Mauritian Endemic Medicinal Plant Extracts Induce G2/M Phase Cell Cycle Arrest and Growth Inhibition of Oesophageal Squamous Cell Carcinoma in Vitro.

Terrestrial plants have contributed massively to the development of modern oncologic drugs. Despite the wide acceptance of Mauritian endemic flowering plants in traditional medicine, scientific evidence of their chemotherapeutic potential is lacking. This study aimed to evaluate the in vitro tumor cytotoxicity of leaf extracts from five Mauritian endemic medicinal plants, namely Acalypha integrifolia Willd (Euphorbiaceae), Labourdonnaisia glauca Bojer (Sapotaceae), Dombeya acutangula Cav. subsp. rosea Friedmann (Malvaceae), Gaertnera psychotrioides (DC.) Baker (Rubiaceae), and Eugenia tinifolia Lam (Myrtaceae). The cytotoxicities of the extracts were determined against six human cancer cell lines, including cervical adenocarcinoma, colorectal carcinoma, oesophageal adenocarcinoma, and oesophageal squamous cell carcinoma. The potent extracts were further investigated using cell cycle analysis and reverse phase protein array (RPPA) analysis. The antioxidant properties and polyphenolic profile of the potent extracts were also evaluated. Gas chromatography mass spectrometry (GC-MS) analyses revealed the presence of (+)-catechin and gallocatechin in E. tinifolia and L. glauca, while gallic acid was detected in A. integrifolia. L. glauca, A. integrifolia, and E. tinifolia were highly selective towards human oesophageal squamous cell carcinoma (KYSE-30) cells. L. glauca and E. tinifolia arrested KYSE- 30 cells in the G2/M phase, in a concentration-dependent manner. RPPA analysis indicated that the extracts may partly exert their tumor growth-inhibitory activity by upregulating the intracellular level of 5'AMP-activated kinase (AMPK). The findings highlight the potent antiproliferative activity of three Mauritian endemic leaf extracts against oesophageal squamous cell carcinoma and calls for further investigation into their chemotherapeutic application.


INTRODUCTION
Oesophageal cancer (OC) is of increasing global concern due to poor prognosis, aggressive disease course, and a lack of efficient selective therapeutics. Oesophageal squamous cell carcinoma and oesophageal adenocarcinoma represent the sixth leading cause of cancer death worldwide and are associated with a 5-year survival of less than 15% and a median overall survival of less than a year [1]. Current treatment options rely primarily on broad-spectrum cytotoxic chemotherapeutics such as cisplatin, fluorouracil, and paclitaxel, which are associated with toxic side effects and limited efficacy, in addition to resistance [2]. Therefore, there is a dire need for searching for novel agents targeted against oesophageal cancer cells.
Plant-sourced bioactive molecules have provided architectural scaffolds for numerous lifesaving clinical agents, including 27% of the approved natural anticancer drugs, since 1980 [3]. Furthermore, over 3,000 global plant taxa have documented ethno-medicinal uses in the treatment of cancer [4].
Mauritius is a biodiversity hotspot located off the southeast coast of the African continent in the Indian Ocean. To date, the ethno-medicinal uses of about 32% of Mauritian endemic plants are documented with limited insight into their anticancer potential [5]. Therefore, this untapped unique resource represents a fertile ground for the bioprospecting of novel oncologic agents.
Thus, the present work evaluated the in vitro antioxidant and tumor cytotoxicity of A. integrifolia, D. acutangula, E. tinifolia, G. psychotrioides, and L. glauca Bojer (Fig. 1), in relation to their polyphenolic content. The effect of the three most potent extracts on cell cycle progression, mode of induced cell death, and their ability to modulate AMPK in oesophageal adenocarcinoma cells was also studied.

Chromatographic determination of phenolic compounds
The GC-MS analysis of trimethylsilylimidazole derivatized extracts of L. glauca, E. tinifolia and A. integrifolia was carried out using an Agilent 7890A gas chromatography system (Agilent Technologies, USA) as described in [7]. The analysis began with the initial oven temperature set at 150°C, which increased at the rate of 10°C/min to 300°C, and was maintained for another 4-5 min to yield a total run of 20 min under constant helium pressure (10 psi). The chromatogram was analyzed by matching the MS spectra of the peaks with those stored in the NIST 2011 Mass Spectral Library.

