueous extract from Jatu-phala-tiga had a potent activity with an IC 50 of 5.0 µg/mL, and extracts from Triphala and Trisamo expressed good activity with IC 50 of 5.7 and 6.1 µg/mL, respectively.

In Vitro Cytotoxicity on HepG2 and Vero Cells
The cytotoxic effects of the ten traditional recipes against HepG2 and Vero cells are shown in Table 4. A benchmark of toxicity levels based on the National Cancer Institute criteria was used. Recipes with CC 50 values <30 µg/mL are considered cytotoxic after 48-72 h of exposure [27]. The results of PBS and 0.5% DMSO treated cells were considered 100% of living cells for aqueous and ethanolic extracts. The cytotoxic effects on HepG2 cells revealed that most of the extracts had CC 50 values greater than the specified criteria. Only three extracts showed cytotoxicity with a CC 50 of <30 µg/mL, indicating mild cytotoxicity [28]. The ethanolic extracts from Benjakul, Benjalotiga, and Trikatuk exhibited CC 50 values of 10.9, 20.1, and 26.5 µg/mL, respectively. A comparison between the effects of aqueous and ethanolic extracts on HepG2 cells showed that the ethanolic extract had a greater toxic effect. The cytotoxic effects on Vero cells demonstrated that ethanolic extract from Chan-tang-ha is more toxic compared to that of the others, which produced CC 50 value <12.5 µg/mL, and Benjalotiga had toxic effects with CC 50 of 15.3 µg/mL for aqueous and 20.5 µg/mL for ethanolic extracts.

Interpretation of Selectivity Index (SI)
The SI was used to calculate risk-benefit assessment and identify progressible extracts that measure the ratio between 50% toxic concentration (CC 50 ) and 50% antiplasmodial concentration (IC 50 ) [29]. An extract with values <1 could be toxic and not applicable; however, a high value will give active extracts without undue risk. As shown in Table 4, both aqueous and ethanolic extracts of Triphala, Trisamo, and Jatu-phala-tiga showed high values of >27 in both HepG2 and Vero cells. The ethanolic extracts of Gaysorn-tangha exhibited high SI values above 20 in HepG2 cells, whereas the ethanolic extracts of Benjathian, Benjagot, and Chan-tang-ha showed high SI values in Vero cells. In addition, the ethanolic extract from other extracts exhibited values greater than that of the aqueous extract in both cell lines.

Hemolysis Results
The hemolytic effects of the ten recipes were determined to indicate hematotoxicity in human erythrocytes, and the results are shown in Figure 2. Triton X-100 (0.5% v/v) was used for complete hemolysis of blood cells, which was then compared with the samples. Cell lysis was measured as the amount of free hemoglobin in the red cell suspension. Aqueous extracts at a fixed dose of 50 µg/mL did not show lytic effects on erythrocytes, except for Benjathian (6.53%) and Trikatuk (2.46%), whereas some ethanolic extracts showed hemolytic effects. The effect of Trisamo and Benjakul recipes reached 100% hemolysis after incubation for 72 h. Benjagot, Gaysorn-tang-ha, Chan-tang-ha, and Benjathian exhibited hemolysis percentages of 45.59, 39.18, 6.09, and 5.79, respectively.

Estimation of Intracellular Reactive Oxygen Species Production
CM-H2DCFDA was used to detect free radical production. The reaction of an oxidant and a CM-H2DCFDA non-fluorescent dye generates a DCF fluorescent component. Oxidant production is indicated by green fluorescence, as shown in Figure 3. Cells treated with DMSO or PBS were considered to have basal levels of oxidant generation. The positive control exhibited the highest oxidant production. The average CTCF values from positive control; aqueous extract of Trisamo; ethanolic extracts of Triphala, Trisamo, and Trikatuk recipes significantly increased ROS levels compared with the levels in negative groups at p < 0.05 by 17.06%, 15.36%, 6.26%, 9.20%, and 3.60%, respectively ( Figure 4). Contrastingly, oxidant generation in the ethanolic extract group of the Benjagot recipe was significantly decreased by 0.50%.

