- Email: [email protected]
Willow bark extract (BNO1455) and its fractions suppress growth and induce apoptosis in human colon and lung cancer cells
Cancer Detection and Prevention 31 (2007) 129–139 www.elsevier.com/locate/cdp
Willow bark extract (BNO1455) and its fractions suppress growth and induce apoptosis in human colon and lung cancer cells Katarina Hostanska PhDa,*, Guido Ju¨rgenliemk PhDb, Gudrun Abel PhDc, Adolf Nahrstedt PhDb, Reinhard Saller MDa a
University Hospital Zu¨rich, Department of Internal Medicine, Institute for Complementary Medicine, FGel 102, Ra¨mistrasse 100, CH-8091 Zu¨rich, Switzerland b University of Muenster, Institute of Pharmaceutical Biology and Phytochemistry, Hittorfstrasse 55, D-48149 Muenster, Germany c Bionorica AG, Phytoneering Company, Department of Research and Development, Kerschensteinerstrasse 11-15, D-92318 Neumarkt, Germany Accepted 3 March 2007
Abstract Background: Recently, there have been extensive efforts to evaluate the chemopreventive role of substances present in natural products. The aim of this study was to examine the effects of the main groups of compounds (salicylalcohol derivates, flavonoids, proanthocyanidins), and salicin isolated from willow bark extract BNO 1455 on proliferation and apoptosis in human colon and cancer cells. Methods: We used human colon cyclooxygenase-2 (COX-2)-positive HT 29 and (COX-2)-negative HCT 116 or lung COX-2 proficient A 549 and low COX-2 expressing SW2 cells. After treatment for 72 h with various concentrations of single substances and acetylsalicylic acid (ASA) as control, inhibition of cell growth and cytotoxicity were measured by colorimetric WST-1 assay and propidium iodide uptake by flow cytometry, respectively. Apoptotic cells were identified by annexin V adhesion using flow cytometry. Results: Studies on dose-dependent effects of BNO 1455 and its fractions showed anti-proliferative activity of all compounds with 50% maximal growth inhibitory concentrations (GI50) between 33.3 and 103.3 mg/ml for flavonoids and proanthocyanidins fractions and 50.0–243.0 mg/ml for salicylalcohol derivates and extract. Apoptosis induction was confirmed by annexin V adherence and analysis of cell morphology based on light scattering characteristics using flow cytometry in all cell lines at GI50. Conclusions: We showed that willow bark extract BNO 1455 an its fractions inhibit the cell growth and promote apoptosis in human colon and lung cancer cell lines irrespective of their COX-selectivity. # 2007 International Society for Preventive Oncology. Published by Elsevier Ltd. All rights reserved. Keywords: Willow bark extract; Salix purpurea; Salicin; Polyphenols; Growth inhibition; Apoptosis; Colon cancer; Lung cancer; Cyclooxygenase-2; Cell lines; Cell growth assay; Cell viability; Cytotoxicity
1. Introduction One of the most important public health and medicinal aims of the new century is the prevention of cancer. For the past two decades, there have been extensive efforts to evaluate the chemopreventive role of substances present in natural products. Several fruits, vegetables and plants with different pharmacological properties are known to be rich sources of chemical constituents that may have potential for the prevention or treatment of several malignancies [1,2]. Willow * Corresponding author. Tel.: +41 61 462 36 41; fax: +41 61 465 90 71. E-mail address: [email protected] (K. Hostanska).
bark (Salix species) has been used throughout the world for centuries as an anti-pyretic and analgesic. Its active constituents, salicin, and its derivates were widely used by 19th century physicians to treat rheumatic fever, different kinds of pain, including back pain, toothache, and headache [3]. In the late 19th century, acetylsalicylic acid (ASA) was synthesised based on the chemical structure of the willow bark constituent, the prodrug salicin (SAL) [4]. Nowadays, synthetic ASA is used not only as an analgesic and antipyretic, but also to prevent myocardial infarctions, strokes, and colorectal cancer. The relation between non-steroidal anti-inflammatory drug (NSAID) use and lung cancer has been examined in several epidemiological studies. Regular
0361-090X/$30.00 # 2007 International Society for Preventive Oncology. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.cdp.2007.03.001
130
K. Hostanska et al. / Cancer Detection and Prevention 31 (2007) 129–139
ASA use appeared to be inversely associated with lung cancer and mortality [5,6]. The mechanisms by which NSAIDs exert anti-tumor effects remain unclear. One potential mechanism for the anti-neoplastic effect of NSAIDs involves inhibition of prostaglandin (PG) synthesis [7]. Further, the NSAIDs might directly influence cell proliferation, cause cell growth arrest, and might trigger cell death apoptosis [8,9]. Pharmacological and clinical studies demonstrated that hydroethanolic willow bark extract had comparable antiinflammatory activities to ASA and showed anti-nociceptive and anti-pyretic activities in various animal models in doses that did not damage the gastric mucosa [10]. Like other herbal extracts, willow bark extracts contain a variety of active constituents such as glycosides of salicylic alcohol and their derivates, aromatic aldehydes and acids, flavonoids and tannins [11,12] so that the precise mode of action of willow bark extract is not yet known. The extract inhibits the production of inflammatory mediators such as cyclooxygenase (COX)-2 mediated PGs, leukotrienes, and cytokines [13] and certain constituents of the bark of Salix species (SAL, tremulacin) possess anti-inflammatory effect in the hen’s egg chorioallantoin membrane test system [14]. In recent years, it has increasingly been recognized that malignancy may not exclusively result from enhanced cell proliferation but also from decreased physiological cell death, apoptosis [15]. Apoptotic induction has been a new target for innovative mechanism-based drug discovery [16]. Certain products from plants are known to induce apoptosis in malignant cells [1,2,17]. It is thus considered important to screen for apoptotic inducers from plants, either in the form of crude extracts or as group of components/and or components isolated from them. Some herbalists recommend willow bark extract as a natural substitute for ASA to achieve these same benefits without side effects. However, there are no data on the use of willow bark extract in the prevention of colorectal/and or lung cancer. This study, therefore, was designed to evaluate the effect of willow bark extract, its component SAL and its isolated phytochemically characterized fractions on human colon and lung cancer cells. Considering the inhibition of PG synthesis as a key determinant in the chemopreventive and anti-neoplastic effect of NSAIDs, we investigated representative cells lines from each group. We selected colon carcinoma HCT 116 which lacks COX transcript and
does not produce PGs even when exogenously stimulated in contrast to HT 29 which produces PGs and small cell lung cancer (SCLC) SW2 cells with negligible COX-2 basal level as compared to COX-2 over expressing non-small cell lung cancer (NSCLC) A 549 cells.
2. Materials and methods 2.1. Samples preparation Willow bark extract was prepared using a drug extract ratio [DER; mass of the herbal drug (starting material) to the mass of the resulting native herbal drug preparation (=native extract)] of 6–12:1 and 70% (v/v) ethanol as extracting agent. According to the certificate of analysis, the extract batch number 01813 was characterized by 16.9% total content of SAL. For fractionation 101 g of a 70% ethanolic Salicis cortex native extract (Salix purpurea; Salicaceae; AE 0665/2, Fa. Bionorica) was separated by LC on SephadexLH201 (column: 90 cm 5.5 cm) in four consecutive portions. Eluates were tested by TLC [silica gel; eluent: ethylacetate/formic acid/water 90:5:5; detection was by Naturstoffreagent (diphenylboryloxyethylamine), vanillin/ hydrochloric acid and anisaldehyde/sulphuric acid]. The Sephadex-column was started using the eluent ethanol/water 3/1 (flow: 20 ml/min; 30 ml/test tube). Fractions containing salicylic alcohol derivatives (fraction 1, F1: 0–1365 ml) and flavonoids (fraction 2, F2: 1365–4220 ml) were combined. At the first detection of dimeric proanthocyanidins (beginning of fraction 3), chromatography was stopped and acetone/water 7/3 was used as eluent until the eluate was colourless and no further compounds could be detected by TLC. The combined eluates (fraction 3, F3) contained the oligomeric and polymeric flavan-3-ols. Total recovery was 98.2% of native extract. All reagents were purchased from Merck (Darmstadt, Germany). For characterization, the three fractions as well as the extract were analysed for salicylic alcohol derivatives (according to S. cortex, Ph. Eur. [18]), total phenolics (phenols reacting with molybdatophosphoric acid), and total tannins (phenols binding to skin powder) (according to the method 2.8.14 Ph. Eur. [18], modified by Glasl [19]), proanthocyanidins (according to Crataegi fructus, Ph. Eur.
Table 1 Results of fractionation and phytochemical characterization
Elution volume (ml) Amount of recovered extract (%) Salicylic alcohol derivates (%) Total phenols (%) Total tannins (%) Proanthocyanidins (%) Flavonoids (%)
Extract
Fraction 1
Fraction 2
Fraction 3
16.9 15.7 8.5 9.2 0.6
0–1365 83.0 20.5 5.7 1.0 0.9 0.1
1365–4220 10.3 3.9 51.5 26.9 30.8 3.3
4220–8500 6.7 0.0 69.8 59.7 73.8 0.3
Percentage of recovered extract is related to the native extract.