In vitro antioxidant capacities of extracts
The in vitro antioxidant activities of the extracts were evaluated using ferric reducing antioxidant power (FRAP) assay; iron chelating assay; superoxide anion radical scavenging assay; and nitric oxide radical inhibition assay, as described in [6]. The modified method proposed by Chu et al. [8] was employed for the DPPH (2-2 diphenyl-1-picrylhydrazyl) radical scavenging activity. Briefly, 100 µl of a methanolic extract (of varying concentration) and 200 µl of 100 µM DPPH dissolved in methanol were incubated for 30 min at room temperature and the absorbance was read at 492 nm. The negative and positive controls contained extract vehicle and gallic acid (or otherwise stated), respectively, instead of the extract. The percentage of metal chelating and free radical scavenging activity of the extracts were calculated with reference to the negative control according to equation 1. A concentration response curve was generated, and the IC 50 value was determined using the GraphPad Prism 6 software (GraphPad Inc., USA). All the experiments were performed in triplicates in three independent assays. MitoTracker Deep Red FM (500 nM, Invitrogen)). The plates were incubated in the dark (30 min), before washing three times with PBS. The staining solution was replaced with normal media, and the plates were imaged on an ImageXpress system (Molecular Devices, UK), taking four images per well. The percentage of cells in each cell cycle phase was determined using the cell cycle application module within the MetaXpress software (Molecular Devices, UK)

Reverse phase protein array (RPPA)
Samples were analyzed by Zeptosens RPPA as described previously [9]. Briefly, tumor lysates were normalized to 2 mg/ml with CLB1 lysis buffer and diluted 1:10 in CSBL1 spotting buffer (Zeptosens-Bayer) prior to preparing a final 4-fold concentration series of 0.2, 0.15, 0.1, and 0.75 mg/ml. The diluted concentration series of each sample was printed onto Zeptosens protein microarray chips (ZeptoChip TM , Zeptosens-Bayer) under environmentally controlled conditions (constant 50% humidity and 14°C) using a non-contact printer (Nanoplotter 2.1e, Gesim). A single 400 picoliter droplet of each lysate concentration was deposited onto the Zeptosens chip (thus representing four spots per single biological replicate). A reference grid of AlexaFluor647 conjugate BSA consisting of four column X 22 rows was spotted onto each subarray; each sample concentration series was spotted in between the reference columns. After array printing, the arrays were blocked with an aerosol of a BSA solution using a custom-designed nebulizer device (ZeptoFOG TM , Zeptosen-Bayer) for 1 h. The protein array chips were subsequently washed in double-distilled water and dried prior to performing the dual antibody immunoassay comprising 24-hour incubation of primary antibody, followed by 2.5-hour incubation with the secondary Alexa-Fluor conjugated antibody detection reagent (anti-rabbit A647 Fab). A selected panel of antibodies, pre-validated for RPPA application, was used. Following secondary antibody incubation and final wash step in the BSA solution, the immune stained arrays were imaged using the ZeptoREADER TM instrument (Zeptosens-Bayer, Germany). For each subarray, five separate images were acquired using different exposure times ranging from 0.5-10 s. Microarray images representing the longest exposure without saturation of fluorescent signal detection were automatically selected for analysis using the ZeptoView TM 3.1 software (Zeptosens-Bayer, Germany). An error weighted least squares linear fit through the fourfold concentration series was used to calculate the median relative fluorescence intensity (RFI) value for each sample replicate. Local normalization of the sample signal to the reference BSA grid was used to compensate for any intra-inter-array/chip variation. Local normalized RFI values were used for the subsequent analysis, and the data are presented as a fold change over DMSO control samples.

Statistical analysis
Statistical analysis was carried out using the GraphPad Prism 6 software. The mean values were compared using One-Way ANOVA. Student's t-test and/or Tukey's multiple comparisons as the Post Hoc test was used to determine the significances of the mean cytotoxic activities of phytochemicals and antioxidants among different extracts and the positive control. All charts were generated using the Microsoft Excel software (version 2010).

Phytochemical analysis
The TPC varied significantly between the investigated leaf extracts (p <0.05), with the highest level measured in L. glauca (Fig. 2). The TPC in terms of gallic acid equivalent ranged from 298.9 ± 9.4 to 110.4 ±13.7 mg GAE/gLW. G. psychotrioides had the most abundant TPrC in terms of the cyanidin chloride equivalent, while a negligible amount was measured in the A. integrifolia leaf extract. The total flavonoid levels ranged between 7.6 mg ± 0.3 and 0.4 ± 0.1 mg QE/gLW, with D. acutangula having the most prominent flavonoid level followed by L. glauca. GC-MS analysis revealed the presence of (+)-catechin and gallocatechin in E. tinifolia and L. glauca, while gallic acid was detected only in A. integrifolia (Fig. 3).

