Cytotoxic Constituents of the Bark of Hypericum roeperianum towards Multidrug-Resistant Cancer Cells

The global cancer burden remains a serious concern with the alarming incidence of one in eight men and one in eleven women dying in developing countries. This situation is aggravated by the multidrug resistance (MDR) of cancer cells that hampers chemotherapy. In this study, the cytotoxicity of the methanol extract (HRB), fractions (HRBa, HRBb, and HRBa1-5), and compounds from the bark of Hypericum roeperianum (HRB) was evaluated towards a panel of 9 cancer cell lines. The mode of action of the HRB and trichadonic acid (1) was also studied. Column chromatography was applied to isolate the constituents of HRB. The cytotoxicity of botanicals and phytochemicals was evaluated by the resazurin reduction assay (RRA). Caspase-Glo assay was used to evaluate the activity of caspases, and reactive oxygen species (ROS) (H2DCFH-DA) were assessed by flow cytometry. Phytochemicals isolated from HRB were trichadonic acid (1), fridelan-3-one (2), 2-hydroxy-5-methoxyxanthone (3), norathyriol (4), 1,3,5,6-tetrahydroxyxanthone (5), betulinic acid (6), 3′-hydroxymethyl-2′-(4″-hydroxy-3″,5″-dimethoxyphenyl)-5′,6′:5,6-(6,8-dihydroxyxanthone)-1′,4′-dioxane (7), and 3′-hydroxymethyl-2′-(4″-hydroxy-3″,5″-dimethoxyphenyl)-5′,6′:5,6-(xanthone)-1′,4′-dioxane (8). Botanicals HRB, HRBa, HRBa2-4, HRBb, and doxorubicin displayed cytotoxic effects towards the 9 tested cancer cell lines. The recorded IC50 values ranged from 11.43 µg/mL (against the P-glycoprotein (gp)-overexpressing CEM/ADR5000 leukemia cells) to 26.75 µg/mL (against HCT116 (p53+/+) colon adenocarcinoma cells) for the crude extract HRB. Compounds 1, 5, and doxorubicin displayed cytotoxic effects towards the 9 tested cancer cell lines with IC50 values varying from 14.44 µM (against CCRF-CEM leukemia cells) to 44.20 µM (against the resistant HCT116 (p53−/−) cells) for 1 and from 38.46 µM (against CEM/ADR5000 cells) to 112.27 µM (against the resistant HCT116 (p53−/−) cells) for 5. HRB and compound 1 induced apoptosis in CCRF-CEM cells. The apoptotic process was mediated by enhanced ROS production for HRB or via caspases activation and enhanced ROS production for compound 1. This study demonstrated that Hypericum roeperianum is a potential source of cytotoxic phytochemicals such as trichadonic acid and could be further exploited in cancer chemotherapy.


Plant Material and Extraction.
e bark of Hypericum roeperianum Schimp. p. ex A. Rich (Guttiferae) was collected in Bangang Wabane (South West Region of Cameroon) in October 2018. No permission was necessary for sample's collection. e identification of the plant was carried out by Dr. Tchiengue Barthelemy at the Cameroon National Herbarium (Yaoundé) where a voucher specimen was deposited under the number 24584/SRF/Cam. Dried barks of the plant (3.0 kg) were powdered and extracted with methanol (MeOH; 3 × 15 L) for 72 h at room temperature to afford a crude extract (HRB; 150.0 g) after filtration with Whatman paper no.1 and evaporation in vacuum, under reduced pressure. A portion of the resulting extract (140.0 g) was, then, exhausted in ethyl acetate to yield 65.0 g of the ethyl acetate extract (EtOAc) (HRBa) and the residue (HRBb; 75 g).

Resazurin Reduction Assay (RRA) for Cell Growth Evaluation.
e RRA was applied to evaluate the cytotoxicity of botanicals, the isolated phytochemicals (1-5, 7, and 8), and doxorubicin on the cell growth as reported earlier [18,28]. Cells treated with various samples at different concentrations were incubated for 72 h in humidified 5% CO 2 atmosphere at 37°C. Cells were further coloured with resazurin and incubated for 1-2 h; the fluorescence was further measured with an Infinite M2000 Pro ™ plate reader (Tecan, Crailsheim, Germany) at 544 nm as the excitation wavelength and 590 nm as the emission wavelength. e IC 50 values represented the concentrations of the sample required to inhibit 50% of cell proliferation and were calculated from a calibration curve by linear regression using Microsoft Excel 2007 [29].

Flow Cytometric Evaluation of Cell Cycle Distribution and Apoptotic Cells.
Various concentrations of botanical HRB, phytochemical 1, and doxorubicin or DMSO (solvent control) were used to treat CCRF-CEM cells (1 × 10 6 cells). Cells were further incubated for 24 h in humidified 5% CO 2 atmosphere at 37°C and analyzed using a BD Accury C6 Flow Cytometer (BD Biosciences, Heidelberg, Germany) by measuring the propidium iodide fluorescence of individual nucleus, as described earlier [10,11]. Experiments were conducted thrice independently with three parallel measurements.

