Phytomedicine 20 (2013) 999 1006

Contents lists available at SciVerse ScienceDirect

Phytomedicine

j ourna l h o mepage: www.elsev ier

Identication of potential anticancer compounds

Chong-Zhi Wanga, Zhiyu Zhanga, Wei-Hua Huangb, Guang-JiChunhao Yua, Rachael Nassa, Jing Zhaob, Wei Duc, Shao-Ping La Tang Center f of Chib State Key Lab ces, Uc Ben May Depd Committee on 37, US

a r t i c

Keywords:Oplopanax horPhytochemistrPolyynesOplopantriol AFalcarindiolAnticancerCell cycleApoptosisStructureactivity relationship

Nort cellsl stud3 co

steroids, and two phenolic acids, of which ve are novel compounds. In this study, we systemicallyevaluated the anticancer effects of compounds isolated from O. horridus. Their antiproliferative effectson a panel of human colorectal and breast cancer cells were determined using the MTS assay. Cell cycledistribution and apoptotic effects were analyzed by ow cytometry. The in vivo antitumor effect wasexamined using a xenograft tumor model. Among the 13 compounds, strong antiproliferative effects

Introductio

Cancer iand other Wment of cawith chemo(Buchholz 2only modercancer are r

Complemthat is gaini

Corresponment of AnestAvenue, MC 40fax: +1 773 83 CorresponMedicine, andMacau SAR, Ch

E-mail add(C.-S. Yuan).

0944-7113/$ http://dx.doi.owere observed from falcarindiol and a novel compound oplopantriol A. Falcarindiol showed the mostpotent antiproliferative effects, signicantly inducing pro-apoptosis and cell cycle arrest in the S andG2/M phases. The anticancer potential of falcarindiol was further veried in vivo, signicantly inhibitingHCT-116 tumor growth in an athymic nude mouse model at 15 mg/kg. We also analyzed the relationshipbetween polyyne structures and their pharmacological activities. We observed that both the terminalhydroxyl group and double bond obviously affected their anticancer potential. Results from this studysupplied valuable information for future semi-synthesis of polyyne derivatives to develop novel cancerchemopreventive agents.

2013 Elsevier GmbH. All rights reserved.

n

s a major public health problem in the United Statesestern countries (Siegel et al. 2012). Current treat-

ncer generally employs surgical resection combinedtherapy using cytotoxic drugs and radiation therapy009; Paoletti et al. 2010). Because these therapies areately successful, novel approaches for the treatment ofequired.entary and alternative medicine (CAM) is an approach

ng more attention for cancer management (Davis et al.

ding author at: Tang Center for Herbal Medicine Research, and Depart-hesia & Critical Care, University of Chicago, 5841 South Maryland28, Chicago, IL 60637, USA. Tel.: +1 773 702 1916;4 0601.ding author at: State Key Laboratory of Quality Research in Chinese

Institute of Chinese Medical Sciences, University of Macau, Taipa,ina. Tel.: +853 8397 4692; fax: +853 2884 1358.resses: SPLi@umac.mo (S.-P. Li), CYuan@dacc.uchicago.edu

2012; Wang et al. 2012a). The emergence of CAM represents anatural experiment of huge dimensions, as millions of Americanshave begun self-medicating with natural products (Bell 2010; Linand Chen 2012). Natural products have been valuable sources ofnew therapeutic candidate compounds (Hait and Hambley 2009;Xu et al. 2011); an analysis of the number of chemotherapeuticagents and their sources indicates that nearly 80% of approveddrugs are derived from natural compounds (Cragg et al. 2009). Withthe advent of new isolation and screening technologies, more activecompounds with enhanced anticancer properties could be found(Kharwar et al. 2011; Randhawa and Alghamdi 2011).

Oplopanax horridus (Sm.) Miq. (Araliaceae), or Devils club, is ashrub distributed around the Pacic Northwest of North America,from Alaska and the southwestern Yukon Territory down to Oregon,Idaho, and Montana (Calway et al. 2012). As an herbal medicine, O.horridus has a rich history of use by the Pacic indigenous peoplesfrom over 38 linguistic groups for the treatment of upwards of 34categories of medical conditions. These medical conditions rangefrom arthritis to cancer (Lantz et al. 2004).

Although recent pharmacological studies have observed that O.horridus possesses antidiabetic, antiviral, antibacterial, and cancer

see front matter 2013 Elsevier GmbH. All rights reserved.rg/10.1016/j.phymed.2013.04.013or Herbal Medicine Research, and Department of Anesthesia & Critical Care, Universityoratory of Quality Research in Chinese Medicine, and Institute of Chinese Medical Scienartment for Cancer Research, University of Chicago, Chicago, IL 60637, USA

Clinical Pharmacology and Pharmacogenomics, University of Chicago, Chicago, IL 606

l e i n f o

ridusy

a b s t r a c t

Oplopanax horridus is a plant native tohas antiproliferative effects on cancerthere has been limited phytochemicaWe recently isolated and identied 1.de /phymed

from Oplopanax horridus

an Dua, Xiao-Dong Wena, Tyler Calwaya,i b,, Chun-Su Yuana,d,

cago, Chicago, IL 60637, USAniversity of Macau, Macao, China

A

h America. Previous reports have demonstrated that this herb but study mostly focused on its extract or fractions. Becausey on this herb, its bioactive compounds are largely unknown.mpounds, including six polyynes, three sesquiterpenes, two

