Epi, a substrate of P-gp, MRP1, and MRP2 [5,12], was selected as a model anticancer drug in this study. The fermented product 8HD exhibits cytotoxic effect on Caco-2 cells (Figure 2A). Thus, this compound is used as an adjuvant with Epi in Caco-2 cells to intensify potency of chemotherapy. An increase in ROS levels, including H₂O₂ and O₂⁻ through co-incubation with 8HD (Figure 3) enhanced Epi cytotoxicity (Figure 2B) and generated a synergistic effect (Figure 2C). Previous investigations have suggested that daidzein, the precursor of 8HD, may exhibit multiple anticancer effects, including proliferation inhibition, P-gp and MRP suppression, and apoptosis induction [1,26–28,30,34]. It is worth stressing that the goal of the present work was to evaluate the effect of ROS production after 8HD and/or Epi treatments on triggering the cellular mechanisms of inhibiting efflux transporter-related MDR and inducing apoptosis. The proposed pathways for reversing pump and non-pump MDR in Caco-2 cells are shown in Figure 7. The detailed mechanisms are discussed as follows.

Intracellular ROS (iROS) have been implicated in P-gp and MRP modulation. ROS might act as negative regulators of P-gp expression [8,9]. An increase in iROS levels decreases P-gp expression, as well as reduces efflux of the P-gp substrate doxorubicin in prostate adenocarcinoma cells [9,35]. However, contradictory studies have demonstrated that high concentrations of ROS confer oxidative stress and enhance P-gp and MRP expressions [11,36]. These conflicting results on ROS levels and expression of MDR transporters inspired our current investigation.

The relationship between the effect of 8HD on ROS production and regulation of P-gp and MRPs has not been reported. In the present study, Epi-related pump resistance was reversed in various degrees by the addition of 8HD through the suppression of P-gp and MRPs. Accordingly, daidzein, the precursor of 8HD, increased intracellular drug levels by modulating P-gp and MRP functions in human cervical carcinoma cells and mice [26,34]. Daidzein was determined as a P-gp/MRP1 inhibitor, significantly increasing the uptake of [(3)H]-atazanavir (ATV) (a P-gp, MRP, and hOATP substrate) in human leukemia cells [27]. In addition, transport of conjugated daidzein is inhibited by the MRP inhibitor leukotriene C4 [24]. Daidzein also inhibits MRP1-, MRP4-, and MRP5-mediated transport and modulates MDR [37]. With these precedents in mind, we suggested that 8HD decreased mRNA expressions of functional MDR transporters (Figure 4A), which might reduce epirubicin efflux and enhance epirubicin accumulation (Figure 4B). The functional involvement of 8HD in reversing MDR transporter mediated Epi resistance was thus verified. We suggest that 8HD might function as a transporter substrate and thus compete with anticancer drug extrusion.

Our study demonstrates that 8HD and/or Epi enhanced ROS production, but reduced the expression of MDR1, MRP1, and/or MRP2, in agreement with other studies regarding the negative correlation between ROS production and MDR regulation [8,9,35]. In addition, we found that the ROS scavenger, NAC, one precursor of intracellular glutathione, abrogated the modulation effect of 8HD and Epi on the MDR transporter expression (Figure 4A). NAC might abolish the negative regulation of P-gp through the inhibition of c-Jun-N-terminal kinase (JNK) signaling pathway [38]. We thus verify that the decrease of MDR1 and MRP1 by 8HD and/or Epi through ROS production circumvented the pump resistance. However, the complicated regulation requires further clarification in the signaling pathway.

Intracellular ROS is likewise involved in apoptosis regulation. Potentiation of oxidative stress is closely linked to apoptotic response induced by a number of antineoplastic agents [11]. Interestingly, doxorubicin, an anthracycline anticancer drug and Epi anomer, induced tumor necrosis factor-alpha (TNF-alpha) that caused ROS generation, leading to mitochondrial dysfunction and apoptosis [11]. Furthermore, ROS played a key role in the cytotoxic effect of genistein on diverse cell lines [39,40]. Oxidative stress is thus confirmed as critical in apoptosis induction, acting as an upstream signaling initiator to trigger programmed cell death in different cell lines.