K. Hostanska et al. / Cancer Detection and Prevention 31 (2007) 129–139
[18]; anthocyanidin method) and flavonoids (according to Calendulae flos, Ph. Eur.) [18]. The results are presented in Table 1. Standard samples acetylsalicylic acid (ASA), salicin (SAL) were purchased from Sigma (Buchs, Switzerland) and flavonoid quercetin.dihydrate (QUE) was obtained from Alexis Corporation (Lausen, Switzerland). All pure compounds were dissolved in DMSO and stored in aliquots at 20 8C. 2.2. Cell lines and culture conditions The human colon adenocarcinoma cell lines HT 29, HCT 116 and A 549 (NSCLC) cells were obtained from American Type Culture Collection (ATCC, Rockville, MD). The small cell lung cancer (SCLC) cell line SW2 was obtained from the Dana–Farber Cancer Institute (Boston, MA). For routine maintenance, all cell lines were grown in RPMI 1640 medium supplemented with 10% fetal calf serum (FCS), 2 mM L-glutamine and penicillin/streptomycin 1%. All cell culture reagents were from Sigma. Cells (HT 29, HCT 116, A 549) were grown as monolayer and SW2 as floating aggregates at 37 8C in humidified atmosphere with 5% CO2. The amount of FCS was reduced to 2% in the assay medium for all cells tested. The final concentration of DMSO in culture medium during the treatment of cells did not exceed 0.5% (v/v). Vehicle treated cells (0.5% DMSO) were always included in control samples. This concentration of 0.5% DMSO did not affect the viability, proliferation, or apoptosis of the investigated cells. 2.2.1. Flow cytometric detection of Bcl-2 gene and COX-1/COX-2 status The basal level of anti-apoptotic gene Bcl-2 and COX-1/ COX-2 was assessed by direct intracellular single or double immunofluorescence staining according to manufacturer’s instructions (BD Biosciences, San Jose, CA). The permeabilized non-treated cells were stained with PEconjugated anti-Bcl-2 antibody or with FITC-conjugated anti-COX-1 and PE-conjugated anti-COX-2 antibodies. Background fluorescence was determined by isotype matched (IgG1) control antibodies. Cells were analyzed with FACScalibur flow cytometer (BD) equipped with CellQuest software. 2.3. Cell growth assay Cells seeded in 96-well plates at a density of 5 103 cells/well were exposed to different concentrations of individual substances in triplicates for 72 h. Cell growth was indirectly measured by use of a colorimetric assay based on the cleavage of the tetrazolium salt WST-1 (Roche Diagnostics, Rotkreuz, Switzerland) in viable cells as described [20]. Absorbance values at 450 and 650 nm (reference wavelength) were measured in a micro-titer plate
131
ELISA reader. Each set of experiments was repeated three times. The test medium alone was used as background control. The growth inhibition rate was calculated as the percentage of parallel untreated controls. As controls, ASA, SAL, and flavonoid QUE in pharmacological concentrations were used. The direct reductive potential of all the substances was tested in a cell-free system. The concentration of substances that reduced cell growth by 50% (GI50) was calculated from the dose–response relationship by regression analysis using Microsoft Excel computer software. 2.4. Analysis of cell viability Plasma membrane integrity of treated cells (2 104/ well) after 72 h was investigated with propidium iodide (PI) at a final concentration of 5 mg/ml for 15 min. Bivariate dot plots of forward scatter (FCS) versus red fluorescence intensity were performed to discriminate nonviable (PI + events) and viable (PI events) cells. In each sample, 104 cells were analyzed on FACScalibur flow cytometer (BD) equipped with CellQuest analysis software. 2.5. Cell death assays 2.5.1. Morphologic assessment of apoptosis To investigate morphological changes of the cells typical for apoptosis, the cells were subjected to FSC and side (SSC) light scatter analysis by flow cytometry. Cell morphology based on light scattering characteristics, cell size according to FSC, and granularity to SSC was investigated simultaneously in the cells stained by annexin V/PI. 2.5.2. Apoptosis assays After treatment of cells (2 104) with BNO, fractions F1, F2, F3, or DMSO vehicle for 72 h, cells were harvested by trypsinization, washed in PBS, and stained with annexin V-FITC conjugate and PI according to the manufacturer’s protocol (Bender Medsystems Diagnostica, Vienna, Austria) as previously described [21]. Cells (104) were then analyzed by flow cytometry using CellQuest acquisition and analysis software. 2.6. Statistical analysis The experiments were repeated at least three times. Cytotoxic and growth inhibitory effect of individual substances on all four cell lines were compared with those of the untreated controls by Student’s two-tailed test. GI50 values were obtained by linear or exponential regression analysis of the dose–response relationship using the Microsoft Excel software. Probability values equal to or less than 0.05 were considered significant. The comparison of histograms was evaluated statistically using the Kolmogorov–Smirnov (K–S) two samples test for overlaid histograms.
132
K. Hostanska et al. / Cancer Detection and Prevention 31 (2007) 129–139
3. Results 3.1. Dose-dependent cell growth inhibition of SAL, BNO 1455, fractions F1, F2 and F3 Growth inhibitory effect of substances was measured by WST-1 assay which measures the dehydrogenase activity of
viable cells. ASA (1–5 mM; 180–900 mg/ml) and flavonol QUE (50 and 100 mM; 16.9 and 33.8 mg/ml) were used as positive controls. ASA caused reduction in cell growth between 14.3 and 88.9% (Fig. 1). There are very few data concerning the effect of the prodrug SAL. Therefore, in addition, the inhibitory effect of SAL (10–100 mM; 2.86– 28.65 mg/ml) was tested in parallel. The most sensitive cell
Fig. 1. Effect of exposure to ASA (1–5 mM), Q (50 mM, 100 mM) on cell growth of HT 29 (A), HCT 116 (B), A 549 (C) and SW2 (D) cells. Cells (5 103/ well) were grown for 72 h in the presence of indicated concentrations of ASA and Q and the cell growth inhibition was assessed by WST-1. Untreated cells (in medium only) and cells incubated in the presence of vehicle (0.5% DMSO) were used as controls. The effect of SAL on the cell growth is shown as individual dose–response curves for each cell line (E). Average growth inhibition values (S.D.) from triplicates of three independent experiments are shown (*p < 0.01 effect of SAL vs. untreated control; t-test).
K. Hostanska et al. / Cancer Detection and Prevention 31 (2007) 129–139
line to the exposure of SAL was lung SW2 with a GI50 of 17 mM. The growth of these cells was reduced by 75% at 20 mM concentration and completely inhibited at 50 mM (Fig. 1). The growth of colon HT 29, HCT 116 and lung A 549 cells was significantly reduced by higher concentration of SAL (50 mM). GI50 values of SAL for these three cell lines were approximately two-fold higher than in SW2 and being very similar (Table 2). The effects of BNO 1455, F1, F2, and F3 were studied as a dose–response experiments after 72 h at concentrations ranging between 25 and 400 mg/ ml. Extract BNO 1455 and F1 reduced the growth of three cell lines (HT 29, HCT 116 and A 549) significantly above 100 mg/ml. The number of SW2 cells was reduced at 50 mg/ ml BNO 1455 and F1 concentration. For a significant reducing effect of F2 and F3 on cells only 50 mg/ml doses were required. The substances induced suppression of growth of individual cell lines from three independent experiments and are presented in Fig. 2. This comparative study showed differences between the patterns of the antiproliferative responses of the four cell lines towards the BNO 1455 extract and fractions F1, F2, and F3. The GI50 was used to compare the sensitivity of individual cell lines to
133
single substances (Table 2). Proanthocyanidins enriched fraction F3 was the most active in all four cell lines with half maximal inhibitory concentrations between 33.3 and 55.0 mg/ml. Extract BNO 1455 exerted inhibitory effect in three cell lines (HT 29, HCT 116 and A 549) with GI50 values between 128.3 and 211.7 mg/ml which correspond to the total SAL concentrations of 21.7–35.8 mg/ml. For the most sensitive SCLC SW2 the GI50 concentration of BNO 1455 was in the same range as for the most active compound F3, i.e. about 50 mg/ml. There were differences in the susceptibility of HT 29 and HCT 116 colon carcinoma cells towards extract BNO 1455 and fraction F1 as mirrored in their GI50 (211.7 mg/ml versus 128.3 mg/ml; p = 0.023 for BNO 1455 and 122.7 mg/ml versus 71.7 mg/ ml; p = 0.048 for F1). F2 exerted comparable effect to F3 in three (HT 29, HCT 116, and SW2) carcinoma cells. The effect of BNO 1455 and F1 differed significantly ( p < 0.001) between the NSCLC A 549 and SCLC SW2 cells. The statistically evaluated differences between the effect of substances on individual cell lines expressed with GI50 values and corresponding cell survival rates are shown in Table 2.
Table 2 Growth inhibitory potency and cytotoxicity of BNO 1455, F1, F2 and F3, and controls ASA and SAL in colon HT 29, HCT 116 (A) and lung A 549, SW2 (B) cancer cells Substance
Part A BNO 1455 F1a F2a F3a F1 F2 F3 ASA (mM) SAL (mM)
GI50 value (mg/ml)
Cytotoxicity at GI50 value (% of viable cells)
HT 29
HCT 116
HT 29
211.7 27.5 a 175.7 22.8 21.8 2.8 14.1 1.8 122.7 20.5 b 71.7 10.4 d 55.0 5.0 d 468.0 72, (2.6 0.4) (11.5 0.9) 103, (40.2 3.1)
128.3 29.3 b 106.5 4.9 15.7 3.5 8.6 1.9 71.7 17.6 d 66.7 12.6 d 33.3 7.6 e 360.0 36, (2.0 0.2) (9.7 0.8) 103, (34.5 2.8)
107.8 2.3
75.6 4.3
103.2 3.7 93.4 2.8 98.5 3.1 88.4 2.8 b 84.3 3.5 b
94.2 2.5 72.9 3.3 89.7 5.2 87.0 3.6b 55.9 4.2b
100 (77.6 3.9) c
Medium only Substance
Part B BNO 1455 F1a F2a F3a F1 F2 F3 ASA (mM) SAL (mM) Medium only
GI50 value (mg/ml)
HCT 116
100 (72.8 5.4) c
Cytotoxicity at GI50 value (% of viable cells)
A 549
SW2
206.7 35.6 a 171.6 29.5 21.3 3.7 13.8 2.4 243.3 20.8 a 103.3 21.6 b, d 51.7 12.6 d 396.0 54, (2.2 0.3) (10.9 0.8) 103, (38.3 2.9)
52.8 8.3 d 43.8 6.9 5.4 0.8 3.5 0.6 50.0 11.2 d 63.3 13.5 d 40.7 5.1 d, e 612.0 90, (3.4 0.5) (4.9 0.5) 103, (17.4 1.9)
A 549 25.4 6.5
84.6 7.2 86.8 3.9 82.1 4.3 101.7 2.5 b 60.4 6.3 b 100 (81.6 4.5) c
SW2 88.5 3.7
95.3 4.1 84.2 2.8 64.3 5.7 92.5 3.2b 64.5 5.7b 100 (79.3 3.1) c
Letters a, b, c, d: values are means S.D. of three experiments performed in triplicate. All the values are statistically different ( p < 0.05), except those marked with the same letters (a, b, c or d). a Corresponding concentrations of the fractions F1, F2 and F3 to the established GI50 of extract BNO 1455. Not involved in statistical evaluation. b Survival of cell treated with ASA (5 mM) and SAL (50 mM). c In parentheses the survival values of untreated cells which were normalized to 100% are expressed.
134
K. Hostanska et al. / Cancer Detection and Prevention 31 (2007) 129–139
Fig. 2. Cell growth inhibitory effect of BNO 1455, F1, F2 and F3. Cell proliferation was determined by WST-1 assay after 72 h treatment. Untreated cells and cells incubated in the presence of vehicle (0.5% DMSO) were used as controls. In all experiments the values of vehicle treated cells did not exceed the intraexperimental variance which was less than 10%. Individual dose–response curves for each cell line from triplicates repeated at least three times are expressed as mean (S.D.). BNO 1455, F1 reduced growth of HT 29, HCT 116 and A 549 significantly ( p < 0.05) at concentrations 100 mg/ml and F2, F3 at concentrations of 50 mg/ml and/or higher. The growth of SW2 was significantly ( p < 0.05) reduced by all four substances at 50 mg/ml concentration.
3.2. Cytotoxicity of BNO 1455, F1, F2 and F3 toward colon HT 29, HCT 116 and lung A 549, SW2 cancer cells Because all substances showed an anti-proliferative effect they were tested, in addition, for their potential toxicity. The tumor cell lines (HT 29, HCT 116, A 549, SW2) were exposed to dose-increasing concentrations (0– 200 mg/ml) of substances for 72 h. We inspected cells during the treatment by inversion microscopy for morphological changes related to the cytotoxicity of compounds. Live and dead cells discrimination were estimated by PI uptake. The viability results (normalized to untreated controls) from flow cytometry detection are presented in Fig. 3. As controls ASA (5 mM) and SAL (50 mM) concentrations were investigated in parallel (Table 2). Sensitivity of the cells to the exposure of substances differed as shown in Fig. 3. We estimated the cell survival according to following criteria. More than 90% viable cells were considered to be unaffected, 80–90% as moderately affected, and values of less than 80% viable cells were ascribed to cytotoxic effects of the compounds. Extract
and F1, F2, and F3 exerted a very low toxic effect on colon HT 29 cells up to highest concentrations of 200 mg/ml. Cell survival was between 78.5 and 102% for individual substances. Survival of colon HCT 116 and lung A 549 cells was affected at concentrations above 100 mg/ml of BNO 1455, F1, F2 and F3. The most sensitive to the cytotoxic effect of tested substances was the SCLC SW2 cell line which survival was affected by all substances at 100 mg/ ml concentration. However, at GI50 of the four substances, the survival of SW2 cells was affected only by the proanthocyanidine enriched fraction F3 (Table 2). 3.3. Activation of apoptosis by the BNO 1455 and fractions F1, F2 and F3 in human colon and lung cancer cells Having established that BNO 1455 extract, F1, F2, and F3 were growth inhibitory to the colon and lung cancer cells, we investigated whether these substances induced apoptosis. A first approach to the mode of cell death by drugs was conducted by flow cytometric cell scatter characteristics.