Antioxidant activities of extracts in in vitro models
The trend in antioxidant potential between the extracts differed in each of the investigated assays ( Table  1). The FRAP value of L. glauca was significantly higher than that of other endemic plant extracts (p< 0.05), and 2.4-fold lower compared to the positive control gallic acid (24.9 mmol Fe 2+ ). All the extracts under  Table 2.

Cell cycle analysis
The cell cycle stage of cells determined using the Cell Cycle Application Module within the MetaXpress software (Molecular Devices, UK) revealed that the extracts affected oesophageal squamous cell carcinoma (KYSE-30) selectively, showing no changes in the oesophageal adenocarcinoma (FLO-1) cell line. Both the E. tinifolia and L. glauca extracts exhibited a concentration-dependent effect on the cell cycle (Fig. 4), causing significant G2/M arrest of KYSE-30 cells, down to 3 µg/ml (p< 0.05). Even though the A. integrifolia extract also caused cell cycle arrest at G2/M, the effect was less concentration-dependent. None of the extracts caused any changes in the cell cycle of FLO-1 cells (Fig. 4).
Interestingly, a similar selectivity pattern was observed in the cell death assays. The number of nuclei per image was determined using the MetaXpress software. Extracts of E. tinifolia and L. glauca induced cell death in KYSE-30 cells in a concentration-dependent manner. However, both extracts had no effect on FLO-1 cells. The calculated IC 50 for E. tinifolia against KYSE-30 was 1.37 µg/ml; IC 50 for L. glauca was 1.77 µg/ml. E. tinifolia at a concentration of 1 µg/ml significantly decreased the KYSE-30 cell number relative to the negative control (p < 0.05), while for L. glauca a significant reduction in the cell number was observed at 3 µg/ml (p < 0.05) (Fig. 5).
The selective effect on KYSE-30 was based on the effects of the extracts on cell morphology (Fig. 6). The three extracts had a marked effect on the cell number, cell and nuclear morphology, being indicative of cytotoxic and cytokinesis defects. However, there were   Values are expressed in mmol Fe 2+ ; 2 IC 50 values are expressed in µg/ml; 3 IC 50 values are expressed in µg/ml; Data represent a mean ± standard deviation (n = 3). Different letters between rows in each column represent significant differences between the extracts (p < 0.05). Asterisks represent significant differences between the extracts and gallic acid (positive control), * p ≤ 0.05, ** p ≤ 0.01, *** p ≤ 0.001, **** p ≤ 0.0001. study showed dose-dependent metal chelating and free radical scavenging activity; the IC 50 values are shown in Table 1. L. glauca had the most potent radical scavenging activity against DPPH and NO radicals. While G. psychotriodes had the highest O 2 .radical scavenging activity, the metal ion chelating activities of A. integrifolia and E. tinifolia were notable compared to those of other investigated extracts.

Cytotoxicity of herbal extracts
The A. integrifolia and E. tinifolia extracts exhibited dose-dependent growth inhibition of the tested cancer cell lines. Both extracts were strongly cytotoxic towards KYSE-30 cells and were 6.9-and 5.6-fold, respectively, less toxic with respect to the immortalized normal RPE-1 cell line. The chemotherapeutic agent etoposide exhibited greater cytotoxicity toward all the tested cancer cell lines, relative to the extracts. However, it is worth noting that etoposide was more cytotoxic tono apparent morphological changes in the FLO-1 cells (Fig. 6).

RPPA results
A selected panel of antibodies, pre-validated for RPPA application, was used to compare the individual levels of the corresponding proteins in cancer cells treated with each of the extracts (at different concentrations) with the mock-treated cells. The comparison of the RPPA analysis revealed that, at the 3-h time point, all three extracts upregulated the level of Thr-172 phosphorylation of the alpha subunit of AMPK in a dose-dependent manner, peaking at 3 µg/ml (Fig. 7), suggesting the AMPK pathway of activation in cancer cells.