However, the relationship between the effect of 8HD on ROS production and apoptosis regulation has not been discussed in the literature. 8HD (7,8,4′-trihydroxyisoflavone) can be biotransformed from daidzein (4′,7-dihydroxyisoflavone), and thus 8HD may induce cancer cell death in mechanisms similar to those of daidzein. Daidzein triggered ROS generation, inhibited cell proliferation and caused significant cell cycle arrest in G1 and G2/M, and induced apoptosis through the mitochondrial caspase-dependent pathway in human cervical cancer, breast cancer, and murine neuroblastoma cells [1,30–32,41]. 8HD showed antioxidative and antiproliferative activities in human promyelocytic leukemia cells [18]. This compound also demonstrated antimutagenic activity and inhibitory activity against tyrosinase [24]. 8HD was verified as a potent suicide substrate of mushroom tyrosinase and showed in vitro cellular tyrosinase and melanogenesis inhibitory activities in mouse melanoma cells and in vivo skin-whitening activity in human volunteers [17]. In addition, this compound possessed DPPH-radical scavenging activity [19]. However, 8HD did not significantly induce apoptosis associated with changes in cell cycle distribution and caspase 3 activation in prostatic cancer cells [20].

With these precedents in mind, our current research clearly shows that 8HD and/or Epi initiated the central cell death signal by overexpressing p53, subsequently triggering the activated proapoptotic proteins, such as Bax. p53 played a key role in promoting cell apoptosis through a positive regulation of Bax and a negative regulation of Bcl-2 [42]. In addition, p53 caused arrest of cell cycle in G₁ and G₂/M by regulating genes such as p21_WAF1[43]. An increase in the percentage of sub-G₁ (apoptosis) and G₂/M (completed DNA replication) fractions of the Caco-2 cell cycle (Figure 5C) also implies that checkpoints are initiated by oxidative stress so that cell cycles are paused for DNA repair and thus prevented cell progression from G₂ into mitosis. Furthermore, it has been reported that the G₁/S checkpoint suppresses the replication of the damaged DNA and that the G₂/M checkpoint stops a cell from entering the M phase for a broken genome [40,44]. If the DNA damage cannot be repaired, the ROS produced by 8HD and/or Epi then increased the Bax/Bcl-2 ratio of the proapoptotic and antiapoptotic Bcl-2 family proteins. These changes were essential in opening the mitochondrial permeability transition pore, disruption of mitochondrial membrane potential, and activation of caspase-9 (Figures 5A and 6).

Two major apoptotic signaling pathways exist, namely, intrinsic or mitochondrial pathway and extrinsic or death receptor pathway. Caspase-8 is one of the initiator caspases for the death receptor pathway, whereas caspase-9 belongs to the initiator caspases for the mitochondrial pathway. Caspase-3 is a downstream effector caspase shared by both pathways [5,42]. In this study, independent Epi treatment significantly affected the caspase-9 and -3 expressions, but had only a marginal effect on caspase-8. This finding was in agreement with previous studies [5,13], confirming that Epi induces the mitochondrial apoptotic pathway. However, incubating Caco-2 cells with 8HD alone or combined with Epi significantly increased the expression levels of caspases-3, -8, and -9, as well as the activities of corresponding caspases in Caco-2 cells. This result suggested that both mitochondrial and death receptor pathways are involved in 8HD-mediated apoptosis. The caspase family of cystein proteases subsequently mediated a series of morphological changes, including chromatin condensation (data not shown) and DNA fragmentation (Figure 5D). A recent study has revealed that gomisin N, a lignan isolated from Schisandra chinensis, enhanced apoptosis induced by TNF-related apoptosis-inducing ligand (TRAIL) through ROS-mediated upregulation of death receptor 4 (DR4) and DR5, causing subsequent activation of caspases-3 and -8 [45]. We speculate that independent 8HD or its combination with Epi may also promote apoptosis mediated by TNF family proteins, such as Fas and TRAIL, through the caspase cascade, including caspases-3 and -8. Whether such a similar signaling pathway is involved needs further examination. To the best of our knowledge, we demonstrate for the first time that 8HD and Epi changed ROS levels; modulated the expression and function of efflux transporters; and induced apoptosis through both the intrinsic and extrinsic pathways in Caco-2 cells. We verified, at least in part, that 8HD and/or Epi activated the crosstalk among ROS levels, MDR transporter modulation, and apoptosis regulation in Caco-2 cells. The proposed schematic diagram for the possible underlying molecular mechanisms is shown in Figure 7.

