Extracts of Tectona grandis and Vernonia amygdalina have anti-Toxoplasma and pro-inflammatory properties in vitro

Tectona grandis (teak) and Vernonia amygdalina (bitter leaf) are plants used in traditional medicine in West Africa. In this study, we tested ethanolic and hydro-ethanolic extracts of bark and leaves of T. grandis and ethanolic extract of leaves of V. amygdalina for their inhibitory effect on Toxoplasma gondii, a protozoan parasite responsible for toxoplasmosis. Ethanolic extract of V. amygdalina leaves had proportional contents of phenols, tannins, flavonoids, and polysaccharides. This extract presented the highest efficacy against T. gondii, the lowest cytotoxicity to mammalian cells, but moderate anti-oxidant activity compared to other plant extracts. Ethanolic extract of T. grandis bark also had elevated anti-T. gondii activity, low cytotoxicity on mammalian cells, and one of the highest anti-oxidant activities. However, the phytochemical content of this extract was not very different from the hydro-ethanolic extract, which had no anti-T. gondii activity. In addition, ethanolic extract of V. amygdalina leaves, but not of T. grandis bark, significantly increased the production of TNF-α and NO by antigen-presenting cells. Both extracts had the tendency to decrease expression of major histocompatibility complex molecules at the surface of antigen-presenting cells, while they did not modulate the percentage of apoptotic cells. A study of signalling pathways would help to determine the mechanisms of action of these plant extracts.


Introduction
Toxoplasma gondii is a protozoan parasite belonging to the Phylum Apicomplexa like Plasmodium, the causal agent of malaria. It is estimated that one third of the world's population is infected with T. gondii. Toxoplasmosis is generally asymptomatic in immune-competent individuals but the persistence of the parasite in cysts located in the brain, muscles and eyes may lead to irreversible damage in case of immune-depression. More importantly, T. gondii is able to cross the placental barrier during pregnancy to infect the foetus. Congenital toxoplasmosis can cause malformations (calcification, hydrocephaly, etc.) or abortion. Toxoplasmosis is considered as a neglected disease worldwide and in the United States of America [32,12]. In this country, T. gondii is the second leading cause of deaths attributable to food-borne illness [5] and toxoplasmosis is one of the five neglected parasitic infections targeted for public health action. Current treatments for toxoplasmosis are the combination of 2,4-diaminopyrimidines (pyrimethamine, trimethoprim) and sulfamides (sulfadiazine, sulfamethoxazole) [20]. Alternative treatments (clindamycin, atovaquone) may be indicated in cases of intolerance to first-line treatments. Since pyrimethamine is potentially teratogenic, spiramycin is administered during the first trimester of pregnancy. Furthermore, 2,4-diaminopyrimidines induce reversible depression of the bone marrow prevented by folinic acid supplementation given concomitantly with the treatment. Due to the lack of specificity and to the limited efficacy of these current treatments, there is a need for new compounds to cure toxoplasmosis with less toxicity and short treatment courses. Several research groups are working on the development of new synthetic compounds against T. gondii (for a review of the last ten years, see reference [19]). Only the bumped kinase inhibitor (BKI)-1294 has been tested in a mouse model of congenital toxoplasmosis with more than 90% reduction in congenital infection [21]. Unfortunately, BKI-1294 seems to have a detrimental effect on fertility. Most of the new drugs approved to treat infectious diseases originate from natural products. For this reason, anti-Toxoplasma activities are screened for in extracts of plants used in traditional medicine in West Africa, and have shown highly variable efficacy in vitro (for review, see reference [29]). In the present work, we studied the anti-Toxoplasma effect of ethanolic and hydro-ethanolic extracts of bark and leaves of Tectona grandis (Lamiaceae), the well-known teakwood used as a material for furniture, and of ethanolic extract of leaves of Vernonia amygdalina (Asteraceae) or bitter leaf, a plant known for its production of beneficial secondary metabolites. This nontoxic plant, consumed as a vegetable, is used in traditional folk medicine to treat or control more than twenty diseases including malaria [31], while T. grandis is mainly used in skin diseases as an antiseptic, woundhealing and anti-acne agent. Nevertheless, methanolic extract of T. grandis bark had leishmanicidal activity on L. major promastigotes [30] and methanolic extract of T. grandis leaves was highly effective against multidrugresistant uropathogenic bacteria in vitro [17]. In the present work, the anti-oxidant, anti-haemolytic, and proinflammatory properties of plant extracts were also studied.

