Introduction

Plant sterols are abundant in vegetables and fruits. Among the phytosterols, sitosterol, campesterol and stigmasterol, are the most important. They are structural analogues of cholesterol. Numerous laboratory studies focused on their potential to reduce cholesterol absorption by the small intestine,1, 2, 3, 4 and cholesterol uptake by LDL.5, 6, 7 In addition, it was reported that phytosterols prevent coronary heart disease,8 and cardiovascular risks.9, 10, 11 However, their potential to reduce the risk of cancer remains controversial.5, 12, 13, 14 Several studies showed antiproliferative properties of plant sterols, on human breast cancer cells,15, 16 prostate cancer cells,17 HT-2918, 19 and HCT116 human colon cancer cells.20 It was suggested that these effects are related to the activation of the sphingomyelin cycle, to cell cycle arrest and/or to the stimulation of apoptosis.18, 19 Induction of apoptosis by β-sitosterol was mediated by increased levels of the proapoptotic Bax protein and caspase induction.20

Phytosterols undergo oxidation during cooking and storage. However, only a few data are available on biological activities of oxyphytosterols.21, 22 It was reported that they cause cellular damage in cultured macrophage-derived cell lines23 and that hydroxysitosterol induces apoptosis of U937 cells, accompanied by a reduction of cellular glutathione.24

Furthermore, numerous studies reported atherogenic25, 26 and cytotoxic properties27, 28, 29, 30 of cholesterol oxides, leading to apoptosis or necrosis in various cell types.24, 31, 32, 33 They may also exert their effects by lowering cholesterol availability for cell membrane formation by inhibiting 3-hydroxy-methylglutaryl coenzyme A reductase (HMG-CoA reductase), a key enzyme in cholesterol synthesis.34 Cholesterol oxides bind to specific membrane receptors or cytosolic binding proteins, leading to the perturbation of cholesterol synthesis.35, 36, 37 It was also shown that 7β-hydroxycholesterol (7β-OHchol) and 7-ketocholesterol may initiate nonapoptotic death in U937 cells.38 These findings highlight the many biological effects of cholesterol oxides. However, the mechanism by which hydroxyphytosterol and hydroxycholesterol induce cell death remains unknown.

In the present study, we aimed at comparing the antiproliferative properties of 7β-hydroxysitosterol (7β-OHsito) and 7β-OHchol in the Caco-2 human colon cancer cells. Cytotoxicity of 7β-OHchol or other cholesterol oxides have been reported on HT-29 and SW620 human colon cancer cells.39 Since 7β-OHsito and 7β-OHchol are the most abundant hydroxysterols34, 40 present in food after cooking and storage, it was of interest to compare their cytotoxic effects. In order to determine whether both molecules induced the same or different apoptotic process in cancer cells, we assessed cell viability, cell cycle analysis, caspases activities and DNA fragmentation. Our data show that both compounds are cytotoxic, but that the pathways leading to cell death are different.

Results

Effects of 7β-OHsito and 7β-OHchol on Caco-2 cell growth

Growth of Caco-2 cells was determined using the sulforhodamine B (SRB) dye method.41 As shown in Figure 1, growth inhibition was observed when cells were exposed to 7β-OHsito and 7β-OHchol in a dose-dependent manner. Cells were more sensitive to 7β-OHchol than to 7β-OHsito (Table 1). A similar number of dead cells was obtained with 60 μM 7β-OHsito than with 30 μM 7β-OHchol. After 32 h of treatment, a 50% cell survival was observed, and all further studies were performed at these concentrations. These compounds are known to induce cell necrosis at higher concentrations.42

Figure 1
figure 1

Effects of 7β-OHsito and 7β-OHchol on Caco-2 cell growth. Cells were exposed to 0.5% ethanol (control), 7β-OHsito (30 and 60 μM) and 7β-OHchol (30 and 60 μM) for 96 h. Data are presented as the mean±S.E. of at least three separate experiments. The asterisk indicates a significant difference between controls and treated (P<0.05)

Full size image
Table 1 Concentration- and time-dependent effect of hydroxysterols on Caco-2 cell death
Full size table

7β-OHsito and 7β-OHchol effects on cell cycle

Exponentially growing Caco-2 cells exposed to both hydroxysterols were analyzed by flow cytometry after staining with propidium iodide (PI)43 (Figure 2). When compared with nontreated controls, 7β-OHsito caused after 32 h a significant accumulation of cells in the S phase. The proportion of cells in the S phase was 60% at 32 h and reached 50% at 48 h. The accumulation of cells treated with 7β-OHchol in the S phase was delayed reaching a proportion of 43% at 72 h.

