Journal of Natural Medicines

, Volume 61, Issue 2, pp 131–137

Anticancer properties of panduratin A isolated from Boesenbergia pandurata (Zingiberaceae)


    • CSIRO Human Nutrition
    • Department of BiologyFaculty of Mathematics and Natural Sciences, Brawijaya University
  • Graham Peter Jones
    • Faculty of Science, School of Agriculture, Food and WineAdelaide University
  • Ian Roland Record
    • CSIRO Human Nutrition
  • Graeme Howie McIntosh
    • CSIRO Human Nutrition
Original Paper

DOI: 10.1007/s11418-006-0100-0

Cite this article as:
Kirana, C., Jones, G.P., Record, I.R. et al. J Nat Med (2007) 61: 131. doi:10.1007/s11418-006-0100-0


Extract of Boesenbergia pandurata (Kaempferia pandurata) (Zingiberaceae) has been used as a replacement for K. rotunda, the main ingredient of a popular traditional tonic called “jamu” especially for women in Indonesia. From our previous study, ethanolic extract of B. pandurata showed strong inhibitory effects on the growth of cancer cells, similar to ethanolic extract of Curcuma longa. C. longa and its bioactive compound, curcumin, have shown potential anticancer activity in in vitro and in vivo studies and have undergone clinical trials. Panduratin A, a chalcone derivative isolated from B. pandurata, was found to inhibit the growth of MCF-7 human breast cancer and HT-29 human colon adenocarcinoma cells with an IC50 of 3.75 and 6.56 µg/ml, respectively. Panduratin A arrested cancer cells labelled with Annexin-V and propidium iodide in the G0/G1 phase and induced apoptosis in a dose-dependent manner. In an animal model study, male Sprague–Dawley rats were fed with AIN diet containing ethanolic extracts prepared from the equivalent of 4% by weight of dried rhizomes of B. pandurata and C. longa. Aberrant crypt foci (ACFs) were induced by two subcutaneous doses (15 mg/kg body weight) of azoxymethane (AOM) 1 week apart. The rats were killed 10 weeks later, and the ACFs were assessed in the colon. At the dose given to rats, it appeared that the extracts were not toxic. Total ACFs were slightly reduced by B. pandurata extract compared to control group but not significantly different. Extract of B. pandurata may have a protective effect against colon cancer but additional studies using different models, such as a breast cancer model, need to be carried out.


Panduratin ABoesenbergia pandurataZingiberaceaeApoptosisCell cycle arrestAzoxymethaneAberrant crypt foci


Medicinal herbs have always been used as traditional primary health care agents, especially in Asian countries, and over the last 20 years, there have been rapid changes in the popularity of the use of natural systems to maintain health and for alternative therapy in Western countries [1]. However, scientific studies on the use of most traditional medicinal plants have not been carried out to assure their efficacy and nontoxicity.

The plant Boesenbergia pandurata Schult (syn. Kaempferia pandurata Roxb.) (Zingiberaceae) is known as “temu kunci” in Indonesia. The fresh rhizomes are used in cooking and traditional medicine to treat diarrhoea, dermatitis, dry cough, and mouth ulcers [2]. In times of shortage, B. pandurata has been used to replace K. rotunda (“kunci pepet”, in Indonesian), the main ingredient in a popular traditional tonic especially for women, locally called “jamu”. The effect of the tonic has never been demonstrated clinically or in animal experiments. Interestingly, an extract of B. pandurata had a very strong growth-inhibitory activity against human HT-29 colon adenocarcinoma and MCF-7 breast cancer cells and was not toxic to the non-transformed human skin fibroblast cells (SF3169), while that of K. rotunda had no inhibitory activity [3]. This plant is also a popular condiment in Thailand.

Chemical studies of the rhizome of B. pandurata have resulted in the isolation and identification of several chalcones such as boesenbergin A, boesenbergin B, cardamonin, panduratin A, and dihydromethoxychalcone, and flavanones such as pinocembrin, pinostrobin, alpinetin and 5-hydroxy-7-methoxyflavanone [4, 5]. The monoterpenoids geranial and neral, as well as flavanones, have also been identified from the rhizomes of B. pandurata [6].