K. Hostanska et al. / Cancer Detection and Prevention 31 (2007) 129–139
135
Fig. 3. Effect of BNO 1455, F1, F2 and F3 on the viability of individual cancer cells. Cells were treated for 72 h, stained with PI (5 mg/ml) and analyzed by flow cytometry as described in Section 2. Survival of cells in medium OR (72.8–81.6%) was normalized to 100%. The survival of vehicle-treated cells did not differ from control cells. The viability of HT 29, HCT 116, A 549 and SW2 cells was 88.4%, 87%, 101.7% and 92.5% for control substance ASA (5 mM). Individual dose–response effect for each cell line from experiments that were repeated twice with similar results are expressed as mean. S.D. values were less than 15%.
Decrease in the cell size coinciding with the decrease in FSC and increase or no change in the granularity of cells coinciding with the increase in SSC are characteristic hallmarks of apoptotic cells. We investigated the effect of substances at concentrations close to their GI50 values (Table 2) with ASA (5 mM) as positive control. The two-parameter histograms plotting cell size versus granularity showed morphological changes typical for apoptosis on colon (HT 29, HCT 116) and lung (A 549, SW2) cancer cells. Morphological changes of apoptosis cell shrinkage and chromatin clumping were well reflected by reduced ability to scatter light in the forward direction and an increase or no change in the 90% light scatter after the treatment with BNO 1455, F1, F2, and F3. These changes were more pronounced by treatment of cells with F2 and F3 (data not shown). The sensitivity of colon (HT 29, HCT 116) and lung (A 549, SW2) cells to apoptosis induction was further examined by flow cytometric determination of annexin V-positive cells at 72 h after drug exposure. An increase in cellular staining with FITC-conjugated annexin V served as an early marker for apoptosis. The cells were simultaneously stained with PI to score for overt loss of cell membrane integrity. The double staining procedure distinguishes early-stage apoptotic cells (annexin V-positive, PI negative) from late-stage apoptotic
cells (annexin V-positive, PI positive). As shown in Fig. 4, treatment of the cancer cells with GI50 concentrations induced apoptotic cell death in all cell lines. Colon carcinoma cells HCT 116 responded by more than 80% of apoptotic cells. The degree of secondary necrotic cells was negligible (max. 4.3%). Spontaneous apoptosis in this cell line was 28%. In HT 29 colon cancer cells, the fraction of early apoptotic cells lay between 46.9 and 59.9% as compared to the untreated control with 16.4% apoptotic cells. Late apoptotic cell fractions occurred between 10 and 26%. SW2 cells exposed for 72 h expressed the mixtures of early apoptotic cells (22.9–31.3%) and secondary necrotic fractions (12.3–32.8%). In untreated controls 12.3% apoptotic and 11% necrotic cells were detected. The apoptotic fractions of A 549 NSCLC cells after the treatment lay between 62.8 and 81.6% compared to 24.3% in untreated controls. Necrotic cell fractions ranged between 9.2 and 16.7%. 3.4. Characterization of basal levels of Bcl-2 protein in four cancer cell lines and constitutive expression of COX-2 in lung SCLC SW2 cells In order to obtain a better understanding of the apoptosis inducing effect of willow bark extract on colon and lung
136
K. Hostanska et al. / Cancer Detection and Prevention 31 (2007) 129–139
Fig. 4. Apoptosis-inducing effect of BNO 1455, F1, F2 and F3 on individual cancer cell lines expressed as percentage of PS-expressing [early apoptotic: annexin V-positive/PI-negative (A+/PI ) and late apoptotic/necrotic: annexin V-positive/PI-positive (A+/PI+)] cells. Cells (20 103) were treated with substances at their half-maximal inhibitory concentrations (GI50) for 72 h. As controls, cells treated with ASA (5 mM) were used. A representative of two experiments with comparable results is shown.
cancer cells, we determined the basal levels of COX-2 isoenzyme and of anti-apoptotic Bcl-2 protein. We determined the constitutive expression of COX-1/COX-2 levels in untreated SW2 cells. With SW2 cells very low expression of COX-2 (4%) was found in contrast to the constitutively expressed COX-1. In addition the basal levels of anti-apoptotic Bcl-2 in untreated cells were quantified by flow cytometric determination. Bcl-2 was abundantly expressed in the SCLC SW2 (92.9%), but was barely detectable in the NSCLC A 549 (6.4%) and colon HCT 116 (16.4%) and was negligible in HT 29 (1.9%) (Table 3).
prodrug SAL is primarily postulated as the active component, but there are other groups of components with pharmacological activity, particularly polyphenols (phenolic acids, hydrolysable tannins, proanthocyanidins, and flavonoids) which have shown anti-oxidant and anti-carcinogenic effects in several studies [1,22]. Besides salicylic alcohol derivates (including SAL), considerable amounts of total Table 3 Basal level of COX-2 receptor and apoptotic genes p53 and Bcl-2 in colon and lung cancer cells Cell lines
COX-2 status
Bcl-2 status (%)
p53 status
4. Discussion
HT 29 HCT 116 A 549 SW2
High expression [9] No expression [9] High expression [29] 4%
0.2 16.4 6.4 92.9
Mutant [33] Wt [33] Wt [34] Mutanta
It is a huge task to identify a potent single chemical and/ or group of compounds from the complex mixture of substances present in a plant. Willow bark extracts have been used world-wide as an anti-pyretic and analgesic. The
a Mutation on the nucleotide 511 of exon 5; codon 171 changed from GAG (Glu) to TAG (Cys): (written personal communication from Dr. Zangemeister-Wittke U., Division of Oncology, University Hospital Zu¨rich).
K. Hostanska et al. / Cancer Detection and Prevention 31 (2007) 129–139
phenols, total tannins, proanthocyanidins, and flavonoids were present in the native extract and its fractions as shown in Table 1. They all could contribute to the pharmacological activity of the native extract. We investigated the response of two colon cancer cells HT 29 (COX-2 expressing), HCT 116 (lacking COX-2 expression) and two lung cancer cells A 549 (NSCLC; COX2 high expression), SW2 (SCLC; low expression of COX-2) to outline similarities and differences between them to the exposure of BNO 1455 and its fractions. Our data demonstrates that willow bark extract and its fractions F1, F2 and F3 exert two profound effects on the human colon cancer cells (HT 29, HCT 116) and lung cancer cells (A 549, SW2): (a) they inhibit their proliferation, and (b) they reduce their survival by inducing cell death by apoptosis. The growth inhibitory effects of four substances were dosedependent (Fig. 2), but there were both differences and similarities between the response of cells to individual fractions as mirrored in their GI50 values (Table 2). Compounds F2 and F3 reduced the growth of all four cancer cells to a comparable degree. Variations of GI50 were only 1.6-fold for both compounds. Our results are in agreement with studies of others on the anti-proliferative effect of different flavonoids and/or proanthocyanidins [22,23] and comparable to the effect of flavonoid QUE used as control in this study (Fig. 1). To our knowledge, this is the first report on growth inhibition and apoptosis in carcinoma cell lines induced by willow bark extract, its fraction of salicyl alcohol derivates F1 and SAL. Because for most NSAIDs, an anti-inflammatory and COX-2 inhibitory effect of willow bark extracts was reported [13], we compared its effect to that of the well established NSAID ASA. In parallel, the effect of SAL, the component used to standardize the commercial extracts, was investigated. We observed a dose-dependent reduction of cell growth (Fig. 1). Based on comparative GI50 values of SAL (17–40 mM) and ASA (2.0–3.4 mM) about five-fold (SW2) or 15-fold higher concentrations (HT 29, HCT 116, A 549) of SAL were required than needed for ASA. In most cases the pharmacological activity of plant extracts is not assigned to single components and often not all active ingredients are known. Therefore, it was of interest to compare the effect of the whole extract BNO 1455 (16.9% SAL) to SAL and ASA. There were differences in the response of cells to BNO 1455. NSCLC A 549 and colon HT 29 cells required about fourfold and colon HCT 116 (two-fold) higher concentration in comparison to gain the same response as highly responsive SW2 SCLC cells. As shown in Table 2, the corresponding values of SAL (8.4–19.5 mg/ml) in the extract were very low. Taking the molecular weight into consideration, there were differences of about two decades of logarithmic scale in comparison to the single compound SAL (4.7–11.2 mg/ml). In addition, the growth inhibitory effect of BNO 1455 seems to be superior to the effect exerted by ASA. Overall about two-fold higher doses of ASA were required in comparison to the whole extract to obtain 50% inhibition of cells growth.
