Food and Chemical Toxicology 50 (2012) 3004–3013

Contents lists available at SciVerse ScienceDirect

Food and Chemical Toxicology

journal homepage: www.elsevier .com/locate/ foodchemtox

Anti-skin cancer properties of phenolic-rich extract from the pericarpof mangosteen (Garcinia mangostana Linn.)

Jing J. Wang a,b, Qing H. Shi a,c, Wei Zhang a,b, Barbara J.S. Sanderson a,⇑a Level 4, Health Science Building, Department of Medical Biotechnology, Flinders Medical Sciences and Technology, School of Medicine, Faculty of Health Science, Flinders University,Registry Road, Bedford Park, Adelaide 5042, Australiab Level 4, Health Science Building, Flinders Centre for Marine Bioproducts Development (FCMBD), Flinders University, Registry Road, Bedford Park, Adelaide 5042, Australiac Department of Biochemical Engineering and Key Laboratory of Systems Bioengineering of Ministry of Education, School of Chemical Engineering and Technology, Tianjin University,Tianjin 300072, China

a r t i c l e i n f o

Article history:Received 8 January 2012Accepted 4 June 2012Available online 13 June 2012

Keywords:Skin cancerMangosteenCell cycle arrestApoptosis inductionMitochondrial pathwayAkt pathway

0278-6915/$ - see front matter � 2012 Elsevier Ltd. Ahttp://dx.doi.org/10.1016/j.fct.2012.06.003

⇑ Corresponding author. Tel.: +618 7221 8556; fax:E-mail addresses: wang0524@flinders.edu.au (J

(Q.H. Shi), wei.zhang@flinders.edu.au (W. Zhang),edu.au (B.J.S. Sanderson).

a b s t r a c t

Skin cancers are often resistant to conventional chemotherapy. This study examined the anti-skin cancerproperties of crude ethanol extract of mangosteen pericarp (MPEE) on human squamous cell carcinomaA-431 and melanoma SK-MEL-28 lines. Significant dose-dependent reduction in% viability was observedfor these cell lines, with less effect on human normal skin fibroblast CCD-1064Sk and keratinocyte HaCaTcell lines. Cell distribution in G1 phase (93%) significantly increased after 10 lg/ml of MPEE versusuntreated SK-MEL-28 cells (78%), which was associated with enhanced p21WAF1 mRNA levels. In A-431cells, 10 lg/ml MPEE significantly increased the sub G1 peak (15%) with concomitant decrease in G1 phaseover untreated cells (2%). In A-431 cells, 10 lg/ml MPEE induced an 18% increase in early apoptosis ver-sus untreated cells (2%). This was via caspase activation (15-, 3- and 4-fold increased caspse-3/7, 8, and 9activities), and disruption of mitochondrial pathways (6-fold decreased mitochondrial membrane poten-tial versus untreated cells). Real-time PCR revealed increased Bax/Bcl-2 ratio and cytochrome c release,and decreased Akt1. Apoptosis was significantly increased after MPEE treatment of SK-MEL-28 cells.Hence, MPEE showed strong anti-skin cancer effect on these two skin cancer cell lines, with potentialas an anti-skin cancer agent.

� 2012 Elsevier Ltd. All rights reserved.

1. Introduction

Skin cancer, including non-melanoma and melanoma, is a grow-ing public health problem due to a dramatically increasing inci-dence (Trakatelli et al., 2007; Doan, 2008). This increase in theincidence of skin cancer is expected to continue due to the agingpopulation and the greater levels of UV radiation reaching theearth surface as a result of the ozone layer depletion (Johnsonet al., 1998; Miller and Weinstock, 1994). Therefore, it is importantto develop novel effective chemopreventive agents that can reduceor control skin cancer.

More than 50% of anti-cancer drugs currently in clinical trialsare derived from, or inspired by, natural products, especiallyterrestrial plants (Cragg and Newman, 2000; Nuijen et al., 2000;Gordaliza, 2007). Mangosteen (Garcinia mangostana Linn.) is atropical fruit available in south-east Asia. Its pericarp has beenused for centuries as traditional medicine by Southeast Asians

ll rights reserved.

+618 7221 8555..J. Wang), qhshi@tju.edu.cn

Barbara.sanderson@flinders.

and South Americans for a great variety of medical conditions,for example, treatment of skin infections, wounds and diarrhea(Mahabusarakam et al., 1987; Pedraza-Chaverri et al., 2008; Obol-skiy et al., 2009). A 50% ethanol extract of mangosteen pericarp hadantioxidative and neuroprotective activities in NG108-15 neuro-blastoma cells against H2O2-induced oxidative stress (Weecha-rangsan et al., 2006). Another ethanol extract of mangosteenpericarp inhibited histamine release, prostaglandin E2 synthesis(Nakatani et al., 2002), and HIV-1 protease (Chen et al., 1996). Amethanol extract of mangosteen pericarp displayed antiprolifera-tive, apoptotic and antioxidative activities on SKBR3 human breastcancer cell line (Moongkarndi et al., 2004a). Xanthones and antho-cyanins are found in the pericarp of mangosteen, with xanthonesbeing the major bioactive compounds (Nguyen et al., 2005). Theantioxidant activity and cytotoxicity of xanthones on some cancercells (e.g. colorectal cancer, hepatoma, leukemia, and small celllung cancer) have been reviewed extensively (Akao et al., 2008;Pedraza-Chaverri et al., 2008; Obolskiy et al., 2009). We previouslydemonstrated that a-mangostin, c-mangostin, and 8-desoxygarta-nin isolated from the pericarp of mangosteen exhibited significantanti-cancer effect on human melanoma SK-MEL-28 cell line via cellcycle arrest in G1 phase and apoptosis induction (Wang et al.,

J.J. Wang et al. / Food and Chemical Toxicology 50 (2012) 3004–3013 3005

2011). However, no information is available on the anti-skin canceractivity of mangosteen pericarp crude extracts.

In this study, we investigated the antioxidant properties of eth-anol extracts of mangosteen pericarp and its anti-skin cancer activ-ity in human squamous cell carcinoma (non-melanoma) A-431 andmelanoma SK-MEL-28 cell lines and the underlying cellular andmolecular mechanisms.

2. Materials and methods

2.1. Chemicals and reagents

Methanol, acetic acid, and ethanol were of analytical grade from Merck (Austra-lia) and all the other reagents were purchased from Sigma–Aldrich (St. Louis, USA)unless stated otherwise.

2.2. Extraction

Fresh mangosteen fruits were purchased from the local market in Adelaide,South Australia, Australia. The standardized extraction method as described previ-ously (Shan and Zhang, 2010) was used to produce the extracts with quality consis-tency and reproducibility. Briefly, the mangosteen fruit was cleaned with MilliQwater. The pericarp was peeled from fruit and ground into fine powder in a juiceblender with a cross blade. The extraction was conducted by mixing mangosteenpericarp with absolute ethanol at 75 �C for 1 h with a weight-to-volume ratio of1:10. After extraction, the mixture was centrifuged at 10,000g for 10 min. Thesupernatant was filtered through 0.22 lm filter (Minisart, Sartorius Stedim Biotech,Germany) and then vacuum dried to yield crude ethanol extract (MPEE). The greentea (Camellia sinensis) sample branded as ‘‘Chinatea’’ was purchased as ground driedplant leaves from a local store and used as an antioxidant comparison standard andits extraction was carried out by mixing green tea powder with MilliQ water at100 �C for 1 h with a weight-to-volume ratio of 1:10. The remaining steps werethe same as for the mangosteen pericarp extraction, except that the supernatantwas freeze dried instead of vacuum dried. The extract was stored in a desiccatorsat �20 �C. When required, the dried MPEE extract was redissolved in ethanol andgreen tea extract was redissolved in MilliQ water. The major xanthone compoundsin the MPEE were identified and quantified by HPLC as a-mangostin (321 mg/gMPEE), b-mangostin (3.88 mg/g MPEE), c-mangostin (81.3 mg/g MPEE), 8-desoxy-gartanin (10.7 mg/g MPEE), 9-hydroxycalabaxanthone (4.84 mg/g MPEE), and gart-anin (18.7 mg/g MPEE) (Shan and Zhang, 2010).