Materials and methods
Compliance with the Nagoya Protocol Togo signed the Nagoya Protocol in 2011 and is party since 2016. We received authorisation from the Institut Togolais de Recherche Agronomique (ITRA) to collect samples and to use plant extracts at the University of Tours (France) for research experiments.

Phytochemical screening
The materials were washed in water to remove dirt and unwanted particles. Washed materials were dried at room temperature for several days. Then ethanolic and hydroethanolic extractions were performed by maceration of plant samples in the solvent (ethanol or mixture of ethanol and distilled water v/v) and with periodic agitation for 72 hours. Afterwards, solvent was evaporated under reduced pressure with a rotary evaporator and the extracts were lyophilised. Yield of dried extract was 7.4% and 7.5% for ethanolic and hydro-ethanolic extracts of T. grandis leaves, 11.1% and 12.4% for ethanolic and hydro-ethanolic extracts of T. grandis bark, and 9.6% for ethanolic extract of V. amygdalina leaves (percentage weight of the starting dried plant material).
Analysis of phytochemical contents was done on lyophilised extracts of leaves and bark of T. grandis and on lyophilised extract of leaves of V. amygdalina suspended in methanol (at 1 mg/mL). Total phenols were quantified by spectrophotometry using the Folin-Ciocalteu method with gallic acid as standard. Total flavonoids were quantified by spectrophotometry using the aluminium chloride colorimetric method with rutin as the standard curve. Polysaccharides were quantified according to Dubois et al. [11] with glucose as the standard curve.

Anti-oxidant activity
The free radical scavenging activity of the plant extracts was measured by the DPPH (2,2-diphenyl-1picrylhydrazyl) and the AAPH (2,2'-azobis 2 amidinopropanedihydrochloride) radical scavenging activity assays. For the first method, plant extracts were prepared in methanol. One volume of 0.1 mM DPPH was mixed with one volume of the plant extracts at different concentrations, ascorbic acid as positive control, or methanol (blank). Absorbance was read at 480 nm after 30 min incubation at room temperature in the dark. The percentage of DPPH free radicals scavenging was calculated as ([Optical Density blank/O.D. plant extract]/O.D. blank) x100. In the second method, the addition of the peroxyl radical initiator AAPH to erythrocytes induces their haemolysis. One volume of sheep blood cell suspension (from the Institut National de la Recherche Agronomique Centre Val-de-Loire) was mixed with one volume of PBS, one volume of 200 mM AAPH, and one volume of the plant extracts at different concentrations or PBS (complete lysis), or ascorbic acid (no lysis). The mixture was incubated at 37°C for 3 hours, washed with PBS, and centrifuged at 1000 g for 10 min.

In vitro inhibition of T. gondii
Plant efficacy on T. gondii was tested in a colometric assay, as previously described [18]. For this, human foreskin fibroblasts (HFF, ATCC number CRL-1634) were seeded in 96-well plates at 2 Â 10 4 cells per well in Dulbecco's Modified Eagle Medium (DMEM without phenol red, Pan Biotech GmbH) with 1% fetal calf serum and 2 mM glutamine at 37°C in 5% CO 2 atmosphere. Ethanolic extracts of V. amygdalina leaves (Va-LE), hydro-ethanolic and ethanolic extracts of T. grandis leaves (Tg-LH and Tg-LE), hydro-ethanolic and ethanolic extracts of T. grandis bark (Tg-BH and Tg-BE) were lyophilised, solubilised in DMSO (stock solution at 250 mg/mL), and diluted in DMEM without phenol red. After 24 h, T. gondii parasites (100 tachyzoites of the T. gondii RH-b-gal strain carrying the E. coli lacZ gene coding for b-galactosidase under the control of the T. gondii sag1 promoter) and plant extracts (1 À 200 mg/mL in quadruplicate) were added to host cells at the same time. In negative control wells, tachyzoites were incubated with same amounts of DMSO alone with final maximum amount of 0.02%. Plant extracts were maintained throughout the parasite invasion and proliferation steps. After 96 h incubation at 37°C in 5% CO 2 atmosphere, release of the b-galactosidase was achieved with lysis solution (Triton X-100 at 0.1%, Sigma) and chlorophenol red-b-D-galactopyranoside (Sigma) was added at 1 mM in 100 mM HEPES pH 8. Optical density of the red colouration was measured at 565 nm. The IC 50 (50% inhibitory concentration) was calculated from the b-galactosidase activity in the presence of plant extracts compared to non-treated parasites.