Members of the Zingiberaceae family have been reported to possess both antioxidant and anti-inflammatory activity. Such antioxidant and anti-inflammatory compounds have often been shown to be effective as anticancer agents [7]. In vitro studies on the extracts of B. pandurata and their isolated compounds have shown some beneficial pharmacological activities. For example methanolic extracts of B. pandurata showed a very strong inhibitory effect in the Epstein-Barr virus (EBV)-activated test [8]. Cardamonin, pinocembrin, panduratin A, and pinostrobin showed strong antimutagen activity in the Ames test using Salmonella typhimurium TA98 [9]. Tuchinda et al. [10] reported that panduratin A had significant anti-inflammatory activity in 12-O-tetradecanoylphorbol 13-acetate (TPA)-induced ear edema in rats. Recently Yun et al. [11] reported that panduratin A isolated from B. pandurata inhibited the growth of HT-29 colon cancer cells and induced apoptosis.

In this study, the anticancer properties of panduratin A were evaluated using human HT-29 colon and MCF-7 breast cancer cells, and the ethanolic extract of B. pandurata was evaluated in the azoxymethane (AOM)-induced aberrant crypt foci (ACF) model in rat. The anticancer effects of the extract were compared with the ethanolic extract of Curcuma longa (turmeric).

ACFs are microscopically identifiable precursor lesions of colon cancer found in rodents exposed to certain procarcinogens but are also in human colonic tissue and are considered to be a useful preneoplastic marker of colon carcinogenesis [12]. The ACF model study provides a rapid and reliable assay system and has been used for the screening of potential chemopreventive agents [13, 14].

Materials and methods

Plant materials and chemicals

Roots and rhizomes of Boesenbergia pandurata Schl. were purchased from the market place in Malang, East Java, Indonesia and were characterised by Chandra Kirana, MAgSc, PhD. A voucher of the specimen was deposited at CSIRO Human Nutrition, Adelaide. Samples were sliced, air-dried and brought to Adelaide, South Australia. Dried roots and rhizomes of C. longa (turmeric) were purchased from Kriomill (KrioKrush Basic Foods P/L, Singapore).

All chromatographic solvents were HPLC grade and were purchased from BDH (Kilsyth, VIC, Australia). Curcumin was purchased from Sigma Chemicals (St Louis, MO).

Fractionation and preparation of panduratin A

Dried sliced roots and rhizomes (100 g) were homogenised and extracted with absolute ethanol (250 ml) overnight at room temperature. The ethanol extract was then vacuum filtered and the extraction repeated three times. The combined ethanol extract was evaporated to dryness at 35°C under vacuum and yielded 13.6% dry extract. The crude ethanol extract was fractionated by preparative reversed-phase LC (Prep Nova-Pak HR Prep LC column, particle size 6 µm, 200×25 mm; equipped with a guard column containing the same material) using two different eluents; fraction A was eluted with 1:1 methanol/water containing 0.025% v/v trifluoroacetic acid (TFA), and fraction B was eluted with 100% methanol. The LC was run with a flow rate of 20 ml/min.

Panduratin A was isolated from fraction B by preparative reversed-phase LC using a mobile phase consisting of a stepwise gradient of 60, 70, 75 and 80% (v/v) aqueous methanol containing 0.025% (v/v) TFA. Isolated panduratin A was then checked for its purity by analytical HPLC (waters) using a modification of the method by He et al. [16]. The HPLC system consisted of a Licrosphere (Supelco, USA) RP C18 column (particle size 5 µm, 250×3.2 mm) equipped with a guard column containing the same material. The HPLC was run under gradient concentrations of 0.25% acetic acid and 100% acetonitrile with a flow rate of 0.77 ml/min at 48°C.

Panduratin A was characterised by nuclear magnetic resonance (1HNMR) (600 MHz, Varian Inova) and mass spectrometry LC ESI-MS (API-300, PE Sciex) and identified by a comparison of the spectral data [15].

Cell lines and cell culture

HT-29 human colon adenocarcinoma cells and MCF-7 human breast cancer cells were purchased from the American Type Culture Collection (ATCC) (Rockville, MD). Tumour cell lines were maintained in monolayer culture in Dulbecco minimum essential medium (DMEM) supplemented with 10% heat-inactivated foetal bovine serum (FBS), 20 mM HEPES (N-2-hydroxyethylpiperazine-N′-2-ethanesulfonic acid) buffer, 0.1% amphostat B and 5 µg/ml gentamycin in a humidified atmosphere of 5% CO2/95% air at 37°C.

DMEM, heat-inactivated FBS, HEPES, gentamycin, amphostat B, PBS, and trypsin were all obtained from Trace, Australia. MTT-tetrazolium salt [3-(4,5-dimethylthiasol-2yl)-2,5-diphenyl tetrazolium bromide], N,N-dimethyl formamide, and lauryl sufate (SDS) were purchased from Sigma (St. Louis, MO).

