Introduction

In the last years, it has been suggested that the daily use of products with antimutagenic and anticarcinogenic activities may be an efficient way for preventing cancer. This approach is known as chemoprevention (Kassie et al. 2002; Majer et al. 2005).

In the diet, several products or compounds of different origins (e.g. cereals and fungi) contain chemopreventive properties. Among such products, there is the medicinal mushroom Agaricus blazei, native to Brazil, which is commonly consumed as a tea. There are different medical indications: it is used against stress, to stimulate the immune system, for improving the life of diabetics, to lower cholesterol, and against osteoporosis, among several other benefits. Due to its biological activities, it has been evaluated for its nutraceutical potential and as medicinal supplement. The constituents of the mushroom fruiting body include steroids (Kawagishi et al. 1988), lipids (Takaku et al. 2001), protein complexes (Ito et al. 1997; Fujimiya et al. 1998) and polysaccharides. Polysaccharides of the β-glucan type are attracting attention due to their antitumor (Mizuno et al. 1990), antiproliferative (Kobayashi et al. 1995), antigenotoxic and antioxidant (Chovartovicova et al. 1999; Angeli et al. 2006), and antimutagenic (Chovartovicova et al. 1999; Krizkova et al. 2006; Oliveira et al. 2007) activities and their capacity to increase the expression of the transcription factors c-Jun/AP1 in breast cancer cells (Talorete et al. 2002), among other properties.

These β-glucans polysaccharides consist of d-glucopyranosyl units and are found in several organisms (lichens, bacteria, fungi, yeast, and cereals). The macrostructure of β-glucan depends on the source and extraction method used (Zekovic et al. 2005). β-glucans extracted from the mushroom A. blazei has a structure comprising a main chain of d-glucose with β (1–3) and β (1–6) linkages (Fig. 1).

Fig. 1
figure 1

Structure of β-glucan extracted from Agaricus blazei, showing β(1 → 3, 1 → 6) linkages

Full size image

The aim of the present work was to evaluate the antigenotoxic effects of β-glucans extracted from A. blazei against DNA damage induced by the pro-carcinogenic compound benzo[a]pyrene (B[a]P), in the human hepatoma cell line HepG2, using the micronucleus test with cytokinesis block and the comet assay.

Materials and methods

Extraction of β-glucan from Agaricus blazei

The polysaccharides were extracted from a 5% aqueous suspension of mushroom (w/v), heated for 5 h, resulting in the acidification of the medium (pH 5). The polymer material was isolated from the aqueous extract after neutralization (0.1 mol/l NaOH), followed by the addition of 1% (m/v) NaCl (where, m is the mass of NaCl and v the extract volume) and precipitated in ethanol (1:5 v/v extract—ethanol). The precipitate was separated by centrifugation in an ethanol-hydrogen peroxide solution (1:1 v/v). Due to the partial solubilization of the material, it was submitted to a second extraction using ethanol (4:1 v/v ethanol—clearing medium). The soluble portion was lyophilized and dissolved in water. The structural characterization was performed by FTIR, 13C NMR, and 1H NMR spectroscopy, demonstrating a β-glucan—protein complex (Gonzaga et al. 2005). β-glucan was isolated from this fraction based on the procedure described by (Yoshioka et al. 1985). The suspension was placed on a shaker (15 h) and centrifuged (6,000 rpm for 60 min). The precipitate was washed in NaCl/thymol (1:1) solution and submitted to dialysis in water. The precipitate was dried in a sand bath (45°C). The structure of β-glucan was confirmed by 1H, 2D-COSY, HMQC, and 13C NMR spectroscopy.

DNA damage inducer

Benzo[a]pyrene—(Fluka)—was diluted in DMSO (JT Baker) and used at a concentration of 20 μM to induce DNA damage (concentration established in a pilot experiment—results not shown). The final concentration of DMSO in the medium did not exceed 1%.