Cytotoxicity assay
HFFs (2 Â 10 4 cells) were cultivated in 96-well plates in 150 mL DMEM with 1% FCS, 2 mM glutamine and the plant extracts at different concentrations, in triplicate. Plates were incubated at 37°C in a 5% CO 2 atmosphere for 96 h. UptiBlue TM (15 mL, Interchim) was then added for 4 h incubation in the dark at 37°C in a 5% CO 2 atmosphere. O.D. was measured at 565 nm and 630 nm to calculate the reduction of the oxidation-reduction indicator. CC 50 (50% cytotoxic concentration) was calculated from the reduction in the presence of plant extracts compared to non-treated cells.
Cell stimulation and tumour necrosis factor (TNF)a/nitric oxide (NO) quantification Cells were seeded at 2 Â 10 5 in 96-well plates in 100 mL DMEM for the RAW 264.7 macrophage cell line (ATCC ® TIB-71 TM ), at 10 6 in 24-well plates in 300 mL DMEM for peritoneal exudate cells (PECs) harvested from Swiss OF1 mice by washing with 5 mL DMEM (directive 2010/63/ EU not applied [chapter 1, article 1.5.]), and at 10 6 in 24well plates in 300 mL RPMI for the SRDC dendritic cell line [28]. The cells were stimulated at 37°C in 5% CO 2 atmosphere for 24 h with plant extracts at different concentrations in 100 or 300 mL medium. DMSO at the lowest dilution was used as a negative control. Levels of tumour necrosis factor (TNF)-a were quantified in the cell culture supernatants of RAW 264.7 and SRDC by using specific sandwich enzyme-linked immunosorbent assay (ELISA) from Affymetrix eBioscience. Levels of nitric oxide (NO) were measured with the Griess assay. For this, two volumes of culture supernatant were mixed with one volume of 0.1% naphthyl ethylenediamine (in water) and one volume of 1% sulfanilamide (in 5% phosphoric acid). A solution of NaNO 2 was used as standard. Absorbance was read at 540 nm.

Statistical analysis
The non-parametric Kruskal-Wallis test followed by the Dunn's multiple comparison test was used for statistical evaluation, and a p value < 0.05 was considered significant (GraphPad Prism 5 software).

Results and discussion
Phytochemical screening of the plant extracts

Anti-Toxoplasma efficacy and cytotoxicity induced by the plant extracts
The plant extracts were tested for their anti-Toxoplasma efficacy in vitro. The concentration inducing 50% reduction in parasite number (IC 50 ) was highly variable from one experiment to another, and the standard error of the mean (S.E.M.) was high (Tab. 2). In our conditions, the ethanolic extract of the leaves of V. amygdalina (Va-LE) was the most effective against T. gondii with an IC 50 of 5.6 ± 0.3 mg/mL. Furthermore, this extract was the lowest cytotoxic on the host cells (HFF) with a CC 50 of 303.6 ± 119.1 mg/mL (selectivity index [CC 50 /IC 50 ] of 50). Aqueous extracts of a plant of the same genus, V. colorata, also showed growth inhibition of T. gondii with IC 50 values of 17-18 mg/mL and with a toxic effect on MRC5 cells at concentrations > 250 mg/mL [6]. When extracts of V. colorata were prepared in organic solvents (ethanol, acetone, dichloromethane), a 5-to 10-fold increase in activity was observed. We have not compared hydroethanolic and ethanolic extracts of V. amygdalina leaves but the results presented by Omoregie et al. [25] have shown that an ethanolic extract of the leaves of this plant had higher anti-Plasmodium activity than hydro-ethanolic and aqueous extracts, with a selectivity index of 6. In a more recent study, treatment of mice with aqueous but not ethanolic extract of V. amygdalina leaves reduced the P. berghei macrogametocyte density by about 50% [1]. Values Juglone (5-hydroxy-1,4-naphthalenedione) has been identified as a compound with antibacterial activity against Listeria monocytogenes and Staphylococcus aureus present in methanol extract of T. grandis bark [23]. A study investigated the antimicrobial properties of isolated constituents of the leaves of T. grandis [22]. Rutin (flavonoid) and ellagic acid (phenol) showed significant activity against Bacillus subtilis, Escherichia coli, Klebsiella pneumonia, and Staphylococcus aureus. Gallic acid (phenol) and sitosterol exhibited activity against these bacteria except S. aureus and E. coli, respectively. Quercitin (flavonoid) showed activity only against K. pneumoniae and S. aureus. The levels of phenols and flavonoids in the ethanolic extract of the bark of T. grandis are similar to those of hydro-ethanolic extract of bark and leaves and may not explain its properties.