137
Products obtained via plant material extraction contain complex mixture of chemicals whose individual biological activities may be considerably different when present in a mixture because of their possible interaction [24,25]. Therefore it can be argued that other constituents (e.g. polyphenols–proanthocyanidins) contributed to the antiproliferative effect of willow bark extract. This observation is in agreement with recently published studies on antiinflammatory activity of Salix extract [26] for which the authors also postulated the contribution of additional, hitherto unknown constituents. Cytotoxicity of substances determined by PI uptake on four cancer cell lines after 72 h treatment was only moderate (Fig. 3). Results of this study showed that the cell growth inhibition is due to the interference with cell proliferation rather than direct cytotoxicity of substances. Discovery of active compounds from natural products with cytocidal (e.g. apoptotic) rather than cytotoxic properties is of great interest for cancer treatment [17]. Apoptosis is a process that regulates cell number or eliminates damaged cells. Deficiencies in apoptosis also contribute to carcinogenesis by creating a permissive environment for genetic instability and accumulation of gene mutations conferring resistance to cytotoxic anticancer drugs and radiation [27]. To further substantiate the cell growth inhibitory effect of BNO 1455 and fractions F1, F2, and F3, the apoptotic cell death was assessed. As shown in Fig. 4 all substances induced apoptosis in the cells. The sensitivity of individual cells to BNO 1455, F1, F2 and F3 as obtained in apoptosis assays correlated well with their growth inhibitory responses. Apoptosis induction by BNO 1455 was accompanied by reduction in cell viability only in A 549 cells. The apoptosis induced by all four substances was preceded by a secondary necrosis in cancer cells. However, in the colon HCT 116 at least and in lung SW2 most necrotic cells were detected (Fig. 4). BNO 1455 and F1-, F2-, F3-induced apoptosis was comparable to that induced by 5 mM (0.9 mg/ml) ASA, but the concentrations of substances (GI50) required for this apoptosis induction were about 4–15-fold lower. ASA and related NSAIDs have attracted considerable attention as compounds that might be of potential benefit in the chemoprevention of cancer. The best defined molecular target for ASA and other NSAIDs is COX-2 which is considered to play an important role in carcinogenesis by regulation of apoptosis. Furthermore NSAIDs exert antiproliferative effects on colon and lung carcinoma cells [7,8,28]. First we looked for any relationship in the growth inhibition and apoptosis between the cells with different constitutive expression of COX-2. The overexpression of COX-2 in NSCLC A 549 [29] and colon HT 29 [9] and its lack in HCT 116 [9] cells were described by others (Table 3). Using flow cytometry, we determined the low basal level of COX-2 on SCLC SW2 cells (Table 3). This is in accordance with the finding that most of SCLC express a low level of COX-2 [30]. We found, that cell lines with deficiencies of COX-2 were more sensitive to the growth inhibitory and
138
K. Hostanska et al. / Cancer Detection and Prevention 31 (2007) 129–139
apoptotic effects of extract and salicylate fraction F1 than cells proficient in production of COX-2 (Tables 2 and 3). These facts support the COX-2-independent effects of BNO 1455 and its F1 fraction and are in agreement with findings that NSAIDs inhibited the proliferation and induced apoptosis in cancer cells by other non-COX-dependent mechanisms [31,32]. To further substantiate the differences we compared the constitutive levels of two important proteins involved in apoptosis namely p53 and Bcl-2. The state of p53 is pivotal for the response of tumour cells to chemo- and radiotherapy and plant-derived compounds which overcome a high apoptotic threshold in tumour cells lacking intact p53 [33]. The colon HT 29, HCT 116 and lung A 549 cells have been characterized [34,35] for their p53 status (Table 3). The SW2 lung cancer cell express a mutated p53 gene (personal written communication Dr. Zangemeister-Wittke U; Oncology Division, University Hospital Zu¨rich). Bcl-2 appears to function as an inhibitor of apoptosis. Over expression of Bcl-2 has been reported in different types of cancer and was considered to be responsible for the blockade of apoptosis [36]. We detected different basal levels of Bcl-2 by flow cytometry (Table 3). Bcl-2 was highly expressed only in SW2 cells in agreement to others [37]. The role of COX-2 in carcinogenesis and in the regulation of apoptosis is quite well established. Previous reports indicate complex, functional, and reciprocal interactions between the COX-2 and p53 systems and forced expression of COX-2 modulate the Bcl-2 expression [33,38]. The differences in responses between HT 29 (COX-positive) and HCT 116 (COX-negative) following exposure to the BNO 1455 and F1 could be attributed not only to the differences of COX-2 level but also to the p53 status of the cells. An association between COX-2 over expression and p53 mutations, or low level of p53 protein has been reported in human cancer cells. The HT 29 cells have mutant p53, while HCT 116 cells have wild-type [33]. NSCLC A 549 cells carry a high expression of COX-2 accompanied by wild-type p53. Despite their p53 status, these cells showed a lower sensitivity to the extract and F1, comparable to colon HT 29 with mutant p53. It has been reported that COX-2 could inactivate wild-type p53 [33]. Interestingly, the SW2 cells which carry mutant p53, high expression of Bcl-2 accompanied with low constitutive level of COX-2 responded mostly to all compounds. It was described that down-regulation of Bcl-2 in COX-2 negative colon cancer cell represents a potential mechanism by which ASA induces apoptosis [37]. In summary, these results suggest that all the compounds tested inhibit the growth and promote apoptosis of four cancer cell lines and these effects depend on the cell lines rather than on the property of drugs. NSAIDs are effective in preventing colon cancer. Currently available drugs, however, have side effects that make them unsuitable for widespread use in the prevention. In this study we showed that willow bark extract exerted a growth reducing effect in cancer cells by triggering
apoptosis. These abilities of willow bark extract suggest that it could be a promising cancer preventive compound. Additional studies are warranted to examine in vivo models in support of the above effects of willow bark.
Conflict of interest None.
References [1] Rao CV, Newmark HL, Reddy BS. Chemopreventive effect of farnesol and lanosterol on colon carcinogenesis. Cancer Detect Prev 2002; 26:419–25. [2] Zhang H, Spitz MR, Tomlinson GE, Schabath MB, Minna JD, Wu X. Modification of lung cancer susceptibility by green tea extract as measured by the comet assay. Cancer Detect Prev 2002; 26:4144–418. [3] Hedner T, Everts B. The early clinical history of salicylates in rheumatology and pain. Clin Rheumatol 1998; 17:17–25. [4] Robbers JE, Speedie MK, Tyler VE. Pharmacognosy and pharmacobiotechnology. Baltimore: Williams & Wilkins, 1996. [5] Moysich KB, Menezes RJ, Ronsani A, Swede H, Reid ME, Cummings KM, et al. Regular aspirin use and lung cancer risk. BMC Cancer 2002; 2:31. [6] Akhmedkhanov A, Toniolo P, Zeleniuch-Jacquotte A, Koenig KL, Shore RE. Aspirin and lung cancer in women. Br J Cancer 2002; 87:49–53. [7] Vane JR. Inhibition of prostaglandin synthesis as a mechanism of action for aspirin-like drugs. Nat New Biol 1971; 231:232–5. [8] Shiff SJ, Koutsos M, Qiao L, Rigas B. Non-steroidal anti-inflammatory drugs inhibit the proliferation of colon adenocarcinoma cells: effects on cell cycle and apoptosis. Exp Cell Res 1996; 222:179–88. [9] Qiao L, Hanif R, Sphicas E, Shiff SJ, Rigas B. Effect of aspirin on induction of apoptosis in HT 29 human colon adenocarcinoma cells. Biochem Pharmacol 1998; 55:53–64. [10] Ma¨rz RW, Kemper F. Willow bark extract—effects and effectiveness. Status of current knowledge regarding pharmacology, toxicology and clinical aspects. Wien Med Wochenschr 2002; 152:354–9. [11] Thieme H. On the tannin content of willow cortex. Pharmazie 1968; 23:212–5. [12] Meier B, Sticher O, Julkunen-Tiitto R. Pharmaceutical aspects of the use of willows in herbal remedies. Planta Med 1988; 54:559–60. [13] Wagner I, Greim C, Laufer S, Heide L, Gleiter CH. Influence of willow bark extract on cyclooxygenase activity and on tumor necrosis factor alpha or interleukin 1 beta release in vitro and ex vivo. Clin Pharmacol Ther 2003; 73:272–4. [14] Albrecht M, Nahrstedt A, Laepke NP, Theisen NL, Baron G. Antiinflammatory activity of flavanol glycosides and salicin derivates from the leaves of Populus tremuloides. Planta Med 1990; 56:660. [15] Reed JC. Mechanisms of apoptosis avoidance in cancer. Curr Opin Oncol 1999; 11:68–75. [16] Senderowicz AM. Targeting cell cycle and apoptosis for the treatment of human malignancies. Curr Opin Cell Biol 2004; 16:670–8. [17] Taraphdar AK, Madhumita R, Bhattacharya RK. Natural products as inducer of apoptosis: implication for cancer therapy and prevention. Curr Sci 2001; 11:1387–96. [18] European Pharmacopeia (Ph Eur), 4th ed. Stuttgart: Deutscher Apotheker Verlag; 2002. [19] Glasl H. Zur Photometrie in der Drogenstandardisierung. DAZ 1983; 123:1979–87. [20] Ishiyama M, Tominaga H, Shiga M, Sasamoto K, Ohkura Y, Ueno K, et al. A combined assay of cell viability and in vitro cytotoxicity with a
K. Hostanska et al. / Cancer Detection and Prevention 31 (2007) 129–139
[21]
[22]
[23]
[24] [25]
[26]
[27] [28]
[29]
highly water-soluble tetrazolium salt, neutral red and crystal violet. Biol Pharm Bull 1996; 1:1518–20. Hostanska K, Hajto T, Fischer J, Mengs U, Weber K, Lentzen H, et al. Selective modulation of phosphatidylserine exposure on sub-populations of human peripheral blood lymphocytes by a plant lectin, Viscum album agglutinin (VAA)-I and its recombinant form (rVAA) in vitro. Cancer Detect Prev 1999; 23:511–23. Kuntz S, Wenzel U, Daniel H. Comparative analysis of the effects of flavonoids on proliferation, cytotoxicity, and apoptosis in human colon cancer cell lines. Eur J Nutr 1999; 38:133–42. Agarwal C, Sharma Y, Agarwal R. Anti-carcinogenic effect of a polyphenolic fraction isolated from grape seeds in human prostate carcinoma DU 145 cells: modulation of mitogenic signalling and cellcycle regulators and induction of G1 arrest and apoptosis. Mol Carcinogen 2000; 28:129–38. Williamson EM. Synergy and other interactions in phytomedicines. Phytomedicine 2005; 8:401–9. Fiebich BL, Chrubasik S. Effects of an ethanolic Salix extract on the release of selected inflammatory mediators in vitro. Phytomedicine 2004; 11:135–8. Khayyal MT, El-Ghazaly MA, Abdallah DM, Okpanyi SN, Kelber O, Weiser D. Mechanisms involved in the anti-inflammatory effect of a standardized willow bark extract. Arzneimittelforschung 2005; 55: 677–87. Green DR, Martin SJ. The killer and executioner: how apoptosis controls malignancy. Curr Opin Immunol 1995; 7:694–703. Hida T, Leyton J, Makheja AN, Ben-Av P, Hla T, Martinez A, et al. Non-small cell lung cancer cyclooxygenase activity and proliferation are inhibited by non-steroidal anti-inflammatory drugs. Anticancer Res 1998; 18:775–82. Hung WC, Chang HC. Methanolic extract of adlay seed suppresses COX-2 expression of human lung cancer cells via inhibition of gene transcription. J Agric Food Chem 2003; 51:7333–7.
139
[30] Yoshimoto A, Kasahara K, Kawashima A, Fujimura M, Nakao S. Characterization of the prostaglandin biosynthetic pathway in nonsmall cell lung cancer: a comparison with small lung cell cancer and correlation with angiogenesis, angiogenic factors and metastases. Oncol Rep 2005; 13:1047–57. [31] Hanif R, Pittas A, Feng Y, Koutsos MI, Qiao L, Staiano-Coico L, et al. Effect of non-steroidal anti-inflammatory drugs on proliferation and on induction of apoptosis in colon cancer cells by a prostaglandin-independent pathway. Biochem Pharmacol 1996; 52: 237–45. [32] Hostanska K, Vuong V, Rocha S, Soengas MS, Glanzmann C, Saller R, et al. Recombinant mistletoe lectin induces p53-independent apoptosis in tumour cells and cooperates with ionising radiation. Br J Cancer 2003; 88:1785–92. [33] Swamy MV, Herzog CR, Rao CV. Inhibition of COX-2 in colon cancer cell lines by celecoxib increases the nuclear localization of active p53. Cancer Res 2003; 63:5239–42. [34] Radhakrihsna Pillai G, Srivastava AS, Hassanein TI, Chauhan DP, Carrier E. Induction of apoptosis in human lung cancer cells by curcumin. Cancer Lett 2004; 208:163–70. [35] Zangemeister-Wittke U, Schenker T, Luedke GH, Stahel RA. Synergistic cytotoxicity of Bcl-2 antisense oligodeoxynucleotides and etoposide, doxorubicin and cisplatin on small cell lung cancer cell lines. Br J Cancer 1998; 78:1035–42. [36] Reed JC. Bcl-2: prevention of apoptosis as a mechanism of drug resistance. Hematol Oncol Clin North Am 1995; 9:451–73. [37] Yu HG, Huang JA, Yang YN, Huang H, Luo HS, Yu JP, et al. The effects of acetylsalicylic acid on proliferation, apoptosis, and invasion of cyclooxygenase-2 negative colon cancer cells. Eur J Clin Invest 2002; 32:838–46. [38] Sheng H, Shao J, Morrow JD, Beauchamp RD, DuBois RN. Modulation of apoptosis and Bcl-2 expression by prostaglandin E2 in human colon cancer cells. Cancer Res 1998; 58:362–6.