Assessment of cell viability and proliferation

Exponentially growing cells were suspended at a density of 2–4×104 cells/ml in DMEM and treated with panduratin A dissolved in dimethylsulphoxide (DMSO) (final concentration of DMSO was 0.5%) and with DMSO alone as the control in 96-well flat-bottom plates for 24, 48 and 72 h. The viability of the cells was assessed by MTT tetrazolium salt assay [17] using a Spectra Max 250 (Microplate Spectrophotometer, Molecular Devices, USA) at 570 nm. The IC50, i.e. the concentration of the extract required to inhibit cell growth by 50%, was determined.

Cell cycle analysis

About 200,000 cells/well were plated in six-well plates overnight to allow cells to attach and were treated with panduratin A dissolved in DMSO at concentrations as above. After 24 h of treatment, cells were harvested by using 0.01% trypsin/EDTA and then resuspended in PBS containing 1% FBS. The cells were fixed by dropwise addition of cold 70% ethanol while vortexing to avoid aggregation and kept at −20°C for at least 30 min. The cells were then washed with PBS twice, suspended in 40 µg/ml PI and 200 µg/ml RNAse (Sigma) in PBS and incubated at room temperature for 30 min. Stained nuclei were analysed for DNA-PI fluorescence using FACScan flow cytometry equipped with FACStation running Cell Quest software (Becton Dickinson, San Jose, CA).

Morphological examination of cancer cells

Cancer cells were grown on cover slips at a density of 100,000 cells/ml in six-well plates and allowed to attach for 24 h and then treated with panduratin A dissolved in DMSO. Five hundred microliters of cell suspension from adhering and floating cells was obtained from the wells, and slides were prepared using a cytocentrifuge (Shandon Southern Products, Cheshire, UK). Slides were air-dried, fixed in methanol and stained using DiffQuick stain (Sigma, St. Louis, MO). One hundred microliters of cell suspension on slides was stained with 1 mM Hoechst 33258 (Sigma). Chromatin condensation and nuclear fragmentation were observed with a fluorescence microscope.

Apoptosis assays

About 200,000 cells/well were plated in six-well plates overnight to allow cells to attach and were treated with panduratin A dissolved in DMSO at concentrations of 3, 9, 18 and 26 µg/ml. After 48 h of treatment, both adhering and floating cells were harvested and then double-labelled with Annexin-V-Fluos (Roche) and propidium iodide (PI) (Sigma) as described by the manufacturer. Cells (10,000) were analysed using a FACScan Flow Cytometry equipped with FACStation running Cell Quest software (Becton Dickinson, San Jose, CA), and total apoptotic cells were determined.

Animal study

Four week-old male outbred Sprague–Dawley rats were purchased from the Animal Research Centre at Murdoch University (Perth, Western Australia) and separated randomly into control and treatment groups. Animals were housed in wire cages to minimise coprophagy and maintained in an air-conditioned environment of 23±2°C with a 12:12 h light–dark cycle. Rats were given free access to diet and water. All experimental procedures involving animals were approved by the Commonwealth Scientific and Industrial Research Organization (CSIRO) Health Sciences and Nutrition Animal Experimentation Ethics Committee and the University of Adelaide Animal Experimentation Ethics Committee before commencing the study.

Experimental design

Beginning at 5 weeks of age, rats were given different experimental diets, which were based on AIN-93 specification [18]. There were three groups of 14 animals. Group 1 diet (control) contained semipurified AIN-93 only, the group 2 diet contained AIN-93 diet mixed with an ethanol extract of B. pandurata and the group 3 diet contained AIN-93 diet mixed with an ethanol extract of C. longa. Extracts were prepared from the equivalent of 4% by weight in diet of dried rhizomes of B. pandurata and of turmeric. The dried material (4%, i.e., 40 g dried rhizomes to make 1 kg diet) was extracted with ethanol three times (material-to-solvent ratio = 1:6) at room temperature. The ethanol extracts were combined and evaporated to reduce the amount of ethanol. Extracts were then mixed with the AIN-93 diet and dried at 35°C in an oven overnight. Diets were made fortnightly and kept in the freezer until required. Diets were replaced every 3 days. At 7 weeks of age, rats received two subcutaneous injections of AOM 1 week apart at a dose rate of 15 mg/kg body weight. Rats were weighed weekly. Rats were sacrificed using halothane/oxygen anesthesia and exsanguinated via the abdominal aorta 10 weeks after the second AOM injection. Colons were removed for ACF counting.