1000 C.-Z. Wang et al. / Phytomedicine 20 (2013) 999 1006

chemopreventive potential, chemical studies of this plant are verylimited, so the active compounds were not previously identied(McCutcheon et al. 1995; Tai et al. 2006). Recently, we conducteda phytochemical isolation of O. horridus root bark and obtained aseries of cobination oftwo polyynthree phencompounds

Regardinvious studiWang et alhorridus is lated the anta panel of hcell line. Wecompound,had the mointestine epcarindiol wusing an in

Materials a

Chemicals a

All solvegrade. Cell ware (FrankSwitzerland15, RPMI-1were obtainand streptoMO). The CeMTS assay kstaining bufDiego, CA). from BD Bio

Plant mater

Root barwas obtainby a botaniCenter for H

Extraction, c

Air-driedwith 80% etwith petrolcessively toseparated bafford comrated by D-silica gel, an13. Structurbination of hydrogen-heronuclear multiple bochemical m(2) oplopan1,11,16-trio(6) oplopanoplopanphe

-daucosterol, (11) -sitosterol, (12) ferulic acid, and (13) caffeicacid (Huang 2012) (Fig. 1).

Cell lines and cell culture

hum-48I-16) w

ssas,nted ied

lifer

poune eariouThe d to olifenufam wagente wm fro

absoent o

le an

HCT seco

witted fith

Tritod in

incualyzound, Od.

sis a

HCT secoddedm wrypsls) w, the

was andnalyeter. d.

xenog

ale is, INthe gmmmpounds. Their structures were elucidated by a com- spectroscopic analyses. Five of them, specically thees oplopantriols A and B (Huang et al. 2010) and theolic glycosides, oplopanphesides A, B, and C, are novel

(Huang et al. 2011).g the anticancer potential evaluation of O. horridus, pre-es focused on its extract or fractions (Tai et al. 2006;. 2010) so the effect of individual compounds from O.argely unknown. In this study, we systemically evalu-iproliferative activities of 13 isolated compounds usinguman colorectal cancer cell lines and a breast cancer

observed that two polyynes, one of which was a novel possessed signicant anticancer activity. Falcarindiolst potent effects while still being safe to normal smallithelial cells. The mechanisms behind the actions of fal-ere explored, and its antitumor potential was veriedvivo xenograft animal model.

nd methods

nd reagents

nts were of high-performance liquid chromatographyculture plasticware was obtained from Falcon Lab-lin Lakes, NJ) and Techno Plastic Products (Trasadingen,). Glutamine, insulin trypsin, McCoys 5A, Leibovitzs L-640 and DMEM media, and phosphate buffered salineed from Mediatech, Inc. (Herndon, VA, USA). Penicillinmycin were obtained from SigmaAldrich (St. Louis,llTiter 96 Aqueous Solution Cell Proliferation Assay, anit, was obtained from Promega (Madison, WI). PI/RNasefer was obtained from BD Biosciences Pharmingen (SanAn annexin V-FITC apoptosis detection kit was obtainedsciences (Rockville, MD).

ials

k of Oplopanax horridus (Sm.) Miq. from Oregon, USAed from Pacic Botanicals, LLC and was authenticatedst. The voucher specimens were deposited in the Tangerbal Medical Research at the University of Chicago.

ompound isolation and structural identication

, powdered root bark of O. horridus was extractedhanol under reux, suspended in water, then extractedeum ether (6090 C), ethyl acetate, and n-butanol suc-

give three fractions. The ethyl acetate fraction wasy silica gel, RP-C18 silica gel, and preparative HPLC topounds 16 and 11. The n-butanol fraction was sepa-101 macroporous resin, normal phase silica gel, RP-C18d preparative HPLC to afford compounds 710, 12, andes of isolated compounds were determined by a com-spectroscopic analyses, including IR, 1H and 13C NMR,ydrogen correlation spectroscopy (H-H COSY), het-multiple quantum coherence (HMQC), heteronuclearnd coherence (HMBC), mass spectroscopic data, andethods. The compounds are identied as (1) falcarindiol,diol, (3) (11S,16S,9Z)-9,17-octadecadiene-12,14-diyne-l,1-acetate, (4) oplopandiol acetate, (5) oplopantriol A,triol B (Huang et al. 2010); (7) oplopanpheside A, (8)side B, (9) oplopanpheside C (Huang et al. 2011); (10)

Theand SW7 (RPM(DMEM(Manaplemehumid

Cell pro

Comintestiday, vwells. exposeCell prthe mamediuMTS rethe plamediuand itsas perc

Cell cyc

TheOn thetreatedincubaxed w0.25% pendeRNase,and anson, MAshlancounte

Apopto

TheOn thewas amediu0.05% ting celgentlynatantV-FITCwere acytomcounte

In vivoimagin

Femanapolunder Use Coan colorectal cancer cell lines HCT-116 (McCoys 5A)0 (Leibovitzs L-15), human breast cancer cell line MCF-40), and rat small intestine epithelial cell line IEC-6ere purchased from American Type Culture Collection

VA, USA) and grown in the indicated media sup-with 10% FBS and 50 IU penicillin/streptomycin in a

atmosphere of 5% CO2 at 37 C.