HepG2 lineage and the experimental protocols

The HepG2 cell line was kindly donated by Dr. Siegfried Knasmuller from the Institute of Cancer Research, Medical University of Vienna, Austria. The cells were stored in foetal bovine serum + 10% DMSO, in a freezer at −80°C. The cultures were maintained in accordance with the protocol proposed by Uhl et al. (1999). The cells were cultivated in 75 cm2 culture flasks (TPP) in MEM medium supplemented with 15% foetal bovine serum, in a 5% CO2 incubator at 37°C, until reaching confluence. After this period, the cultures were trypsinized (0.1%), 2 × 105 cells/well were seeded on a 24-well plate, and the plates were incubated for 24 h, before performing the treatments. Three different β-glucan concentration were used: 7, 21, and 63 μg/ml; B1, B2, and B3, respectively, determined by pilot experiment (results not shown). Concerning the simultaneous treatment, the cells were exposed to β-glucan and to B[a]P simultaneously for 24 h. In the pre-treatment, the cells were exposed to β-glucan for 24 h before the addition of B[a]P and incubation for another 24 h. In the presimultaneous treatment, the cells were exposed to β-glucan for 24 h, followed by a change in culture medium and the addition of B[a]P. Unlike in the pre-treatment protocol, β-glucan was added again to the culture with B[a]P.

Alkaline single gel electrophoresis (comet assay)

Regarding the comet assay, the protocol used was that proposed by Uhl et al. (1999, 2000) according to the premises proposed by Tice et al. (2000). Briefly, a cell suspension (20 μl, 104 cells) was mixed with 0.5% low-melting-point agarose (100 μl) at 37°C, and distributed over agarose-coated microscope slides. The slides were covered and kept at 4°C for 20 min. Subsequently, the coverslips were removed and the slides were immersed in lysis solution (pH 10) containing 89.9 ml lysis buffer (2.5 M NaCl, 100 mM Na2EDTA, 10 mM Tris), 1 ml Triton-X 100 and 10 ml DMSO, protected from light, for 1 h at 4°C. After lysis, the slides were transferred to an electrophoresis chamber and immersed in alkali buffer. After alkali unwinding (1 mM Na2EDTA and 0.3 mM NaOH), pH 13.5, for 40 min. The electrophoresis were carried out under standard (25 V, 300 mA, 20 min). Subsequently, the slides were rinsed two times with 400 mM Tris buffer (pH 7.5), dried, fixed in 100% ethanol for 10 min, and stored at 4°C until analysis. Stained with 80 μl ethidium bromide solution (0.02% in water) and analysed with fluorescence microscopy (Nikon, model 027012). Every treatment was carried out with three independent repetitions. Concerning the comet assay, 100 cells were analysed visually (Kobayashi et al. 2005), and classified according to the following criteria: (class 0) cells with undetectable damage—without tail; (class 1) cells with tails whose length was less than the diameter of the nucleus; (class 2) cells with tails whose length was one to two times larger than the diameter of the nucleus; (class 3) cells with tails whose length was more than twice the diameter of the nucleus. Afterward, the scores were calculated by totalling each class value of the 100 cells examined for each treatment. Apoptotic cells that showed a totally fragmented nucleus were not considered in the analysis (Speit and Hartmann 2005). In parallel to the comet assay, 20 μl from the cell suspension were collected and mixed with 20 μl of trypan blue (Gibco). Afterward, the solution was placed in a Neubauer counting chamber and the cells were classified as living cells (those with no colouration) and dead cells (those with blue colouration). In all experiments, the viability of the cells was determined and only cultures in which survival was > 80% were evaluated for comet formation.

Cytokinesis-block micronucleus assay

The cytokinesis-block micronucleus assay was performed according to the protocol of Darroudi and Natarajan (1991). Thus, 2.5 × 105 cells were cultivated in 25-cm2 flasks for 2 days, in a 5% CO2 incubator at 37.5°C and 96% relative humidity. After 2 days of growth, the medium was removed and the cells were exposed to the compounds or to the solvent (DMSO—negative control) and to the different treatment protocols. In all cases, the cells were exposed for 24 h to the test compounds. The treated cells were washed and cytochalasin-B (3 μg/ml—Sigma) was added to obtain binucleated cells. The cells were harvested after 28 h of cytochalasin-B exposure, treated with cold sodium citrate and centrifuged (at 800 rpm for 8 min). The fixative methanol/acetic acid (3/1) was added to the pellet at the same time as a drop of formaldehyde, which was performed three times. For the detection of micronuclei, the cells were stained with Giemsa (5%). The nuclear division index (NDI) was calculated counting the number of binucleated cells in 1,000 cells, and dividing by 100 (Eastmond and Tucker 1989).