Anti-oxidant properties of the plant extracts
The anti-oxidant activities of the plant extracts were evaluated by the widely used method of electron transfer from anti-oxidant to DPPH (Fig. 1A). Ascorbic acid, known for its anti-oxidant effect, induced 75% inhibition of oxidation at all concentrations tested (31.5 to 250 mg/mL).
All plant extracts presented a dose-dependent anti-oxidant effect but with lower efficacy than ascorbic acid, even though the differences are not statistically significant. At the lowest concentrations (31.5 and 62.5 mg/mL), ethanolic extracts of the leaves (Tg-LE) and the bark (Tg-BE) of T. grandis were the most effective. At higher concentrations (125 and 250 mg/mL), hydro-ethanolic and ethanolic extracts of T. grandis bark (Tg-BH and Tg-BE) had the highest anti-oxidant effects. The ethanolic extract of V. amygdalina (Va-LE) and the hydro-ethanolic extract of T. grandis leaves (Tg-LH) were the least effective at any concentration tested. A second method was used to study the anti-oxidant properties of the extracts. The addition of the peroxyl radical initiator AAPH to erythrocytes causes oxidation of lipids and proteins in the cell membrane and induces haemolysis, inhibited by anti-oxidant molecules (Fig. 1B). All preparations, even ascorbic acid, should be added at high concentrations to inhibit lysis of Values are means ± S.E.M. of three independent experiments. Tg-BE: T. grandis À bark ethanolic extract; Tg-BH: T. grandis À bark hydro-ethanolic extract; Tg-LE: T. grandis À leaf ethanolic extract; Tg-LH: T. grandis À leaf hydro-ethanolic extract; Va-LE: V. amygdalina À leaf ethanolic extract. CC 50 , 50% cytotoxic concentration on HFF; IC 50 , 50% inhibitory concentration on T. gondii growth; SI, selectivity index. erythrocytes. In these conditions, ethanolic and hydroethanolic extracts of bark of T. grandis (Tg-BE and Tg-BH) were more efficient than ascorbic acid at 62.5 and 125 mg/ mL. At the concentration of 250 mg/mL, only ethanolic extract of T. grandis bark (Tg-BE) was as effective as ascorbic acid, while ethanolic and hydro-ethanolic extracts of leaves of T. grandis (Tg-LE and Tg-LH) presented the lowest activity. At 500 mg/mL, all extracts except Tg-LH almost reached the percentage of oxidation inhibition of ascorbic acid. These results indicate that anti-oxidant activity demonstrated with an artificial method such as DPPH assay might be slightly different in an assay more similar to natural conditions like in the presence of erythrocytes. Our results on the anti-oxidant activity of the ethanolic extract of T. grandis leaves are consistent with the study of Diallo et al. [8]. The authors induced a model mimicking haemolytic anaemia due to Plasmodium falciparum by injecting phenyl hydrazine at 40 mg/kg for 2 days to wistar rats. Ethanolic extract of T. grandis leaves administered at 1 g/kg by the oral route for 5 days inhibited anaemia by increasing osmotic resistance of erythrocytes and concentration of haemoglobin [8]. It has been shown that the anti-oxidant activity of the constituents of T. grandis leaves depends on the number and the position of hydroxyl groups with the highest activity for the flavonoids quercitin and rutin, followed by the phenols gallic and ellagic acids [22]. Polysaccharides of T. grandis bark have also demonstrated anti-oxidant activity with IC 50 values from 0.1 to 2.4 mg/mL for scavenging of free radical, hydroxyl radical, and NO [27].