ation analysis

nds were dissolved in DMSO. Cancer cells or rat smallpithelial cells were seeded in 96-well plates. After 1s concentrations of compounds were added to thenal concentration of DMSO was 0.5%. Controls werea culture medium containing 0.5% DMSO without drugs.ration was evaluated using an MTS assay according tocturers instructions. Briey, after 48 h treatment, theas replaced with 100 l of fresh medium and 20 l oft (CellTiter 96 Aqueous Solution) in each well, and thenas returned to the incubator for 12 h. A 60-l aliquot ofm each well was transferred to an ELISA 96-well platerbance at 490 nm was recorded. Results are expressedf control (0.5% DMSO control set at 100%).

alysis

-116 cells were seeded in 24-well tissue culture plates.nd day, the medium was changed and the cells wereh different concentrations of falcarindiol. The cells wereor 48 h before harvesting. Next, the cells were gently80% ethanol in a freezer for 2 h and then treated withn X-100 for 5 min in an ice bath. The cells were resus-300 l of PBS containing 40 g/ml PI and 0.1 mg/mlbated in a dark room for 20 min at room temperatureed using a FACScan ow cytometer (Becton Dickin-tain View, CA) and FlowJo 7.1.0 software (Tree Star,R). For each measurement, at least 10,000 cells were

ssay

-116 cells were seeded in 24-well tissue culture plates.nd day, the medium was changed and the falcarindiol. After treatment for 48 h, the cells oating in theere collected. The adherent cells were detached within. Then the culture medium containing FBS (and oat-as added to inactivate the trypsin. After being pipettedcells were centrifuged for 5 min at 1500 g. The super-

removed and the cells were stained with annexin PI according to the manufacturers instructions. Cellszed immediately after staining using a FACScan owFor each measurement, at least 20,000 cells were

graft tumor model and xenogen bioluminescence

athymic nude mice (46 weeks of age, Harlan, Indi-) were used. The use and care of animals was carried outuidelines approved by the Institutional Animal Care andittee. Subconuent HCT-116-Luc cells were harvested

C.-Z. Wang et al. / Phytomedicine 20 (2013) 999 1006 1001

Fig. 1. Chemic hese cpolyynes; Com ds (12

and resuspeby 0.4% trycutaneous 50 l PBS wthe same daat doses of injected wicarried out were subjecSciences, Houlation. d-Lat 100 mg/kas a substracollected bphotographformed wit

Statistical a

Data areANOVA waresults. In sgroups. The

Results

Effects of 13proliferation

We evalusing two 480 and o2012). As s

lifertratil growedal structures of compounds isolated from the root bark of Oplopanax horridus. Tpounds (7)(9), sesquiterpenes; Compounds (10) and (11), steroids; and Compoun

nded in PBS. Before inoculation, cell viability was testedpan blue exclusion assay (viable cells > 90%). For sub-injection, approximately 1 106 HCT-116-Luc cells inere injected into both anks of each mouse. Starting

antiproconcencer cel13 shoy, falcarindiol was intraperitoneally (IP) administered15 mg/kg (body weight) every day. Control mice wereth the vehicle. Animal whole body optical imaging wasas described previously (He et al. 2011). Briey, animalsted to Xenogen IVIS 200 imaging system (Caliper Lifepkinton, MA) for imaging weekly after tumor cell inoc-uciferin sodium salt (Gold Biotechnology, St. Louis, MO)g body weight in 0.1 ml sterile PBS was administered IPte before imaging. The acquired pseudo images werey superimposing the emitted light over the grayscales of the animal. Quantitative image analysis was per-h Xenogens Living Image V2.6.1 software.

nalysis

presented as mean standard error (SE). A one-ways employed to determine statistical signicance of theome cases, Students t-test was used for comparing two

level of statistical signicance was set at p < 0.05.

compounds on colorectal and breast cancer cell

uated the antiproliferative effects of 13 compoundshuman colorectal cancer cell lines HCT-116 and SW-ne human breast cancer cell line MCF-7 (Huanghown in Fig. 2, the 13 compounds exhibited different

cer cells at cells. Howegrowth inh

In HCT-1liferative efinhibited byfor compou2 and 3 shocell growthp < 0.01). Coerative effecell growthinhibited calmost com

Similar ecells, but thline was loweffects of co480 cells, buMCF-7 cellsmost poten

Antiproliferand SW-480

Since twtive to polyby the two at low concompounds can be separated into four groups: Compounds (1)(6),) and (13), phenolic acids.

ative effects on the three cancer cell lines. At the testedons (10300 M), compounds 711 did not inhibit can-wth in either of the three cell lines. Compounds 12 and

some antiproliferative effects on two colorectal can-

300 M, but such effects were not observed in MCF-7ver, compounds 16 showed different potential for cellibition in the three cancer cell lines.16 cells, compounds 4 and 6 showed moderate antipro-fects and at 30 and 100 M, HCT-116 cell growth was

32.7% and 98.8% for compound 4, and 26.0% and 96.5%nd 6, respectively (all p < 0.01 vs. control). Compoundswed stronger effects; when treated with 30 M, cancer

was inhibited by 91.7% and 89.8%, respectively (bothmpounds 1 and 5 showed the most potent antiprolif-cts. At 10 M, compound 5 (oplopantriol A) inhibited

by 76.4% (p < 0.01), while compound 1 (falcarindiol)ell growth by 98.1% (p < 0.01). Falcarindiol at 10 Mpletely inhibited HCT-116 cell growth (Fig. 2A).ffects were also observed in SW-480 colorectal cancere inhibition potential of compounds 16 on this celler than that of HCT-116 cells (Fig. 2B). Antiproliferativempounds 16 on MCF-7 cells were weaker than on SW-t all six compounds showed dose-dependent effects on. Moreover, falcarindiol and oplopantriol A showed thet antiproliferative effects (Fig. 2C).