2.3. Total phenolics measurement

Total phenolics content of MPEE or green tea extract was determined by themethod as described previously (Singleton and Rossi, 1965) with minor modifica-tion. Briefly, 1 ml MilliQ water was added into the 200 ll of the extract followedby adding 50 ll Folin–Ciocalteu reagent. The mixture was vortexed and incubatedfor 7 min at room temperature (RT). Then 290 ll of sodium carbonate (20%) wasadded and mixed evenly. The absorbance was measured at 760 nm by UV–vis Spec-trophotometer (UV mini 1240, Shimadzu Corporation, Japan) after 1 h incubation.Quantitation of total phenolics content was based on the standard curve of gallicacid. Results were expressed in terms of gallic acid equivalence (GAE) in mg perdry pericarp weight in gram.

2.4. 2,2-diphenyl-1-picrylhydrazyl (DPPH) radical scavenging activity

The extract was tested for its ability to scavenge DPPH� radicals by DPPH radicalscavenging assay as described previously (Blois, 1958). Three milliliters of 0.1 mMDPPH� in ethanol was added to 0.5 ml of MPEE extract or green tea extract and incu-bated for 30 min in the dark at RT. Absorbance was read at 517 nm using a UV–visSpectrophotometer. The percentage of free radical scavenging activity wascalculated using the formula

%DPPH inhibition¼ð1�Absorbance in the presence of extract=Absorbance in the absence of extractÞ�100

The antioxidant activity of the extract was expressed as IC50 (the concentrationof the extract required to inhibit DPPH radicals by 50%), as calculated using Graph-Pad Prism 5 software (San Diego, CA, USA).

2.5. Ferric reducing antioxidant power (FRAP) assay

The FRAP assay was performed as previously described (Benzie and Strain,1996) to test the reducing activity of MPEE and green tea extract. Briefly, the extractwas centrifuged for 10 min at 2830g prior to use and the supernatant was saved.

FRAP reagent consisted of 25 ml of 300 mM acetate buffer at pH 3.6, 2.5 ml of10 mM 2,4,6-tripyridyl-s-triazine and 2.5 ml of 20 mM FeCl3. FRAP reagent(900 ll) was heated at 37 �C for 10 min, then 90 ll of Milli-Q water and 30 ll ofsample were added to it and mixed well. The absorbance of the reaction mixtureat 593 nm was measured spectrophotometrically after incubation at 37 �C for4 min. FeSO4�7H2O of 1 mM was used as the standard solution (standard solutionranged from 0 to 1 mM). The antioxidant activity of the tested MPEE or green teaextract was expressed as FRAP value/g dry pericarp weight (DW) as shown in theequation below:

FRAP value ðlmol=LFeSO4 equivalents ðFEÞ=g DWÞ¼ ½ðAbsorbance at 593 nm=slopeÞ � dilution factor�� volume of supernatant ðmlÞ=equivalent DW ðgÞ=1000

2.6. Oxygen radical absorbance capacity (ORAC) assay

The assay was carried out as described previously (Huang et al., 2005) withminor modification. The ORAC assay was carried out on a microplate reader (DTX880, Beckman Coulter, USA). Trolox, a water-soluble analog of vitamin E, was usedas a control standard. Briefly, the extract and trolox were diluted in a 96-wellmicroplate. Trolox was prepared in concentrations of 100, 75, 50, 25, 12.5 and0 lM while the dilution factors of 20, 200, 400, 800, 1600 and 3200� were appliedfor the extract sample. In the analysis, 180 ll of fluorescein working solution, 30 llof phosphate working buffer and 30 ll of diluted sample/trolox were loaded intoeach well. The reaction was carried out at 37 �C and the plate was incubated for10 min before adding 2,20-azobis (2-amidinopropane) dihydrochloride (AAPH).After adding AAPH, the plate was shaken to mix the reagents in each well beforerecording the initial fluorescence (f0). Fluorescence readings were taken at 0 s (s)(f0) and then every 90 s thereafter (f1, f2, f3, . . . ) for 50 cycles. The final ORAC valueswere calculated by using a regression equation between trolox concentration andthe net area under the curve by MATLAB program and were expressed as micromoletrolox equivalents (TE) per gram of dry pericarp weight (lmol TE/g DW). All thereaction mixtures were prepared in duplicate, and at least three independent assayswere performed for each sample.

2.7. Cell culture

Human melanoma SK-MEL-28, squamous cell carcinoma A-431, and skin fibro-blast CCD-1064Sk cell lines were purchased from the American Type Culture Collec-tion. Human normal keratinocytes (HaCaT) cell line was a kind gift from theDepartment of Haematology, Flinders University, Adelaide, Australia. SK-MEL-28and A-431 cells were cultured in DMEM, HaCaT in RPMI 1640, and CCD-1064Skin IMEM, all supplemented with 10% heat-inactivated fetal bovine serum (FBS;Invitrogen Corporation, Australia) in the presence of 100 U/ml penicillin and0.1 mg/ml streptomycin (Thermo Scientific, Melbourne, Australia). All the cultureswere maintained in a fully humidified incubator with 5% CO2 at 37 �C. Cells usedin experiments were under the passage number of 20 and free of mycoplasma con-tamination as detected by PCR (data not known).

2.8. Treatment preparation

MPEE was dissolved in absolute ethanol, and diluted with the respective mediato the desired treatment concentrations. Ethanol was used as vehicle control and itsvolume was consistently kept as 1% (v/v) for each treatment.

2.9. Cell proliferation assay

Crystal violet assay was carried out to measure the cell viability as describedpreviously (Wang et al., 2011). EC50 was determined as 50% viability relative tothe untreated control, as calculated using GraphPad Prism 5 software. 5-Fluoroura-cil (5-FU) and dacarbazine (DTIC) were used as positive control for A431 and SK-MEL-28, respectively, because 5-FU is a commonly used drug for human squamouscarcinoma, and DTIC is the only drug for melanoma approved by both the UnitedStates Food and Drug Administration and the European Agency for the Evaluationof Medicinal Products (Tsao et al., 2004; Garbe and Eigentler, 2007). The morpho-logical changes of cells were observed under the microscope and recorded by theOlympus 1X71 phase contrast inverted fluorescence microscope and analySIS im-age capture software (magnification 100�).

2.10. DNA cell cycle analysis

Cell cycle distribution of the untreated/treated cells was measured by flowcytometry using propidium iodide staining as described previously (Wang et al.,2011).

3006 J.J. Wang et al. / Food and Chemical Toxicology 50 (2012) 3004–3013

2.11. Apoptosis analysis

Apoptosis induction was quantitatively determined by flow cytometry usingthe Annexin V-conjugated propidium iodide Apoptosis detection kit (BD Biosci-ences, USA) based on the manufacturer’s instructions as described previously(Wang et al., 2011).

2.12. Caspase 3/7, 8 and 9 assays

Caspase activity was measured by using Caspase-Glo� 3/7, 8 and 9 assay kits(Promega Corporation, Australia). Briefly, SK-MEL-28 and A-431 cells (1 � 104

cells/well) were seeded in a luminometer plate and incubated for 24 h at 37 �C. Cellswere then treated with different concentrations of individual compounds for 48 h.To each well, 50 ll of caspase-3/7, -8 or -9 reagents were added and the lumines-cence was recorded every 15 min for 1 h at 28 �C in a microplate reader (Synergy4, Bioteck, Millennium Science).

Prior to the Caspase assay, cell number in the same well after treatment wasmeasured by the Cell Titer Blue assay (Promega Corporation, Australia) to normalizethe result. Briefly, 10 ll of the staining was added to each well and incubated for 1 hat 37 �C. The fluorescence was recorded at 560(excitation)/590(emission) in amicroplate reader.

The results are presented as relative luminescence unit (RLU)/cell/min.

2.13. Mitochondrial membrane potential (DWm) detection

Change in DWm was quantitatively determined by staining the cells with thecationic dye, JC-1, following the manufacturer’s instructions (Sigma, St. Louis,USA), as described previously (Wang et al., 2011).

