Plumbagin Modulates Leukemia Cell Redox Status

Plumbagin is a plant naphtoquinone exerting anti-cancer properties including apoptotic cell death induction and generation of reactive oxygen species (ROS). The aim of this study was to elucidate parameters explaining the differential leukemia cell sensitivity towards this compound. Among several leukemia cell lines, U937 monocytic leukemia cells appeared more sensitive to plumbagin treatment in terms of cytotoxicity and level of apoptotic cell death compared to more resistant Raji Burkitt lymphoma cells. Moreover, U937 cells exhibited a ten-fold higher ROS production compared to Raji. Neither differential incorporation, nor efflux of plumbagin was detected. Pre-treatment with thiol-containing antioxidants prevented ROS production and subsequent induction of cell death by apoptosis whereas non-thiol-containing antioxidants remained ineffective in both cellular models. We conclude that the anticancer potential of plumbagin is driven by pro-oxidant activities related to the cellular thiolstat.

healthy donors. We selected the U937 (most sensitive) and Raji (less sensitive) cells for a comparative mechanistic study. As PBMCs were not affected by the treatment, we also confirmed the excellent differential anti-cancer potential of plumbagin. Altogether, we observed that the pro-oxidant regulation is independent of a differential intake/uptake of the compound by the two cancer cell models. We rather demonstrate the differential ability of the compound to elicit ROS in U937 and Raji cells as well as to impact the intracellular GSH pool. We finally suggest a differential expression of redox-related factors as potential regulators of the observed differential susceptibility.

Plumbagin Reduces Leukemia Cell Viability
Evaluation of the effect of plumbagin on the viability of different leukemia cell lines by trypan blue exclusion assay revealed that this compound presents a cytotoxic effect towards all cell lines tested (Table 1). U937 cells appear as the most sensitive cell line with an IC 50 ranging from 0.82 ± 0.04 μM to 0.66 ± 0.02 μM observed between 24 and 72 h of treatment. Raji cells were less sensitive with an IC 50 value of 5.06 ± 0.22 μM and 2.66 ± 0.03 μM respectively after 24 and 72 h treatment. Even at the highest concentration tested (10 μM), PBMCs were not affected by plumbagin treatment. For further mechanistic studies of the effects of plumbagin, we selected U937 and Raji cells to perform a comparative analysis using IC 50 concentrations at 24 h, respectively, 1 µM for U937 and 5 µM for Raji cells. Table 1. Cytotoxic effect of plumbagin on different human leukemia cell lines compared to PBMCs from healthy donors. IC 50 values were determined by three independent trypan-blue assays after 24, 48 and 72 h of treatment. The data are the mean of at least three independent experiments ± SD. N.C. stands for "not cytotoxic" (viability > 80%) for a concentration up to 10 μM.

Cell Lines
IC 50

Plumbagin Induces Apoptotic Cell Death
Considering the elevated levels of cytotoxicity, we analyzed the type of cell death triggered by plumbagin in U937 and Raji cells by fluorescence microscopy after staining with Hoechst and propidium iodide (PI). 24 h of treatment at a concentration of 1 μM (U937) and 5 μM (Raji) induced the appearance of nuclear morphological alterations compatible with apoptosis in both cell lines ( Figure 1A,B). This finding was further confirmed by the analysis of the exposure of phosphatidylserine by Annexin V/PI assay ( Figure 1C,D). Results pointed out that both cell lines died by an apoptotic process in a dose-dependent manner. These results were confirmed by Western-blot analysis that showed caspase cleavage and decrease Mcl-1 and Bcl-2 anti-apoptotic protein expression levels, starting from the respective IC 50 concentration of plumbagin in U937 and Raji cells (Figure 2A). Pre-treatment with the pan-caspase activity inhibitor (Z-VAD-FMK) prevented the 17-10 kDa caspase-3 fragment formation. This result confirms that plumbagin induces cell death by a caspase-dependent apoptotic process ( Figure 2B). These results have been confirmed by fluorescence microscopy analysis after Hoechst staining (data not shown). Cell population corresponding to early and late (in secondary necrosis) apoptotic cells are respectively in the lower and upper right quadrants. All results presented are the mean ± SD of at least three independent experiments. The values in both tables correspond to percentage of apoptotic cells of at least three independent experiments. * p < 0.05, ** p < 0.01 and *** p < 0.001 compared to non-treated cells, respectively.