Assessment of Apoptosis by Annexin V/PI Staining.
e CCRF-CEM cells (1 × 10 6 ; 1 ml) were also treated with HRB, compound 1 and doxorubicin for 24 h (in humidified 5% CO 2 atmosphere at 37°C), and apoptosis was further assessed by flow cytometry using the flouresceinisothiocynate-(FITC-) conjugated annexin V/PI assay kit (eBioscience ™ Annexin V; Invitogen, San Diego, USA), as previously published [10,11]. Briefly, treated cells were centrifuged at 1200 rpm for 5 min, then washed twice with ice-cold PBS, resuspended in 500 µl binding buffer, and stained with 5 µl FITC-conjugated annexin V (10 mg/mL) and 10 µl PI (50 mg/ml). After 15 min incubation at room temperature (RT) in the dark, cells were analyzed using a BD Accury C6 Flow Cytometer (BD Biosciences). Cells stained with only annexin V were evaluated as being in early apoptosis. Cells stained with both annexin V and propidium iodide were evaluated as being in late apoptosis or in a necrotic stage.

Statistics.
Statistical analyses were performed with Graph pad prism 5 software. Representative data from three independent experiments are shown as mean value ± S.E.M. One-way Analysis Variance (ANOVA) followed by post hoc Tukey's test was used to determine the significance of the Evidence-Based Complementary and Alternative Medicine difference between mean values relative to the control. e p value was calculated to determine significant differences (p value < 0.05).

Cytotoxicity of Phytochemicals and Doxorubicin.
e cytotoxicity of crude extracts, fractions, and phytochemicals 1-5, 7, 8, and doxorubicin was investigated using RRA towards 9 cancer cell lines and normal AML12 hepatocytes (Tables 1 and  2). e degree of resistance (D.R.) of the tested samples was determined as the ratio of the IC 50 value of the resistant cell line divided by that of the corresponding parental sensitive cell line (Tables 1 and 2). Collateral sensitivity or hypersensitivity was deduced if the D.R. was below 1 while normal sensitivity was defined as a D.R. of 1 or around 1; cross resistance was considered as a D.R. above 1. e selectivity index (S.I.) was also calculated as the ratio of the IC 50 value in normal AML12 hepatocytes by the corresponding values in HepG2 hepatocarcinoma cells (Tables 1 and 2). e obtained IC 50 values ranged from 11.43 µg/mL (against the P-gp-overexpressing CEM/ADR5000 leukemia cells) to 26 Collateral sensitivity (D.R. below 1) of CEM/ADR5000 cells, BCRP-expressing MDA-MB-231 cells, and HCT116 (p53 −/− ) cells to the mother botanical, HRB compared to their sensitive counterparts CCRF-CEM cells, MDA-MB-231 cells, and HCT116 (p53 +/+ ) cells, respectively, was observed. Hypersensitivity of all resistant cell lines to fraction HRBa4 and HRBb compared to their sensitive parental cell lines was also recorded ( Table 1). Collateral sensitivity of BCRP-expressing MDA-MB-231 cells and U87MG.ΔEGFR cells to phytochemicals 1, 3-5, and 8 compared to their sensitive counterparts MDA-MB-231 cells and U87MG cells, respectively, was also observed ( Table 2). Concerning the most active compound 1, a little cross resistance of HCT116 (p53 −/− ) cells compared to their sensitive counterparts HCT116 (p53 +/+ ) was observed with a D.R. of 2.55; however, this value was lower than that obtained with doxorubicin (D.R. of 3.73) (Table 2). Apart against HRBb, the S.I. of all samples was above 2 in HepG2 as compared with normal AML12 hepatocytes (Table 1) . Compounds 1, 3, 5, 8, and doxorubicin had an S.I. above 2 in HepG2 as compared with normal AML12 hepatocytes (Table 1). Regarding the recorded IC 50 values, trichadonic acid (1) had the best activity and was consequently selected, together with the crude extract, HRB, for further mechanistic studies.