DISCUSSION
The role played by the chemical scaffolds provided by terrestrial plants for lifesaving drugs is immense [3]. To date, only about 15% of the global plant species have been appraised for their curative potential [10]. In Mauritius, about 32% of the endemic flora have ethnomedicinal uses, although there has been limited scientific validation of their efficacies [5]. The majority Data represent mean IC 50 values ± standard deviation (n = 3). Selectivity index determined as a ratio of the IC 50 immortalized RPE-1 normal cell to the IC 50 of cancer cell lines is indicated. ND = at 50 and 25 µg/ml the extracts and etoposide, respectively, failed to induce 50% growth; hence, no IC 50 value was determined. Different letters between the rows in each column represent significant differences between the extracts (p< 0.05). Asterisks represent significant differences between the extracts and etoposide (positive control), *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, ****p ≤ 0.0001.   of the untapped endemic species, both medicinal and non-medicinal, continue to represent a valuable source of novel chemotypes that await discovery. The therapeutic benefits of plants are ascribed to their secondary metabolites, which are broadly grouped into alkaloids, terpenoids, and phenolics, of which the latter represent an interesting class [11].
Given the involvement of oxidative stress in the multistage carcinogenic process [25], polyphenol-rich extracts are expected to prevent or halt the progression of cancerous cell growth. Polyphenolics are reported to attenuate in vitro cancerous cell growth via diverse mechanisms [26]. The in vitro antiproliferative potential of selected Mauritian endemic Eugenia and Syzygium species against breast cancer cells (MDA-MB and MCF-7) has been described earlier [27]. However, this study reports for the first time on the growth inhibitory activity of Mauritian endemic leaf extracts against cervical adenocarcinoma, colorectal carcinoma, and oesophageal squamous cell carcinoma. The leaf extracts of A. integrifolia, L. glauca, and E. tinifolia exhibited dose-dependent growth inhibition against the selected cancer cell lines ( Table 2). However, oesophageal squamous cell carcinoma KYSE-30 cells were the most sensitive to the extracts' treatment. As per the United States National Cancer Institute cytotoxicity guidelines, crude extracts having an IC 50 value below 20 µg/ml are considered active against tested cell lines [28]. In the present study, the calculated IC 50 values ranged between 6.42 and 9.89 µg/ml when KYSE-30 cells were co-cultured with A. integrifolia, L. glauca, and E. tinifolia extracts for 24 h. Furthermore, the selective cytotoxicity of the extracts toward KYSE-30 cells was fivefold or greater compared to immortalized non-malignant normal cells.
Guided by the AlamarBlue™ assay results, the effects of the three extracts on cell death and cell cycle stages of KYSE-30 and FLO-1 cells were further investigated across a concentration dilution series using fluorescence staining and high-content image analysis. The findings indicated that extract treatment for 48 h induces considerable dose-dependent cell death in KYSE-30 cells but not in FLO-1 cells (Figs. 5, 6). A similar trend was also observed in cell cycle analysis, as extracts treatment induced accumulation of KYSE-30 cells in the G2/M phase, but no effect on the FLO-1 cell cycle was apparent (Fig. 4). The blockade of KYSE-30 cells through G2/M was concentration-dependent for both E. tinifolia and L. glauca, while in the case of A. integrifolia the effect was less influenced by the extract dose.
In cancerous cells, signaling components of different cellular pathways are often mutated [29]. In this vein, activation of AMPK and subsequent inhibition of the mTOR pathway is an area of active research [30]. It was previously suggested that activators of AMPK, which include plant polyphenols, can interfere with tumor growth via cell cycle arrest [31] and apoptosis [32,33] in cancerous cells. The RPPA analysis of the canonical signaling pathways that regulate the cell cycle progression and survival pathways, which in turn regulate cell cycle progression and survival, revealed that all three extracts increase the level of Thr-172 phosphorylation of the alpha subunit of AMPK in KYSE-30 cells (Fig.  7), suggesting its activation. This study showed for the first time that the antiproliferative effect of Mauritian endemic medicinal plant leaf extracts is related to their influence on the AMPK pathway modulation.

CONCLUSIONS
In conclusion, the specific selectivity of A. integrifolia, E. tinifolia, and L. glauca to KYSE-30 oesophageal AMPK alpha P Thr172/total AMPK FLO-1 AMPK alpha P Thr172/total AMPK KYSE-30 Extract concentration, µg/ml The results represent the ratio of phosphorylated to total AMPK levels and are expressed as a fold change over DMSO control for each cell type squamous cell carcinoma relative to oesophageal adenocarcinoma (FLO-1) calls for further investigation into the anticancer effect of these medicinal plants. This is further substantiated by the G2/M phase cell cycle arrest, leading to cell death. AMPK pathway modulation following the exposure of KYSE-30 cells to the extracts provides novel insight into the mechanism of action of these leaf extracts as a potential chemotherapeutic treatment. This is particularly important since esophageal cancer is among the leading causes of cancer mortality worldwide as treatment is limited due to adverse systemic effects, limited efficacy, and emergence of drug resistance. Clinical studies with molecularly targeted therapies have so far been disappointing, with little improvement in patient outcomes. Hence, there is an urgent need to search for new effective treatments for oesophageal cancer. Although preliminary characterization indicated the presence of (+)-catechin and gallocatechin, in-depth phytochemical identification is warranted in order to identify the AMPK modulator.