Hemicellulase-treated A. blazei (AB fraction H, ABH) was supplied by Japan Applied Microbiology Research Institute Ltd (Yamanashi, Japan). Mycelia of A. blazei Murill were digested with 0.1% hemicellulases from Trichoderma viride, Trichoderma harzianum, Aspergillus tamari and Aspergillus niger for 1 h at 45°C. After inactivation of the enzymes at 70°C, the filtrate of the degradation product was lyophilized. ABH consists of 63.3% carbohydrates, 30.9% proteins, 0.3% lipids and minor components. The samples were examined for endotoxin by limulus amebocyte lysate (LAL) assay (Seikagaku Corp., Tokyo, Japan). The endotoxin contents of samples were <0.02 EU per 1 mg of sample. The culture medium containing 0.1% ABH had a pH of 7.1–7.4 and iso-osmotic pressure (285 mOsm l⁻¹), determined by the freezing point depression method (Auto & Stat OM-6030; ARKRAY Inc., Kyoto, Japan). For comparison with ABH, mycelia of A. blazei Murill (Japan Applied Microbiology Research Institute Ltd) were water-extracted without hemicellulase. After heat treatment, the extract was lyophilized as described above.

Cell growth was assessed by formazan formation induced by reduction of exogenous tetrazolium salt. Cells (1 × 10⁴ per well for non-adherent cells or 4 × 10⁵ per well for adherent cells) were cultured in flat-bottomed 96-well plates (BD Biosciences, San Jose, CA). After incubation for the indicated times, 2-(2-methoxy-4-nitrophenyl)-3-(4-nitrophenyl)-5-(2,4-disulfophenyl)-2H-tetrazolium monosodium salt and 1-methoxy-5-methylphenazinium methylsulfate mixture (TetraColor ONE™; Seikagaku Corp., Tokyo, Japan) was added to each well. The optical density was measured at 450 and 630 nm (Molecular Devices Corp., Sunnyvale, CA) after 2 h. In another experiment, non-adherent cells (2 to 4 × 10⁵ per well) were cultured in 24-well plates (BD Biosciences). The viable cell number in the cultures was counted by trypan blue dye exclusion method at the indicated times. The viability was determined as percentage of trypan blue-negative cells to total cells.

Apoptosis was assessed by determining phosphatidylserine exposure on the cell membrane and DNA fragmentation. Cells were labeled with fluorescein isothiocyanate-conjugated annexin V and propidium iodide (PI) in binding buffer (Biovision Inc., Mountain View, CA), and analyzed using FACSCalibur™ with CellQuest™ software (BD Biosciences). On the other hand, cells were incubated with 500 μg ml⁻¹ proteinase K (Sigma-Aldrich, St Louis, MO), 500 μg ml⁻¹ RNase (Sigma-Aldrich), and 1% SDS at 37°C for 30 min. Total DNA was extracted from these cultures using a DNA extraction kit (Wako Pure Chemical Industries Ltd, Osaka, Japan). DNA samples were electrophoresed in 1.5% agarose gels and visualized by ethidium bromide staining.

Cells were fixed with 70% ethanol (Wako) at 4°C for 2 h. After washing and centrifugation, the precipitates were resuspended in 50 μg ml⁻¹ PI solution including 250 μg ml⁻¹ RNase, and incubated at 37°C for 30 min. DNA content in the cells was determined using FACSCalibur™ with CellQuest™ software. In cell cycle analyses, the signals of aggregating cells were removed on an FL-width/FL-area dot plot. In one experiment, CCRF-CEM cells were enriched at early S, mid-S or G₂/M and G₁ phases of the cell cycle with a high concentration of thymidine as described previously (18). Briefly, the cells were cultured in the presence of 2 mM thymidine (Sigma-Aldrich) for 24 h, and then cultured without thymidine for 12 h. The cells were further incubated with 2 mM thymidine for 14 h, followed by a 4, 10 or 16 h recovery period without thymidine for cell enrichment at early S, mid-S or G₂/M and G₁ phases, respectively.

Aliquots of cells (5 × 10⁵) were stained with a carboxyfluorescein-labeled fluoromethyl ketone peptide (FAM-LEHD-FMK; Chemicon International Inc., Temecula, CA), which permeates the cell membrane and binds covalently to active caspase-9, in culture medium at 37°C for 1 h according to the manufacturer's instructions. After washing and PI staining, the caspase activity was analyzed by flow cytometry.