Willow bark extract (BNO1455) and its fractions suppress growth and induce apoptosis in human colon and lung cancer cells Katarina Hostanska PhDa,*, Guido Ju¨rgenliemk PhDb, Gudrun Abel PhDc, Adolf Nahrstedt PhDb, Reinhard Saller MDa a
University Hospital Zu¨rich, Department of Internal Medicine, Institute for Complementary Medicine, FGel 102, Ra¨mistrasse 100, CH-8091 Zu¨rich, Switzerland b University of Muenster, Institute of Pharmaceutical Biology and Phytochemistry, Hittorfstrasse 55, D-48149 Muenster, Germany c Bionorica AG, Phytoneering Company, Department of Research and Development, Kerschensteinerstrasse 11-15, D-92318 Neumarkt, Germany Accepted 3 March 2007
Abstract Background: Recently, there have been extensive efforts to evaluate the chemopreventive role of substances present in natural products. The aim of this study was to examine the effects of the main groups of compounds (salicylalcohol derivates, flavonoids, proanthocyanidins), and salicin isolated from willow bark extract BNO 1455 on proliferation and apoptosis in human colon and cancer cells. Methods: We used human colon cyclooxygenase-2 (COX-2)-positive HT 29 and (COX-2)-negative HCT 116 or lung COX-2 proficient A 549 and low COX-2 expressing SW2 cells. After treatment for 72 h with various concentrations of single substances and acetylsalicylic acid (ASA) as control, inhibition of cell growth and cytotoxicity were measured by colorimetric WST-1 assay and propidium iodide uptake by flow cytometry, respectively. Apoptotic cells were identified by annexin V adhesion using flow cytometry. Results: Studies on dose-dependent effects of BNO 1455 and its fractions showed anti-proliferative activity of all compounds with 50% maximal growth inhibitory concentrations (GI50) between 33.3 and 103.3 mg/ml for flavonoids and proanthocyanidins fractions and 50.0–243.0 mg/ml for salicylalcohol derivates and extract. Apoptosis induction was confirmed by annexin V adherence and analysis of cell morphology based on light scattering characteristics using flow cytometry in all cell lines at GI50. Conclusions: We showed that willow bark extract BNO 1455 an its fractions inhibit the cell growth and promote apoptosis in human colon and lung cancer cell lines irrespective of their COX-selectivity. # 2007 International Society for Preventive Oncology. Published by Elsevier Ltd. All rights reserved. Keywords: Willow bark extract; Salix purpurea; Salicin; Polyphenols; Growth inhibition; Apoptosis; Colon cancer; Lung cancer; Cyclooxygenase-2; Cell lines; Cell growth assay; Cell viability; Cytotoxicity
1. Introduction One of the most important public health and medicinal aims of the new century is the prevention of cancer. For the past two decades, there have been extensive efforts to evaluate the chemopreventive role of substances present in natural products. Several fruits, vegetables and plants with different pharmacological properties are known to be rich sources of chemical constituents that may have potential for the prevention or treatment of several malignancies [1,2]. Willow * Corresponding author. Tel.: +41 61 462 36 41; fax: +41 61 465 90 71. E-mail address: [email protected] (K. Hostanska).
bark (Salix species) has been used throughout the world for centuries as an anti-pyretic and analgesic. Its active constituents, salicin, and its derivates were widely used by 19th century physicians to treat rheumatic fever, different kinds of pain, including back pain, toothache, and headache [3]. In the late 19th century, acetylsalicylic acid (ASA) was synthesised based on the chemical structure of the willow bark constituent, the prodrug salicin (SAL) [4]. Nowadays, synthetic ASA is used not only as an analgesic and antipyretic, but also to prevent myocardial infarctions, strokes, and colorectal cancer. The relation between non-steroidal anti-inflammatory drug (NSAID) use and lung cancer has been examined in several epidemiological studies. Regular
0361-090X/$30.00 # 2007 International Society for Preventive Oncology. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.cdp.2007.03.001
130
K. Hostanska et al. / Cancer Detection and Prevention 31 (2007) 129–139
ASA use appeared to be inversely associated with lung cancer and mortality [5,6]. The mechanisms by which NSAIDs exert anti-tumor effects remain unclear. One potential mechanism for the anti-neoplastic effect of NSAIDs involves inhibition of prostaglandin (PG) synthesis [7]. Further, the NSAIDs might directly influence cell proliferation, cause cell growth arrest, and might trigger cell death apoptosis [8,9]. Pharmacological and clinical studies demonstrated that hydroethanolic willow bark extract had comparable antiinflammatory activities to ASA and showed anti-nociceptive and anti-pyretic activities in various animal models in doses that did not damage the gastric mucosa [10]. Like other herbal extracts, willow bark extracts contain a variety of active constituents such as glycosides of salicylic alcohol and their derivates, aromatic aldehydes and acids, flavonoids and tannins [11,12] so that the precise mode of action of willow bark extract is not yet known. The extract inhibits the production of inflammatory mediators such as cyclooxygenase (COX)-2 mediated PGs, leukotrienes, and cytokines [13] and certain constituents of the bark of Salix species (SAL, tremulacin) possess anti-inflammatory effect in the hen’s egg chorioallantoin membrane test system [14]. In recent years, it has increasingly been recognized that malignancy may not exclusively result from enhanced cell proliferation but also from decreased physiological cell death, apoptosis [15]. Apoptotic induction has been a new target for innovative mechanism-based drug discovery [16]. Certain products from plants are known to induce apoptosis in malignant cells [1,2,17]. It is thus considered important to screen for apoptotic inducers from plants, either in the form of crude extracts or as group of components/and or components isolated from them. Some herbalists recommend willow bark extract as a natural substitute for ASA to achieve these same benefits without side effects. However, there are no data on the use of willow bark extract in the prevention of colorectal/and or lung cancer. This study, therefore, was designed to evaluate the effect of willow bark extract, its component SAL and its isolated phytochemically characterized fractions on human colon and lung cancer cells. Considering the inhibition of PG synthesis as a key determinant in the chemopreventive and anti-neoplastic effect of NSAIDs, we investigated representative cells lines from each group. We selected colon carcinoma HCT 116 which lacks COX transcript and
does not produce PGs even when exogenously stimulated in contrast to HT 29 which produces PGs and small cell lung cancer (SCLC) SW2 cells with negligible COX-2 basal level as compared to COX-2 over expressing non-small cell lung cancer (NSCLC) A 549 cells.
2. Materials and methods 2.1. Samples preparation Willow bark extract was prepared using a drug extract ratio [DER; mass of the herbal drug (starting material) to the mass of the resulting native herbal drug preparation (=native extract)] of 6–12:1 and 70% (v/v) ethanol as extracting agent. According to the certificate of analysis, the extract batch number 01813 was characterized by 16.9% total content of SAL. For fractionation 101 g of a 70% ethanolic Salicis cortex native extract (Salix purpurea; Salicaceae; AE 0665/2, Fa. Bionorica) was separated by LC on SephadexLH201 (column: 90 cm 5.5 cm) in four consecutive portions. Eluates were tested by TLC [silica gel; eluent: ethylacetate/formic acid/water 90:5:5; detection was by Naturstoffreagent (diphenylboryloxyethylamine), vanillin/ hydrochloric acid and anisaldehyde/sulphuric acid]. The Sephadex-column was started using the eluent ethanol/water 3/1 (flow: 20 ml/min; 30 ml/test tube). Fractions containing salicylic alcohol derivatives (fraction 1, F1: 0–1365 ml) and flavonoids (fraction 2, F2: 1365–4220 ml) were combined. At the first detection of dimeric proanthocyanidins (beginning of fraction 3), chromatography was stopped and acetone/water 7/3 was used as eluent until the eluate was colourless and no further compounds could be detected by TLC. The combined eluates (fraction 3, F3) contained the oligomeric and polymeric flavan-3-ols. Total recovery was 98.2% of native extract. All reagents were purchased from Merck (Darmstadt, Germany). For characterization, the three fractions as well as the extract were analysed for salicylic alcohol derivatives (according to S. cortex, Ph. Eur. [18]), total phenolics (phenols reacting with molybdatophosphoric acid), and total tannins (phenols binding to skin powder) (according to the method 2.8.14 Ph. Eur. [18], modified by Glasl [19]), proanthocyanidins (according to Crataegi fructus, Ph. Eur.
Table 1 Results of fractionation and phytochemical characterization
Elution volume (ml) Amount of recovered extract (%) Salicylic alcohol derivates (%) Total phenols (%) Total tannins (%) Proanthocyanidins (%) Flavonoids (%)
Extract
Fraction 1
Fraction 2
Fraction 3
16.9 15.7 8.5 9.2 0.6
0–1365 83.0 20.5 5.7 1.0 0.9 0.1
1365–4220 10.3 3.9 51.5 26.9 30.8 3.3
4220–8500 6.7 0.0 69.8 59.7 73.8 0.3
Percentage of recovered extract is related to the native extract.
K. Hostanska et al. / Cancer Detection and Prevention 31 (2007) 129–139
[18]; anthocyanidin method) and flavonoids (according to Calendulae flos, Ph. Eur.) [18]. The results are presented in Table 1. Standard samples acetylsalicylic acid (ASA), salicin (SAL) were purchased from Sigma (Buchs, Switzerland) and flavonoid quercetin.dihydrate (QUE) was obtained from Alexis Corporation (Lausen, Switzerland). All pure compounds were dissolved in DMSO and stored in aliquots at 20 8C. 2.2. Cell lines and culture conditions The human colon adenocarcinoma cell lines HT 29, HCT 116 and A 549 (NSCLC) cells were obtained from American Type Culture Collection (ATCC, Rockville, MD). The small cell lung cancer (SCLC) cell line SW2 was obtained from the Dana–Farber Cancer Institute (Boston, MA). For routine maintenance, all cell lines were grown in RPMI 1640 medium supplemented with 10% fetal calf serum (FCS), 2 mM L-glutamine and penicillin/streptomycin 1%. All cell culture reagents were from Sigma. Cells (HT 29, HCT 116, A 549) were grown as monolayer and SW2 as floating aggregates at 37 8C in humidified atmosphere with 5% CO2. The amount of FCS was reduced to 2% in the assay medium for all cells tested. The final concentration of DMSO in culture medium during the treatment of cells did not exceed 0.5% (v/v). Vehicle treated cells (0.5% DMSO) were always included in control samples. This concentration of 0.5% DMSO did not affect the viability, proliferation, or apoptosis of the investigated cells. 2.2.1. Flow cytometric detection of Bcl-2 gene and COX-1/COX-2 status The basal level of anti-apoptotic gene Bcl-2 and COX-1/ COX-2 was assessed by direct intracellular single or double immunofluorescence staining according to manufacturer’s instructions (BD Biosciences, San Jose, CA). The permeabilized non-treated cells were stained with PEconjugated anti-Bcl-2 antibody or with FITC-conjugated anti-COX-1 and PE-conjugated anti-COX-2 antibodies. Background fluorescence was determined by isotype matched (IgG1) control antibodies. Cells were analyzed with FACScalibur flow cytometer (BD) equipped with CellQuest software. 2.3. Cell growth assay Cells seeded in 96-well plates at a density of 5 103 cells/well were exposed to different concentrations of individual substances in triplicates for 72 h. Cell growth was indirectly measured by use of a colorimetric assay based on the cleavage of the tetrazolium salt WST-1 (Roche Diagnostics, Rotkreuz, Switzerland) in viable cells as described [20]. Absorbance values at 450 and 650 nm (reference wavelength) were measured in a micro-titer plate
131
ELISA reader. Each set of experiments was repeated three times. The test medium alone was used as background control. The growth inhibition rate was calculated as the percentage of parallel untreated controls. As controls, ASA, SAL, and flavonoid QUE in pharmacological concentrations were used. The direct reductive potential of all the substances was tested in a cell-free system. The concentration of substances that reduced cell growth by 50% (GI50) was calculated from the dose–response relationship by regression analysis using Microsoft Excel computer software. 2.4. Analysis of cell viability Plasma membrane integrity of treated cells (2 104/ well) after 72 h was investigated with propidium iodide (PI) at a final concentration of 5 mg/ml for 15 min. Bivariate dot plots of forward scatter (FCS) versus red fluorescence intensity were performed to discriminate nonviable (PI + events) and viable (PI events) cells. In each sample, 104 cells were analyzed on FACScalibur flow cytometer (BD) equipped with CellQuest analysis software. 2.5. Cell death assays 2.5.1. Morphologic assessment of apoptosis To investigate morphological changes of the cells typical for apoptosis, the cells were subjected to FSC and side (SSC) light scatter analysis by flow cytometry. Cell morphology based on light scattering characteristics, cell size according to FSC, and granularity to SSC was investigated simultaneously in the cells stained by annexin V/PI. 2.5.2. Apoptosis assays After treatment of cells (2 104) with BNO, fractions F1, F2, F3, or DMSO vehicle for 72 h, cells were harvested by trypsinization, washed in PBS, and stained with annexin V-FITC conjugate and PI according to the manufacturer’s protocol (Bender Medsystems Diagnostica, Vienna, Austria) as previously described [21]. Cells (104) were then analyzed by flow cytometry using CellQuest acquisition and analysis software. 2.6. Statistical analysis The experiments were repeated at least three times. Cytotoxic and growth inhibitory effect of individual substances on all four cell lines were compared with those of the untreated controls by Student’s two-tailed test. GI50 values were obtained by linear or exponential regression analysis of the dose–response relationship using the Microsoft Excel software. Probability values equal to or less than 0.05 were considered significant. The comparison of histograms was evaluated statistically using the Kolmogorov–Smirnov (K–S) two samples test for overlaid histograms.