Estimation of Intracellular Reactive Oxygen Species Production
CM-H2DCFDA was used to detect free radical production. The reaction of an oxidant and a CM-H2DCFDA non-fluorescent dye generates a DCF fluorescent component. Oxidant production is indicated by green fluorescence, as shown in Figure 3. Cells treated with DMSO or PBS were considered to have basal levels of oxidant generation. The positive control exhibited the highest oxidant production. The average CTCF values from positive control; aqueous extract of Trisamo; ethanolic extracts of Triphala, Trisamo, and Trikatuk recipes significantly increased ROS levels compared with the levels in negative groups at p < 0.05 by 17.06%, 15.36%, 6.26%, 9.20%, and 3.60%, respectively ( Figure 4). Contrastingly, oxidant generation in the ethanolic extract group of the Benjagot recipe was significantly decreased by 0.50%.     Table 4 shows the SI values of the 20 crude extracts on the HepG2 and Vero cell lines. Nine SI values that were not determined included the values of aqueous extracts of Gaysorn-tang-ha, Benjagot, and Benjakul on both HepG2 and Vero cells, and aqueous extracts of Trikatuk, Benjatian, and Chan-tang-ha on Vero cells, because SI values cannot be calculated. The aqueous extract of Triphala showed the highest values in HepG2 cells, whereas the ethanolic extract showed a high constant value of 176.13 in Vero cells. It is imperative to review the formula used to calculate SI indices. The values of IC50 produced the same antiplasmodial activity, which was 5.7 ± 0.2 for aqueous and 4.4 ± 1.3 µg/mL for ethanolic extracts, whereas the values of cytotoxicity aqueous extract were greater than ethanolic in both HepG2 and Vero cell lines. Therefore, an aqueous extract of Triphala was selected for testing in mouse models.

Chemical Profiling of Aqueous Extract of Triphala
A total of 83 metabolites in the negative ion mode and 89 metabolites in the positive ion mode were identified from the aqueous extract of Triphala using LC-QTOF-MS. Table  5 shows several compounds identified from the Mass Hunter METLIN database library by matching their accurate masses. The peak chromatograms are presented in Figure 5. Figures 6 and 7 showed proposed fragmentation patterns in negative and positive modes, respectively.

Chemical Profiling of Aqueous Extract of Triphala
A total of 83 metabolites in the negative ion mode and 89 metabolites in the positive ion mode were identified from the aqueous extract of Triphala using LC-QTOF-MS. Table 5 shows several compounds identified from the Mass Hunter METLIN database library by matching their accurate masses. The peak chromatograms are presented in Figure 5. Figures 6 and 7 showed proposed fragmentation patterns in negative and positive modes, respectively.        Table 6 shows the percentage of parasitemia and suppression in mice infected with P. berghei. Although the most effective suppression was exhibited in the chloroquine group, which produced 100% suppression, the administration of the extract at all doses revealed a significant (p < 0.05) increase in percent suppression when compared with that of the negative control. Different doses of the extract reduced the percentage of parasitemia in a dose-dependent manner. The extract at 600 mg/kg showed the best suppressive effect (75.47%), followed by 44.13% at the dose of 400 mg/kg and 42.72% at the dose of 200 mg/kg.

Acute Toxicity Test
The mice that received the aqueous extract of Triphala at 2 g/kg did not experience any death or abnormalities of the eyes, fur, or skin and did not show behavioral changes such as general activity, tremors, convulsions, ataxia, or diarrhea, indicating that the LD 50 value of the extract was greater than 2 g/kg. Both groups of mice showed a percent increase in body weight, and no significant difference (p < 0.05) was observed between the two groups ( Table 7). The relative organ weights of the liver and kidney in mice that received the extract showed no significant difference (p < 0.05) compared to the control group (Table 8).
In addition, the extract did not induce significant (p < 0.05) changes in the liver and kidney enzymes, including BUN, creatinine, AST, ALT, and ALP (Table 9). All values are expressed as the mean ± SEM. There were no statistically significant differences at p < 0.05. All values are expressed as the mean ± SEM. There were no statistically significant differences at p < 0.05. All values are expressed as the mean ± SEM. There were no statistically significant differences between groups at p < 0.05. BUN, blood urea nitrogen; CREA, creatinine enzyme; AST, aspartate aminotransferase; ALT, alanine transaminase; ALP, alkaline phosphatase. Figure 8 shows the histological examination of the control and treatment groups, which showed normal structures of the liver and kidney in both groups. The liver sections did not show dilation of the central vein, cytoplasmic vacuolization, inflammatory cell infiltration, or congestion in the hepatic sinusoids. Kidney histological changes in extract-treated mice exhibited no obvious damage compared to the control group. Figure 8b,d show normal renal tubules and glomeruli in the glomerular basement membrane. All values are expressed as the mean ± SEM. There were no statistically significant differences between groups at p < 0.05. BUN, blood urea nitrogen; CREA, creatinine enzyme; AST, aspartate aminotransferase; ALT, alanine transaminase; ALP, alkaline phosphatase.