Plumbagin Induces Different Levels of Intracellular ROS in U937 vs. Raji Cells
Published data demonstrated the capacity of plumbagin to elicit ROS in cancer cells [26,27,49]. Analysis of intracellular ROS production in U937 and Raji cells exposed to plumbagin was performed by flow cytometry analysis after staining with 2',7'-dichlorodihydrofluorescein diacetate (H 2 DCFDA). As shown in Figure 3A,B, plumbagin induces ROS production in both cell lines tested. U937 cells show a very robust increase in intracellular ROS already after 15 min treatment at 1 μM plumbagin. Raji cells, in contrast, show a much milder intracellular ROS increase compared to the positive control hydrogen peroxide (H 2 O 2 , 50 μM). The level of ROS production induced by plumbagin is ten-fold higher in U937 compared to Raji cells. The incubation of Raji cells with a lower, sub-apoptogenic, concentration of plumbagin (1 μM) showed a comparable level of ROS production. The flow cytometry analysis did not reveal any cell subpopulations differently responsive to ROS production, therefore indicating a homogenous ability of cells to increase ROS production upon treatment. Moreover, no significant changes in ROS production were observed for longer incubation times with both cell lines. These results suggest that the differential plumbagin-induced ROS production is an early event.

Differential ROS Generation Is not a Consequence of a Different Uptake/Efflux of Plumbagin
The observed ROS generation could be the consequence of a differential internalization of plumbagin. As plumbagin is a fluorescent pigment [45], we then investigated differential internalization. Analysis of the intracellular fluorescence of plumbagin by flow cytometry (see "Experimental" section) revealed that this compound accumulates similarly in U937 and Raji cell lines as a function of incubation time ( Figure 4A). Besides, in the drug-efflux test, no decrease of the plumbagin-generated intracellular fluorescence was observed in both cell lines up to 90 min of incubation in plumbagin-free culture medium (recovery). Fluorescence maintained very high levels up to 4 h (not shown) without any difference depending on the cell models. Untreated cells did not show any modulation of fluorescence over the time as expected (data not shown). This aspect may reveal specific aspects related to the intracellular metabolism of the compound and may deserve future investigations, beyond the scope of this study. Altogether, we can exclude a differential internalization or efflux of the compound as the responsible factor for lower ROS levels in Raji. Samples were collected and analyzed by flow cytometry without additional staining; (B) Plumbagin efflux assay was performed after 30 min of incubation with the same concentration of plumbagin used for the uptake assay, followed by recovery in plumbagin-free medium (see "Experimental" section) [50]. U937 and Raji cells were collected at indicated times and their auto-fluorescence was analyzed by flow cytometry. Untreated cells did not show any modulation of fluorescence over the time as expected (data not shown). Results are the mean ± SD of at least three independent experiments.

Thiol-Containing Antioxidants Prevent Plumbagin-Induced Apoptosis
As ROS are known to play a key role in apoptosis induction [51], we investigated the effect of antioxidants on ROS generation and apoptotic cell death induction. Using BSO, a GSH depletor, and H 2 O 2 as positive controls for ROS generation, our analysis showed that the pre-treatment with DTT [52], NAC [12,15] and Trolox [53] buffers plumbagin-dependent ROS production whereas the metal chelator Tiron [53,54], a hydroxyl radical and superoxide scavenger, remained ineffective as antioxidant agent in both models ( Figure 5A,C). Then, we estimated the percentage of apoptosis by analyzing the loss of mitochondrial membrane potential (see "Experimental" section), a marker of the mitochondrial apoptotic pathway [12,50]. Only pre-treatment with DTT or NAC, two thiol-containing antioxidants, prevented cell death, while non-thiol antioxidants, Tiron and Trolox, did not affect apoptosis, although Trolox was able to buffer ROS formation ( Figure 5B,D). Cancer cells typically develop alterations of their oxidative status, by showing altered expression patterns of enzymes whose function might depend on thiol modulation [40,42,43,[55][56][57]. Our findings indicate an ability of plumbagin to eventually lead to the modulation of important intracellular functions especially those dominated by thiol modulation, which include also enzymes controlling and/or modulating the cellular redox status in cancer cells.