132
K. Hostanska et al. / Cancer Detection and Prevention 31 (2007) 129–139
3. Results 3.1. Dose-dependent cell growth inhibition of SAL, BNO 1455, fractions F1, F2 and F3 Growth inhibitory effect of substances was measured by WST-1 assay which measures the dehydrogenase activity of
viable cells. ASA (1–5 mM; 180–900 mg/ml) and flavonol QUE (50 and 100 mM; 16.9 and 33.8 mg/ml) were used as positive controls. ASA caused reduction in cell growth between 14.3 and 88.9% (Fig. 1). There are very few data concerning the effect of the prodrug SAL. Therefore, in addition, the inhibitory effect of SAL (10–100 mM; 2.86– 28.65 mg/ml) was tested in parallel. The most sensitive cell
Fig. 1. Effect of exposure to ASA (1–5 mM), Q (50 mM, 100 mM) on cell growth of HT 29 (A), HCT 116 (B), A 549 (C) and SW2 (D) cells. Cells (5 103/ well) were grown for 72 h in the presence of indicated concentrations of ASA and Q and the cell growth inhibition was assessed by WST-1. Untreated cells (in medium only) and cells incubated in the presence of vehicle (0.5% DMSO) were used as controls. The effect of SAL on the cell growth is shown as individual dose–response curves for each cell line (E). Average growth inhibition values (S.D.) from triplicates of three independent experiments are shown (*p < 0.01 effect of SAL vs. untreated control; t-test).
K. Hostanska et al. / Cancer Detection and Prevention 31 (2007) 129–139
line to the exposure of SAL was lung SW2 with a GI50 of 17 mM. The growth of these cells was reduced by 75% at 20 mM concentration and completely inhibited at 50 mM (Fig. 1). The growth of colon HT 29, HCT 116 and lung A 549 cells was significantly reduced by higher concentration of SAL (50 mM). GI50 values of SAL for these three cell lines were approximately two-fold higher than in SW2 and being very similar (Table 2). The effects of BNO 1455, F1, F2, and F3 were studied as a dose–response experiments after 72 h at concentrations ranging between 25 and 400 mg/ ml. Extract BNO 1455 and F1 reduced the growth of three cell lines (HT 29, HCT 116 and A 549) significantly above 100 mg/ml. The number of SW2 cells was reduced at 50 mg/ ml BNO 1455 and F1 concentration. For a significant reducing effect of F2 and F3 on cells only 50 mg/ml doses were required. The substances induced suppression of growth of individual cell lines from three independent experiments and are presented in Fig. 2. This comparative study showed differences between the patterns of the antiproliferative responses of the four cell lines towards the BNO 1455 extract and fractions F1, F2, and F3. The GI50 was used to compare the sensitivity of individual cell lines to
133
single substances (Table 2). Proanthocyanidins enriched fraction F3 was the most active in all four cell lines with half maximal inhibitory concentrations between 33.3 and 55.0 mg/ml. Extract BNO 1455 exerted inhibitory effect in three cell lines (HT 29, HCT 116 and A 549) with GI50 values between 128.3 and 211.7 mg/ml which correspond to the total SAL concentrations of 21.7–35.8 mg/ml. For the most sensitive SCLC SW2 the GI50 concentration of BNO 1455 was in the same range as for the most active compound F3, i.e. about 50 mg/ml. There were differences in the susceptibility of HT 29 and HCT 116 colon carcinoma cells towards extract BNO 1455 and fraction F1 as mirrored in their GI50 (211.7 mg/ml versus 128.3 mg/ml; p = 0.023 for BNO 1455 and 122.7 mg/ml versus 71.7 mg/ ml; p = 0.048 for F1). F2 exerted comparable effect to F3 in three (HT 29, HCT 116, and SW2) carcinoma cells. The effect of BNO 1455 and F1 differed significantly ( p < 0.001) between the NSCLC A 549 and SCLC SW2 cells. The statistically evaluated differences between the effect of substances on individual cell lines expressed with GI50 values and corresponding cell survival rates are shown in Table 2.
Table 2 Growth inhibitory potency and cytotoxicity of BNO 1455, F1, F2 and F3, and controls ASA and SAL in colon HT 29, HCT 116 (A) and lung A 549, SW2 (B) cancer cells Substance
Part A BNO 1455 F1a F2a F3a F1 F2 F3 ASA (mM) SAL (mM)
GI50 value (mg/ml)
Cytotoxicity at GI50 value (% of viable cells)
HT 29
HCT 116
HT 29
211.7 27.5 a 175.7 22.8 21.8 2.8 14.1 1.8 122.7 20.5 b 71.7 10.4 d 55.0 5.0 d 468.0 72, (2.6 0.4) (11.5 0.9) 103, (40.2 3.1)
128.3 29.3 b 106.5 4.9 15.7 3.5 8.6 1.9 71.7 17.6 d 66.7 12.6 d 33.3 7.6 e 360.0 36, (2.0 0.2) (9.7 0.8) 103, (34.5 2.8)
107.8 2.3
75.6 4.3
103.2 3.7 93.4 2.8 98.5 3.1 88.4 2.8 b 84.3 3.5 b
94.2 2.5 72.9 3.3 89.7 5.2 87.0 3.6b 55.9 4.2b
100 (77.6 3.9) c
Medium only Substance
Part B BNO 1455 F1a F2a F3a F1 F2 F3 ASA (mM) SAL (mM) Medium only
GI50 value (mg/ml)
HCT 116
100 (72.8 5.4) c
Cytotoxicity at GI50 value (% of viable cells)
A 549
SW2
206.7 35.6 a 171.6 29.5 21.3 3.7 13.8 2.4 243.3 20.8 a 103.3 21.6 b, d 51.7 12.6 d 396.0 54, (2.2 0.3) (10.9 0.8) 103, (38.3 2.9)
52.8 8.3 d 43.8 6.9 5.4 0.8 3.5 0.6 50.0 11.2 d 63.3 13.5 d 40.7 5.1 d, e 612.0 90, (3.4 0.5) (4.9 0.5) 103, (17.4 1.9)
A 549 25.4 6.5
84.6 7.2 86.8 3.9 82.1 4.3 101.7 2.5 b 60.4 6.3 b 100 (81.6 4.5) c
SW2 88.5 3.7
95.3 4.1 84.2 2.8 64.3 5.7 92.5 3.2b 64.5 5.7b 100 (79.3 3.1) c
Letters a, b, c, d: values are means S.D. of three experiments performed in triplicate. All the values are statistically different ( p < 0.05), except those marked with the same letters (a, b, c or d). a Corresponding concentrations of the fractions F1, F2 and F3 to the established GI50 of extract BNO 1455. Not involved in statistical evaluation. b Survival of cell treated with ASA (5 mM) and SAL (50 mM). c In parentheses the survival values of untreated cells which were normalized to 100% are expressed.
134
K. Hostanska et al. / Cancer Detection and Prevention 31 (2007) 129–139
Fig. 2. Cell growth inhibitory effect of BNO 1455, F1, F2 and F3. Cell proliferation was determined by WST-1 assay after 72 h treatment. Untreated cells and cells incubated in the presence of vehicle (0.5% DMSO) were used as controls. In all experiments the values of vehicle treated cells did not exceed the intraexperimental variance which was less than 10%. Individual dose–response curves for each cell line from triplicates repeated at least three times are expressed as mean (S.D.). BNO 1455, F1 reduced growth of HT 29, HCT 116 and A 549 significantly ( p < 0.05) at concentrations 100 mg/ml and F2, F3 at concentrations of 50 mg/ml and/or higher. The growth of SW2 was significantly ( p < 0.05) reduced by all four substances at 50 mg/ml concentration.
3.2. Cytotoxicity of BNO 1455, F1, F2 and F3 toward colon HT 29, HCT 116 and lung A 549, SW2 cancer cells Because all substances showed an anti-proliferative effect they were tested, in addition, for their potential toxicity. The tumor cell lines (HT 29, HCT 116, A 549, SW2) were exposed to dose-increasing concentrations (0– 200 mg/ml) of substances for 72 h. We inspected cells during the treatment by inversion microscopy for morphological changes related to the cytotoxicity of compounds. Live and dead cells discrimination were estimated by PI uptake. The viability results (normalized to untreated controls) from flow cytometry detection are presented in Fig. 3. As controls ASA (5 mM) and SAL (50 mM) concentrations were investigated in parallel (Table 2). Sensitivity of the cells to the exposure of substances differed as shown in Fig. 3. We estimated the cell survival according to following criteria. More than 90% viable cells were considered to be unaffected, 80–90% as moderately affected, and values of less than 80% viable cells were ascribed to cytotoxic effects of the compounds. Extract
and F1, F2, and F3 exerted a very low toxic effect on colon HT 29 cells up to highest concentrations of 200 mg/ml. Cell survival was between 78.5 and 102% for individual substances. Survival of colon HCT 116 and lung A 549 cells was affected at concentrations above 100 mg/ml of BNO 1455, F1, F2 and F3. The most sensitive to the cytotoxic effect of tested substances was the SCLC SW2 cell line which survival was affected by all substances at 100 mg/ ml concentration. However, at GI50 of the four substances, the survival of SW2 cells was affected only by the proanthocyanidine enriched fraction F3 (Table 2). 3.3. Activation of apoptosis by the BNO 1455 and fractions F1, F2 and F3 in human colon and lung cancer cells Having established that BNO 1455 extract, F1, F2, and F3 were growth inhibitory to the colon and lung cancer cells, we investigated whether these substances induced apoptosis. A first approach to the mode of cell death by drugs was conducted by flow cytometric cell scatter characteristics.