Discussion
Over the years, drug resistance has become one of the biggest problems in infectious diseases, including malaria. Partial resistance to artemisinin has emerged and spread, leading to delayed parasite clearance after treatment with ACT [30]. The development of antimalarial agents has been urgently needed; hence, phytomedicine became an interesting idea [31]. This study investigated the antiplasmodial activity of ten traditional recipes that provided scientific justification for malaria treatment. Solvent choices for extraction were selected according to the traditionally used phytomedicine in Thailand, which has otherwise been used as a polar and slightly nonpolar solvent for the study of biological activities. An aqueous solution is considered the best choice because of its low cost, nontoxicity, and health safety, and ethanol can be added to increase the solubilization of polar substances [32]. In addition, different solvent polarities affect the extract yields. The Benjalotiga and Chan-tang-ha recipes produced higher ethanolic yields than that of aqueous

Discussion
Over the years, drug resistance has become one of the biggest problems in infectious diseases, including malaria. Partial resistance to artemisinin has emerged and spread, leading to delayed parasite clearance after treatment with ACT [30]. The development of antimalarial agents has been urgently needed; hence, phytomedicine became an interesting idea [31]. This study investigated the antiplasmodial activity of ten traditional recipes that provided scientific justification for malaria treatment. Solvent choices for extraction were selected according to the traditionally used phytomedicine in Thailand, which has otherwise been used as a polar and slightly nonpolar solvent for the study of biological activities. An aqueous solution is considered the best choice because of its low cost, nontoxicity, and health safety, and ethanol can be added to increase the solubilization of polar substances [32]. In addition, different solvent polarities affect the extract yields. The Benjalotiga and Chan-tang-ha recipes produced higher ethanolic yields than that of aqueous solutions. This result is in agreement with previous studies, where they found that the compounds present in trees can be extracted using organic solvents [33]. This finding suggests that alcohol can improve the solubilization of nonpolar molecules from wood [34].
Antiplasmodial properties of herbal recipes revealed that high activity against P. falciparum was mostly present in ethanolic extracts with an IC 50 value ranging from 2.8-8.5 µg/mL. In comparison, the Gaysorn-tang-ha recipe showed the highest activity, followed by Trikatuk, Triphala, Benjalotiga, Benjagot, Chan-tang-ha, Jatu-phala-tiga, Benjakul, and Benjathian recipes. This suggests that the solubility of potentially active substances in semi-polar solutes may be greater than that in polar solvents, which is consistent with previous research. Potent antiplasmodial compounds are concentrated in the medium-polar solvent [35]. In addition, single ethanolic extracts from Dracaena loureiri Gagnep, Myristica fragrans Houtt, and Piper chaba Hunt were reported to possess antiplasmodial activity with IC 50 of 10.5, 8.9, and 5.3 µg/mL, respectively [19,36]. In contrast, only three aqueous extracts from Triphala, Trisamo, and Jatu-phala-tiga exhibited good activity with an IC 50 of 5.7 ± 0.2, 6.1 ± 0.7, and 5.0 ± 0.3 µg/mL, respectively, which is in agreement with previous evidence. The ingredients in water extracts of Triphala, Trisamo, and Jatu-phala-tiga recipes such as Terminalia bellerica (Gaertn) Roxb., Phyllanthus emblica L., and Terminalia chebula Retz were reported antiplasmodial activities with an IC 50 of 14.3, 14.4, and 15.4 µg/mL, respectively [37]. Therefore, our findings imply that the antiplasmodial activity of these recipes is caused by the combined effect or synergism of the ingredients.
Historically, plants have been considered to have pharmacological properties that respond to the presence of phytocompounds [38]. This study investigated eight phytoconstituents based on the main components present in medicinal plants [39]. Flavonoids, a group of natural substances with aromatic organic structures, are potential sources of antimalarial compounds [40]. Its mechanism of action is believed to be its interference with functional biomolecules, such as protein, enzymes, DNA, etc., under cellular oxidative stress and inhibition of fatty acid biosynthesis during the intraerythrocytic cycle [40]. Terpenoids play a key role in the eradication of malarial parasites. Artemisinin, a sesquiterpene lactone compound, is currently the most effective antimalarial drug derived from the medicinal plant [41]. They create radical ions that can damage various