Statistical analysis

Statistical analyses were carried out with the Prism 4.0 statistical program. The Mann–Whitney test was performed to verify differences among the treatments in the comet assay, and ANOVA was performed, followed by Dunnett’s multiple comparisons test for the micronucleus test; significance was set at the 5% level for all tests.

Results

β-Glucan was devoid of genotoxic and mutagenic effect (Fig. 2). Increased DNA fragmentation, as determined by the comet assay, and alterations in cell viability (trypan blue) were not observed (Fig. 2a). Regarding the micronucleus test, there was neither an increase in the formation of micronuclei nor a delay in the cell cycle as demonstrated by the NDI (Fig. 2b).

Fig. 2
figure 2

Effects of β-glucan from Agaricus blazei on DNA fragmentation (a) and micronucleus formation (b) in the cell line HepG2. Bars indicate mean ± standard deviation of three independent repetitions. Control (DMSO-1%); B[a]P, 20 μM; β-glucan—B1, 7 μg/ml; B2, 21 μg/ml; B3, 63 μg/ml. Asterisk indicates statistically different from control (P < 0.05)

Full size image

Figure 3a and b represent the effects of the combination of three concentrations of β-glucan with the damage-inducing agent (B[a]P), in the simultaneous treatment, determined by the comet assay, micronucleus test, and trypan blue exclusion method for cell viability, respectively. In the comet and micronucleus assays, the three concentrations of β-glucan tested showed a dose-dependent protective effect. The mean score of the comet assay for B[a]P was approximately 120 arbitrary units; with the highest concentration of β-glucan combined with B[a]P, this value decreased to approximately 65 arbitrary units. Similar results were obtained for the micronucleus test, where B[a]P alone induced a micronucleus frequency higher than 50; when combined with the highest concentration of β-glucan, this frequency decreased to approximately 35, which was evident from the dose–response curve.

Fig. 3
figure 3

Effects of β-glucan from Agaricus blazei on DNA fragmentation (a) and micronucleus formation induced by benzo[a]pyrene (b) in the cell line HepG2 (simultaneous treatment). Bars indicate mean ± standard deviation of three independent repetitions. Control (DMSO-1%); B[a]P, 20 μM; β-glucan—B1, 7 μg/ml; B2, 21 μg/ml; B3, 63 μg/ml. Asterisk indicates statistically different from control (P < 0.05)

Full size image

Different results were observed in the two assays with pre-treatment (Fig. 4). Except for the highest concentration of β-glucan, which was effective in reducing DNA damage, the two lower concentrations did not differ from the positive control. In the comet assay, the positive control score was approximately 150 arbitrary units, which decreased to 120 when combined with the highest concentration of β-glucan (Fig. 4a). Similar results were obtained in the micronucleus assay, where the highest concentration protected against DNA damage induced by B[a]P (Fig. 4b). The mean frequency of micronuclei obtained was 47 versus 35, in the positive control combined with β-glucan.

Fig. 4
figure 4

Effects of β-glucan from Agaricus blazei on DNA fragmentation (a) and micronuclei formation induced by benzo[a]pyrene (b) in the cell line HepG2 (pre-treatment). Bars indicate mean ± standard deviation of three independent repetitions. Control (DMSO-1%); B[a]P, 20 μM; β-glucan—B1, 7 μg/ml; B2, 21 μg/ml; B3, 63 μg/ml. Asterisk indicates statistically different from control (P < 0.05)

Full size image

Finally, with aim of assessing the additive effect in the simultaneous treatment and pre-treatment, we pre-treated the cells with β-glucan; however, the presence of β-glucan in the medium was continued when B[a]P was added. There was a decrease in DNA fragmentation in a dose-dependent manner (Fig. 5a) as well as in micronucleus formation (Fig. 5b). The score in the comet assay for B[a]P was approximately 115 arbitrary units, which declined to 55 when combined with the highest β-glucan concentration. The micronucleus assay revealed the same result, where the positive control showed a mean micronucleus frequency of 55 which dropped to approximately 25 when combined with the highest β-glucan concentration. The micronucleus frequency in this latter treatment was lower than that seen in the negative control.

