Introduction

Vitamin K, an essential nutrient, has an important role in normal blood coagulation systems [1]. It acts as a cofactor of a number of plasma proteins such as prothrombin and factors VII, IX, and X. In recent years, several investigations have shown that vitamin K also possesses anticancer activity [2, 3]. Hemostatic proteins, such as thrombin, fibrin, and tissue factors play dual roles in both thrombosis formation and cancer progression [4, 5]. Chemically synthesized vitamin K3 (2-methyl-1,4-naphthoquinone; VK3) has a potent anticancer effect against various types of carcinoma, including hepatic, oral cavity, pharyngeal, mammary, breast, bladder, and blood cancers in vitro [3].

The cellular metabolism of VK3 results in the formation of reactive oxygen species (ROS) including H2O2, O2 −·, and OH, the highly deleterious hydroxyl radical, is generated in the presence of metal ions. The generated ROS cause various alterations including the induction of macromolecular damage, disruption of calcium homeostasis, depletion of cellular thiol levels, increases in lipid peroxidation, breaks in DNA, and cell death [6–10]. Regarding apoptosis, two pathways have been proposed; one is a mitochondrial-based pathway [11] and the other operates via the endoplasmic reticulum (ER) [12] in normal pancreatic acinar cells.

Our goal in this study is to clarify the cytotoxic mechanism of VK3-induced cell death in breast cancer cells. Using the MCF-7 breast cancer cell line, we observed that VK3 induces apoptosis involving DNA fragmentation in MCF-7 cells. This apoptosis was initiated via ROS production, and then followed by mitochondrial dysfunction and activation of the caspase-9-related apoptosis pathway. We also demonstrated that the intensity of apoptosis induced by VK3 is twice as high in the breast cancer cell line MCF-7, than in normal mammary gland HMEpC epithelial cells.

Materials and methods

Cells and reagents

MCF-7 human breast cancer cells (passage up to 20) were provided by the Cell Resource Center for Biomedical Research, at the Institute of Development, Aging, and Cancer of Tohoku University. They were routinely maintained in a monolayer on plastic culture flasks at 37°C using minimum essential medium supplemented with 10% fetal calf serum, 100 IU/ml penicillin, and 100 μg/ml streptomycin. HMEpC human mammary epithelial cells (passage up to 15) were purchased from Toyobo Co. Ltd. (Osaka, Japan) and cultivated in the same manner as the MCF-7 cells.

Cell Counting Kit-F was obtained from Dojindo Laboratories (Kumamoto, Japan). VK3 was purchased from Nacalai Tesque Inc. (Kyoto, Japan). Ac-LEVD-CHO, an inhibitor of caspase-4, was bought from Calbiochem (Darmstadt, Germany), z-DEVD-fluoromethylketone (FMK), an inhibitor of caspase-7, was obtained from Alexis Biochemicals (San Diego, CA), z-LEHD-FMK, an inhibitor of caspase-9 was purchased from R&D systems (Minneapolis, MN). Calpain-specific substrate Suc-LLVY- aminomethylcoumarin (AMC) was purchased from Bachem AG (Bubendorf, Germany). They were dissolved in dimethyl sulfoxide (DMSO) just before use. The final concentration of DMSO in the medium did not exceed 0.1% and did not affect cell viability (data not shown).

Anti-caspase-7 and 9 antibodies were obtained from StressGen Biotechnologies (Ann Arbor, MI), and anti-caspase-4 antibody was acquired from Calbiochem. Peroxidase-conjugated anti-mouse and anti-rabbit IgG were purchased from Amersham Pharmacia Biotech (Arlinogton Heights, IL).

Cell viability assay

Cell Counting Kit-F was used to assess cell viability [13]. Briefly, cells were seeded onto 96 well microplates at a density of 1 × 104 cells/well and incubated for 24 h. The cells were treated with VK3 for a further 24 h. They were then washed with PBS and allowed to react with calcein-AM for 30 min. The released fluorescent calcein was measured in a CytoFluor® Plate Reader (PerSeptive Biosystems, Foster City, CA) using an excitation wavelength of 490 nm/emission 515 nm.