Figure 2
figure 2

7β-OHsito and 7β-OHchol effects on cell cycle. Cell cycle phases were analyzed for 72 h. Cells were treated with ethanol (0.5%), 7β-OHsito (60 μM) and 7β-OHchol (30 μM). At each indicated time point, cells were harvested, stained with PI and submitted to flow cytometry analysis. Results are presented as the percentage of labelled cells at each cell cycle phase. Data are presented as the mean±S.E. of at least three separate experiments

Full size image

In the case of 7β-OHsito, the accumulation of cells in the S phase at 48 h was associated with a simultaneous increase of hypodiploid cells (sub-G0/G1), which may indicate an apoptotic process.42 A similar number of hypodiploid cells was not observed with 7β-OHchol before 72 h of treatment.

7β-OHsito increases caspase-3 and -9 levels in Caco-2 cells

In order to ensure that cell death induced by both hydroxysterols was the result of apoptosis, we used a colorimetric substrate, Ac-DEVD-pNA, which mimics the cleavage site of an endogenous substrate PARP.44 In cells exposed to 7β-OHsito, caspase-3 activity reached a maximum value at 32 h (Figure 3). This was not observed in control cells or in cells treated with 7β-OHchol.

Figure 3
figure 3

Effect of hydroxysterols on caspase-3 activity. Caco-2 cells were exposed to 0.5% ethanol (control), to 60 μM 7β-OHsito or to 30 μM 7β-OHchol for 72 h. Results are indicated as nmol of pNA released/mg of total protein. Data are presented as the mean±S.E. of at least three separate experiments. The asterisk indicates a significant difference between controls and treated (P<0.05)

Full size image

Several specific substrates were also used in order to investigate whether caspase-8 or -9 were activated after treatment with hydroxysterols. As shown in Figure 4, caspase-9 activity was significantly enhanced in cells exposed to 7β-OHsito for 32 h. In contrast, when cells were treated with 7β-OHchol, caspase-9 was activated only after 56 h. Caspase-8 activity remained at a basal level in cells treated with both hydroxysterols (data not shown). Thus, caspase-3 and -9 were simultaneously activated in cells treated with 7β-OHsito, while with 7β-OHchol, only a retarded and small increase of caspase-9 activity was observed.

Figure 4
figure 4

Effect of hydroxysterols on caspase-9 activity. Caco-2 cells were exposed to 0.5% ethanol (control) or to 60 μM 7β-OHsito or to 30 μM 7β-OHchol. At each time point, cells were isolated and prepared for measurement of caspase-9 activities. Values were adjusted to total cell protein and results are presented as the percent of caspase-9 activity. The basic level was considered at 100% (nontreated control cells). All data are represented as the mean±S.E. of at least three separate experiments. The asterisk indicates a significant difference between controls and treated (P<0.05)

Full size image

Detection of DNA fragmentation induced by 7β-OHsito and 7β-OHchol treatment

Apoptosis was also evaluated by determination of DNA fragmentation.45 DNA integrity was evaluated by gel electrophoresis (Figure 5). In the presence of 7β-OHsito, DNA fragmentation was detected already after 56 h of treatment, but DNA fragmentation was delayed in cells treated with 7β-OHchol and appeared only after 72 h.

Figure 5
figure 5

Effect of hydroxysterols on DNA fragmentation. Caco-2 cells were treated with 0.5% ethanol or with 60 μM 7β-OHsito or 30 μM 7β-OHchol. At different time points, cells were harvested and lysed. Fragmented DNA was separated from genomic DNA by electrophoresis on agarose gel. MW, molecular weight marker; 1, positive control; controls: 2, at 32 h; 3, at 48 h; 4, at 56 h; 5, at 72 h; 60 μM 7β-OHsito: 6, at 32 h; 7, at 48 h; 8, at 56 h; 9, at 72 h; 30 μM 7β-OHchol: 10, at 32 h; 11, at 48 h; 12, at 56 h; 13, at 72 h. Note that DNA fragmentation was observed only for 60 μM 7β-OHsito at 56 h (8) and 72 h (9) and for 30 μM 7β-OHchol at 72 h (13)

Full size image

Effect of pancaspase inhibitor Z-VAD.fmk on 7β-OHsito- and 7β-OHchol- treated cells

To determine whether cell death occurs independent of caspase-3 and -9 activities, Caco-2 cells were treated with the pancaspase inhibitor Z-VAD.fmk before treatment with hydroxysterols.