Fig. 5
figure 5

Effects of β-glucan from Agaricus blazei on DNA fragmentation (a) and micronucleus formation induced by benzo[a]pyrene (b) in the cell line HepG2 (presimultaneous treatment). Bars indicate mean ± standard deviation of three independent repetitions. Control (DMSO-1%); B[a]P, 20 μM; β-glucan—B1, 7 μg/ml; B2, 21 μg/ml; B3, 63 μg/ml. Asterisk indicates statistically different from control (P < 0.05)

Full size image

Discussion

Several compounds in the diet have shown anticarcinogenic and antimutagenic potential through different mechanisms, where inhibition of phase I enzymes (CYP 450 family), activation of phase II enzymes (GST family), induction of apoptosis, and protection against free radicals are the main ones (Lee and Park 2003). Stimulation of DNA repair, immunostimulatory effects, inhibition of cyclooxygenase, caloric restriction, and reduced transit of mutagens in the gastrointestinal tract are also some mechanisms by which foods can exert their protective effect (Ferrari and Torres 2003).

The present study shows that β-glucan did not induce any harmful effect on the genetic material, exerting additionally a substantial protective effect against a mutagenic agent that could be present in the human diet, depending on eating habits. β-glucan revealed a more intensive protective effect with simultaneous treatment, which indicates the binding of β-glucan directly to B[a]P, since it is already known that some dietary fibres are able to bind to aromatic compounds (Ferguson et al. 1993; Williams et al. 1999). However, we cannot ignore the fact that β-glucan may act as a free radical scavenger (Chovartovicova et al. 1999; Angeli et al. 2006), being able to bind to free radicals produced by the activation of B[a]P. The activation of such compounds by the enzymes of the P450 family generates free radicals which may interact noxiously with DNA (Briede et al. 2004).

The protective effect shown by β-glucan in pre-treatment indicates that it could act through enzyme modulation as already described in another work by our group (Angeli et al. 2006). Thus, these data reinforce this hypothesis, because this finding indicates the possibility of change in the expression and/or activity of some enzymes. This action could be supported by the involvement of receptors on the cell membrane which activate signalling pathways, expression of transcriptional factors and, consequently, modulation of gene expression; this is corroborated by the fact that polysaccharides extracted from the mushrooms Lentinus edodes and A. blazei show an in vivo inhibitory effect on the expression of isoenzymes of the P450 1A family (Hashimoto et al. 2002; Okamoto et al. 2004), which catalyse the first stage in the formation of the reactive metabolites of B[a]P. Although this hypothesis needs further investigations, since this parameters were not studied in this cell line.

When comparing the presimultaneous protocol to the simultaneous protocol, a significant statistical difference in results is noticed between them, demonstrating a greater protective effect when the cells were also exposed to β-glucan for 24 h in the absence of B[a]P, and therefore an additive effect. Thus, the greatest protective effect observed in the presimultaneous treatment, could be explained by an inhibitory effect of β-glucan on enzymes of the P450 family, as well as its role in the removal of reactive oxygen species formed in the activation of B[a]P, or even binding directly to B[a]P, inhibiting its entrance into the cell and metabolism. In general, the results of this study demonstrate that β-glucan (polysaccharide) extracted from A. blazei protects human hepatic cells against the genotoxic and mutagenic effects of B[a]P—a carcinogenic compound found commonly in the environment. This finding indicates to us that β-glucans could decrease the deleterious effects of this compound in real situations of human exposure, working therefore as a chemopreventive agent, which could be included in the diet of humans by consumption of A. blazei tea or as a dietetic supplement with β-glucan. Another intriguing finding of this study is the possible modulation of phase I enzymes by β-glucan, making it an interesting supportive agent in chemotherapy. According to Walter-Sack and Klotz (1996), the inhibition of CYP 450 enzymes enhances the efficacy of chemotherapy, since chemotherapeutic agents may take longer to be excreted.

The data here presented, is in accordance with other papers regarding protective effect of both, A. blazei extracts and β-glucan from different origins (Bellini et al. 2006; Oliveira et al. 2007). Thus presenting, almost none deleterious effects in physiological relevant doses and protective effect against a wide range of mutagenic compounds (Mantovani et al., in press), making A. blazei, a interesting source for chemopreventive compounds.