2.14. Real time-reverse transcription PCR (qRT-PCR)

Real time reverse transcription-PCR (qRT-PCR) was performed to determinemRNA levels of a number of cell cycle and apoptosis-related genes on the A431and SK-MEL-28 cells subjected to the indicated treatments. Total cellular RNAwas isolated using TRIzol solution (Invitrogen, Australia) according to the manufac-turer’s instructions. The quality of RNA was determined using 1.2% formaldehyde–agarose gel. The DNA contamination from RNA preparation was removed by usingAmbion TURBO DNA-free DNase treatment reagent (Applied Biosystems, Australia).Reverse transcription was carried out with 1 lg of total RNA in a reaction volume of20 ll using a High Capacity cDNA Reverse Transcription kit (Applied Biosystems,Australia) following the manufacturer’s instructions. Real time PCR was carriedout in 20 ll of PCR mixture consisting of 10 ll of 2� GoTaq� qPCR Master Mix(Promega Corporation, Australia) and 10 ll of primers and 2 ll cDNA sample on aRotorgene 3000 Thermocycler (Corbett Life Science, Sydney, NSW, Australia). Eachsample was amplified in triplicate. The mRNA level of the beta-actin house keepinggene was also determined by qRT-PCR in each cDNA sample to normalize theexpression of genes of interest (GOI). The primers used are listed in Table 1. ThePCR condition for each particular amplicon was optimized to establish a linearrelationship between threshold cycle (Ct) values and the input amount of cDNA(a correlation coefficient >95%) and to achieve PCR efficiency around 100%(± 10%). Cycling was followed by Melt Curve analysis to confirm the specificity of

Table 1Primer sequences used for target genes amplification.

Name Primer Sequence (50–30)

Bax F: AACCATCATGGGCTGGAR: CGCCACAAAGATGGTCA

Bcl-2 F: TGGATGACTGAGTACCTGAR: TGAGCAGAGTCTTCAGAGA

Cytochrome c F: CCAGTGCCACACCGTTGAAR: TCCCCAGATGATGCCTTTGTT

Cyclin D1 F: CGTGGCCTCTAAGATGAAGGR: TGCGGATGATCTGTTTGTTC

P21WAF1 F: GACACCACTGGAGGGTGACTR: CAGGTCCACATGGTCTTCCT

Akt1 F: TCTATGGCGCTGAGATTGTGR: CTTAATGTGCCCGTCCTTGT

NFjB F: CCACAAGACAGAAGCTGAAGR: AGATACTATCTGTAAGTGAAC

IjBa F: ACACTAGAAAACCTTCAGATGR: ACACAGTCATCATAGGGCAG

Beta-actin (house-keeping gene) F: TTGCCGACAGGATGCAGAAGR: GCCGATCCACACGGAGTACT

the PCR product. Quantitative analysis was performed using Q-Gene software (Si-mon, 2003). The expression of levels of GOI was normalized to the expression levelsof beta-actin, which exhibited comparable expression across different groups (datanot shown). The ratio of GOI/beta-actin was compared among samples, and the foldchange of GOI expression was obtained by setting the values from the untreatedcells to one.

2.15. Statistical analysis

Data are expressed as mean (± SEM). The experiments were repeated at leastthree independent times. Statistical analysis of the data was carried out usingone way ANOVA, followed by Tukey’s HSD post hoc test (equal variances) orDunnett’s T3 post hoc test (unequal variances). These tests were performed usingSPSS software (version 17). Difference was considered statistically significant whenthe P-value was less than or equal to 0.05 (significant) and 0.01 (highly significant).MATLAB ver. 7.0, GraphPad Prism 5.0 and Microsoft Excel 2003 were also used forstatistics analysis and graphical evaluation.

3. Results

3.1. Antioxidant properties of MPEE

One hundred grams of mangosteen pericarp powder yielded23.2 g of crude ethanol extract. The vacuum dried extract was ayellowish paste.

The total phenolic content in the ethanol extracts was64.3 ± 3.2 mg GA equivalent per g dry pericarp (GAE/g DW), lessthan that of the green tea, a standard antioxidant, with a value of72.5 ± 5.8 GAE/g DW. The total phenolics content in the green teawas similar with previous study which reported the total phenolicscontent of 62.1 mg GAE/g DW for green tea water extract (Vinokuret al., 2006).

The antioxidant activities of MPEE were measured by the DPPH,FRAP and ORAC assays and the results are summarized in Table 2.The IC50 value of DPPH was similar to green tea, while the FRAPand ORAC values were less.

3.2. Anti-proliferative activities of MPEE

MPEE significantly inhibited the proliferation of A-431 and SK-MEL-28 cells in a dose-dependent manner and a time-dependentmanner at higher concentrations (Fig. 1). MPEE showed lesscytotoxicity on human normal skin cell lines than human skincancer cells, as evidenced by higher EC50 values of 12.62 (onCCD-1064Sk) and 8.32 lg/ml (on HaCaT) compared to 5.07 (on

Accession Number Product Size (bps)

L22473.1 133

M14745.1 139

NM_018947.5 136

NM_053056.2 215

NM_001220778.1 172

NM_005163.2 113

NM001165412.1 149C

C NM020529.2 110

NM_001101.3 101

Table 2Antioxidant properties of MPEE and green tea.

Samples FRAP value (lM FE/g DW) ORAC value (lmol TE/g DW) DPPH IC50 value (lg/ml)

MPEE 500 ± 31 562 ± 54 7.32 ± 0.4Green tea extract 1800 ± 102 1200 ± 82 6.41 ± 0.21

Ferric reducing antioxidant power (FRAP), expressed as lM of FeSO4 equivalents (FE) per gram of dry weight (DW). Oxygen radical absorbance capacity (ORAC), expressed aslmol of trolox equivalents (TE) per gram of dry weight (DW). DPPH IC50 value is the concentration scavenging 50% DPPH radicals. Values represent the mean ± SEM of threedifferent measurements.

J.J. Wang et al. / Food and Chemical Toxicology 50 (2012) 3004–3013 3007

A-431) and 6.89 lg/ml (on SK-MEL-28) after 48-h treatment withMPEE.

After 48 h treatment, morphological alterations in both celllines were observed compared with the untreated cells (Fig. 2). Un-treated cells were polygonal in ordinary shape, while treated cells,especially after treatment with 10 lg/ml of MPEE, were shown tobe retracted and rounding in shape with some cells detached fromthe surface and floating in the media.

Significant decreases in cell viability were observed after treat-ment with the positive controls of 5-FU and DTIC (data not shown).The 50% effective concentration (EC50) values are summarized inTable 3. EC50 value for DTIC treatment could not be obtained forSK-MEL-28 cell line due to potential solvent effects (DTIC was dis-solved in 0.02 N acetic acid). The maximum concentration possibleto use without significantly altering the cell culture medium pHwas 100 lg/ml. At this concentration of DTIC, 25% cell killing wasobserved on SK-MEL-28 cell lines, which was consistent with aprevious study (Lillehammer et al., 2007). Additionally, 5-FU andDTIC showed similar toxicity on both cancer and normal cells asevidenced by similar EC50 values on both cancer and normal celllines.

3.3. Cell cycle arrest induced by MPEE

Treatment of A-431 cells with MPEE at 10 lg/ml resulted in asignificant increase in the sub G1 peak (14%; P < 0.05) with a con-comitant decrease in G1 phase compared to the untreated control(2%) (Fig. 3a). Treatment of SK-MEL-28 at the same concentrationresulted in a significant increase in cells in the G1 phase (93%;P < 0.05) compared to the untreated control (77%) (Fig. 3b).

3.4. Apoptosis induction by MPEE

Apoptotic cells were measured as early and late apoptotic cells(Fig. 4). Treatment of A-431 cells with MPEE induced a dose-dependent induction of apoptosis at the early and late stages ofapoptosis. The percentage of early apoptotic cells after treatmentwith MPEE significantly increased from 2% (untreated control) to18% (10 lg/ml of MPEE) (P < 0.01; Fig. 4a). Although of a smallerfold change, a significant increase in early apoptosis was also ob-served after treatment with 10 lg/ml of MPEE in SK-MEL-28 cells(P < 0.05; Fig. 4b). No significant increase in late apoptosis wasfound after treatment with MPEE on these two skin cancer celllines.