As illustrated in Figure 6, for cells treated with 7β-OHsito, cell death was retarded in the presence of Z-VAD.fmk. In contrast, the caspase inhibitor showed no significant effects on untreated cells or on cells treated with 7β-OHchol. This suggests that caspase-mediated death pathways are involved in the case of 7β-Ohsito, whereas other death pathways seem to be involved with 7β-OHchol.

Figure 6
figure 6

Effect of pancaspase inhibitor Z-VAD.fmk on 7β-OHsito- and 7β-OHchol-treated cells. Subconfluent Caco-2 cells were treated with 0.5% ethanol or 60 μM 7β-OHsito or 30 μM 7β-OHchol in the presence or absence of 50 μM Z-VAD.fmk for 96 h. The medium was changed every 24 h. Cell growth was measured in the presence or absence of 7β-OHsito 60 μM and 7β-OHchol 30 μM. Data are presented as the mean±S.E. of at least three separate experiments. The asterisk indicates a significant difference between controls and treated (P<0.05)

Full size image

Discussion

This study presents new findings on the antiproliferative mechanism of 7β-OHsito versus 7β-OHchol on human colon cancer cells. Caco-2 cells are more sensitive to 7β-OHchol than to 7β-OHsito. These compounds caused the same death rate if used at 30 and 60 μM, respectively.

Only 7β-OHsito altered cell cycle traverse leading to an accumulation in S phase, and a simultaneous decrease of cells engaged in the G2/M phase. In parallel, cells underwent apoptosis as characterized by the accumulation of cells containing hypodiploid DNA.42, 43 In contrast, 7β-OHchol did not alter the cell cycle, and delayed apoptosis by several hours. On the one hand, 7β-OHsito induced an apoptotic process mediated by caspase-9 and -3 activation, and on the other, cells treated with 7β-OHchol never showed activation of caspase-3, and only a low activation of caspase-9. However, DNA fragmentation was similarly enhanced in cells treated with both hydroxysterols. By using the broad-spectrum caspase inhibitor Z-VAD.fmk,31, 46, 47 we demonstrated that caspases activation plays an important role in the antiproliferative effect initiated by 7β-OHsito. This was not the case for cells treated with 7β-OHchol, which were not affected by the caspase inhibitor. These results indicate that different apoptotic processes are involved in the antiproliferative effects of the two compounds.

Another study has shown that both compounds exhibited a similar cytotoxicity pattern in U937 cells, whereas hydroxyphytosterols were less toxic.24 Cholesterol oxides and a mixture of β-sitosterol/campesterol oxides were examined for their cytotoxic effects on a macrophage-derived cell line (C57BL/6). All compounds exhibited similar cytotoxicity as indicated by LDH leakage, cell viability and mitochondria dehydrogenase activity, although oxyphytosterols exerted less severe effects than cholesterol oxides.23

In the presence of the pancaspase inhibitor, Z-VAD.fmk, apoptotic events were not completely suppressed.31 Hence, 7β-OHsito may trigger apoptosis through postmitochondrial caspase cascades and DNA fragmentation.48, 49 In contrast, 7β-OHchol showed no effect on cell cycle and induced cell death mechanisms leading to DNA fragmentation that were not delayed by the inhibitor Z-VAD.fmk. During the treatment with 7β-OHchol, a low but significant level of caspase-9 activity was measured, which did not induce an apoptotic process involving the activation of caspase-3. It was recently reported that when U937 leukemia cells were exposed to 7β-OHchol, they showed apoptotic characteristics such as permeability to PI, condensed and/or fragmented nuclei. It was suggested that these cells were apoptotic but that a different cell death pathway controlled by autophagic process was induced.38 It is possible that autophagy delays apoptosis, as was described for HT-29 cells treated with sulindac sulfide.50 It is worth noting that these events are not to be considered necessarily as separate events.51