ACF analysis

Colons were removed, flushed with buffer and opened from caecum to anus and then fixed flat between two pieces of filter paper in 10% buffered formaldehyde. After a minimum of 24 h in formalin they were kept in 70% ethanol until ACFs were counted. Each colon was placed for 7–10 min in a petri dish containing 0.2% methylene blue dissolved in PBS. The colon was then placed mucosa-side up on a microscope slide, and ACFs were scored using a low power light microscope (10×).

Statistical analysis

Data are reported as mean±SEM, and comparisons were analysed by one-way analysis of variance (ANOVA), with Bonferroni post-hoc test to identify between-group differences (P<0.05) using Graphpad Prism (version 2.0).

Results and discussion

Our previous in vitro study showed that an ethanolic extract of B. pandurata had inhibitory activity on the growth of cancer cells similar to that of C. longa. C. longa is a member of Zingiberaceae that has been extensively studied and shown to possess anticancer activity [2]. An initial analysis of the bioactive compounds in ethanol extracts of B. pandurata and C. longa using the MTT tetrazolium salt assay revealed that fraction B of B. pandurata was much more active in inhibiting the growth of HT-29 colon cancer cells than fraction A (Fig. 1). By comparison, fractions A and B of C. longa both showed inhibitory activity, although fraction A was more active than fraction B. Curcuminoids (curcumin, demethoxycurcumin and bisdemethoxycurcumin), the active compounds of C. longa, were identified in fraction A by HPLC assay. A chalcone derivative, panduratin A (2,6-dihydroxy-4-methoxyphenyl)-[3′-methyl-2′-(3″-methylbut-2″-enyl)-6′-phenylcyclohex-3′-enyl] methanone, was identified in fraction B, isolated and characterised as described in the methods. Panduratin A has the molecular formula C26H30O4, which includes 12 aliphatic carbons (3CH3, 2CH2, 3CH, 2C=CH−), 12 aromatic carbons and one methoxyl carbon [15]. The NMR data of panduratin A were compared with and shown to be similar to published data [9, 15].
Fig. 1

The growth inhibitory influence of fractions on HT-29 colon cancer cells at 72 h. Fraction A (A) was eluted with 1:1 methanol/water containing 0.025% v/v trifluoroacetic acid using preparative reversed-phase LC, and Fraction B (B) was eluted with 100% methanol of B. pandurata (BP) and C. longa (CL). Viable cells were determined using MTT tetrazolium salt assay as described under Materials and methods. Values are means±SE of three independent experiments

The inhibitory activity of panduratin A was assessed using HT-29 colon and MCF-7 breast cancer cells. Figure 2 shows the viability of the cell lines treated with panduratin A for 72 h at concentrations of 3, 9, 11, and 18 µg/ml. It appeared that panduratin A was more sensitive to MCF breast cancer than HT-29 colon cancer cells. The IC50 of panduratin A in HT-29 cells was 6.56 µg/ml and in MCF-7 cells was 3.75 µg/ml. Panduratin A showed a cytostatic effect (100% inhibition) on both HT-29 and MCF-7 cells at a concentration of 9 µg/ml.
Fig. 2

The viability and IC50 of MCF-7 and HT-29 cells after 72 h treatment at different concentrations of panduratin A. Viable cells were determined by the MTT method. The results are expressed as a percentage of the control (cells in 0.5% DMSO). Values are means±SE of three independent experiments

Cell-cycle analysis of HT-29 cancer cells treated with panduratin A at 3, 9, 18 and 26 µg/ml for 24 h showed alterations in the distribution of DNA content (Figs. 3 and 4). There was an increase in the G0/G1 phase from 33% in untreated cells to 71% at 26 µg/ml. The proportion of cells in the S phase was slightly reduced from 18.7% in untreated cells to 10.9% at 26 µg/ml. By comparison, there was a decrease in the G2/M phase from 36.8% in untreated cells to 15.4% at a concentration of 26 µg/ml. Unlike panduratin A, which arrested the cells in the G0/G1 phase, curcumin had a blocking effect in the G2/M phase of the cell cycle in many types of cancer cells [19].
Fig. 3

HT-29 cancer cells in different phases of the cell cycle after 24-h treatment with different concentrations of panduratin A. The cells were fixed and stained with propidium iodide, and the DNA content was analysed by FACScan flow cytometry. Values are means±SE of three independent experiments
Fig. 4