3.5. Caspase activation by MPEE

Significant 15-, 3- and 4-fold increases of activities of Caspase 3/7, 8, 9, respectively, were observed after treatment of A-431 cellswith MPEE (P < 0.01, Fig. 5a). No significant increase of activitiesfor caspase 8 or 9 was observed for SK-MEL-28 cells, althoughcaspase 3/7 did increase (P < 0.05, Fig. 5b).

3.6. Disruption of mitochondrial membrane potential (Dwm) by MPEE

Significant increases in DWm were observed for both skin can-cer cell lines. Treatment with 10 lg/ml of MPEE for 48 h signifi-cantly increased% cells positive for JC-1 monomers from 9% inuntreated cells to 44% on both cell lines (Fig. 6) (P < 0.05).

3.7. Effect of MPEE on the expression of cell cycle and apoptosis-relatedgenes

As determined by qRT-PCR, MPEE was found to inhibit prolifer-ation of the skin cancer cells by altering the expression of the cellcycle and apoptosis-related genes (Fig. 7). Compared with the un-treated control, MPEE at 10 lg/ml induced a 12.5- and 3.5-fold in-crease in the mRNA expression of cytochrome c in A-431 (P < 0.05)and SK-MEL-28 cells (P < 0.01), respectively (Fig. 7a and c). In A-431 cells, treatment with MPEE at 10 lg/ml induced a 5.4-fold de-crease in the mRNA expression of Akt1 (Fig. 7a; P < 0.01), and a 2.4-fold increase in the ratio of Bax/Bcl-2 (Fig. 7b; P < 0.05) comparedwith the untreated control. However, mRNA expression of thesethree genes was not significantly altered by MPEE in SK-MEL-28cells. In SK-MEL-28 cells, a 10.6-fold increase in mRNA level ofp21WAF1 was found after treatment with MPEE at 10 lg/ml(Fig. 7c; P < 0.05), while that of cyclin D1 remained unchanged(Fig. 7c). No significant changes in the mRNA levels of NFjB andIjBa were found after treatment of MPEE on A-431 or SK-MEL-28 cell lines (Fig. 7a and c).

4. Discussion

Antioxidants play an important role in protection againstreactive oxygen species to maintain human health, including theprevention of cancer progression (Rajkumar et al., 2011a;Fridovich, 1998). There are strong correlations between total phe-nolic compounds and antioxidant activity in different kinds offruits (Sellappan et al., 2002; Gorinstein et al., 2004). Under theextraction condition in this study, the total phenolics content inMPEE was 64.3 mg GAE/g dry pericarp weight. Patthamakanokp-orn et al. (2008) reported phenolics content for mangosteen fruitextract at only 4.2 mg GAE/g dry fruit weight. This differencemay be explained firstly, the mangosteen extract was from thewhole fruit (including flesh and pericarp) in Patthamakanokporn’sstudy, while the mangosteen extract was from the pericarp in ourstudy. In vegetables and fruits, the external and other lignocellu-losic parts (skins and seeds) present higher phenolic compoundsthan their corresponding inner part (Cruz et al., 2004). The totalphenolics level in the fruit flesh may be lower than in the pericarp.We found the total phenolics level in the fruit flesh was 29.2 mgGAE/g dry flesh weight (data not shown). Also, in Patthamakanokp-orn’s study, the extraction was carried out with 40% ethanol for 1 hat RT, compared to the optimal extraction temperature in our studyof 75 �C. Our study showed the total phenolics level was 16.4 mgGAE/g DW when the extraction was carried out with 50% ethanolfor 24 h at room temperature (data not shown). Additionally, the

(a) A-431

0

20

40

60

80

100

120

Cel

l Via

bilit

y (%

)

24hr 48hr 72hr

Concentration (µg/ml)

*

** **

**

****

**

** **

(b) SK-MEL-28

0

20

40

60

80

100

120

0 1%E 1.25 2.5 5 10 20

0 1%E 1.25 2.5 5 10 20

Cel

l Via

bilit

y(%

)

24hr 48hr 72hr

Concentration (µg/ml)

* ** **

****

** ** ****

(c) CCD-1064Sk and HaCaT

0

20

40

60

80

100

120

0 1%E 1.25 2.5 5 10 20

Cel

l Via

bilit

y (%

)

CCD-1064Sk HaCaT

Concentration (µg/ml)

*

*

**

**

Fig. 1. Cell viability determined by the crystal violet assay after treatment withmangosteen pericarp ethanol extract (MPEE) for 24, 48, and 72 h of (a) A-431 and(b) SK-MEL-28 cells and for 48 h of (c) CCD-1064Sk and HaCaT. One percent ofethanol (1% E; v/v) was used as vehicle control. Data are shown as% viabilitycompared to the untreated control and are presented as the mean ± SEM of threeindependent experiments. Treatments significantly different from the untreatedcontrol at P < 0.05 are presented as ⁄ and at P < 0.01 as ⁄⁄.

3008 J.J. Wang et al. / Food and Chemical Toxicology 50 (2012) 3004–3013

yield of MPEE was 23.2% (w/w) in this current study. A previousstudy reported that the yield of extracts from the mangosteen peri-carp was 2.0%, 6.1%, 6.2% and 22.3% (w/w) when extraction wascarried out at RT for 7 days with 50% ethanol, 95% ethanol, ethylacetate and water, respectively (Weecharangsan et al., 2006).Therefore, higher total phenolics recovery was obtained whenextraction was carried out at higher temperature in the samesolvent system. This may be because the increase in temperature

accelerates mass transfer and improves the extraction yield (Wanget al., 2008). The current result was in agreement with the studyfrom Le Floch et al. (1998) who reported that increased total phen-olics amount and extraction efficiency with increasing tempera-ture was observed from the extraction of olive leaves.

Mangosteen displays a high level of antioxidant activities, com-pared with other fruit and vegetables (Leong and Shui, 2002;Chomnawang et al., 2007). The antioxidant activities of differentmangosteen extracts and its major compounds have been reviewed(Akao et al., 2008; Pedraza-Chaverri et al., 2008; Obolskiy et al.,2009). In the current study, the antioxidant properties of MPEEwere evaluated using three different antioxidant assays (FRAP,ORAC, and DPPH) based on different principles. FRAP assay mea-sures the total antioxidant concentration based on the total reduc-ing capacity of electron-donating antioxidants (Benzie and Strain,1996). ORAC assay measures the effectiveness of the antioxidantsin a sample which reflects the antioxidant activity of a particularsample against peroxyl radicals (Huang et al., 2005). DPPH assaymeasures the ability to scavenge DPPH free radicals (Blois, 1958).Therefore, different antioxidant levels can be expected from thesethree methods. Current results showed that MPEE had similarscavenging activity of DPPH radical with the green tea extract,but lower activity than green tea by the FRAP and ORAC assays.The lower ORAC value exhibited by MPEE might be due to itslow solubility in the phosphate buffer used in the assay, sinceMPEE was extracted using absolute ethanol. It is likely that notall DPPH-scavenging antioxidants in MPEE are reducing agentsand peroxyl radicals-scavenging agents as detected by the FRAPand ORAC assay. Consistent with our study, Chomnawang et al.(2007) reported similar DPPH IC50 value (6.1 lg/ml) from the man-gosteen extract. However, another study reported the DPPH IC50

values of 30.8, 58.5, 77.8 and 35.0 lg/ml of different mangosteenpericarp extracts using 50% ethanol, 95% ethanol, ethyl acetateand water, respectively (Weecharangsan et al., 2006). A previousstudy (Patthamakanokporn et al., 2008) reported values for man-gosteen extract of ORAC and FRAP values at 122.7 and 31.8 lMTE/g DW. Compared to these previous studies, MPEE from ourstudy displayed higher yield (23.2%) and possessed higher antiox-idant potential (see Table 2), indicating the current extractionmethod was more efficient as more total phenolics were recovered.This may be due to the use of different solvents and different tem-peratures resulting in extraction of different amounts and propor-tions of active compounds (Wang et al., 2008; Rajkumar et al.,2011b).