Cholesterol oxides are potent regulators of cholesterol metabolism.52 They can enter in cells directly and are recognized by the oxysterol binding protein (OBP) localized in the endoplasmic reticulum37, 53 or by other OBPs such as liver X receptor.54, 55 This interaction may lead to the regulation of sterol metabolism by transcriptional responses.37 In this case, HMG-CoA reductase may be inhibited leading to the inhibition of cholesterol synthesis, with the consequence that cholesterol availability will be limited for cell growth and cells might undergo cell death. In addition to the antiproliferative properties of cholesterol oxides, it is known that these compounds are able to induce oxidative stress and consequently, they can play a direct role in atherogenesis or induce apoptosis of cancer cells.30, 56, 57 Oxidative stress induced by hydroxylated plant sterols and related products was not described.

At a first glance, it is difficult to explain why we observed caspase-9 activation and DNA fragmentation with 7β-OHchol, without any activation of caspase-3. However, it was previously shown that other inhibitors of HMG-CoA reductase, such as lovastatin, induce apoptosis in various cell types in which the role of caspase-3 is also controversial.58, 59 Caspase-independent cell death might be mediated by apoptosis-inducing factor (AIF), a mitochondrial flavoprotein, which is translocated from the mitochondria to the cytosol and to the nucleus when apoptosis has started. AIF has multiple effects such as activation of endonucleases, and phosphatidyl serine exposure at the cell surface. It also induces cytochrome c release,60, 61, 62 which may explain the delayed caspase-9 activation observed with 7β-OHchol in our present study.

Concerning 7β-OHsito, no data are available for its action at the cell surface. This β-sitosterol derivative acts as a cholesterol-lowering agent by inhibiting cholesterol absorption through displacement of cholesterol from micelles, but not by inhibiting cholesterol synthesis. In fact, inhibition of cholesterol absorption favors intracellular cholesterol synthesis without any negative effect on HMG-CoA reductase activity.63 It can be assumed that the additional alkyl group in the side chain prevents the plant sterols from an interaction with the sites responsible for the feedback regulation of cholesterol synthesis.64 It is known that the alkyl side chain render plant sterols more hydrophobic and consequently, they pass through the cell membrane more easily and provoke excitability inducing membrane-bound or intracellular enzymes.13, 19

In summary, the effects of 7β-OHsito and 7β-OHchol both induce death of Caco-2 cells. However, the two compounds target different death pathways in spite of their structural similarity. In the case of 7β-OHsito, the intracellular caspase cascade is activated, whereas in the case of 7β-OHchol, the process is more complex and may involve various caspase-independent factors, including cholesterol metabolism and its products, oxidative stress compounds and/or autophagic processes. These differences may be of interest for the development of new anticancer strategies.

Materials and Methods

Cell culture

Human colon adenocarcinoma Caco-2 cells were purchased from the European Collection of Animal Cell Culture (ECACC Salisbury, UK). They were cultured in 75 cm2 Falcon flasks in Dulbecco's modified Eagle's medium (DMEM) containing 25 mM glucose and supplemented with 10% heat-inactivated (56°C) horse serum, 100 U/ml penicillin, 100 μg/ml streptomycin, 1% nonessential amino acids (Gibco, Invitrogen Corp., Cergy Pontoise, France). Cells were incubated at 37°C in a humidified atmosphere with 5% CO2 and subcultured to preconfluency after trypsinization (0.5% trypsin/2.6 mM EDTA).

For all experiments, cells were seeded at 1 × 104 cells/well in 96-well plates and 1 × 106 cells in culture dishes (10 mm diameter). The culture medium was DMEM supplemented with 3% heat-inactivated horse serum, 5 μg/ml transferrin, 5 ng/ml selenium and 10 μg/ml insulin (Gibco, France).