ative effects of two potent compounds on HCT-116 cells

o human colorectal cancer cell lines were more sensi-ynes, and cell growth was almost absolutely inhibitedmost potent compounds falcarindiol and oplopantriol Aentrations, we tested the antiproliferative effects of the

1002 C.-Z. Wang et al. / Phytomedicine 20 (2013) 999 1006

80

100

8 9 10 11 12 13

n (

%)

HCT-11610 M 30 M 100 M 300 M

A

8 9 10 11 12 13

M 300 M

8 9 10 11 12 13

M 300 M

Fig. 2. Effects . Human colorectal cancer cell lines (A) HCT-116 and (B) SW-480, and (C)human breast solated compounds. Cells were treated with 10300 M of compounds for48 h, and cell p ntrol in percentage and expressed as average standard error of the threeexperiments (

two compothe dose-deby treatmen

For oploconcentratiliferation bSW-480 celeration by oplopantrioeffects. Whcell prolifer(Fig. 3A). Alless antiprsignicantlcarindiol is O. horridus.

Effects of fal

Antiprollated compwhether facycle arrestusing ow the effects at concentr2 M of falboth the S apared to 270

20

40

60

1 2 3 4 5 6 7

Pro

life

rati

o

Compound

0

20

40

60

80

100

1 2 3 4 5 6 7

Pro

life

rati

on

(%

)

Compound

SW-48010 M 30 M 100

B

0

20

40

60

80

100

1 2 3 4 5 6 7

Pro

life

rati

on

(%

)

MCF-710 M 30 M 100

C

Compound

of 13 compounds isolated from Oplopanax horridus on proliferation of cancer cells cancer cell line MCF-7 were employed to evaluate the antiproliferative effects of iroliferation was assayed by the MTS method. Results were normalized to each cosolvent vehicle set at 100%). Compound number is the same as Fig. 1.

unds in even lower concentrations. As shown in Fig. 3,pendent effects of the two compounds were observedt with concentrations of 120 M.pantriol A, the newly identied compound, treatmentons of 5, 10, and 20 M, inhibited HCT-116 cell pro-y 69.0%, 74.1%, and 88.2%, respectively (Fig. 3A). Inls, treatment with 10 and 20 M inhibited cell prolif-19.2% and 51.1%, respectively (Fig. 3B). Compared tol A, falcarindiol showed more potent antiproliferativeen treated with 2 and 5 M of falcarindiol, HCT-116ation was inhibited by 69.7% and 98.1%, respectivelythough at the same concentrations, falcarindiol showedoliferative effects on SW-480 cells, its potential isy higher than that of oplopantriol A (Fig. 3B). Thus, fal-most potent antiproliferative compound identied from

carindiol on colorectal cancer cell cycle distribution

iferative evaluation suggested that among the 13 iso-ounds, falcarindiol is the most active. To examinelcarindiols cell growth inhibition was because of cell

at a specic phase, cell cycle proles were determinedcytometry after staining with PI. As shown in Fig. 4,of falcarindiol on the cell cycle prole were observedations as low as 1 M. Treatment of HCT-116 cells withcarindiol for 48 h increased the percentages of cells innd G2/M phases to 39.2% and 31.8%, respectively, com-.7% and 21.6% in the vehicle treated cells (all p < 0.01).

0

20

40

60

80

100

0 5 10 15 20

Pro

life

rati

on

(%

)

Concentr ation (M)

HCT-116

Oplopantriol A

Falcarindiol

A

0

20

40

60

80

100

0 5 10 15 20

Pro

life

rati

on

(%

)

SW-48 0

Oplopantriol A

Falcarindiol

B

Concentr ation (M)

Fig. 3. Antiproliferative effect of falcarindiol (compound 1) and oplopantriol A (com-pound 5) on human colorectal cancer cells. (A) HCT-116 and (B) SW-480 cells weretreated with 120 M of falcarindiol or oplopantriol A for 48 h, and cell prolifera-tion was assayed by the MTS method. Results were normalized to each control inpercentage and expressed as average standard error of the three experiments.

C.-Z. Wang et al. / Phytomedicine 20 (2013) 999 1006 1003

Fig. 4. Cell cyc -480ethanol and st Repre(B) Percentage as thvs. control.

When treatthe G2/M p

Treatmebut differenfalcarindiolwere in thep < 0.01). Wwere in the cells were ifalcarindiolS and G2/M

Apoptotic in

To exploinhibits celetry after dcolorectal cearly and laptosis or neand PI (lowannexin V aor necrotic right quadr

As showing treatmewith 6 M totic cells incells (contro(control: 5.signicantlycell lines. Infalcarindiol116 than in

or erindi

valuase-tic nule analysis of HCT-116 and SW-480 cells treated with falcarindiol. HCT-116 and SWained with propidium iodide. DNA content was determined by ow cytometry. (A)

of each cell cycle phase with various treatments or with control. Data are presented

ment concentration was increased, most cells were inhase.nt of SW-480 cells also changed their cell cycle proletly than that of the HCT-116 cells (Fig. 4A). 2 M of

increased both the S and G2/M phases, but most cells G2/M phase (45.4% compared to 11.3% in the control,