DNA fragmentation

The cells were plated onto 100 mm tissue culture dishes and treated with VK3 for 24 h. They were then harvested with 100 μl lysis buffer (10 mM Tris–HCl, 10 mM EDTA, and 0.5% Triton X-100, adjusted to pH 7.4) for 20 min at 4°C. After centrifugation (15,000g for 5 min at 4°C), the supernatant was treated with 20 μg RNase A for 60 min at 37°C and subsequently with 20 μg proteinase K for 60 min at 50°C. After phenol–chloroform extraction, the fragmented DNA was precipitated with 0.4 M NaCl/50% isopropanol at −20°C and then electrophoresed on a 2% agarose gel containing 0.1 μg ethidium bromide and visualized under a 312 nm UV transilluminator.

Measurement of the population of apoptosis cells using sub-G1 analysis

The sub-G1 population was measured to quantify the number of apoptotic cells. After 24-h cultivation, adhered cells were trypsinized, washed with PBS, and fixed with 70% ethanol for 1 h at −20°C. After being washed with phosphate-citrate buffer (0.2 M Na2HPO4 and 0.1 M citric acid), the cells were treated with staining solution (PBS containing 3 μg/ml propidium iodide and 10 μg/ml RNase A) for 5 min at room temperature. The DNA content was then analyzed by flow cytometry (FACSCalibur™, Becton, Dickinson and Company, Franklin Lakes, NJ).

Measurement of caspase and calpain activity

Caspase activation was evaluated by the cleavage of procaspase proteins. To do so, immunoblot analysis was performed for the detection of cleaved caspase fragments as described previously [14]. Additional competitive experiments with caspase inhibitors were also applied to confirm the involvement of the three caspases during apoptosis. Briefly, cells were plated on a 96-well plate and cultured for 24 h. After treatment with caspase inhibitors for 30 min, they were incubated with VK3 for more 6 h. The cell viability was assessed using Cell Counting Kit-F.

Calpain activity was determined using its specific substrate (Suc-LLVY-AMC) as described previously [14]. The enzyme activity of μ- and m-calpain was measured in the presence of 100 μM and 1 mM calcium, respectively.

Measurement of ROS

Reactive oxidative species generation was quantified with 2′,7′-dichlorofluorescein diacetate (DCFH-DA) as a substrate for measuring intracellular oxidant production [15]. Briefly, MCF-7 cells were incubated with 15 μM VK3 at 37°C in suspension culture for various time periods. After removing the VK3, the cells were incubated with 20 μM DCFH-DA for 20 min. After removing the DCFH-DA and washing the cells with ice-cold PBS, the fluorescence was measured by excitation at 480 nm/emission 530 nm.

Intracellular ATP level

The intracellular ATP level was quantified by luciferase reagents using a luciferin/luciferase assay kit (Cellno, TOYOINK™, Tokyo, Japan) according to the manufacturer’s instructions. Briefly, a 100 μl of luciferin/luciferase reagent solution was poured into the sample tubes, and the luminescent signal was measured with an MTP-700 Lab luminometer (CORONA Scientific Co. Ltd., Japan).

Measurement of mitochondrial membrane potential

J-aggregate-forming lipophilic cation (JC-1) was used to evaluate mitochondrial damage [16]. The uptake of JC-1 into mitochondria and the subsequent formation of J-aggregates is caused by high mitochondrial potential, and thus, this cationic dye is used to evaluate mitochondrial dysfunction.

Cells were seeded onto cover slips and treated with 15 μM VK3 for 20, 40, 60, or 120 min. They were incubated with 5 μg/ml JC-1 at 37°C for a further 30 min. After washing the cells thrice with PBS, they were applied to fluorescence microscopic observation. The shifting of mitochondrial membrane potential was also quantified using flow cytometry (FACSCalibur™ flow cytometry system, Becton, Dickinson and Co., Franklin Lakes, NJ). Briefly, after incubation with VK3, the cells were incubated with JC-1 for 15 min. After washing them twice with PBS, the adherent cells were trypsinized, and the suspension was subjected to flow cytometry. J-aggregates are detected as orange fluorescence, whereas J-monomers are detected as green fluorescence.

Statistical analysis

The data are expressed as the mean ± SD. Statistical significance was compared with Dunnett’s or Williams’ multiple comparison method using an one-way or two-way analytical model. P values of <0.05 were considered to be statistically significant.