In addition to the antioxidant effect, MPEE also exhibited cyto-toxic effects on two skin cancer cell lines (A-431 and SK-MEL-28)tested in the present study. According to the American NationalCancer Institute, the EC50 value of a crude extract should be lessthan 30 lg/ml to be considered as promising for further purifica-tion (Stuffness and Pezzuto, 1990). It is noteworthy that the cur-rent EC50 values obtained for MPEE were within this threshold onboth tested cell lines, and hence can be considered as a potentialanti-skin cancer drug candidate. The result is consistent with aprevious study in which a crude methanolic extract from themangosteen pericarp exhibited an anti-proliferative effect on ahuman breast cancer (SKBR3) cell line with an EC50 value of9.25 lg/ml (Moongkarndi et al., 2004a). However, different celllines showed different sensitivities to each specific mangosteenpericarp extract, as evidenced by different EC50 values. For exam-ple, Chiang et al. (2004) reported that an aqueous extract of man-gosteen pericarp showed cytotoxicity on leukemia K562 and Rajicells with EC50 values of 61.0 and 152.9 lg/ml, respectively. Like-wise, another study conducted by Watanapokasin et al. (2010) re-ported that different EC50 values were obtained after treatmentwith mangosteen pericarp ethyl acetate extract on four differentcolon cancer cell lines: 7.5 lg/ml for COLO 205 cells, 17.7 lg/ml

Fig. 2. Morphological alterations of A-431 and SK-MEL-28 cells after 48-h treatment with MPEE was observed by phase contrast inverted microscope (magnification 100�).Scale bar is 100 lm. The images shown are from a representative experiment of three replicates with similar results.

Table 3EC50 values (lg/ml) based on the results obtained using the crystal violet assay as perthe methods section.

Cell lines MPEE Commercial drugs (48 h)

24 h 48 h 72 h

A-431 6.66 5.07 5.29 1.28 (5-FU)SK-MEL-28 6.89 6.84 5.82 >100 (DTIC)CCD-1064Sk ND 12.62 ND 1.22 (5-FU); >100 (DTIC)HaCaT ND 8.32 ND 0.31 (5-FU); 50 (DTIC)

ND, not determined; 5-FU, 5-fluorouracil; DTIC, dacabazine.

(a) A-431

0

20

40

60

80

100

0 2.5 5 10

% C

ell D

istri

butio

n

Sub G1 G0/G1 S G2/M

**

Concentration (µg/ml)

(b) SK-MEL-28

0

20

40

60

80

100

0 2.5 5 10

% C

ell D

istri

butio

n

SubG1 G0/G1 S G2/M

Concentration (µg/ml)

**

Fig. 3. Effect of MPEE on cell cycle progression after 48-h treatment of (a) A-431and (b) SK-MEL-28 cells was determined by propidium iodide staining and analyzedfor DNA content by flow cytometry. Data were obtained from 20,000 events andpresented as the percentage of cells in the sub G1, G0/G1, S and G2/M phases. Thevalues are shown as the mean ± SEM of five independent experiments. Treatmentssignificantly different from the untreated control at P < 0.05 are presented as ⁄ andat P < 0.01 as ⁄⁄.

J.J. Wang et al. / Food and Chemical Toxicology 50 (2012) 3004–3013 3009

for CX-1 cells, 10.0 lg/ml for MIP-101 cells, and 16.1 lg/ml forSW620 cells. Furthermore, an EC50 value (lg/ml) of 15.45 was re-ported after treatment of human breast cancer SKBR3 with man-gosteen pericarp ethanol extract (Moongkarndi et al., 2004b). Thisvalue was 2- to 3-folds greater than the ones in SK-MEL-28 andA-431 cells from our study, indicating mangosteen pericarp etha-nol extract was more effective on skin cancer cell lines thanbreast cancer cell line. Different responses of different types ofcells to each specific xanthone compound were summarised byShan et al. (2011). For example, the EC50 values of a-mangostinranged from 0.7 to 19.6 lg/ml on a range of cell lines (Laphook-hieo et al., 2006; Shan et al., 2011; Watanapokasin et al., 2011).These results therefore provide important information for thespecific anti-cancer activity of mangosteen pericarp extract intwo types of human skin cancer cells. Furthermore, our previousstudy reported that the EC50 values of a-mangostin, c-mangostin,and 8-desoxygartanin on SK-MEL-28 cell line were 5.92, 3.55,3.83 lg/ml, respectively (Wang et al., 2011). These values aresimilar to that of MPEE in the current study (6.84 lg/ml). Theseresults indicate that xanthones are likely to be the principleactive compounds in MPEE, given that the total of majorxanthone content (a-mangostin, b-mangostin, c-mangostin,8-desoxygartanin, gartanin, and 9-hydroxycalabaxanthone) inthe extract is approximately 44% (Shan and Zhang, 2010).

The EC50 value of MPEE to A-431 cell line was similar to that of5-FU (1.28 lg/ml) but for SK-MEL-28 cell line it was much lowerthan that of DTIC (EC25 100 lg/ml). Importantly, MPEE was lesstoxic to human normal skin cell lines of CCD-1064Sk and HaCaTthan to the skin cancer lines. In contrast, the toxicity of 5-FU andDTIC on human normal skin cells was equal to or greater than thatto skin cancer cells, as evidenced by the EC50 values summarized inTable 3. These results clearly support MPEE as an excellent candi-date for skin cancer chemotherapy.

Cell cycle deregulation is an important feature of most ofcancers and therefore control of cell cycle progression is suggestedto be an effective strategy to inhibit cancer growth (Grana andReddy, 1995; Pavletich, 1999). Treatment of SK-MEL-28 with MPEEinduced significant cell cycle arrest in G1 phase (P < 0.01).Treatment of A-431 cells significantly increased the sub G1 peak

(a) A-431

0

5

10

15

20

25

0 2.5 5 10

% C

ell P

opul

atio

n

Early apoptosisLate apoptosis/Necrosis

Concentration (µg/ml)

**

(b) SK-MEL-28

0

5

10

15

20

25

0 2.5 5 10

% C

ell P

opul

atio

n

Early apoptosisLate apoptosis/Necrosis

Concentration (µg/ml)

*

Fig. 4. Apoptotic effect of MPEE on (a) A-431 and (b) SK-MEL-28 cells after 48-htreatment was determined by Annexin V-conjugated propidium iodide stainingthrough flow cytometry. Data were obtained from 20,000 events and presented asthe percentage of early apoptotic cells (Annexin positive) and late apoptotic/necrotic cells (Annexin positive/PI positive). The values are shown as the mean ± -SEM of five independent experiments. Treatments significantly different from theuntreated control at P < 0.05 are presented as ⁄ and at P < 0.01 as ⁄⁄.

(a) A-431

0

5

10

15

20

25

30

35

40

0150

RLU

/Cel

l/Min

Caspase 3/7

Caspase 8

Caspase 9

Concentration (µg/ml)

**

****

(b) SK-MEL-28

0

2

4

6

8

10

12

14

16

18

0150

RLU

/Cel

l/Min

Caspase 3/7

Caspase 8

Caspase 9

Concentration (µg/ml)

*

Fig. 5. Caspase 3/7, 8 and 9 activity was determined on (a) A-431and (b) SK-MEL-28cells after 48-h treatment with MPEE using luminescent kits as described in themethod. Data were calculated as relative luminescence unit (RLU)/cell/min. Thevalues are shown as the mean ± SEM of three independent experiments. Treatmentssignificantly different from the untreated control at P < 0.05 are presented as ⁄ andat P < 0.01 as ⁄⁄.