proteins, including sarco-endoplasmic reticulum Ca 2+ -ATPase, and inhibit PfATP6, leading to the breakage of mitochondrial and parasitic membranes [42]. Alkaloids are a broad class of biological compounds with antiplasmodial properties. Quinoline alkaloids are well-known compounds in malaria research, such as quinine and quinidine. For quinine, the structure has been modified to improve efficacy and reduce toxicity. Its mechanism is related to hemoglobin breakdown pathways. [43,44]. Tannins have a positive effect on antiplasmodial prophylaxis [45,46]. Saponins and coumarins have been reported as attractive compounds against malaria [47,48]. Therefore, plant secondary metabolites are strongly correlated with antiplasmodial activities, and it is suggested that the good activity of the traditional recipes in this study may be owing to the action of one individual or synergistic effects of phytocompounds. Traditional herbal medicines not only provide benefits but also generate potentially harmful effects or side effects from the plants [49]. This encouraged us to investigate the toxic effects of these medicinal plants.
An in vitro cell-based approach was used as the model for toxicity screening. The liver is the primary site for drug-induced toxicity, whereas the kidney is the primary organ involved in drug clearance for oral drug delivery [50]. Consequently, we investigated the cytotoxic effects of ten traditional recipes on both Hep-G2 and Vero cells and found that all extracts showed CC 50 against Hep-G2 at concentrations greater than 30 µg/mL, except for the ethanolic extracts from Benjakul, Benjalotiga, and Trikatuk, whereas Benjalotiga and the aqueous extract of Chan-tang-ha exhibited CC 50 against Vero cells at concentrations below 30 µg/mL. These results suggest that the extracts with a CC 50 below 30 µg/mL exhibited cytotoxicity. This finding is consistent with those of previous studies [51,52]. The ethanolic extract of Benjakul possesses anticarcinogenic activity, which is a response of at least three cytotoxic components of plumbagin, piperine, and 6-gingerol. Likewise, the toxic effects of Trikatuk may be caused by its components. For the Benjalotiga recipe, Santalol, the major constituent of Santalum album L., was reported to have antitumor properties on human hepatocellular carcinoma cell lines, and Silvestrol and episilvestrol were announced as potential anticancer agents against human oral epidermoid carcinoma [53,54]. Taccalonolides isolated from plants of the genus Tacca, such as Tacca chantrieri, are a new class of microtubule-stabilizing anticancer agents [55]. This finding suggests that the cytotoxicity of the recipes might be caused by toxic compounds that are deposited in plants. However, the toxic effects of these recipes on Vero cells exhibited CC 50 greater than those on HepG2 cells, except for Benjalotiga. It was implied that ethanolic extracts from Trikatuk and Benjakul displayed selective toxicity towards cancer cell lines owing to the contribution of antitumor or anticancer compounds.
Furthermore, drug-induced hemolysis is a serious toxicity liability, particularly in malaria. The hemolytic toxicity of the recipes was quantified to forecast the direct harmful effects. Aqueous extracts except for Benjathian and Trikatuk at a fixed dose of 50 µg/mL did not promote the breakdown of red blood cells, and only three aqueous extracts from Triphala, Trisamo, and Jatu-phala-tiga exhibited high SI values. These results suggest that aqueous extracts from Triphala, Trisamo, and Jatu-phala-tiga are good candidates for further evaluation in animal models. In contrast, some ethanolic extracts exhibited hemolytic effects. Trisamo and Benjakul recipes produced 100% hemolysis after incubation for 72 h. Benjagot, Gaysorn-tang-ha, Chan-tang-ha, and Benjathian exhibited percent hemolysis at 45.59 ± 6.81, 39.18 ± 9.62, 6.09 ± 9.14, and 5.79 ± 9.66, respectively. The extracts may be attributed to the presence of toxic substances that affect hematopoietic cells. In addition, this finding may imply that the antiplasmodial activity of ethanolic extracts of Gaysorn-tang-ha, Benjagot, Chan-tang-ha, Trisamo, Benjakul and Benjatian with IC 50 of 2.8, 6.8, 7.2, 7.7, 8.5 and 15.5 µg/mL, respectively could be due to hemolytic activity at 50 µg/mL.
Based on these findings, we propose that extracts not inducing hemolysis must be considered for use in medicine, whereas extracts that show hemolysis effects of more than 10% should be identified as toxic components or used with caution.