Cell culture treatments with 7β-OHsito, 7β-OHchol and pancaspase inhibitor Z-VAD.fmk

Cells were treated with 7β-OHsito and 7β-OHchol at 30 and 60 μM. The β-sitosterol was purified by preparative adsorption chromatography starting from a commercial mixture of phytosterols (unsaponifiable fraction of Soya oil). The purity of the obtained β-sitosterol was higher than 95%. The impurities consisted of traces of campesterol. Cholesterol was a Sigma product (Sigma-Aldrich, Steinheim, Germany). The β-sitosterol and cholesterol were oxidized according to the method suggested by Schroepfer et al.65 After purification on silica column, the 7β-OHsito and 7β-OHchol were obtained with purities of about 95%. Ethanol 0.5% (v/v) was used for control group and as solvent for sterols. Untreated and treated cells were exposed to 50 μM of Z-VAD.fmk (in 0.1% dimethyl sulfoxide (DMSO)), (Sigma-Aldrich, Germany) for 1 h prior to exposure to sterols. Control cells were treated with both 0.5% ethanol and 0.1% DMSO. In all experiments, cells were incubated in a humidified atmosphere (5% CO2 at 37°C) for 1–8 days. Culture medium was changed every 24 h or 48 h.

Cell growth

Cells (1 × 104/well) were seeded in 96-well plates. Cells were exposed to different compounds 24 h after seeding and incubated for different times. Cell culture was stopped by the addition of 50 μl trichloroacetic acid (50%, v/v) and cell proteins were stained with 200 μl SRB (0.4% (w/v)) diluted in 1% acetic acid (Sigma-Aldrich, Germany). Cells were rinsed three times with 1% acid acetic. A measure of 200 μl/well of 10 mM Tris-HCl (pH 10.5) was added and absorbance was measured at 490 nm. The relationship between cell number (protein content/well) and absorbance is linear from 0 to 200 000 cells/well. For estimation of cell survival, cells were stained with Trypan Blue Dye (Gibco, France) (1/1, v/v) and the number of stained and nonstained cells was determined by optical microscopy.

Cell cycle analysis

Cell cycle distribution was analyzed by labeling cells with PI and assays were carried out as described previously.43 Briefly, 1 × 106 cells were seeded in 10 mm plates and harvested by trypsinization (0.5% trypsin/2.6 mM EDTA) at different time points after initial treatment with hydroxysterols. Then, cells were centrifuged and fixed in 1 ml methanol : PBS (9 : 1, v/v), washed twice in PBS and resuspended in 200 μl PBS containing 0.25 mg/ml RNAse A and 0.1 mg/ml PI (Sigma-Aldrich, Germany). After incubation in the dark at 37°C for 30 min, the fluorescence of 10 000 cells was analyzed using a FACSCcan flow cytometer and CellQuest software (Becton Dickinson, San Jose, CA, USA).

Determination of caspase-3, -8 and -9 activities

Cells (1 × 106) were harvested, washed twice in PBS and stored at −20°C. Caspase-3 and -8 activity was detected by using Caspase-3 and Caspase-8 Assay Colorimetric Kits (Sigma-Aldrich, Germany). These assays were based on the hydrolysis of the peptide substrate Ac-DEVD-pNA by caspase-3 and of the peptide substrate Ac-IETD-pNA by caspase-8, resulting in the release of a pNA moiety. The concentration of the pNA released was calculated from the absorbance values at 405 nm and calibration curve of defined pNA solutions. Results were adjusted to the total protein content, and values were expressed as nmol pNA/mg of total protein.

For caspase-9 activity measurement, the Caspase-9 Colorimetric Assay (R&D Systems, Minneapolis, USA) was used. This assay was similar to the procedure described above, except for the peptide substrate, which was LEHD-pNA. Results were expressed as fold increase in caspase activity of apoptotic cells over that of noninduced cells.

Determination of DNA fragmentation

Detection of apoptotic DNA from high molecular weight, intact, genomic DNA was performed as described previously.66 In brief, 1 × 106 were seeded in 10 mm plates, treated with hydroxysterols or ethanol and at each time point, cells were collected by trypsinization and centrifugation, and then stored at −80°C. Apoptotic DNA was recovered from cells by using the Suicide Track™ DNA ladder kit (Oncogene Research Products, Cambridge, UK). The samples were loaded into a 1.5% (w/v) agarose gel containing ethidium bromide and prepared in standard Tris-acetate-EDTA (pH 8.5) (50 × ) diluted 1 × . DNA fragments were separated by electrophoresis (100 V, 30 min) and were monitored under UV light by transluminator Gel Doc 2000 and analyzed with Quantity One 1-D Analysis Software (BioRad Laboratories, Marnes-la-Coquette, France).

Statistical analysis

All experiments were performed at least three times. Data are reported as mean±S.E. Statistical differences between control group and treated cells were evaluated using the Student's t-test. Differences were considered significant at P<0.05.