Plumbagin Decreases the Intracellular GSH Level
Next, we investigated the impact of plumbagin on GSH as it has been shown that GSH depletion is a common feature in apoptotic cell death [58]. To elucidate the mechanisms explaining the differential sensitivity, we measured the level of total GSH, which is 30% higher in Raji compared to sensitive U937 cells ( Figure 6A). These differential GSH levels could explain the plumbagin-induced ROS production observed in the two selected models (Figures 3 and 5). Then, we analyzed the GSH/GSSG ratio after stimulation with plumbagin using BSO-and NAC-treated cells as controls. Plumbagin decreases the GSH/GSSG ratio in both cell lines in a dose-dependent manner. NAC treatment per se did not significantly alter the GSH/GSSG ratio observed in control cells. In Raji cells, a 40% decrease is observed at 5 μM compared to a decrease of 60% in U937 cells ( Figure 6B). The analysis of reduced GSH by O-phthalaldehyde (OPA) assay revealed that a dose of 5 μM of plumbagin is requested to decrease the GSH content by 25% in Raji cells whereas 1 μM is sufficient to reach the same level in U937 cells. 24 h pre-treatment with NAC completely abrogates the previously described depletion of GSH by plumbagin ( Table 2). Table 2. Intracellular GSH level was determined using OPA probe after 1 h of treatment with plumbagin. Cells pre-treated during 24 h with NAC (10 mM) were used as control. Results are the mean ± SD of at least three independent experiments (values are indicated as a ratio of fluorescence values between treated and control cells). * p < 0.05, and ** p < 0.01 compared to non-treated cells, respectively.
The isolated peripheral blood mononuclear cells (PBMCs) were cultured at 37 °C and 5% CO 2 for 24 h before use.

Fluorescence Microscopy
After plumbagin treatment, cells were stained with Hoechst 33342 (Calbiochem) and propidium iodide (Sigma-Aldrich) during 20 min at 37 °C. Labeled cells were analyzed with an inverted Cell M Olympus Microscope (Olympus, Aartselaar, Belgium) and Cell M software.

Apoptosis Assays
Apoptosis was assessed and estimated by three different assays: (1) analysis of nuclear fragmentation (Hoechst staining and fluorescence microscope observation, performed as previously described [59]; (2)

Western-Blot
Total proteins extracts were subjected to sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE, 12%) and transferred onto nitrocellulose membranes (Hybond™-P membrane, GE Healthcare). Membranes were pre-hybridized with 5% non-fat milk in PBS 1X containing 0.1% (v/v) Tween 20 (PBS-T) overnight at 4 °C or 1 h at room temperature. Membranes hybridizations with primary antibodies directed against caspase-3, caspase-7, caspase-8, caspase-9 (Cell Signaling, Bioké, Leiden, The Netherlands), Bcl-2 (Calbiochem) and β-actin (Sigma) used as loading control, were carried out in PBS-T containing 5% milk or 5% bovine serum albumin (BSA) for 1 h at room temperature or overnight at 4 °C, according to the providers' protocols. Etoposide-treated U937 cells (VP16, 100 μM, 4 h) were used as apoptosis positive control and equal loading of samples was controlled using β-actin. After incubation with primary antibodies, membranes were washed and probed with the corresponding secondary (horseradish peroxidase conjugated) antibodies following manufacturers' instructions for 1 h at room temperature. Proteins of interest were visualized with ECL Plus Western Blotting Detection Reagents (GE Healthcare) using the ImageQuant LAS 4000 Mini (GE Healthcare).
In the presence of ROS, the non fluorescent cell permeant DCFDA is converted in highly fluorescent 2',7'-dichlorofluorescein (DCF). 50 μM of H 2 O 2 for 15 or 30 min were used as an inducer of ROS production (positive control). Relative intracellular ROS levels were depicted as mean fluorescence intensity (MFI).