Total apoptosis in HT-29 cancer cells after 48-h treatment with different concentrations of panduratin A. The cells were fixed and double-labelled with Annexin-V-Fluos and propidium iodide and analysed by FACScan flow cytometry. Values are means±SE of three independent experiments

HT-29 cells treated with panduratin A showed features of apoptosis such as membrane blebbing, chromatin condensation and/or nuclear fragmentation and apoptotic bodies when the cells were stained with Hoechst 33258. Flow cytometry analysis using Annexin-V stain further revealed that panduratin A induced apoptosis in HT-29 cells. At 48 h, 2.2% of cells treated with 11 µg/ml of panduratin A underwent apoptosis and the number of apoptotic cells increased to 16.7% when treated with 26 µg/ml. Panduratin A induced apoptosis by activation of caspase-3 associated with PARP cleavage [11].

In a 13-week experiment, body weights of rats did not differ among the dietary treatment groups over the experiment period (Fig. 5). The final body weights (means±SE) were as follows: 509±8 in control group, 513±11 in rats fed with the equivalent of 4% B. pandurata and 519±8 in rats fed with the equivalent of 4% C. longa. In addition, the weight of spleens and livers of rats did not differ among the groups (data not shown), suggesting that the extracts at the concentrations employed in this study were not toxic. Diet containing 2% turmeric has been reported to inhibit benzo[a]pyrene (BP)-induced forestomach tumours in mice [20]. However, our previous study using a short-term (6 weeks) AOM-induced colon cancer in rats showed that diet containing 2% C. longa failed to affect the formation of ACFs (data not shown). In this study, diet containing extracts from the equivalent of 4% by weight of C. longa significantly reduced the formation of total ACFs (P<0.01), with three to four aberrant per foci (P<0.01). There was a reduction in the formation of ACFs in rats fed with diet containing 4% by weight of B. pandurata, but it was not significantly different compared to the control group (Fig. 6).
Fig. 5

Body weight (g) of rats fed with control AIN diet (AIN) was not different from those fed with diet containing AIN mixed with extract prepared from the equivalent of 4% by weight of dried rhizomes of B. pandurata (BP) or C. longa (CL) throughout 13 weeks of experiment (n=14)
Fig. 6

ACF formation in the colon of rats treated with different diets: AIN-93 (AIN), AIN with added extract of Boesenbergia pandurata (BP) and Curcuma longa (CL). Extracts of BP and CL were prepared from 4% of dried material of B. pandurata and C. longa respectively. Values are means±SD, n=14. Asterisk indicates significant difference from control group of the same column (P<0.01)

The antiproliferative activity of panduratin A, and possibly other bioactive compounds in B. pandurata, to block cells in the G0/G1 cell-cycle phase and to prevent apoptosis may have benefits for reducing the formation of ACFs in this in vivo study. In addition, extracts of B. pandurata have also been shown to induce mammalian phase-II chemoprotective and antioxidant enzymes [21], and panduratin A has been reported to have antimutagenic activity due to induction of quinone reductase (QR) [9]. Phase II enzymes, such as gluthatione S-transferase (GST) and QR, are important in cellular defence and metabolism, including detoxification of electrophilic species and thereby prevention of carcinogenesis [22]. In this respect, panduratin A may offer similar protective effects. Another anticancer activity of the extract that was reported by Yun et al. [11] was that panduratin A inhibited the expression of COX-2 in HT-29 cells. The upregulation of COX-2 has been correlated with pathological processes of inflammation and tumour development [23].

The results of this study show that panduratin A isolated from B. pandurata has benefits as a chemoprevention agent to inhibit the proliferation of cancer cells by blocking cells in G0/G1 phase and by inducing apoptosis. The extract of B. pandurata reduced the formation of ACFs in a colon cancer model study although the difference was not significant. To the best of our knowledge, this was the first animal study conducted on B. pandurata. Further studies need to be carried out to elucidate the anticancer activity of B. pandurata. As panduratin A appeared to be more sensitive toward MCF-7 breast cancer cells than HT-29 colon cancer, different animal models such as a breast cancer study would be beneficial for improving the understanding of the pharmacological actions of this plant and its active compounds.


The authors thank Dr Yoji Hayasaka of the Australian Wine Research Institute for LC-MS data, Phil Clements of the Department of Chemistry at the University of Adelaide for NMR data and Dr Lindsay Dent of the Department of Molecular Biosciences of the University of Adelaide for permission to use the FACScan equipment. This project was funded by AusAID and The Adelaide University and CSIRO Collaborative Grant Program 2000.

Copyright information

© The Japanese Society of Pharmacognosy and Springer 2006