Antitumof falca

To eluciferathymith an increase in treatment concentration, more cellsS phase. After treatment with 8 M of falcarindiol, 72.9%n the S phase (Fig. 4B). Thus, in both cancer cell lines,

signicantly increased the number of cancer cells in the phases.

duction effects of falcarindiol on colorectal cancer cells

re the potential mechanism through which falcarindioll growth, cell apoptosis was assayed by ow cytom-ouble staining with annexin V and PI in both humanancer cell lines. Annexin V can be detected in both thete stages of apoptosis. PI enters the cell in late apo-crosis. Viable cells were negative for both annexin Ver left quadrant); early apoptotic cells were positive fornd negative for PI (lower right quadrant); late apoptoticcells displayed both positive annexin V and PI (upperant).n in Fig. 5, apoptotic cells increased signicantly follow-nt with 18 M of falcarindiol for 48 h. After treatmentof falcarindiol, the percentage of early and late apop-creased to 41.6% and 27.8%, respectively, for HCT-116l: 5.1% and 4.9%); and 20.8% and 20.7% for SW-480 cells

8% and 1.8%). The results demonstrate that falcarindiol induces cell apoptosis in both human colorectal cancer

addition, similar to its antiproliferative effects (Fig. 3),s apoptotic induction activity is more potent in HCT-

SW-480 cells (Fig. 5).

istered witevery day. Tcence imagresults at wtime pointsrized in Figgroup exhiintensities crevealed thgrowth in tWeeks 3 anweek 2 (bo

A rat smevaluate thfalcarindiol120 M, fthe controlare 101.5% absolutely when treatesuggest thathe HCT-11normal inte

Discussion

Oplopantus, O. japon cells were treated with 18 M of falcarindiol for 48 h, then xed insentative histograms of the DNA content in each experimental group.e mean standard error of triplicate experiments. *p < 0.05, ** p < 0.01

ffects in xenograft tumor model and safety evaluationol

ate the in vivo antitumor potential of falcarindiol, reyagged HCT-116 cells were inoculated into the anks ofde mice. Beginning on day 1, animals were also admin-

h falcarindiol at 15 mg/kg or vehicle intraperitoneallyumor growth was measured by xenogeny biolumines-ing on a weekly basis. Representative xenogen imagingeeks 04 are shown in Fig. 6A. Tumor size at indicated

as assessed by imaging signal intensities is summa-. 6B. The data showed that the falcarindiol treatmentbited signicantly decreased xenogeny imaging signalompared with the control group. Quantitative analysisat falcarindiol signicantly inhibited xenograft tumorhe 2nd week after falcarindiol administration (p < 0.05).d 4 exhibited more signicant antitumor effects thanth p < 0.01).all intestine epithelial cell line, IEC-6, was used to

e safety of falcarindiol. Fig. 6C shows the activities of on the proliferation of IEC-6 cells. At concentrations ofalcarindiol did not inhibit cell growth. Compared with

(100%), the cell viabilities of falcarindiol on IEC-6 cellsat 20 M, and 85.5% at 30 M. Cell growth was almostinhibited in both of the two colorectal cancer cell linesd with 10 M of falcarindiol (Fig. 3). Thus, these resultst falcarindiol showed signicant antitumor activity in6 xenograft tumor model and was relatively safe tostinal cells.

ax is a small genus, consisting of three species: O. ela-icus, and O. horridus. Because O. elatus and O. japonicus

1004 C.-Z. Wang et al. / Phytomedicine 20 (2013) 999 1006

Fig. 5. Apopto -480with annexin ytome(x-axis). (B) Pe he mecontrol.

are found icountries wand depth of the two 2012). Relathe former still limited

CompareO. horridus hgon. Wild rhorridus is juseveral polhorridus (Kcal studies antidiabeticused in pre(Wang et asingle comp

Becausebark (Lantzcal studies2011). Thiron their cgroups: pol79), steroipounds 12 aoplopantriothree sesquand (9) oplo

In this stcancer activsteroids didcentrationsOn the oth

lifers wextracus obotenes, itic analysis of HCT-116 and SW-480 cells treated with falcarindiol. HCT-116 and SWV/propidium iodide (PI) before the extent of apoptosis was determined by ow crcentage of viable early apoptotic and late apoptotic cells. Data are presented as t

n the East Asian countries of China, Korea, and Japan,ith a long history of herbal medicine use, the breadthof research on the phytochemistry and pharmacologyspecies is higher than that of O. horridus (Calway et al.tively more literature was found to report progress ontwo species; however, anticancer related reports were

antipropoundbark epreviomost ppolyyn even for the two Asian species.d to the sporadic distribution of the two Asian species,as a more widespread distribution from Alaska to Ore-

esources of O. horridus are plentiful, but research on O.st at the beginning stage. Based on the literature search,

yynes and volatile compounds were isolated from O.obaisy et al. 1997; Calway et al. 2012). Pharmacologi-on O. horridus were only focused on its antibacterial,, and antimalignancy properties, and the componentsvious studies were mostly herbal extracts or fractionsl. 2010; Calway et al. 2012). The biological activities ofounds isolated from O. horridus are largely unknown.