The cytosolic protein fraction was prepared as described previously (19). Briefly, cells were suspended in extraction solution (20 mM HEPES, pH 7.5, 10 mM KCl, 1.9 mM MgCl₂, 1 mM EDTA, 1 mM EGTA) containing protease inhibitor cocktail (Santa Cruz Biotechnology Inc., Santa Cruz, CA) and incubated on ice for 20 min. The homogenate was centrifuged at 1000× g for 5 min to pellet out nuclei and intact cells. The supernatant was further centrifuged at 14000× g for 30 min to collect the cytosolic (supernatant) fraction. Aliquots of 30 μg of the protein were separated by SDS–PAGE and transferred onto PVDF membranes. The membranes were blocked with 5% non-fat dry milk in Tris-buffered saline, and then incubated with 0.2 μg ml⁻¹ anti-cytochrome c antibody. Blots were developed with horseradish peroxidase-conjugated secondary antibody, and proteins were visualized by enhanced chemiluminescence staining (Amersham Biosciences, Piscataway, NJ).

To investigate the effects of ABH in more detail, we examined whether ABH caused apoptosis in these cells. Five leukocytic cell lines were cultured in the presence of 0.1% ABH for 3 days as described above. FACS analyses showed populations of annexin V-stained cells with different staining properties of PI in ABH-treated Jurkat and BALL-1 cells (Fig. 3A). On the other hand, these populations were detectable but were markedly smaller in CCRF-CEM and U937 cells cultured with 0.1% ABH, which had a significant antiproliferative effect on these two cell lines. In addition, ABH did not induce apoptosis in THP-1 cells under the same conditions. Another characteristic feature of apoptosis, DNA fragmentation, which is detected as DNA ladder formation by electrophoretic separation on agarose gels, was also observed in ABH-treated BALL-1 cells (Fig. 3B, lane 3). However, fragments of apoptotic DNA were not detectable in ABH-treated CCRF-CEM cells (lane 5) in this assay, as was the case with ABH-treated THP-1 (lane 7) and all untreated cells (lanes 2, 4 and 6). These results indicated that ABH induced apoptosis in some cell lines, but that it also had other effects on cell growth, because it strongly suppressed the proliferation of CCRF-CEM and U937 cells without severe cytotoxic effects. Consequently, we postulated that ABH exerted an influence on cell cycle progression. To confirm this, five leukocytic cell lines, which were cultured with ABH as in the above assay, were fixed every 12 h and stained for DNA with PI. FACS analyses demonstrated the cycle phases of individual cells, allowing the measurement of the DNA contents of the cells (Fig. 4). ABH induced alterations in the distribution of cell cycle phases in four cell lines (Jurkat, BALL-1, CCRF-CEM and U937) during incubation. Their dominant peaks on histograms shifted from G₁ to early S phase in the first 24 h. The majority or a portion of these cells moved synchronously in S phase during the next 12 h. On the other hand, the distribution pattern of cell cycle phases of ABH-treated THP-1 remained virtually constant throughout the experimental period, as was the case with all untreated cell lines (data not shown). Although the receptivity of these cell lines to ABH in alteration of cell cycle progression corresponded to that in cell proliferation, these results raised an additional question of whether this alteration resulted in inhibition of cell proliferation. Therefore, CCRF-CEM cells were pre-treated with a high concentration of thymidine, followed by different recovery periods without thymidine to enrich the cells in early S, mid-S or G₂/M and G₁ phases. These cells were then left untreated or were cultured with ABH as before (Fig. 5). The cells in early S phase at the onset (0 h) progressed synchronously in this phase during the first 6 h without ABH, and the bulk of the cells entered G₂/M phase within 12 h, and formed a dominant peak of G₁ phase, as seen on the histogram at 18 h (Fig. 5A, top). However, the cells in early S phase at the onset moved more slowly in this phase in culture with ABH, and almost all cells were still in S and G₂/M phases even after 30 h (Fig. 5A, bottom). Similarly, the cell cycle progression of cells in mid-S phase at the onset was markedly delayed in cultures to which ABH was added, and the cells did not form a dominant peak of G₁ phase even after 24 h unlike those cultured without ABH (Fig. 5B). In contrast, the cells in G₂/M phase at the onset made a transition to G₁ phase even in the presence of ABH to the same extent as those cultured without ABH, and then the G₁ peak formed by these cells shifted to early S phase (Fig. 5C). These results suggested that ABH delayed cell cycle progression in S phase resulting in concentration of the cells in this phase. The inhibition of cell proliferation was induced, at least in part, by exposure to ABH in S phase and subsequent delay of cell cycle progression in responsive cells.