Results

Cytotoxicity of VK3 in normal and tumor breast epithelial cells

Figure 1 shows the effects of VK3 on the cell viability of MCF-7 and HMEpC cells using cytosolic esterase activity as an indicator. VK3 reduced the enzyme activity of both the cells in a concentration-dependent manner. The IC50 value was 14.2 μM for MCF-7 and 27.9 μM for HMEpC cells. Since VK3 affects MCF-7 cells twice as much as HMEpC cells, this suggests the potential usefulness of VK3 chemotherapy against breast cancer.

Fig. 1
figure 1

Concentration-dependent cytotoxicity of VK3 in MCF-7 (filled square) and HMEpC (open square) cells. The cells were treated with 0–50 μM VK3 for 24 h, and then cell viability was assessed via cytosolic esterase activity. Values are expressed as the mean ± SD (n = 5). Statistical significance from the control: *P < 0.05, **P < 0.01

Full size image

VK3-induced apoptosis of MCF-7

Then, we evaluated the action mechanism of VK3-induced cell death in MCF-7. Agarose gel electrophoresis revealed that VK3 caused concentration-dependent apoptosis involving DNA fragmentation (Fig. 2a), especially at 20 μM VK3 exposure for 24 h. The DNA fragmentation was also quantified using flow cytometry analysis. As shown in Fig. 2b, the sub-G1 population was increased by the treatment of VK3 for 24 h at concentrations of 1–20 μM. These results indicate that VK3 induces apoptosis, although necrosis might occur at the higher concentration of 30 μM in MCF-7 cells.

Fig. 2
figure 2

Vitamin K3-induced apoptotic DNA fragmentation in MCF-7 cells. a Cells were treated with 10–30 μM drugs for 24 h, and then the nuclear DNA was electrophoresed onto 2% agarose gel. M 100-bp DNA ladder marker. b Quantitative evaluation of apoptosis in VK3-treated MCF-7 cells. The cells were cultured in GM with 0––20 μM VK3 for 24 h followed by fixation with 70% ethanol. The sub-G1 population was then analyzed using flow cytometry. Values represent mean ± SD (n = 8) **P < 0.01 versus control, as assessed by Williams’ multiple comparison test

Full size image

In addition, we investigated caspase dependency during this apoptosis. Immunoblot analysis was applied to detect caspase cleavage. Fifteen micromoles of VK3-induced cleavage of caspase-7 and 9 from 6 h onwards in MCF-7 cells, whereas caspase-4 cleavage was not observed during the 24-h treatment period (Fig. 3a). An additional cell viability assay was done using their specific caspase inhibitors (Fig. 3b). VK3-induced cell death was significantly reduced with specific inhibitors of caspase-7 and 9, whereas the caspase-4 inhibitor did not have any effects. Further, we measured the cytosolic calpain activity, which is activated by cleaved caspase-4 in the ER stress-related caspase cascade [17]. Interestingly, both m-calpain and μ-calpain were time-dependently activated in MCF-7 cells; however, VK3 did not affect their activity (Fig. 4). Taken together, our current observations suggest that mitochondrial damage followed by the induction of caspase-7 and 9, at least partly contributes to VK3-induced apoptosis in MCF-7 cells.

Fig. 3
figure 3

Examination for the participation of various caspase cascades during VK3-induced MCF-7 apoptosis. a Caspase cleavage was visualized using Western blotting analysis. Cells were treated with 15 μM VK3. Cleaved caspase-4 (top), 7 (middle), and 9 (bottom) were detected with their specific antibodies. b The protecting effect of caspase-specific inhibitors on VK3-induced cytotoxicity. The cells were incubated with (filled triangle) or without (open circle) 50 μM caspase-4, 7, and 9 inhibitors for 30 min followed by treatment with VK3 for 6 h. Cell viability was then assessed using Calcein-AM. Values represent the mean ± SD (n = 6). Statistical significance was assessed by two-way ANOVA; **P < 0.01

Full size image
Fig. 4
figure 4

Effect of VK3 on cytosolic calpain activity in MCF-7 cells. The cells were treated with (filled triangle) or without (open circle) 15 μM VK3 for 0, 6, 9, 12, 15, or 18 h. They were then incubated with substrate buffer for 60 min and supplied for subsequent m- (a) and μ-calpain (b) analysis. Values represent the mean ± SD (n = 8). Statistical significance was assessed by two-way ANOVA: *P < 0.05, **P < 0.01

Full size image

The effects of VK3 on intracellular ROS and ATP levels and mitochondrial membrane potentials in MCF-7 cells