the commonly used part of O. horridus is the root et al. 2004), we recently performed phytochemi-

on the root bark of this plant (Huang et al. 2010,teen compounds were isolated and identied. Basedhemical structures, they can be divided into fouryynes (compounds 16), sesquiterpenes (compoundsds (compounds 10 and 11), and phenolic acids (com-nd 13). Among these compounds, the two polyynes (5)l A, and (6) oplopantriol B (Huang et al. 2010), and theiterpenes (7) oplopanpheside A, (8) oplopanpheside B,panpheside C (Huang et al. 2011) are novel compounds.udy, we systemically evaluated the 13 compounds anti-ities. An MTS assay showed that the sesquiterpenes and

not show any antiproliferative effects at the tested con-. Phenolic acids showed weak antiproliferative effects.er hand, all six polyynes showed dose-dependent

strong antipPolyyne

ecological finto the mbiosynthesiHowever, iwere produin future st

The two480 were mwere selectexpression.p53. Tumorcellular resMutations resistance t2004). Our oplopantriolines in a comore sensit

Since faeffects amoantiprolifershowed thais differentat a lowerconcentratiin the G2/M cells were treated with 18 M of falcarindiol for 48 h, then stainedtry. (A). Representative scatter plots of PI (y-axis) versus annexin Van standard error of triplicate experiments. *p < 0.05, **p < 0.01 vs.

ative effects of varying strengths. These polyyne com-re isolated from the hydrophobic fraction of the roott (Huang et al. 2010). This was consistent with ourservations that the hydrophobic fraction showed thet antiproliferative activity (Sun et al. 2010a,b). Twoncluding the novel compound oplopantriol A, showed

roliferative activity (Fig. 2).

natural products are interesting for their wide variety ofunctions, and surprising mode of biosynthesis. Researchetabolism of polyynes has revealed that they can bezed by plants and fungi (Minto and Blacklock 2008).t is uncertain whether the polyynes from O. horridusced by the plants or phytofungi. This will be observedudies.

human colorectal cancer cell lines HCT-116 and SW-ore sensitive to the potent polyynes (Fig. 3). Thus, theyed for further observation. The two cell lines differ in p53

HCT-116 has p53 wild-type, while SW-480 has mutated suppressor gene p53 is thought to be important in theponse to chemotherapeutic agent-induced cell death.in p53 have been shown to correlate with increasedo chemotherapeutic agents in cancer cells (Din et al.,results showed that the two polyynes falcarindiol andl A signicantly inhibited cell growth in these two cellncentration-dependent manner and that HCT-116 wasive than SW-480.lcarindiol showed the most potent antiproliferativeng all of the isolated compounds, evaluation of itsative mechanism was carried out. Cell cycle assayt the response to falcarindiol in the two cell lines on. Falcarindiol arrested HCT-116 cells in the S phase

concentration but in the G2/M phase at a higheron. Interestingly, falcarindiol arrested SW-480 cells

phase at a lower concentration, but in S phase at a

C.-Z. Wang et al. / Phytomedicine 20 (2013) 999 1006 1005

Fig. 6. In vivo l. (A) Fof athymic mic ontroxenogen bioluat the indicateSafety evaluatwas determine

higher concmore sensianticancer aand apopto(Palozza et

In vivo ation. After mgrowth wathermore, adetermine line. Falcariat 10 M, thcell growthshowed canfor normal

Based oactivities, compoundsthe four chobserved bgroups. Phepared with potent. Theeffects on athe core stactivity. Prefalcarindioland Ulrich-colorectal cantitumor observation using a xenograft model and safety evaluation of falcarindioe subcutaneously (n = 10/group), and the tumor sizes after treatment with solvent c

minescence imaging. Representative xenogen imaging results are shown. (B) Quantitativd time points are represented with imaging signal intensities (in photons/second/cm2/stion of falcarindiol on normal intestinal cells. The IEC-6 rat small intestine epithelial cells wed.

entration. Apoptotic assay showed that HCT-116 wastive than SW-480 cells, suggesting that falcarindiolsctivity is in part due to the induction of cell cycle arrestsis and that p53 may participate in the mechanismal. 2009; Wang et al. 2012b).ntitumor evaluation supported the in vitro observa-ice received 15 mg/kg of falcarindiol, HCT-116 tumor

s signicantly inhibited from weeks 24 (Fig. 6). Fur- rat small intestine epithelial cell line was used tothe safety of falcarindiol on the normal intestinal cellndiol did not show signicant antiproliferative effectse concentration at which HCT-116 and SW-480 cancer

was absolutely inhibited, suggesting that falcarindiolcer cell selective inhibition activity and was thus safecells.n their chemical structures and observed biologicalwe explored the structureactivity relationship of

from O. horridus on cancer chemoprevention. Withinemical groups, cell growth inhibition effects were noty the compounds in the sesquiterpene and steroidnolic acids showed lower antiproliferative effects com-the polyynes, while the effects of caffeic acid are more

polyyne group showed signicant antiproliferativell three cancer cell lines. This result suggested thatructure of polyynes is important for antiproliferativevious studies show that polyyne compounds including

possess antitumor and antibiotic activities (WagnerMerzenich 2013). In this study, we observed anti-ancer activity of isolated compounds and identied

two polyynrectal cancand clinicaof falcarindfermentatio

With recompoundsby similaritpounds 1, 3the terminathree compture. If the group, its angroup at Cincreased sihydroxyl grfor maintaisupplied imcompounds