K. Hostanska et al. / Cancer Detection and Prevention 31 (2007) 129–139
135
Fig. 3. Effect of BNO 1455, F1, F2 and F3 on the viability of individual cancer cells. Cells were treated for 72 h, stained with PI (5 mg/ml) and analyzed by flow cytometry as described in Section 2. Survival of cells in medium OR (72.8–81.6%) was normalized to 100%. The survival of vehicle-treated cells did not differ from control cells. The viability of HT 29, HCT 116, A 549 and SW2 cells was 88.4%, 87%, 101.7% and 92.5% for control substance ASA (5 mM). Individual dose–response effect for each cell line from experiments that were repeated twice with similar results are expressed as mean. S.D. values were less than 15%.
Decrease in the cell size coinciding with the decrease in FSC and increase or no change in the granularity of cells coinciding with the increase in SSC are characteristic hallmarks of apoptotic cells. We investigated the effect of substances at concentrations close to their GI50 values (Table 2) with ASA (5 mM) as positive control. The two-parameter histograms plotting cell size versus granularity showed morphological changes typical for apoptosis on colon (HT 29, HCT 116) and lung (A 549, SW2) cancer cells. Morphological changes of apoptosis cell shrinkage and chromatin clumping were well reflected by reduced ability to scatter light in the forward direction and an increase or no change in the 90% light scatter after the treatment with BNO 1455, F1, F2, and F3. These changes were more pronounced by treatment of cells with F2 and F3 (data not shown). The sensitivity of colon (HT 29, HCT 116) and lung (A 549, SW2) cells to apoptosis induction was further examined by flow cytometric determination of annexin V-positive cells at 72 h after drug exposure. An increase in cellular staining with FITC-conjugated annexin V served as an early marker for apoptosis. The cells were simultaneously stained with PI to score for overt loss of cell membrane integrity. The double staining procedure distinguishes early-stage apoptotic cells (annexin V-positive, PI negative) from late-stage apoptotic
cells (annexin V-positive, PI positive). As shown in Fig. 4, treatment of the cancer cells with GI50 concentrations induced apoptotic cell death in all cell lines. Colon carcinoma cells HCT 116 responded by more than 80% of apoptotic cells. The degree of secondary necrotic cells was negligible (max. 4.3%). Spontaneous apoptosis in this cell line was 28%. In HT 29 colon cancer cells, the fraction of early apoptotic cells lay between 46.9 and 59.9% as compared to the untreated control with 16.4% apoptotic cells. Late apoptotic cell fractions occurred between 10 and 26%. SW2 cells exposed for 72 h expressed the mixtures of early apoptotic cells (22.9–31.3%) and secondary necrotic fractions (12.3–32.8%). In untreated controls 12.3% apoptotic and 11% necrotic cells were detected. The apoptotic fractions of A 549 NSCLC cells after the treatment lay between 62.8 and 81.6% compared to 24.3% in untreated controls. Necrotic cell fractions ranged between 9.2 and 16.7%. 3.4. Characterization of basal levels of Bcl-2 protein in four cancer cell lines and constitutive expression of COX-2 in lung SCLC SW2 cells In order to obtain a better understanding of the apoptosis inducing effect of willow bark extract on colon and lung
136
K. Hostanska et al. / Cancer Detection and Prevention 31 (2007) 129–139
Fig. 4. Apoptosis-inducing effect of BNO 1455, F1, F2 and F3 on individual cancer cell lines expressed as percentage of PS-expressing [early apoptotic: annexin V-positive/PI-negative (A+/PI ) and late apoptotic/necrotic: annexin V-positive/PI-positive (A+/PI+)] cells. Cells (20 103) were treated with substances at their half-maximal inhibitory concentrations (GI50) for 72 h. As controls, cells treated with ASA (5 mM) were used. A representative of two experiments with comparable results is shown.
cancer cells, we determined the basal levels of COX-2 isoenzyme and of anti-apoptotic Bcl-2 protein. We determined the constitutive expression of COX-1/COX-2 levels in untreated SW2 cells. With SW2 cells very low expression of COX-2 (4%) was found in contrast to the constitutively expressed COX-1. In addition the basal levels of anti-apoptotic Bcl-2 in untreated cells were quantified by flow cytometric determination. Bcl-2 was abundantly expressed in the SCLC SW2 (92.9%), but was barely detectable in the NSCLC A 549 (6.4%) and colon HCT 116 (16.4%) and was negligible in HT 29 (1.9%) (Table 3).
prodrug SAL is primarily postulated as the active component, but there are other groups of components with pharmacological activity, particularly polyphenols (phenolic acids, hydrolysable tannins, proanthocyanidins, and flavonoids) which have shown anti-oxidant and anti-carcinogenic effects in several studies [1,22]. Besides salicylic alcohol derivates (including SAL), considerable amounts of total Table 3 Basal level of COX-2 receptor and apoptotic genes p53 and Bcl-2 in colon and lung cancer cells Cell lines
COX-2 status
Bcl-2 status (%)
p53 status
4. Discussion
HT 29 HCT 116 A 549 SW2
High expression [9] No expression [9] High expression [29] 4%
0.2 16.4 6.4 92.9
Mutant [33] Wt [33] Wt [34] Mutanta
It is a huge task to identify a potent single chemical and/ or group of compounds from the complex mixture of substances present in a plant. Willow bark extracts have been used world-wide as an anti-pyretic and analgesic. The
a Mutation on the nucleotide 511 of exon 5; codon 171 changed from GAG (Glu) to TAG (Cys): (written personal communication from Dr. Zangemeister-Wittke U., Division of Oncology, University Hospital Zu¨rich).
K. Hostanska et al. / Cancer Detection and Prevention 31 (2007) 129–139
phenols, total tannins, proanthocyanidins, and flavonoids were present in the native extract and its fractions as shown in Table 1. They all could contribute to the pharmacological activity of the native extract. We investigated the response of two colon cancer cells HT 29 (COX-2 expressing), HCT 116 (lacking COX-2 expression) and two lung cancer cells A 549 (NSCLC; COX2 high expression), SW2 (SCLC; low expression of COX-2) to outline similarities and differences between them to the exposure of BNO 1455 and its fractions. Our data demonstrates that willow bark extract and its fractions F1, F2 and F3 exert two profound effects on the human colon cancer cells (HT 29, HCT 116) and lung cancer cells (A 549, SW2): (a) they inhibit their proliferation, and (b) they reduce their survival by inducing cell death by apoptosis. The growth inhibitory effects of four substances were dosedependent (Fig. 2), but there were both differences and similarities between the response of cells to individual fractions as mirrored in their GI50 values (Table 2). Compounds F2 and F3 reduced the growth of all four cancer cells to a comparable degree. Variations of GI50 were only 1.6-fold for both compounds. Our results are in agreement with studies of others on the anti-proliferative effect of different flavonoids and/or proanthocyanidins [22,23] and comparable to the effect of flavonoid QUE used as control in this study (Fig. 1). To our knowledge, this is the first report on growth inhibition and apoptosis in carcinoma cell lines induced by willow bark extract, its fraction of salicyl alcohol derivates F1 and SAL. Because for most NSAIDs, an anti-inflammatory and COX-2 inhibitory effect of willow bark extracts was reported [13], we compared its effect to that of the well established NSAID ASA. In parallel, the effect of SAL, the component used to standardize the commercial extracts, was investigated. We observed a dose-dependent reduction of cell growth (Fig. 1). Based on comparative GI50 values of SAL (17–40 mM) and ASA (2.0–3.4 mM) about five-fold (SW2) or 15-fold higher concentrations (HT 29, HCT 116, A 549) of SAL were required than needed for ASA. In most cases the pharmacological activity of plant extracts is not assigned to single components and often not all active ingredients are known. Therefore, it was of interest to compare the effect of the whole extract BNO 1455 (16.9% SAL) to SAL and ASA. There were differences in the response of cells to BNO 1455. NSCLC A 549 and colon HT 29 cells required about fourfold and colon HCT 116 (two-fold) higher concentration in comparison to gain the same response as highly responsive SW2 SCLC cells. As shown in Table 2, the corresponding values of SAL (8.4–19.5 mg/ml) in the extract were very low. Taking the molecular weight into consideration, there were differences of about two decades of logarithmic scale in comparison to the single compound SAL (4.7–11.2 mg/ml). In addition, the growth inhibitory effect of BNO 1455 seems to be superior to the effect exerted by ASA. Overall about two-fold higher doses of ASA were required in comparison to the whole extract to obtain 50% inhibition of cells growth.