As predicted above, VK3 induces apoptosis in MCF-7 cells through mitochondrial failure. In order to confirm the hypothesis, we directly evaluated the effect of VK3 on mitochondrial function. Mitochondrial dysfunction has been shown to lead to the generation of ROS, which could be a contributing factor to cell death [18, 19]. Therefore, we investigated whether ROS are generated in MCF-7 cells. When MCF-7 cells were treated with 15 μM VK3, a significant increase in intracellular ROS (Fig. 5a) and a decrease in ATP (Fig. 5b) occurred in a time-dependent manner. Therefore, we observed the mitochondrial membrane potential in the presence and absence of VK3, using JC-1 cationic dye (Fig. 6). The control cells showed heterogeneous staining of the cytoplasm with both red and green fluorescence coexisting in the same cells (Fig. 6a–c). Exposure of the MCF-7 cells to 15 μM VK3 for 6 h induced marked changes in mitochondrial potential as evident from the disappearance of red fluorescence and the increase in green fluorescence in most cells with a predominantly peripheral distribution. Several cells were devoid of red fluorescence, which is an indication of the loss of mitochondrial potential and the severity of cell damage caused by 15 μM VK3 treatment for 6 h (Fig. 6d). These alterations induced mitochondrial dysfunction and a loss of mitochondrial membrane potential.

Fig. 5
figure 5

Several intracellular changes associated with mitochondrial dysfunction. a VK3-induced ROS generation in MCF-7 cells. The cells were treated with (filled triangle) or without (open circle) 15 μM VK3 for 20, 40, 60, or 120 min. Then, they were reacted with 20 μM DCFH-DA for 20 min. Values represent the mean ± SD (n = 4) **P < 0.01 versus control assessed by Williams’ multiple comparison test. b Rapid ATP decrease in VK3-treated MCF-7 cells. Cells were treated with (filled triangle) or without (open circle) 15 μM VK3. Intracellular ATP levels were measured using a luciferase-based bioluminescent assay. Values represent the mean ± SD (n = 4) **P < 0.01 versus control, as assessed by Dunnett’s multiple comparison test

Full size image
Fig. 6
figure 6

VK3 caused the loss of mitochondrial membrane potential. Fluorescence microscopic observation of JC-1-stained MCF-7 cells: control at 0 (a) and 6 h (c), 15 μM VK3 treatment for 0 (b) and 6 h (d). Functional mitochondria containing J-aggregates were stained orange, whereas damaged mitochondria containing J-monomers were stained green. var: 10 μm. e Quantitative analysis of JC-1-stained cells using flow cytometry. The cells were incubated with 15 μM VK3 for 6 and 9 h and then, they were stained with JC-1 for 15 min. Values represent the mean ± SD (n = 5). **P < 0.01 versus control, as assessed by Williams’ multiple comparison test

Full size image

The mitochondrial damage was quantified by flow cytometry. JC-1-stained cells revealed a significantly increased green ratio in a time-dependent manner (Fig. 6e). These results indicate that VK3 induces mitochondria dysfunction, probably through ROS, and resulting in apoptosis.

Discussion

In this study, the authors evaluated the potential of VK3 as a chemotherapeutic agent against breast cancer cells. Indeed, in the MCF-7 human breast cancer cell line, VK3 caused a dose-dependent cytotoxicity at an IC50 value of 14.2 μM (Fig. 1). This cytotoxicity was induced through nuclear DNA fragmentation (Fig. 2) suggesting that VK3 causes cell death in MCF-7 by apoptosis. Our previous report indicated that VK3 also possesses cytotoxic activity in the Hep G2 human hepatoblastoma cell line with a similar IC50 value [20]. On the other hand, VK3 has a lower cytotoxicity against HMEpC cells with an IC50 of 27.9 μM (Fig. 1b), which is twice as high as that of MCF-7 cells. Therefore, these observations suggest that VK3 has the potential to be used as a wide-spectrum chemotherapeutic agent against recalcitrant solid carcinoma. Salido et al. [21] suggested that ROS generated by VK3 induce growth suppression that includes the direct suppression of mitochondrial function. In general, mitochondrial defects have long been suspected to play an important role in the development and progression of cancer. Indeed, due to the inherent inefficiency of glycolytic ATP generation, malignant cells require a high amount of glucose to fulfill their cellular energy requirements. Therefore, the differences in energy metabolism between normal and cancer cells constitute a biochemical basis to speculate that therapeutic strategies might be developed to selectively kill cancer cells via their inherently compromised respiratory state [22–25]. Taken together, VK3 might induce severer cytotoxicity in carcinoma cells than in normal cells.