Conclusion

To identthe anticanincluding stwo phenodid not shtions. Phenthe six polyirey luciferase-tagged HCT-116 cells were injected into both anksl or 15 mg/kg/day of falcarindiol were measured on a weekly basis by

e analysis of xenogen bioluminescence imaging. Average tumor sizeseradian) as mean standard error. *p < 0.05, **p < 0.01 vs. control. (C)re treated with 130 M of falcarindiol for 48 h, and cell proliferation

es, including a novel compound, that are potent colo-er chemopreventive compounds. For future animall studies, as an alternative approach, large quantitiesiol and other polyynes could be obtained throughn on an industrial scale (Radic and Strukelj 2012).gard to this polyyne group, there are three pairs of: compounds 1 and 2; 3 and 4; and 5 and 6 groupedy of structure. Regarding their biological effects, com-, and 5 showed more potent activity, suggesting thatl double bond increases anticancer activity. Among theounds, oplopantriol A can be considered a basic struc-hydroxyl group at C-1 was esteried with an acetatetiproliferative effect was decreased, but if the hydroxyl-1 was eliminated, its antiproliferative effect wasgnicantly. However, compared to the derivation of theoup at C-1, the terminal double bond is more importantning anticancer potential. This structureactivity assayportant information for semi-synthesis of this group of

to nd novel anticancer agents.

ify active anticancer compounds in Oplopanax horridus,cer potentials of 13 compounds isolated from this plant,ix polyynes, three sesquiterpenes, two steroids andlic acids, were evaluated. Sesquiterpenes and steroidsow any antiproliferative effects at tested concentra-olic acids showed weak antiproliferative effects, whileynes showed dose-dependent antiproliferative effects

1006 C.-Z. Wang et al. / Phytomedicine 20 (2013) 999 1006

in human colorectal and breast cancer cells. Two polyynes, fal-carindiol and oplopantriol A, signicantly inhibited cell growth,and falcarindiol had the most potent effects. An in vivo xenografttumor model supported the in vitro observation that 15 mg/kgof falcarindiol signicantly inhibited HCT-116 tumor growth. Astructureactivity relationship assay showed that elimination ofthe hydroxyl group at C-1 increased anticancer activity, and thatthe terminal double bond is the key structure maintaining the anti-cancer potential of polyyne compounds. Mechanism observationsuggested that falcarindiol induced cancer cell death in part by cellcycle arrest and induction of pro-apoptosis, and that the tumor sup-pressor protein p53 may participate in its cancer chemoprevention.

Acknowledgements

This work was supported in part by the NIH/NCCAMAT004418 and AT005362, NIH GM074197, NIH/NCI CA149275,DOD W81XWH-10-1-0077, and the University of Macau grant(UL015/09-Y1).

References

Bell, R.M., 2010. A review of complementary and alternative medicine practicesamong cancer survivors. Clinical Journal of Oncology Nursing 14, 365370.

Buchholz, T.A., 2009. Radiation therapy for early-stage breast cancer after breast-conserving surgery. The New England Journal of Medicine 360, 6370.

Calway, T., Du, G.J., Wang, C.Z., Huang, W.H., Zhao, J., Li, S.P., Yuan, C.S., 2012.Chemical and pharmacological studies of Oplopanax horridus, a North Americanbotanical.

Cragg, G.M., Gdevelopin

Davis, E.L., Oh,and patienuse: a syst

Din, F.V., Dunloof aspirin e91, 38138

Hait, W.N., Ham1263126

He, B.C., Gao, JWagner, EN., Zhou, QWnt/beta-cancer. Mo

Huang, W.H., Ph.D. Thes

Huang, W.H., Zphesides Ahorridus. C

Huang, W.H., Zhang, Q.W., Wang, C.Z., Yuan, C.S., Li, S.P., 2010. Isolation and iden-tication of two new polyynes from a North American ethnic medicinal plant Oplopanax horridus (Smith) Miq. Molecules 15, 10891096.

Kharwar, R.N., Mishra, A., Gond, S.K., Stierle, A., Stierle, D., 2011. Anticancercompounds derived from fungal endophytes: their importance and future chal-lenges. Natural Product Reports 28, 12081228.

Kobaisy, M., Abramowski, Z., Lermer, L., Saxena, G., Hancock, R.E., Towers, G.H.,Doxsee, D., Stokes, R.W., 1997. Antimycobacterial polyynes of Devils Club(Oplopanax horridus), a North American native medicinal plant. Journal of Nat-ural Products 60, 12101213.

Lantz, T.C., Swerhun, K., Turner, N.J., 2004. Devils club (Oplopanax horridus): Anethnobotanical review. Herbal Gram 62, 3348.

Lin, J.G., Chen, Y.H., 2012. The role of acupuncture in cancer supportive care. TheAmerican Journal of Chinese Medicine 40, 219229.

McCutcheon, A.R., Roberts, T.E., Gibbons, E., Ellis, S.M., Babiuk, L.A., Hancock, R.E.,Towers, G.H., 1995. Antiviral screening of British Columbian medicinal plants.Journal of Ethnopharmacology 49, 101110.

Minto, R.E., Blacklock, B.J., 2008. Biosynthesis and function of polyacetylenes andallied natural products. Progress in Lipid Research 47, 233306.