137
Products obtained via plant material extraction contain complex mixture of chemicals whose individual biological activities may be considerably different when present in a mixture because of their possible interaction [24,25]. Therefore it can be argued that other constituents (e.g. polyphenols–proanthocyanidins) contributed to the antiproliferative effect of willow bark extract. This observation is in agreement with recently published studies on antiinflammatory activity of Salix extract [26] for which the authors also postulated the contribution of additional, hitherto unknown constituents. Cytotoxicity of substances determined by PI uptake on four cancer cell lines after 72 h treatment was only moderate (Fig. 3). Results of this study showed that the cell growth inhibition is due to the interference with cell proliferation rather than direct cytotoxicity of substances. Discovery of active compounds from natural products with cytocidal (e.g. apoptotic) rather than cytotoxic properties is of great interest for cancer treatment [17]. Apoptosis is a process that regulates cell number or eliminates damaged cells. Deficiencies in apoptosis also contribute to carcinogenesis by creating a permissive environment for genetic instability and accumulation of gene mutations conferring resistance to cytotoxic anticancer drugs and radiation [27]. To further substantiate the cell growth inhibitory effect of BNO 1455 and fractions F1, F2, and F3, the apoptotic cell death was assessed. As shown in Fig. 4 all substances induced apoptosis in the cells. The sensitivity of individual cells to BNO 1455, F1, F2 and F3 as obtained in apoptosis assays correlated well with their growth inhibitory responses. Apoptosis induction by BNO 1455 was accompanied by reduction in cell viability only in A 549 cells. The apoptosis induced by all four substances was preceded by a secondary necrosis in cancer cells. However, in the colon HCT 116 at least and in lung SW2 most necrotic cells were detected (Fig. 4). BNO 1455 and F1-, F2-, F3-induced apoptosis was comparable to that induced by 5 mM (0.9 mg/ml) ASA, but the concentrations of substances (GI50) required for this apoptosis induction were about 4–15-fold lower. ASA and related NSAIDs have attracted considerable attention as compounds that might be of potential benefit in the chemoprevention of cancer. The best defined molecular target for ASA and other NSAIDs is COX-2 which is considered to play an important role in carcinogenesis by regulation of apoptosis. Furthermore NSAIDs exert antiproliferative effects on colon and lung carcinoma cells [7,8,28]. First we looked for any relationship in the growth inhibition and apoptosis between the cells with different constitutive expression of COX-2. The overexpression of COX-2 in NSCLC A 549 [29] and colon HT 29 [9] and its lack in HCT 116 [9] cells were described by others (Table 3). Using flow cytometry, we determined the low basal level of COX-2 on SCLC SW2 cells (Table 3). This is in accordance with the finding that most of SCLC express a low level of COX-2 [30]. We found, that cell lines with deficiencies of COX-2 were more sensitive to the growth inhibitory and
138
K. Hostanska et al. / Cancer Detection and Prevention 31 (2007) 129–139
apoptotic effects of extract and salicylate fraction F1 than cells proficient in production of COX-2 (Tables 2 and 3). These facts support the COX-2-independent effects of BNO 1455 and its F1 fraction and are in agreement with findings that NSAIDs inhibited the proliferation and induced apoptosis in cancer cells by other non-COX-dependent mechanisms [31,32]. To further substantiate the differences we compared the constitutive levels of two important proteins involved in apoptosis namely p53 and Bcl-2. The state of p53 is pivotal for the response of tumour cells to chemo- and radiotherapy and plant-derived compounds which overcome a high apoptotic threshold in tumour cells lacking intact p53 [33]. The colon HT 29, HCT 116 and lung A 549 cells have been characterized [34,35] for their p53 status (Table 3). The SW2 lung cancer cell express a mutated p53 gene (personal written communication Dr. Zangemeister-Wittke U; Oncology Division, University Hospital Zu¨rich). Bcl-2 appears to function as an inhibitor of apoptosis. Over expression of Bcl-2 has been reported in different types of cancer and was considered to be responsible for the blockade of apoptosis [36]. We detected different basal levels of Bcl-2 by flow cytometry (Table 3). Bcl-2 was highly expressed only in SW2 cells in agreement to others [37]. The role of COX-2 in carcinogenesis and in the regulation of apoptosis is quite well established. Previous reports indicate complex, functional, and reciprocal interactions between the COX-2 and p53 systems and forced expression of COX-2 modulate the Bcl-2 expression [33,38]. The differences in responses between HT 29 (COX-positive) and HCT 116 (COX-negative) following exposure to the BNO 1455 and F1 could be attributed not only to the differences of COX-2 level but also to the p53 status of the cells. An association between COX-2 over expression and p53 mutations, or low level of p53 protein has been reported in human cancer cells. The HT 29 cells have mutant p53, while HCT 116 cells have wild-type [33]. NSCLC A 549 cells carry a high expression of COX-2 accompanied by wild-type p53. Despite their p53 status, these cells showed a lower sensitivity to the extract and F1, comparable to colon HT 29 with mutant p53. It has been reported that COX-2 could inactivate wild-type p53 [33]. Interestingly, the SW2 cells which carry mutant p53, high expression of Bcl-2 accompanied with low constitutive level of COX-2 responded mostly to all compounds. It was described that down-regulation of Bcl-2 in COX-2 negative colon cancer cell represents a potential mechanism by which ASA induces apoptosis [37]. In summary, these results suggest that all the compounds tested inhibit the growth and promote apoptosis of four cancer cell lines and these effects depend on the cell lines rather than on the property of drugs. NSAIDs are effective in preventing colon cancer. Currently available drugs, however, have side effects that make them unsuitable for widespread use in the prevention. In this study we showed that willow bark extract exerted a growth reducing effect in cancer cells by triggering
apoptosis. These abilities of willow bark extract suggest that it could be a promising cancer preventive compound. Additional studies are warranted to examine in vivo models in support of the above effects of willow bark.
Conflict of interest None.
References [1] Rao CV, Newmark HL, Reddy BS. Chemopreventive effect of farnesol and lanosterol on colon carcinogenesis. Cancer Detect Prev 2002; 26:419–25. [2] Zhang H, Spitz MR, Tomlinson GE, Schabath MB, Minna JD, Wu X. Modification of lung cancer susceptibility by green tea extract as measured by the comet assay. Cancer Detect Prev 2002; 26:4144–418. [3] Hedner T, Everts B. The early clinical history of salicylates in rheumatology and pain. Clin Rheumatol 1998; 17:17–25. [4] Robbers JE, Speedie MK, Tyler VE. Pharmacognosy and pharmacobiotechnology. Baltimore: Williams & Wilkins, 1996. [5] Moysich KB, Menezes RJ, Ronsani A, Swede H, Reid ME, Cummings KM, et al. Regular aspirin use and lung cancer risk. BMC Cancer 2002; 2:31. [6] Akhmedkhanov A, Toniolo P, Zeleniuch-Jacquotte A, Koenig KL, Shore RE. Aspirin and lung cancer in women. Br J Cancer 2002; 87:49–53. [7] Vane JR. Inhibition of prostaglandin synthesis as a mechanism of action for aspirin-like drugs. Nat New Biol 1971; 231:232–5. [8] Shiff SJ, Koutsos M, Qiao L, Rigas B. Non-steroidal anti-inflammatory drugs inhibit the proliferation of colon adenocarcinoma cells: effects on cell cycle and apoptosis. Exp Cell Res 1996; 222:179–88. [9] Qiao L, Hanif R, Sphicas E, Shiff SJ, Rigas B. Effect of aspirin on induction of apoptosis in HT 29 human colon adenocarcinoma cells. Biochem Pharmacol 1998; 55:53–64. [10] Ma¨rz RW, Kemper F. Willow bark extract—effects and effectiveness. Status of current knowledge regarding pharmacology, toxicology and clinical aspects. Wien Med Wochenschr 2002; 152:354–9. [11] Thieme H. On the tannin content of willow cortex. Pharmazie 1968; 23:212–5. [12] Meier B, Sticher O, Julkunen-Tiitto R. Pharmaceutical aspects of the use of willows in herbal remedies. Planta Med 1988; 54:559–60. [13] Wagner I, Greim C, Laufer S, Heide L, Gleiter CH. Influence of willow bark extract on cyclooxygenase activity and on tumor necrosis factor alpha or interleukin 1 beta release in vitro and ex vivo. Clin Pharmacol Ther 2003; 73:272–4. [14] Albrecht M, Nahrstedt A, Laepke NP, Theisen NL, Baron G. Antiinflammatory activity of flavanol glycosides and salicin derivates from the leaves of Populus tremuloides. Planta Med 1990; 56:660. [15] Reed JC. Mechanisms of apoptosis avoidance in cancer. Curr Opin Oncol 1999; 11:68–75. [16] Senderowicz AM. Targeting cell cycle and apoptosis for the treatment of human malignancies. Curr Opin Cell Biol 2004; 16:670–8. [17] Taraphdar AK, Madhumita R, Bhattacharya RK. Natural products as inducer of apoptosis: implication for cancer therapy and prevention. Curr Sci 2001; 11:1387–96. [18] European Pharmacopeia (Ph Eur), 4th ed. Stuttgart: Deutscher Apotheker Verlag; 2002. [19] Glasl H. Zur Photometrie in der Drogenstandardisierung. DAZ 1983; 123:1979–87. [20] Ishiyama M, Tominaga H, Shiga M, Sasamoto K, Ohkura Y, Ueno K, et al. A combined assay of cell viability and in vitro cytotoxicity with a
K. Hostanska et al. / Cancer Detection and Prevention 31 (2007) 129–139
[21]
[22]
[23]
[24] [25]
[26]
[27] [28]
[29]
highly water-soluble tetrazolium salt, neutral red and crystal violet. Biol Pharm Bull 1996; 1:1518–20. Hostanska K, Hajto T, Fischer J, Mengs U, Weber K, Lentzen H, et al. Selective modulation of phosphatidylserine exposure on sub-populations of human peripheral blood lymphocytes by a plant lectin, Viscum album agglutinin (VAA)-I and its recombinant form (rVAA) in vitro. Cancer Detect Prev 1999; 23:511–23. Kuntz S, Wenzel U, Daniel H. Comparative analysis of the effects of flavonoids on proliferation, cytotoxicity, and apoptosis in human colon cancer cell lines. Eur J Nutr 1999; 38:133–42. Agarwal C, Sharma Y, Agarwal R. Anti-carcinogenic effect of a polyphenolic fraction isolated from grape seeds in human prostate carcinoma DU 145 cells: modulation of mitogenic signalling and cellcycle regulators and induction of G1 arrest and apoptosis. Mol Carcinogen 2000; 28:129–38. Williamson EM. Synergy and other interactions in phytomedicines. Phytomedicine 2005; 8:401–9. Fiebich BL, Chrubasik S. Effects of an ethanolic Salix extract on the release of selected inflammatory mediators in vitro. Phytomedicine 2004; 11:135–8. Khayyal MT, El-Ghazaly MA, Abdallah DM, Okpanyi SN, Kelber O, Weiser D. Mechanisms involved in the anti-inflammatory effect of a standardized willow bark extract. Arzneimittelforschung 2005; 55: 677–87. Green DR, Martin SJ. The killer and executioner: how apoptosis controls malignancy. Curr Opin Immunol 1995; 7:694–703. Hida T, Leyton J, Makheja AN, Ben-Av P, Hla T, Martinez A, et al. Non-small cell lung cancer cyclooxygenase activity and proliferation are inhibited by non-steroidal anti-inflammatory drugs. Anticancer Res 1998; 18:775–82. Hung WC, Chang HC. Methanolic extract of adlay seed suppresses COX-2 expression of human lung cancer cells via inhibition of gene transcription. J Agric Food Chem 2003; 51:7333–7.
139
[30] Yoshimoto A, Kasahara K, Kawashima A, Fujimura M, Nakao S. Characterization of the prostaglandin biosynthetic pathway in nonsmall cell lung cancer: a comparison with small lung cell cancer and correlation with angiogenesis, angiogenic factors and metastases. Oncol Rep 2005; 13:1047–57. [31] Hanif R, Pittas A, Feng Y, Koutsos MI, Qiao L, Staiano-Coico L, et al. Effect of non-steroidal anti-inflammatory drugs on proliferation and on induction of apoptosis in colon cancer cells by a prostaglandin-independent pathway. Biochem Pharmacol 1996; 52: 237–45. [32] Hostanska K, Vuong V, Rocha S, Soengas MS, Glanzmann C, Saller R, et al. Recombinant mistletoe lectin induces p53-independent apoptosis in tumour cells and cooperates with ionising radiation. Br J Cancer 2003; 88:1785–92. [33] Swamy MV, Herzog CR, Rao CV. Inhibition of COX-2 in colon cancer cell lines by celecoxib increases the nuclear localization of active p53. Cancer Res 2003; 63:5239–42. [34] Radhakrihsna Pillai G, Srivastava AS, Hassanein TI, Chauhan DP, Carrier E. Induction of apoptosis in human lung cancer cells by curcumin. Cancer Lett 2004; 208:163–70. [35] Zangemeister-Wittke U, Schenker T, Luedke GH, Stahel RA. Synergistic cytotoxicity of Bcl-2 antisense oligodeoxynucleotides and etoposide, doxorubicin and cisplatin on small cell lung cancer cell lines. Br J Cancer 1998; 78:1035–42. [36] Reed JC. Bcl-2: prevention of apoptosis as a mechanism of drug resistance. Hematol Oncol Clin North Am 1995; 9:451–73. [37] Yu HG, Huang JA, Yang YN, Huang H, Luo HS, Yu JP, et al. The effects of acetylsalicylic acid on proliferation, apoptosis, and invasion of cyclooxygenase-2 negative colon cancer cells. Eur J Clin Invest 2002; 32:838–46. [38] Sheng H, Shao J, Morrow JD, Beauchamp RD, DuBois RN. Modulation of apoptosis and Bcl-2 expression by prostaglandin E2 in human colon cancer cells. Cancer Res 1998; 58:362–6.