We next investigated the mechanism of action of VK3-induced apoptosis in MCF-7 cells. Generally, three main caspase cascades induce irreversible apoptosis: the ‘death ligand’-mediated pathway through caspase-8, the mitochondrial damage-induced pathway through caspase-9, and the ER stress-related pathway through caspase-12 [26]. The caspase molecules transmit apoptotic signals in a manner that transcends species differences [27]. In humans, caspase-4 mediates apoptosis specifically in response to ER stress in Alzheimer’s disease [17, 28] and in neuronal ceroid lipofuscinoses, commonly known as Batten disease [29]. Caspase-7 is considered to be the downstream and ultimate apoptosis factor in the mitochondrial pathway in MCF-7 cells [30].

Based on the above information, we measured the activation of various caspases in MCF-7 cells (Fig. 3). VK3 activated caspase-7 and 9, whereas it did not cleave the caspase-4 molecule. In addition, inhibition of these caspases recovered cell viability (Fig. 3b). These findings indicate that the mitochondrial pathway plays a central role in VK3-induced apoptosis. In order to check this hypothesis, we examined mitochondrial function in VK3-treated MCF-7 cells. VK3 generated intracellular ROS quickly and decreased the intracellular ATP level (Fig. 5). Finally, a fluorescent dye assay using JC-1 (Fig. 6) directly detected that VK3 caused mitochondrial dysfunction accompanied by the disappearance of membrane potential, which is essential for mitochondrial energy production.

It is unclear whether VK3 accumulates in breast cancer cells. Although VK3 is quite lipophilic and conceivably distributes to various organ, the distribution to breast cancer may not exceed to other organs. However, previous reports showed that VK3 is rapidly conjugated to intracellular glutathione via nucleophilic reaction to form a stable conjugate and the loss of viability of Hep G2 cells upon exposure to VK3 is preceded by rapid depletion of intracellular glutathione [31, 32]. This glutathione decrement was caused by the conjugation of glutathione to VK3 or the oxidation of glutathione to the disulfide form [33]. The S conjugate, thiodione, is hydrophilic and thus does not permeate the plasma membrane by passive diffusion.

In addition, our personal data also showed that the intracellular ATP level was not decreased by the addition of VK3 to HMEpC cells (data not shown). This suggests VK3 preferentially affects breast adenocarcinoma rather than normal mammary gland. Because this report focused on the mechanism of cytotoxicity induced by VK3 in cancer cells, we did not describe this observation in the current manuscript.

Park et al. [34] reported similar apoptosis results in VK3-treated cardiomyoblast cells, i.e., VK3-induced apoptosis caused by mitochondrial dysfunction due to generated ROS. In general, mitochondria have two essential functions during apoptosis: (1) energy production in the form of ATP, which is required by all cells that die via an apoptotic pathway and (2) proapoptotic proteins are sequestered into the intermembrane space of the cytosol, in which they trigger downstream apoptotic signaling pathways [17, 35–39]. In addition, apoptosis is accompanied by signs of mitochondrial dysfunction, including a loss of the inner mitochondrial transmembrane potential; the release of soluble intermembrane proteins, including cytochrome c; and caspase-9 activation [40–42].

Several investigators have shown that ROS are also induced by ER stress-related calcium ion leakage [12] and the subsequent activation of calpains and caspase-4, which are signal factors related to apoptosis [17]. In contrast, Obeng and Boise [43] suggested that caspase-4 activation is not required for the induction of ER stress-induced apoptosis. VK3-induced apoptosis in MCF-7 cells might be not necessary for the activation of the ER. Our results only revealed that VK3 generated ROS, causing mitochondrial dysfunction (Figs. 5a, 6). Before this, caspase-9 activation followed by the caspase-7 activation occurred. In contrast, caspase-4 activation was not observed suggesting that VK3-induced apoptosis is mainly triggered by mitochondrial damage and is independent of the ER stress-related pathway. Further investigation is needed to confirm the effects of VK3 on mitochondrial function.

In conclusion, VK3 caused apoptosis in breast tumor cells via the mitochondria-related pathway. Its potency is as twice high in MCF-7 cells than in normal breast epithelial cells, and thus, it might be useful for chemotherapy in breast cancer patients.