Anticancer potential of Thevetia peruviana fruit methanolic extract

Abstract

Background: Thevetia peruviana (Pers.) K. Schum or Cascabela peruviana (L.) Lippold (commonly known as ayoyote, codo de fraile, lucky nut, or yellow oleander), native to Mexico and Central America, is a medicinal plant used traditionally to cure diseases like ulcers, scabies, hemorrhoids and dissolve tumors. The purpose of this study was to evaluate the cytotoxic, antiproliferative and apoptotic activity of methanolic extract of T. peruviana fruits on human cancer cell lines.

Methods: The cytotoxic activity of T. peruviana methanolic extract was carried out on human breast, colorectal, prostate and lung cancer cell lines and non-tumorigenic control cells (fibroblast and Vero), using the MTT assay. For proliferation and motility, clonogenic and wound-healing assays were performed. Morphological alterations were monitored by trypan blue exclusion, as well as DNA fragmentation and AO/EB double staining was performed to evaluate apoptosis. The extract was separated using flash chromatography, and the resulting fractions were evaluated on colorectal cancer cells for their cytotoxic activity. The active fractions were further analyzed through mass spectrometry.

Results: The T. peruviana methanolic extract exhibited cytotoxic activity on four human cancer cell lines: prostate, breast, colorectal and lung, with values of IC₅₀ 1.91 ± 0.76, 5.78 ± 2.12, 6.30 ± 4.45 and 12.04 ± 3.43 μg/mL, respectively. The extract caused a significant reduction of cell motility and colony formation on all evaluated cancer cell lines. In addition, morphological examination displayed cell size reduction, membrane blebbing and detachment of cells, compared to non-treated cancer cell lines. The T. peruviana extract induced apoptotic cell death, which was confirmed by DNA fragmentation and AO/EB double staining. Fractions 4 and 5 showed the most effective cytotoxic activity and their MS analysis revealed the presence of the secondary metabolites: thevetiaflavone and cardiac glycosides.

Conclusion: T. peruviana extract has potential as natural anti-cancer product with critical effects in the proliferation, motility, and adhesion of human breast and colorectal cancer cells, and apoptosis induction in human prostate and lung cancer cell lines, with minimal effects on non-tumorigenic cell lines.

Keywords: Anti-proliferative activity; Apoptosis; Cardiac glycosides; Cytotoxic activity; Flavonoid; Human cancer cells; Motility.

Fig. 1

Clonogenic assay of four human cancer cell lines treated with methanol T. peruviana extract. a Image showing colonies produced by human cancer lines following plating of 100 cells and 10–14 days incubation. Cells were treated with IC₅₀ value corresponding for each cell line; b cells were quantified and error bar indicates mean ± SEM of three independent experiments. The level of significance was determined using Student’s t-Test with ^nsrepresenting ρ > 0.05, **** indicates ρ < 0.0001, ** indicates ρ < 0.01 and * indicates ρ < 0.05

Fig. 2

Morphological changes on four human cancer cell lines during treatment with T. peruviana methanol extract. Human cancer cells were treated at IC₅₀ corresponding value for each cell line and monitored over a period of 24 h. Black arrows indicated blebbing membranes

Fig. 3

Viability of four human cancer cells exposed to T. peruviana methanol extract over a period of 24 h. Human cancer cells were treated at IC₅₀ value corresponding for each cell line. Cell viability was evaluated using the trypan blue exclusion method. Each data point represents values from three independent experiments (n = 3). Error bar indicates mean ± SEM. Different letters represent statistically significant differences determined by one way ANOVA (ρ < 0.05)

Fig. 4

Classification of normal and cancer cell lines exposed to T. peruviana fruit methanolic extract according to independent component analysis (ICA). The distribution of the cell lines (panel A, projections with 95% confidence ellipses) and variables (panel B, projection of variable loadings, with maximum loading indicated by a circle) is shown in the space spanned by the two leading independent components. The clear unsupervised discrimination among the six cell lines reflects the greater effect of T. peruviana extract on tumor cell lines (lung cells, L, ●, prostate cells, P, ■, breast cells, B, ♦, colorectal cells, C, ▲), while normal cells are less affected (fibroblast cells, Fb, □, Vero cells, V, ○). The independent component 1 is clearly separating cancer cells from normal cells, mainly due to the effect of the extract on the motility (WH) and membrane permeability (MP), while the independent component 2 is separating samples mainly by the IC₅₀ value observed for each cell line

Fig. 5

DNA fragmentation of human cancer cell lines treated with T. peruviana methanol extract. Lung, prostate, breast and colorectal cancer cell lines were treated with the corresponding IC₅₀ value for each cell line during a 24 h period; m, molecular-weight marker; C, untreated control; E, cancer cell line treated with T. peruviana methanol extract at its corresponding IC₅₀ value and, D, cancer cell line treated with doxorubicin (10 μg/mL). Representative result from three independent experiments is shown

Fig. 6

AO/EB double stain of human cancer cell lines after a treatment with T. peruviana methanol extract. a. Prostate and lung cells were treated with IC₅₀ of T. peruviana methanol extract. Images represent the control (untreated cells), treated cells and, cell treated with doxorubicin (10 μg/mL) as positive control. Cells were stained with acridine orange and ethidium bromide (AO/EB) after 4 h of treatment. White arrows indicate live (L), early apoptotic (EA), late apoptotic (LA) or necrotic (N) cells. b. Error bar indicates mean ± SEM of three independent experiments. +Ext and +Dox, both indicate cells treated with extract or doxorubicin. Different letters represent statistically significant differences determined by one way ANOVA (ρ < 0.05) between bars with same color by cell line. Non-letter bars have no statistical difference