In previous findings, the antiparasitic effect of artesunate has been linked with DNA damage by increasing ROS production [56]. In our study, we observed oxidant generation of the extracts that had the ability to kill the parasite. As shown in Figure 3, our results illustrated that aqueous extracts of Trisamo and ethanolic extracts of Triphala, Trikatuk, and Trisamo induced significantly higher oxidant levels, suggesting that parasite death might be related to enhancing of oxidant levels and that increasing the oxidant can be targeted to oxidative damage to intracellular proteins, lipids and nucleic acids [57]. This finding is in accordance with previous research. Phenolic compounds in Terminalia species could potentially behave as either antioxidants or prooxidants, depending on their concentration, redox state, and the ratio between compounds [58]. The antioxidant effect is due to acting with a variety of free radicals, whereas prooxidant properties are related to the presence of transition metal ions such as copper or iron in the extract. The ability to reduce the metal ions may exert a redox cycling mechanism resulting in the formation of prooxidants [58].
In order to investigate the antimalarial activity and toxicity in an animal model, an aqueous extract of Triphala was selected for the in vivo evaluation. However, additional candidates such as Trisamo and Jatu-phala-tiga can be chosen as candidates in further studies because extracts with SI values of ≥10 can be assumed as potential samples for further investigation [59]. ICR mice were used because they are susceptible to infection by P. berghei ANKA [60]. The aqueous extract of Triphala reduced the parasite load up to 75.47% at a concentration of 600 mg/kg. Although the standard drug eliminated the infection, the extract at all doses significantly suppressed parasitemia (p < 0.05) compared to that of the infected control. This finding implies that the extract possesses good antimalarial activity against P. berghei ANKA. This activity may have been derived from the active compounds deposited in the extract. Gallic acid increases ROS production in macrophages, which may enhance phagocytic activity, and ellagic acid has antimalarial activity [61][62][63]. In addition, previous studies found gallic acid, ellagic acid, and chebulinic acid to be the major constituents of the Triphala recipe [11]. There was consistency with our results by LC-QTOF-MS analysis. Furthermore, Triphala has been used to maintain appropriate homeostasis in the body. Thus, the reduction in parasites might be owing to indirect effects. Triphala possesses antioxidant, free-radical scavenging, and immunomodulatory activities [63,64]. Therefore, the extract may activate the mechanism of cell-mediated immunity or humoral-mediated immunity because it is responsible for the stimulation of the immune system [63].
The harmful effects of the aqueous extract of Triphala showed that the extract did not cause acute toxicity in ICR mice. The indications for safety were normal behavior and absence of alteration in body weight, organ weight, liver-kidney function level, and histology compared with that of the control mice. Changes in body and organ weights are important parameters for assessing toxicity. Weight loss may indicate that a substance is causing damage to the body. Organ weight is a sensitive indicator of chemical or druginduced organ damage [65]. Thus, the results of this study indicate that the extract is safe for the liver and kidneys. In addition, the results of liver-kidney functions and histological examination showed no significant changes compared to the control group. Therefore, we conclude that the aqueous extract of Triphala is safe.

Conclusions
This study confirmed that all ethanolic and aqueous extracts from Triphala, Trisamo, and Jatu-phala-tiga could kill P. falciparum. However, high SI values in both HepG2 and Vero cells were present in aqueous extracts from Triphala, Trisamo, and Jatu-phala-tiga recipes. The aqueous extract of Triphala exhibited good antimalarial activity in a mouse model, and a single oral dose of 2 g/kg was safe in acute toxicity tests.

Institutional Review Board Statement:
The study protocol was reviewed and approved by the Human Ethics Committee of Walailak University prior to the recruitment of any participant (approval number: WUEC-22-153-01) and followed the tenets of the Declaration of Helsinki. Written informed consent was obtained from all participants before data and blood sample collection to establish models of blood-stage malaria infection and hemolysis, and the study protocol was reviewed and approved by the Animal Ethics Committee of Walailak University, National Research Council of Thailand (NRCT) (protocol number: WU-ACUC-65049).
Informed Consent Statement: Informed consent was obtained from all the subjects involved in the study.

Data Availability Statement:
The data associated with this study are included in this published article. Additional files are available from the corresponding authors upon request.