Epi (Pharmorubicin) was purchased from Pfizer Inc. (New York, NY, USA). All cell culture media and reagents were purchased from Gibco BRL (Grand Island, NY, USA) or Hyclone (Logan, UT, USA). Most of the other chemical reagents were purchased from either Merck (Darmstadt, Germany) or Sigma-Aldrich (St. Louis, MO, USA). 8HD is a gift from Professor T.S. Chang. 8HD was isolated from soy germ koji fermented by A. oryzae BCRC 32288 (Food Industry Research and Development Institute, Hsinchu, Taiwan) in the lab of Professor T.S. Chang. The structure has been identified as 7,8,4′-trihydroxyisoflavone (Figure 1B) [15,16].

Caco-2 cell line was obtained from the Food Industry Research and Development Institute. Caco-2 cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM, Gibco) supplemented with 10% fetal bovine serum (Hyclone) and 1% penicillin and streptomycin (Hyclone) at 37 °C in a humidified atmosphere of 5% CO₂ and 95% air.

Caco-2 cells were seeded in 96-well plates at a density of 3.5 × 10⁴ cells/well, allowed to attach overnight, and then treated with different Epi concentrations with or without 8HD (0, 10, 25, 50, and 100 μM). After 48 h incubation, 100 μL of MTT (0.2 mg/mL) was added to each well, and the cells were reincubated for an additional 4 h. Dimethylsulfoxide (100 μL) was added to each well to dissolve the formazan. Absorbance (OD₅₄₅) was measured at 545 nm using a microplate reader (MRX; Dynatech Laboratories, Chantilly, VA, USA). Cell viability (%) was calculated by dividing the number of cells incubated with Epi and/or 8HD by the number of cells incubated with only DMEM (control). Potency of single administration of Epi, as well as that of combined Epi and 8HD, on cell-growth inhibition is expressed as IC₅₀, defined as the drug concentration necessary to inhibit cell growth by 50%. Data are the means ± standard deviation (SD) of three experiments.

In addition, the combination index (CI) in isobologram analysis was used to evaluate the synergistic, antagonistic, or additive effects of the combined treatment of 8HD and Epi. The CI is calculated using the formula: CI = a/A + b/B, in which “a” is the concentration of Epi required to achieve 50% growth inhibition in the combined treatment; “A” is the concentration of Epi that produces an identical effect alone; “b” is the concentration of 8HD which reaches a 50% growth inhibition in the combination; “B” is the concentration of 8HD that causes the same effect alone [46]. CI < 1 represents synergy, CI = 1 stands for additivity, and CI > 1 corresponds to antagonism [47].

Cells were seeded in 6-well plates at a density of 2 × 10⁵ cells/well, allowed to attach overnight, and exposed to 0.3 μg/mL of Epi and/or 25 μM of 8HD for 48 h. Cells were then incubated in the dark with DCFH-DA (20 μM) or DHE (5 μM) at 37 °C for 1 h. The cells were then harvested and immediately analyzed using a flow cytometer (Cell Lab Quanta SC MPL (abbreviated as Quanta SC); Beckman Coulter, Fullerton, CA, USA). Data acquisition and analysis were performed using commercial software (Quanta SC) [40].