Modulation of antigen-presenting cells by plant extracts
We have observed that ethanolic extract of T. grandis bark and V. amygdalina leaves had the highest anti-Toxoplasma effect. An increase in TFN-a levels was measured in the serum of P. berghei-infected mice treated with ethanol extract of V. amygdalina leaves [26]. We thus aimed to determine whether the plant extracts could modulate responses of antigen-presenting cells like macrophages and dendritic cells. To answer this question, mouse macrophages of the RAW 264.7 cell line and mouse dendritic cells of the SRDC cell line were stimulated with 10 mg/mL of each plant extract. V. amygdalina but not T. grandis extract significantly increased the production of TNF-a by the two types of cells ( Fig. 2A), and the production of NO by SRDC (Fig. 2B). In addition to cytokine production, an increase in the surface expression of MHC molecules of classes I and II is a marker of maturation of antigen-presenting cells. In contrast to what we expected, expression of both MHC molecules of classes I and II decreased on SRDC in response to Tg-BE and Va-LE. However, the differences with the cells stimulated with the solvent (DMSO) are not statistically significant (Fig. 2C). Polyphenols could be responsible for this decrease. Indeed, it has been shown that apple polyphenol extract decreased HLA-DR (MHC class II) expression on THP-1-derived human dendritic cells in vitro [14]. The authors suggested regulation through the ubiquitinproteasome pathway. We have previously shown that glycosylphosphatidylinositols purified from T. gondii tachyzoites were able to increase MHC expression and antigen presentation on non-infected cells [7]. We suggested that their recognition and destruction as infected cells could be a strategy developed by the parasite to escape the immune system. T. grandis and V. amygdalina extracts may have immunomodulatory agents able to counteract the parasite effects on antigenpresenting cells. Finally, we studied programmed cell death in response to the plant extracts. A recent review identifies the natural compounds able to target endoplasmic reticulum (ER) stress, modulated by inflammatory factors and resulting in apoptosis or autophagy [15]. These compounds include polyphenols, alkaloids, and saponins and some of them could be abundant in the plant extracts we tested here. Our results show that incubation of primary antigen presenting cells (PECs) with ethanolic extract of T. grandis bark or V. amygdalina leaves did not modulate the basal percentage of apoptotic cells (Fig. 2D). As primary cells, PECs are more susceptible to death than cell lines. Therefore, this last result comforts the low cytotoxic effect of these extracts observed on HFF.

Concluding remarks
Plant extracts may act at different levels to produce anti-parasitic activities. Aside from direct inhibition of T. gondii proliferation, T. grandis and V. amygdalina extracts induced the production of TNF-a and NO, two proinflammatory molecules known to play an important role in resistance to T. gondii in vivo. The anti-oxidant activity of plant extracts is a parameter systematically studied. Our results indicated that the V. amygdalina extract with the highest parasite inhibitory effect had one of the lowest antioxidant activities. In the literature, anti-T. gondii activity was not related to antioxidant activity. For example, treatment with vanillin of T. gondii-infected mice induced a higher survival rate than treatment with resvan, a vanillin derivative with much higher anti-oxidant activity (19%, 85%, and 94% for vanillin, resvan, and ascorbic acid, respectively) [24]. Disappointingly, the direct effect of vanillin or resvan against T. gondii has not been investigated in vitro. However, the anti-oxidant properties could be important in the overall treatment of toxoplasmosis. Indeed, significant decreases in the total anti-oxidant ability were recorded 7 days after T. gondii infection in rats [4]. In T. gondii seropositive patients without symptoms, decreases of glutathione activity and increases of malondialdehyde related to lipid peroxidation reflected similar declines in the response to oxidative stress [13]. Dincel et al. have also shown that oxidative stress due to reactive oxygen/nitrogen species and NO production at pathological levels play a role in worsening neurodegeneration in the process of Toxoplasma encephalitis [9]. An anti-oxidant organoselenium compound reversed the alterations (splenomegaly, increase in reactive oxygen species levels and adenosine deaminase activity) found in the spleen of T. gondii-infected mice [10]. On the contrary, diet supplementation with vitamin E and selenium was associated with higher cyst numbers in the brains of infected mice [16]. Thus, strong antioxidants can inhibit the formation of free radicals involved in parasite resistance, leading to lower therapeutic efficacy. Markers of oxidative stress must be followed in case of infection with T. gondii and treatment with different plant extracts could be considered to obtain a balance between pro-and anti-oxidant factors in order to fight the parasite without deleterious side effects.