Discussion
Taking into account the rapid development of resistance by cancer cell lines, the use of MDR phenotypes when screening phytochemicals is an interesting approach. Collateral or normal sensitivity (D.R. below or equal to 1) of MDR cells to phytochemicals combined to their good cytotoxicity could be better criteria to select substances for clinical studies. In the present work, four MDR cells lines, CEM/ADR5000 cells, MDA-MB-231-BCRP cells, HCT116 (p53 − / − ) cells, and U87.MGΔEGFR cells, were used, and their susceptibilities to isolated phytochemicals were compared with those of their parental sensitive counterparts, CCRF-CEM cells, MDA-MB-231 cells, HCT116 (p53 +/+ ) cells, and U87.MG cells, respectively (Tables 1 and 2). Interestingly, collateral sensitivity of CEM/ADR5000 cells, BCRP-expressing MDA-MB-231 cells, and HCT116 (p53 −/− ) cells to HRB was achieved, as well as the hypersensitivity of all resistant cell lines to fraction HRBa4 and HRBb compared to their sensitive parental cell lines (Table 1). Collateral sensitivity of BCRP-expressing MDA-MB-231 cells and U87MG.ΔEGFR cells to phytochemicals 1, 3-5, and 8 was observed, suggesting that they might be useful to fight drug resistance in breast cancer and glioblastoma (Table 2). is clearly indicates that these botanicals and phytochemicals can be exploited in the fight against recalcitrant cancers. e IC 50 values below 20 μg/mL or below 10 μM after incubation between 48 and 72 h have been set for promising cytotoxic botanicals and phytochemicals, respectively [39,40]. Importantly, IC 50 values below 20 µg/mL were obtained with HRB against 8/9 tested cancer cells lines, HRBa2 and HRBa4 against 7/9 cancer cell lines, HRBa against 4/9 cancer cell lines, and HRBa3 against 3/9 cell lines (Table 1). It can, therefore, be confirmed that these botanicals are interesting cytotoxic agents. However, IC 50 values below the established threshold were not achieved with phytochemicals, though terpenoid 1 and xanthone 5 had cytotoxic effects towards the 9 tested cancer cell lines. However, their good selectivity indexes still suggest that they can still be good candidates to tackle cancers, especially when drug resistance is observed. To the best of our knowledge, the cytotoxicity of the crude extract and compounds 1, 2, 3, 7, and 8 on the studied cell lines is being reported, herein, for the first time. Betulinic acid (6) is a well-known cytotoxic compound and has previously been found active towards the cancer cell lines tested in the present work, with IC 50 values ranging from 7.65 µM (in CEM-ADR5000 cells) to 44.17 µM (in HepG2 cells) [41]. Although it was not further tested, herein, compound 6 can be ranked amongst the best active principles of Hypericum roeperianum. Also, the cytotoxicity of norathyriol (4) in JB6 P+ mouse skin epidermal cells was reported [42]. 1,3,5,6-Tetrahydroxyxanthone (5) had low     [43].
Apoptosis is a programmed cell death and is also the most investigated mechanism of action of antiproliferative drugs. In this study, it was found that both HRB and compound 1 induced apoptosis in CCRF-CEM cells with cell cycle arrest in the Go/G1 phase (Figures 2 and 3). Modulation of caspases activities is one of the events observed in the apoptotic process in cancer cell lines [44], making these enzymes a target for cytotoxic drug discovery [8,45]. However, no significant increase in the activity of initiator caspases (caspases 8 and 9) or in that of the activator caspases (caspase 3/7) was observed (Figure 4). Phytochemical 1 induced 2.52-fold, 2.62-fold, and 2.23-fold increase of the activity of caspases 3/7, 8, and 9, respectively (Figure 4), suggesting that this molecule is a caspase modulator. Botanical HRB and compound 1 were also shown to induce increase of ROS by up to 71.92% and 68.12%, respectively (2 × IC 50 ; Figure 5); this is an indication that one of the modes of action of this triterpenoid also includes the enhancement of ROS production in cancer cells.
Regarding the structure-activity relationship, it appears that pentacyclic triterpene 1 is different from 2 by the presence of the carboxyl group (-COOH) in C-13 ( Figure 1); the presence of this −COOH group significantly enhanced the cytotoxic activity of triterpene 1, with IC 50 values ranging from 14.44 µM to 44.20 µM in cancer cells tested whilst no IC 50 value at up to 100 µM was recorded with triterpene 2 (Table 1). Betulic acid, another pentacyclic triterpene with-COOH in C-17, previously displayed good cytotoxicity against all cancer cell lines tested in this work [41], illustrating the importance of the carboxyl function in the cytotoxicity of pentacyclic triterpenes. Concerning xanthones, though 5 was active in all tested cancer cell lines,    (Table 1). It was previously shown that additional hetrocycle in xanthone combined to prenylation improved the cytotoxicity of xanthones, with cudraxanthone I (additional hetrocycle combined to -C-8prenylation) displaying significant cytotoxic effects (IC 50 value below 10 µM) against all cancer cell lines tested in the present study [46]. is observation is also confirmed with another prenylated xanthone bearing additional herocycle, xanthone V1 [47]. e difference between the two xanthonolignoids 7 and 8 is the presence of two hydroxyl (-OH) groups in C-6 and C-7 (Compound 8) (Figure 1). is difference seems to influence the selectivity, as compound 8 was active against HCT116 (p53 +/+ ) cells, with an IC 50 value of 37.79 µM compared to the IC 50 value above 91.74 µM obtained for compound 7 (against the same cell line ( Table 1).

Data Availability
All data generated or analyzed during this study are included in this published article and its supplementary information files.

Conflicts of Interest
e authors declare that they have no conflicts of interest.

Authors' Contributions
SBT, FD, M-GGF, ATM, and VK carried out the experiments; IC recorded NMR data; FD, GTMB, and JDSM elucidated the chemical structures; ATM and VK wrote the manuscript; and all authors read and approved the final manuscript.