A recent study demonstrated that β-glucan extracted from A. blazei induced mitochondria-mediated apoptosis in a human ovarian HRA cell line (15). Therefore, we examined release of mitochondrial cytochrome c into the cytosol, which is one of early events in induction of mitochondria-mediated apoptosis (20), in cultures with ABH. BALL-1, CCRF-CEM and THP-1 cells, which were differentially sensitive to ABH, were cultured with 0.1% ABH, and the cytosolic fraction was prepared every 24 h. Western blotting analysis indicated that ABH treatment induced release of cytochrome c into the cytosol in a time-dependent manner in BALL-1 cells (Fig. 6A). The release of cytochrome c was slight in the first 24 h, but it increased to a large extent subsequently. On the other hand, ABH-treated CCRF-CEM cells showed the minimal release at a later stage in culture, and no release of cytochrome c was observed in ABH-treated THP-1 cells throughout the experimental period. Next, these cells were stained with PI and a fluorescent-conjugated peptide (FAM-LEHD-FMK), which binds active caspase-9, as the release of mitochondrial cytochrome c leads to activation of caspase-9, followed by activation of caspase-3, which initiates a set of events that culminate in apoptosis (21). FACS analyses detected larger populations of caspase-9-active BALL-1 cells in culture with ABH for 3 days (Fig. 6B) but not 1 day (data not shown). As expected, the degrees of caspase-9 activation corresponded to those of release of cytochrome c in ABH-treated CCRF-CEM and THP-1 cells.

The study by Kobayashi et al. cited above also indicated that β-glucan extracted from A. blazei induced activation of p38 MAPK in HRA cells, and that p38 MAPK-specific inhibitor suppressed β-glucan-induced apoptosis (15). p38 MAPK has been shown to respond strongly to various stress stimuli, such as osmotic shock, lipopolysaccharide, inflammatory cytokines and irradiation (22–24), and studies using the inhibitor suggested that p38 activation is responsible for apoptosis induced by these stressors (25). Therefore, we also analyzed p38 activation after treatment with ABH in five leukocytic cell lines. These cells were incubated with 0.1% ABH for up to 64 h. The permeabilized cells were reacted with anti-phospho-p38 MAPK mAb, and the levels of phosphorylated p38 MAPK were measured by flow cytometry (Fig. 7). ABH induced strong p38 activation, which peaked within 15 min and returned to the basal levels in 60 min in both Jurkat and BALL-1 cells. On the other hand, ABH also activated p38 MAPK in CCRF-CEM and U937 cells, although the levels were very much lower than in Jurkat and BALL-1 cells. CCRF-CEM cells showed extremely weak and slow p38 activation in response to ABH, peaking at 16 h. However, these four cell lines showed p38 activation even after 64 h in culture. In contrast, ABH-treated THP-1 cells showed no p38 activation up to 64 h in culture. Next, we examined whether MEK and MAPK inhibitors attenuated the ABH-induced effects on cell growth. Four cell lines responsive to ABH were pre-treated with 10 μM SB203580, a p38 MAPK-specific inhibitor, 4 μM SP600125, a JNK-specific inhibitor or 10 μM PD98059, a MEK-specific inhibitor, and then cultured with or without 0.1% ABH for 3 days. Their cell proliferation was examined using tetrazolium salt as described above, and expressed relative to cell growth in cultures without ABH (Fig. 8A). These concentrations of inhibitors had no effect on growth of the four cell lines by themselves (data not shown). However, SB203580 weakened the antiproliferative effects of ABH on all of these cell lines, blocking the antiproliferative effects in CCFR-CEM and U937 cells almost completely. On the other hand, SP600125 and PD98059 had almost no influence on the ABH-induced effects, or rather increased the effect slightly, except in U937 cells. Increased concentrations of these two inhibitors (up to 50 μM) also did not improve the proliferation of ABH-treated cells (data not shown). FACS analyses confirmed that pre-treatment with SB203580 moderately or significantly inhibited ABH-induced apoptosis in Jurkat or BALL-1 cells, respectively (Fig. 8B). It was interesting to note that SB203580 also markedly improved the ABH-induced delay of cell cycle progression in S phase to almost the same degree as in cultures without ABH in CCRF-CEM cells, which had been synchronized in S phase at the onset of the experiment (Fig. 8C). These results indicated that ABH induced continual p38 activation, the levels of which differed between cell lines, and that this p38 activation was involved in not only apoptosis but also in the delay of cell cycle progression in S phase induced by ABH.