0

10

20

30

40

50

60

0 2.5 5 10

% C

ells

Pos

itive

for J

C-1

Mon

omer

s

A-431

SK-MEL-28

Concentration (µg/ml)

* *

Fig. 6. Loss of mitochondrial membrane potential was examined by JC-1 dyethrough flow cytometry after 48-h treatment of A-431 and SK-MEL-28 cells withMPEE. Data were obtained from 20,000 events and presented as the percentage ofcells positive for JC-1 monomer. The values are shown as the mean ± SEM of fiveindependent experiments. Treatments significantly different from the untreatedcontrol at P < 0.05 are presented as ⁄ and at P < 0.01 as ⁄⁄.

3010 J.J. Wang et al. / Food and Chemical Toxicology 50 (2012) 3004–3013

(hypodiploid debris) (P < 0.01) with a concomitant decrease in G1

phase, indicating induction of apoptosis. The current study wasconsistent with our previous study which showed the major com-pounds in the pericarp of mangosteen induced G1 phase arrest (byc-mangostin and 8-deoxygartanine) and sub G1 phase arrest (bya-mangostin) on SK-MEL-28 cells (Wang et al., 2011). In contrast,Matsumoto et al. (2005) reported S phase arrest by c-mangostin,although they found G1 phase arrest by a-mangostin andb-mangostin in human colon cancer DLD-1 cells. This differencemay be due to the difference in response of different cell types.The mRNA expression levels of cell cycle regulating genes at theG1/S boundary (Cyclin D1) and the universal cyclin-dependentkinase inhibitor (CDKI) (p21WAF1) were determined by qRT-PCRto understand the molecular mechanisms. A significant increasein the expression of p21WAF1 (P < 0.01) was observed after treat-ment of SK-MEL-28 with MPEE. The MPEE-induced increase inp21WAF1 could be p53-independent, because SK-MEL-28 cellspossess mutant p53 (Soengas et al., 2001). However, only a slightdecrease in the expression of cyclin D1 was observed after treat-ment with 10 lg/ml of MPEE. The expression of the other cyclins(e.g. A/E) and cyclin-dependent kinases (CDK) (e.g. CDK2/4/6)which regulate G1/S transition can be studied in the future tofurther elucidate this mechanism.

Mangosteen compounds can mediate anti-cancer properties byinduction of apoptosis on some cancer types (Akao et al., 2008;

Pedraza-Chaverri et al., 2008; Obolskiy et al., 2009). In the currentstudy, MPEE induced significant apoptosis on the two skin cancer

(a) A-431

0

2

4

6

8

10

12

14

16

18

Akt1 Cytochrome C NFκB IκBα

Rel

ativ

e G

ene

Expr

essi

on(A

rbitr

ary

Uni

ts)

Genes of Interest

0 µg/ml5 µg/ml10 µg/ml

** **

*

(b) A-431

0

0.5

1

1.5

2

2.5

3

0150

Bax

/Bcl

-2 R

atio

(Arb

itrar

y U

nits

)

Concentration (µg/ml)

*

0

2

4

6

8

10

12

14

16

Akt1 Cytochrome C Cyclin D1 P21WAF1 NFκB IκBα

Rel

ativ

e G

ene

Expr

essi

on(a

rbitr

ary

units

)

Genes of Interest

0 µg/ml5 µg/ml10 µg/ml

*

**

*

(c) SK-MEL-28

Fig. 7. Effect of MPEE on target gene expressions on A-431 (a and b) and SK-MEL-28(c) cells after 48-h treatment were determined by qRT-PCR. The values are shown asrelative gene expression compared to the untreated control and are presented asthe mean ± SEM of three independent experiments. Treatments significantlydifferent from the untreated control at P < 0.05 are presented as ⁄ and at P < 0.01as ⁄⁄.

J.J. Wang et al. / Food and Chemical Toxicology 50 (2012) 3004–3013 3011

cell lines, most markedly for A-431 (P < 0.01). To understand themechanisms of MPEE-induced apoptosis, two distinct apoptoticpathways were studied: the extrinsic pathway involving the death

receptor signaling and the intrinsic pathway involving the mito-chondrial cascades (Henry-Mowatt et al., 2004). Activation of cas-pase-3/7 (effector caspases) is involved in both pathways, whilecaspase 8 and 9 (initiator caspases) are involved in extrinsic andintrinsic pathway, respectively (Ashkenazi and Dixit, 1998; Greenand Amarante-Mendes, 1998). MPEE significantly increased thelevels of caspase 3/7, 8 and 9 enzyme activities in A-431 cells(P < 0.01). However, only caspase 3/7 activation was observed forSK-MEL-28 (P < 0.05). In the future, caspase inhibitors could beused to confirm the activation of caspases. Also, Western blotanalysis could be used to study the levels and active forms of thecaspase proteins. Significant loss of mitochondrial membranepotential was also observed after treatment of both skin cancer celllines with MPEE (P < 0.01). The results support that the apoptosisinduced by MPEE occurs through both the extrinsic and intrinsicpathways on A-431 cells, but only the intrinsic pathway onSK-MEL-28 cells. Different responses to xanthones by different celllines have also been reported by Matsumoto et al., who founda-mangostin induced caspase-independent apoptosis via the mito-chondrial pathway in colon cancer cells (Matsumoto et al., 2005).In contrast, a-mangostin induced apoptosis via caspase 3 and -9activation and disruption of mitochondrial membrane potentialin human leukemia HL60 cells (Matsumoto et al., 2004). Addition-ally, gaudichaudione A, a cytotoxic xanthone, induced mitochon-drial destabilization (Wu et al., 2002). Therefore, althoughxanthones induced apoptosis via different pathways in differentcancer cell types, current studies support that xanthones preferen-tially target the mitochondria for apoptosis induction.

To understand the molecular mechanism of apoptosis inducedby MPEE, the genes involved in apoptosis were analyzed by theqRT-PCR. Release of mitochondrial cytochrome c is an importantsignaling event in the intrinsic apoptotic activation pathway. Asignificant increase in the mRNA expression of cytochrome c wasobserved for both skin cancer cell lines (P < 0.01). Bcl-2 familyare key regulators of the intrinsic pathway, which either suppressor promote changes in mitochondrial membrane permeabilityrequired for release of cytochrome c (Green and Reed, 1998; Grosset al., 1999). Overexpression of anti-apoptotic Bcl-2 probablyoccurs in most of cancers (Amundson et al., 2000) and a low ratioof pro-apoptotic Bax and anti-apoptotic Bcl-2 implies apoptosisresistance (Raisova et al., 2001). Treatment with MPEE enhancedthe ratio of Bax/Bcl-2 in A-431 cells. This supports that the releaseof cytochrome c from the mitochondria into the cytosol by MPEEresults from down-regulation of Bcl-2 and/or upregulation of Baxin A-431 cells. However, the Bax/Bcl-2 ratio was not affected byMPEE in SK-MEL-28, indicating these two genes are not involvedin MPEE-induced apoptosis in this cell line. This is in agreementwith findings of Matsumoto et al. (2004). Further study on othergenes of Bcl-2 family (e.g. Puma, Bad, Mcl-1, etc.) is required tounderstand the mechanisms.

In addition to apoptosis induction, the inhibition of the survivalpathway is another important approach to overcome cancer. Akt, aserine/threonine protein kinase, is involved in cell survival andgrowth, and is constitutively activated through multiple mecha-nisms in most of cancers (Pap and Cooper, 1998). Among its threeisoforms (Akt1, 2, and 3), Akt1 plays a key role in maintaining cellsurvival (Chen et al., 2001; Nicholson and Anderson, 2002; Gonz-alez and McGraw, 2009). Therefore, Akt1 pathway can be animportant molecular target for cancer prevention and treatment(Crowell et al., 2007). MPEE significantly decreased the mRNA levelof Akt1 in A-431 cells (P < 0.01). Inhibition of Akt1 was also ob-served on SK-MEL-28 cells, although not statistically significant.The results indicate that MPEE inhibited proliferation of skincancer cells by down-regulating the activities of Akt and furtherstudies are needed to elucidate which steps in the survival path-way are involved.