Palozza, P., Torelli, C., Boninsegna, A., Simone, R., Catalano, A., Mele, M.C., Picci, N.,2009. Growth-inhibitory effects of the astaxanthin-rich alga Haematococcuspluvialis in human colon cancer cells. Cancer Letters 283, 108117.

Paoletti, X., Oba, K., Burzykowski, T., Michiels, S., Ohashi, Y., Pignon, J.P., Rougier, P.,Sakamoto, J., Sargent, D., Sasako, M., Van Cutsem, E., Buyse, M., 2010. Benetof adjuvant chemotherapy for resectable gastric cancer: a meta-analysis. JAMA303, 17291737.

Radic, N., Strukelj, B., 2012. Endophytic fungi: the treasure chest of antibacterialsubstances. Phytomedicine 19, 12701284.

Randhawa, M.A., Alghamdi, M.S., 2011. Anticancer activity of Nigella sativa (blackseed) a review. The American Journal of Chinese Medicine 39, 10751091.

Siegel, R., Naishadham, D., Jemal, A., 2012. Cancer statistics. CA: A Cancer Journalfor Clinicians 62, 1029.

Sun, S., Du, G.J., Qi, L.W., Williams, S., Wang, C.Z., Yuan, C.S., 2010a. Hydrophobic con-stituents and their potential anticancer activities from Devils Club (Oplopanaxhorridus Miq.). Journal of Ethnopharmacology 132, 280285.

Li, X.Lities poneneung,oxidangy 108, H., Ulhines.Z., Auce liqidus: cC.Z., Ccance669..Z., Dette, , possernatiohen, Xides: inal of Journal of Natural Medicines 66, 249256.rothaus, P.G., Newman, D.J., 2009. Impact of natural products ong new anti-cancer agents. Chemical Reviews 109, 30123043.

B., Butow, P.N., Mullan, B.A., Clarke, S., 2012. Cancer patient disclosuret-doctor communication of complementary and alternative medicineematic review. The Oncologist 17, 14751481.p, M.G., Stark, L.A., 2004. Evidence for colorectal cancer cell specicityffects on NF kappa B signalling and apoptosis. British Journal of Cancer8.bley, T.W., 2009. Targeted cancer therapeutics. Cancer Research 69,

7 (discussion 1267)..L., Zhang, B.Q., Luo, Q., Shi, Q., Kim, S.H., Huang, E., Gao, Y., Yang, K.,.R., Wang, L., Tang, N., Luo, J., Liu, X., Li, M., Bi, Y., Shen, J., Luther, G., Hu,., Luu, H.H., Haydon, R.C., Zhao, Y., He, T.C., 2011. Tetrandrine inhibitscatenin signaling and suppresses tumor growth of human colorectallecular Pharmacology 79, 211219.2012. Chemical Investigation on Root Barks of Oplopanax horridus.is, University of Macau, Macau, 268 pp.hang, Q.W., Meng, L.Z., Yuan, C.S., Wang, C.Z., Li, S.P., 2011. Oplopan-C, three new phenolic glycosides from the root barks of Oplopanaxhemical and Pharmaceutical Bulletin 59, 676679.

Sun, S., activcom

Tai, J., Chanticolo

Wagneron C

Wang, Cmanhorr

Wang, for 657

Wang, CsonnRb1Inte

Xu, Z., CcharJour., Wang, C.Z., Williams, S., Yuan, C.S., 2010b. Improving anticancerof Oplopanax horridus root bark extract by removing water-solublets. Phytotherapy Research 24, 11661174.

S., Cheah, S., Chan, E., Hasman, D., 2006. In vitro anti-proliferative andt studies on Devils Club Oplopanax horridus. Journal of Ethnopharma-, 228235.rich-Merzenich, G. (Eds.), 2013. Evidence and Rational Based Researche Drugs. Springer, New York, p. 525 (Chapter 1, p. 125).ng, H.H., Mehendale, S.R., Shoyama, Y., Yuan, C.S., 2010. High perfor-uid chromatographic analysis and anticancer potential of Oplopanaxomparison of stem and berry extracts. Fitoterapia 81, 132139.alway, T., Yuan, C.S., 2012a. Herbal medicines as adjuvantsr therapeutics. The American Journal of Chinese Medicine 40,

u, G.J., Zhang, Z., Wen, X.D., Calway, T., Zhen, Z., Musch, M.W., Bis-M., Chang, E.B., Yuan, C.S., 2012b. Ginsenoside compound K, notsses potential chemopreventive activities in human colorectal cancer.nal Journal of Oncology 40, 19701976.., Zhong, Z., Chen, L., Wang, Y., 2011. Ganoderma lucidum polysac-mmunomodulation and potential anti-tumor activities. The AmericanChinese Medicine 39, 1527.

Identification of potential anticancer compounds from Oplopanax horridusIntroductionMaterials and methodsChemicals and reagentsPlant materialsExtraction, compound isolation and structural identificationCell lines and cell cultureCell proliferation analysisCell cycle analysisApoptosis assayIn vivo xenograft tumor model and xenogen bioluminescence imagingStatistical analysis

ResultsEffects of 13 compounds on colorectal and breast cancer cell proliferationAntiproliferative effects of two potent compounds on HCT-116 and SW-480 cellsEffects of falcarindiol on colorectal cancer cell cycle distributionApoptotic induction effects of falcarindiol on colorectal cancer cellsAntitumor effects in xenograft tumor model and safety evaluation of falcarindiol

DiscussionConclusionAcknowledgementsReferences