Cells were pretreated with or without the antioxidant NAC (5 mM) for 30 min, and exposed to Epi (0.3 μg/mL) and/or 8HD (25 μM) for 48 h. Total RNA was extracted from cells using the Total RNA Extraction Miniprep System (Viogene, Taipei, Taiwan) according to the manufacturer’s instructions. RNA yield and purity were assessed using NanoDrop 2000 (Thermo, Wilmington, DE, USA). cDNA was prepared from total RNA using a High-capacity RNA-to-cDNA kit (Applied Biosystems; Foster City, CA, USA), following the manufacturer’s protocol. Gene-specific primers (Table 1) of MDR1, MRP1, MRP2, Bcl-2, Bax, and p53, as well as caspases-3, -8, and -9, were developed through multiple sequence alignment. GAPDH served as an internal control. Quantitative PCR was performed using the StepOne Real-Time PCR system (Applied Biosystems) and SYBR Green PCR Master Mix (Applied Biosystems). The cycling program was set as follows: denaturation at 95 °C for 10 min, followed by 40 cycles of 95 °C for 15 s and 60 °C for 1 min. Results were normalized to GAPDH level. mRNA expression ratio was calculated as the mRNA expression level compared with the cell control. Each experiment was performed in triplicate.

Cells were seeded in six-well plates at a density of 2 × 10⁵ cells/well, allowed to attach overnight, pretreated with 25 or 50 μM of 8HD for 1 h. Then, 0.3 μg/mL of Epi was added to the culture medium for 3 h. The cells were collected and suspended in PBS at 37 °C. Flow cytometric analysis was then conducted using a flow cytometer (Quanta SC) equipped with an argon ion laser and operated at 488 nm. At least 10,000 cells were analyzed in each sample. Within each experiment, determinations were performed in triplicate [5].

Cells (2 × 10⁵ cells/well) were incubated for 48 h with or without 0.3 μg/mL of Epi and 25 μM of 8HD. The cells were then treated with DiOC₆ (10 μM) for 1 h at 37 °C. Then cells were harvested and immediately analyzed using a flow cytometer (Quanta SC). DiOC₆ was excited at 488 nm, and fluorescence was analyzed at 525 nm (FL-1) after logarithmic amplification [31].

Cells at a density of 2 × 10⁵ cells/well were incubated for 48 h with or without 0.3 μg/mL of Epi and 25 μM of 8HD. Cells were harvested by centrifugation and fixed gently with 70% of ice-cold ethanol overnight at −20 °C. The cells were collected by centrifugation. Then, the cells were stained with 1 mg/mL propidium iodide (PI), incubated at 25 °C for 30 min in the dark, and analyzed using a flow cytometer (Quanta SC).

DNA fragmentation was detected by agarose gel electrophoresis. 2 × 10⁵ cells/well were gently scraped and harvested by centrifugation after Epi and/or 8HD treatments for 48 h. The cells were lysed with lysis buffer (50 mM Tris-HCl, 10 mM EDTA, 0.5% Triton X-100, 0.5 mg/mL Proteinase K) and then incubated at 56 °C for 24 h. DNA was extracted with phenol, chloroform, and isoamyl alcohol (25:24:1), prior to electrophoresis on 2% agarose gel at 50 V. The gel was stained with SYBR^® Safe (Invitrogen, Carlsbad, CA, USA) and digitally photographed using a gel documentation system (UVIdoc; UVItec Limited, Cambridge, UK).

The activities of caspases-3, -8, and -9 were detected with Caspase-Glo 3, Caspase-Glo 8, and Caspase-Glo 9 Assay Kits (Promega, Madison, WI, USA), respectively. Cells (2 × 10⁵ cells/well) were harvested after treatment with 0.3 μg/mL of Epi and/or 25 μM of 8HD for 48 h. Fifty microlitres of the cell suspension was then mixed with 50 μL of caspase-3, -8, and -9 reagents containing the corresponding luminogenic substrate of Ac-DEVD-pNA, Ac-LETD-pNA, and Ac-LEHD-pNA, respectively, at room temperature for 30 min. Released aminoluciferin luminescence levels were detected using a luminometer (MiniLumat LB9506; Berthold, Bad Wildbad, Germany) [48].

All data are expressed as means ± SD. Student’s t-test was used to analyze differences between two treatment groups. Statistical analysis was also conducted using one-way ANOVA and Dunnett’s multiple comparison tests. Significant difference was set at p < 0.05.