3012 J.J. Wang et al. / Food and Chemical Toxicology 50 (2012) 3004–3013

In conclusion, the results demonstrated that MPEE from peri-carp of mangosteen exhibits a potent antioxidant activity andanti-cancer activity on these two distinctive types of skin cancercell lines. The anti-cancer activity was specific for the tested skincancer cells and was through apoptosis induction via intrinsicand /or extrinsic pathways and proliferation inhibition via downregulation of Akt pathway. The crude extract is likely to have im-proved therapeutic applications compared to existing drugs dueto selectivity resulting in more tumour cell death and less side-ef-fects. To better understand the selectivity of MPEE on these twospecific skin cancer cells, the primary counterparts of these cells(melanocytes and keratinocytes) can be used in future study. Also,further study could be carried out using more cancer cell lines withdifferent genetic background to understand more about the anti-skin cancer activity of MPEE. A long term goal is to study thein vivo therapeutic efficacy of MPEE.

Conflict of Interest

The authors declare that there are no conflict of interest.

Role of funding source

The study sponsors did not have any involvement in the studydesign, nor in the collection, analysis and interpretation of data;in the writing of the manuscript; nor in the decision to submitthe manuscript for publication.

Acknowledgement

Jing J. Wang gratefully acknowledges Flinders University forawarding an EIPRS scholarship. The research was supported byKlein Research Institute Ltd., Adelaide, Australia and FlindersUniversity Underwriting Grant. The authors are thankful toMrs. Sheree Bailey and Mr. Eugene Ng from the Department ofImmunology, Allergy and Arthritis, Flinders University, Adelaide,Australia, for their helpful assistance with Flow Cytometry; toGeorge Mayne from the Department of Surgery, Flinders Univer-sity, Adelaide, Australia, for his kind help with qRT-PCR methodsand analysis.

References

Akao, Y., Nakagawa, Y., Iinuma, M., Nozawa, Y., 2008. Anti-cancer effects ofxanthones from pericarps of mangosteen. Int. J. Mol. Sci. 9, 355–370.

Amundson, S.A., Myers, T.G., Scudiero, D., Kitada, S., Reed, J.C., Fornace Jr., A.J., 2000.An informatics approach identifying markers of chemosensitivity in humancancer cell lines. Cancer Res. 60, 6101–6110.

Ashkenazi, A., Dixit, V.M., 1998. Death receptors: signaling and modulation. Science281, 1305–1308.

Benzie, I.F., Strain, J.J., 1996. The ferric reducing ability of plasma (FRAP) as ameasure of ‘‘antioxidant power’’: the FRAP assay. Anal. Biochem. 239, 70–76.

Blois, M.S., 1958. Antioxidant determinations by the use of a stable free radical.Nature 181, 1199–1200.

Chen, S.X., Wan, M., Loh, B.N., 1996. Active constituents against HIV-1 protease fromGarcinia mangostana. Planta Med. 62, 381–382.

Chen, W.S., Xu, P.Z., Gottlob, K., Chen, M.L., Sokol, K., Shiyanova, T., Roninson, I.,Weng, W., Suzuki, R., Tobe, K., Kadowaki, T., Hay, N., 2001. Growth retardationand increased apoptosis in mice with homozygous disruption of the Akt1 gene.Genes Dev. 15, 2203–2208.

Chiang, L.C., Cheng, H.Y., Liu, M.C., Chiang, W., Lin, C.C., 2004. In vitro evaluation ofantileukemic activity of 17 commonly used fruits and vegetables in Taiwan.Food Sci. Technol. 37, 539–544.

Chomnawang, M.T., Surassmo, S., Nukoolkarn, V.S., Gritsanapan, W., 2007. Effect ofGarcinia mangostana on inflammation caused by Propionibacterium acnes.Fitoterapia 78, 401–408.

Cragg, G.M., Newman, D.J., 2000. Antineoplastic agents from natural sources:achievements and future directions. Expert Opin. Investig. Drugs 9, 2783–2797.

Crowell, J.A., Steele, V.E., Fay, J.R., 2007. Targeting the AKT protein kinase for cancerchemoprevention. Mol. Cancer Ther. 6, 2139–2148.

Cruz, J.M., Dominguez, H., Parajo, J.C., 2004. Assessment of the production ofantioxidants from winemaking waste solids. J. Agric. Food Chem. 52,5612–5620.

Doan, P.S., 2008. Primary cutaneous malignant melanoma. Aviat. Space Environ.Med. 79, 919–921.

Fridovich, I., 1998. Oxygen toxicity: a radical explanation. J. Exp. Biol. 201, 1203–1209.Garbe, C., Eigentler, T.K., 2007. Diagnosis and treatment of cutaneous melanoma:

state of the art 2006. Melanoma Res. 17, 117–127.Gonzalez, E., McGraw, T.E., 2009. The Akt kinases: isoform specificity in metabolism

and cancer. Cell Cycle 8, 2502–2508.Gordaliza, M., 2007. Natural products as leads to anticancer drugs. Clin. Transl.

Oncol. 9, 767–776.Gorinstein, S., Zachwieja, Z., Katrich, E., Pawelzik, E., Haruenkit, R., Trakhtenberg, S.,

Martin-Belloso, O., 2004. Comparison of the contents of the main antioxidantcompounds and the antioxidant activity of white grapefruit and his new hybrid.LWT – Food Sci. Technol. 37, 337–343.

Grana, X., Reddy, E.P., 1995. Cell cycle control in mammalian cells: role of cyclins,cyclin dependent kinases (CDKs), growth suppressor genes and cyclin-dependent kinase inhibitors (CKIs). Oncogene 11, 211–219.

Green, D.R., Amarante-Mendes, G.P., 1998. The point of no return: mitochondria,caspases, and the commitment to cell death. Results Probl. Cell Differ. 24, 45–61.

Green, D.R., Reed, J.C., 1998. Mitochondria and apoptosis. Science 281, 1309–1312.Gross, A., McDonnell, J.M., Korsmeyer, S.J., 1999. BCL-2 family members and the

mitochondria in apoptosis. Genes Dev. 13, 1899–1911.Henry-Mowatt, J., Dive, C., Martinou, J.C., James, D., 2004. Role of mitochondrial

membrane permeabilization in apoptosis and cancer. Oncogene 23, 2850–2860.Huang, D., Ou, B., Prior, R.L., 2005. The chemistry behind antioxidant capacity

assays. J. Agric. Food Chem. 53, 1841–1856.Johnson, T.M., Dolan, O.M., Hamilton, T.A., Lu, M.C., Swanson, N.A., Lowe, L., 1998.

Clinical and histologic trends of melanoma. J. Am. Acad. Dermatol. 38, 681–686.Laphookhieo, S., Syers, J.K., Kiattansakul, R., Chantrapromma, K., 2006. Cytotoxic and

antimalarial prenaylated xanthones from Cratoxylum cochinchinense. Chem.Pharm. Bull (Tokyo) 54, 745–747.

Le Floch, F., Tena, M.T., Rios, A., Valcarcel, M., 1998. Supercritical fluid extraction ofphenol compounds from olive leaves. Talanta 46, 1123–1130.

Leong, L.P., Shui, G., 2002. An investigation of antioxidant capacity of fruits inSingapore markets. Food Chem. 76, 69–75.

Lillehammer, T., Engesaeter, B.O., Prasmickaite, L., Maelandsmo, G.M., Fodstad, O.,Engebraaten, O., 2007. Combined treatment with Ad-hTRAIL and DTIC or SAHAis associated with increased mitochondrial-mediated apoptosis in humanmelanoma cell lines. J. Gene Med. 9, 440–451.

Mahabusarakam, W., Wiriyachitra, P., Taylor, W., 1987. Chemical constituents ofGarcinia mangostana. J. Nat. Prod. 50, 474–478.

Matsumoto, K., Akao, Y., Ohguchi, K., Ito, T., Tanaka, T., Iinuma, M., Nozawa, Y., 2005.Xanthones induce cell-cycle arrest and apoptosis in human colon cancer DLD-1cells. Bioorg. Med. Chem. 13, 6064–6069.