ABH showed a weak cytotoxic effect in CCRF-CEM and U937 cells despite its strong antiproliferative effect (Fig. 2A). It induced low levels of p38 activation in CCRF-CEM and U937 cells unlike BALL-1 and Jurkat cells, in which high levels of p38 activation resulted in both alteration of cell cycle progression and apoptosis. Therefore, we hypothesized that weak p38 activation by ABH did not result in apoptosis. However, we also suspected that these two cell lines were not susceptible to apoptosis regardless of the degree of p38 activation. Consequently, we attempted to increase the activity of p38 MAPK in these two cell lines. TNF-α has a wide variety of physiological activities. The ligation of TNF-α and its receptor on the cells has been shown to induce p38 activation (23). In addition, it has been shown that TNF-α triggers death receptor-mediated apoptosis (26). However, it also has anti-apoptotic effects, activating NF-κB (27). In fact, TNF-α (10 nM) had only a minor impact on growth of CCRF-CEM and U937 cells by itself (Fig. 9A). FACS analyses also showed that TNF-α induced little apoptosis in these two cell lines (Fig. 9B). However, the combination of 0.1% ABH and 10 nM TNF-α significantly augmented the cytotoxic effects on these cell lines. This combination decreased the cell viability of the cultures to far lower levels (Fig. 9A), and generated larger populations of annexin V-stained cells (Fig. 9B). Furthermore, the combination of ABH and TNF-α significantly augmented the activation of caspase-9 in CCRF-CEM cells (Fig. 9C), although TNF-α alone had little effect on the activation of caspase-9 in CCRF-CEM cells, and ABH also hardly activated caspase-9 in these cells (Fig. 6B). On the other hand, the growth of THP-1 cells was unaffected even by the same combination of ABH and TNF-α. Based on these results, the levels of p38 activation were measured in cultures with ABH and TNF-α (Fig. 10A). TNF-α induced sufficient levels of p38 MAPK activation immediately in these two cell lines, although it induced little apoptosis. On the other hand, p38 activation induced by combination of ABH and TNF-α was far stronger and longer lasting. These two cell lines showed a transient decrease in p38 MAPK activity after the rapid and strong activation, but the cells then showed restoration of the extremely strong activation at 16 and 64 h. Next, we examined whether this p38 activation was correlated with the enhanced cytotoxicity induced by combination of ABH and TNF-α. SB203580 significantly attenuated the enhanced cytotoxic effect on CCRF-CEM and U937 cells cultured with ABH and TNF-α (Fig. 10B). FACS analyses also confirmed that it blocked apoptosis induced by ABH and TNF-α (Fig. 10C). Although this enhancement of the cytotoxic effect was not due to upregulation of TNF type 1 receptor (TNFR1) on the cells by ABH (data not shown), the mechanism remains to be elucidated in detail. However, these results suggested that the degrees of p38 activation induced by ABH, which were dependent on tumor cells and other stimuli, determined subsequent cellular responses on cell growth. We envisaged that some cell factors defined the responsiveness of p38 MAPK to ABH. Many molecules, such as TNF receptor family and Toll like receptors (28,29), protein kinase C (PKC) subtypes (30) and small G proteins (31), have been suggested to play a role in p38 activation. The quantitative or functional differences in these molecules between cell lines may result in differences in responsiveness to ABH. ABH is a crude extract from A. blazei mycelia and consists of multiple components. The components responsible for the ABH-induced effects demonstrated in the present study have not yet been identified. However, as hemicellulase-degradation of A. blazei enhanced the inhibitory effects on cell proliferation, some low molecular weight carbohydrate components may be involved in these effects, acting on the receptors, signal proteins or transcriptional factors that have not yet been clarified.