Matsumoto, K., Akao, Y., Yi, H., Ohguchi, K., Ito, T., Tanaka, T., Kobayashi, E., Iinuma,M., Nozawa, Y., 2004. Preferential target is mitochondria in alpha-mangostin-induced apoptosis in human leukemia HL60 cells. Bioorg. Med. Chem. 12,5799–5806.

Miller, D.L., Weinstock, M.A., 1994. Nonmelanoma skin cancer in the United States:incidence. J. Am. Acad. Dermatol. 30, 774–778.

Moongkarndi, P., Kosem, N., Kaslungka, S., Luanratana, O., Pongpan, N., Neungton,N., 2004a. Antiproliferation, antioxidation and induction of apoptosis byGarcinia mangostana (mangosteen) on SKBR3 human breast cancer cell line. J.Ethnopharmacol. 90, 161–166.

Moongkarndi, P., Kosem, N., Luanratana, O., Jongsomboonkusol, S., Pongpan, N.,2004b. Antiproliferative activity of Thai medicinal plant extracts on humanbreast adenocarcinoma cell line. Fitoterapia 75, 375–377.

Nakatani, K., Atsumi, M., Arakawa, T., Oosawa, K., Shimura, S., Nakahata, N.,Ohizumi, Y., 2002. Inhibitions of histamine release and prostaglandin E2synthesis by mangosteen, a Thai medicinal plant. Biol. Pharm. Bull. 25,1137–1141.

Nguyen, L.H., Venkatraman, G., Sim, K.Y., Harrison, L.J., 2005. Xanthones andbenzophenones from Garcinia griffithii and Garcinia mangostana.Phytochemistry 66, 1718–1723.

Nicholson, K.M., Anderson, N.G., 2002. The protein kinase B/Akt signalling pathwayin human malignancy. Cell Signal. 14, 381–395.

Nuijen, B., Bouma, M., Manada, C., Jimeno, J.M., Schellens, J.H., Bult, A., Beijnen, J.H.,2000. Pharmaceutical development of anticancer agents derived from marinesources. Anticancer Drugs 11, 793–811.

Obolskiy, D., Pischel, I., Siriwatanametanon, N., Heinrich, M., 2009. Garciniamangostana L.: a phytochemical and pharmacological review. Phytother. Res.23, 1047–1065.

Pap, M., Cooper, G.M., 1998. Role of glycogen synthase kinase-3 in thephosphatidylinositol 3-Kinase/Akt cell survival pathway. J. Biol. Chem. 273,19929–19932.

Patthamakanokporn, O., Puwastien, P., Nitithamyong, A., Sirichakwal, P.P., 2008.Changes of antioxidant activity and total phenolic compounds during storage ofselected fruits. J. Food Compost. Anal. 21, 241–248.

Pavletich, N.P., 1999. Mechanisms of cyclin-dependent kinase regulation: structuresof Cdks, their cyclin activators, and Cip and INK4 inhibitors. J. Mol. Biol. 287,821–828.

Pedraza-Chaverri, J., Cardenas-Rodriguez, N., Orozco-Ibarra, M., Perez-Rojas, J.M.,2008. Medicinal properties of mangosteen (Garcinia mangostana). Food Chem.Toxicol. 46, 3227–3239.

J.J. Wang et al. / Food and Chemical Toxicology 50 (2012) 3004–3013 3013

Raisova, M., Hossini, A.M., Eberle, J., Riebeling, C., Wieder, T., Sturm, I., Daniel, P.T.,Orfanos, C.E., Geilen, C.C., 2001. The Bax/Bcl-2 ratio determines thesusceptibility of human melanoma cells to CD95/Fas-mediated apoptosis. J.Invest. Dermatol. 117, 333–340.

Rajkumar, V., Guha, G., Kumar, R.A., 2011a. Antioxidant and anti-cancer potentialsof Rheum emodi Rhizome extracts. Evid. Based Complement. Alternat. Med..http://dx.doi.org/10.1093/ecam/neq048. Article ID: 697986.

Rajkumar, V., Guha, G., Kumar, R.A., 2011b. Antioxidant and anti-neoplasticactivities of Picrorhiza kurroa extracts. Food Chem. Toxicol. 49, 363–369.

Sellappan, S., Akoh, C.C., Krewer, G., 2002. Phenolic compounds and antioxidantcapacity of Georgia-grown blueberries and blackberries. J. Agric. Food Chem. 50,2432–2438.

Shan, T., Ma, Q., Guo, K., Liu, J., Li, W., Wang, F., Wu, E., 2011. Xanthones frommangosteen extracts as natural chemopreventive agents: potential anticancerdrugs. Curr. Mol. Med. 11, 666–677.

Shan, Y., Zhang, W., 2010. Preparative separation of major xanthones frommangosteen pericarp using high-performance centrifugal partitionchromatography. J. Sep. Sci. 33, 1274–1278.

Simon, P., 2003. Q-Gene: processing quantitative real-time RT-PCR data.Bioinformatics 19, 1439–1440.

Singleton, V.L., Rossi, J.A., 1965. Colorimetry of total phenolics withphosphomolybdic–phostungstic acid reagents. Am. Soc. Enol. Viticult. 16,144–158.

Soengas, M.S., Capodieci, P., Polsky, D., Mora, J., Esteller, M., Opitz-Araya, X.,McCombie, R., Herman, J.G., Gerald, W.L., Lazebnik, Y.A., Cordon-Cardo, C., Lowe,S.W., 2001. Inactivation of the apoptosis effector Apaf-1 in malignantmelanoma. Nature 409, 207–211.

Stuffness, M., Pezzuto, J.M., 1990. Assays Related to Cancer Drug Discovery.Methods in Plant Biochemistry: Assays for Bioactivity. Academic Press, London.

Trakatelli, M., Ulrich, C., Del Marmol, V., Euvrard, S., Stockfleth, E., Abeni, D., 2007.Epidemiology of nonmelanoma skin cancer (NMSC) in Europe: accurate andcomparable data are needed for effective public health monitoring andinterventions. Br. J. Dermatol. 156, 1–7.

Tsao, H., Atkins, M.B., Sober, A.J., 2004. Management of cutaneous melanoma. N.Engl. J. Med. 351, 998–1012.

Vinokur, Y., Rodov, V., Reznick, N., Goldman, G., Horev, B., Umiel, N., Friedman, H.,2006. Rose petal tea as an antioxidant-rich beverage: cultivar effects. J. Food Sci.71, S42–S47.

Wang, J.J., Sanderson, B.J., Zhang, W., 2011. Cytotoxic effect of xanthones frompericarp of the tropical fruit mangosteen (Garcinia mangostana Linn.) on humanmelanoma cells. Food Chem. Toxicol. 49, 2385–2391.

Wang, L., Yang, B., Du, X., Yi, C., 2008. Optimisation of supercritical fluid extractionof flavonoids from Pueraria lobata. Food Chem. 108, 737–741.

Watanapokasin, R., Jarinthanan, F., Jerusalmi, A., Suksamrarn, S., Nakamura, Y.,Sukseree, S., Uthaisang-Tanethpongtamb, W., Ratananukul, P., Sano, T., 2010.Potential of xanthones from tropical fruit mangosteen as anti-cancer agents:caspase-dependent apoptosis induction in vitro and in mice. Appl. Biochem.Biotechnol. 162, 1080–1094.

Watanapokasin, R., Jarinthanan, F., Nakamura, Y., Sawasjirakij, N., Jaratrungtawee,A., Suksamrarn, S., 2011. Effects of alpha-mangostin on apoptosis induction ofhuman colon cancer. World J. Gastroenterol. 17, 2086–2095.

Weecharangsan, W., Opanasopit, P., Sukma, M., Ngawhirunpat, T., Sotanaphun, U.,Siripong, P., 2006. Antioxidative and neuroprotective activities of extracts fromthe fruit hull of mangosteen (Garcinia mangostana Linn.). Med. Princ. Pract. 15,281–287.

Wu, X., Cao, S., Goh, S., Hsu, A., Tan, B.K., 2002. Mitochondrial destabilisation andcaspase-3 activation are involved in the apoptosis of Jurkat cells induced bygaudichaudione A, a cytotoxic xanthone. Planta Med. 68, 198–203.