Plumbagin Intracellular Uptake and Efflux
Exponentially growing Raji and U937 cells were exposed to 5 μM plumbagin. Plumbagin uptake was assessed by measuring plumbagin intracellular fluorescence from compound-loaded cells after different incubation times (15,30,45,60,90 and 120 min). At the end of these specific incubation times, cells were collected, centrifuged and re-suspended in fresh medium for further analysis. Efflux of plumbagin was evaluated in the following way, as previously described for doxorubicin [50]. After 30 min of incubation with plumbagin (5 μM), plumbagin-containing medium was removed and cells were re-suspended in fresh medium for recovery [50]. Fluorescence was evaluated immediately (T = 0 min) and after 15, 30, 60, and 90 min. Plumbagin fluorescence was evaluated at the indicated times by flow cytometry using a FACSCalibur, tuned at 488 nm, at standard pass filters; FL2 (FL2 = 585/42 nm). Data were recorded using the CellQuest software and further analyzed with FlowJo.

Analysis of GSH Content
Reduced (GSH) and oxidized (GSSG) glutathione measurements were performed using the GSH/GSSG-Glo™ Assay kit (Promega, Leiden, The Netherlands). Briefly, after treatment with plumbagin, 5 × 10 5 cells are collected, centrifuged and resuspended in 1 mL of pre-warmed Hank's Buffered Salt Solution (HBSS). Cells treated with BSO served as a positive control of depletion of GSH content [60]. A volume of 25 μL of the cell suspension is transferred into wells of a 96-well plate. An equivalent volume of appropriate lysis buffer is then added. GSH/GSSG-Glo assay is then performed following manufacturer's instructions. The analysis of cellular GSH content was carried out by staining of the cells with o-phtalaldehyde (OPA), a permanent fluorescent probe. OPA is a direct tool that can interact with small thiol groups (e.g., GSH) in order to form adducts with them. Briefly, Raji and U937 cells were treated with 1 or 5 μM of plumbagin for 1 h. At the end of the incubation time, cells were washed with PBS and incubated with 50 μM OPA for 20 min. OPA fluorescence was evaluated by spectrofluorimetry (SpectraMax Gemini EM, Molecular Devices, Sunnyvale, CA, USA).

Statistical Analysis
Results from at least three independent experiments were analyzed for statistical significant differences using the Student's t-test. They are expressed as the mean ± SD. p-values <0.05 (*), <0.01 (**) and <0.001 (***) were considered as statistically significant.

Conclusions
Plumbagin is a natural compound that exerts differential cytotoxicity towards leukemia cancer cells resulting from its modulatory activities on the cellular redox state, however the actual cellular targets of plumbagin in the redox control remain still under debate. ROS increase is commonly detected upon plumbagin treatment. Recently, several studies have pointed at the specific ability of plumbagin to modulate the intracellular thiols. These findings would imply the modulation of the intracellular thiolstat as relevant to trigger the anti-cancer effects of plumbagin, rather then the generation of ROS, which therefore would appear as a merely additional side effect, in the fact not essential for its anticancer activity. Our data seem to support this latter model (Table 3), as apoptosis induced by plumbagin in our sensitive (U937) and less sensitive (Raji) hematopoietic cancer cell models can be prevented only by antioxidants containing thiol species. There is evidence that plumbagin may directly interact with GSH, by likely a nucleophilic addition, which in turn may contribute to GSH depletion [22,45]. The small intracellular thiol GSH is paradigmatic for a huge group of additional and more complex intracellular thiols potentially targetable by plumbagin, which also includes many structural proteins and enzymes. Tubulin is among the proteins/enzymes known to be bound by plumbagin [61]. Remarkably, thiol modulation is particularly relevant also for the multi-step activation of the pro-apoptotic Bcl-2 family members [62][63][64]. Strong evidence suggests the modulation of specific Bax (Bcl-2-associated X protein) cysteine residues is critical for the acquisition of its suitable conformation, oligomerization and translocation/insertion into the mitochondrial membrane [65]. It would be relevant in the future to investigate any potential ability of plumbagin to directly interact and thereby modulate i.e., Bax activation. Such interactions can further contribute to protein derivatization [45]. Glutathione-S-transferases, which control detoxification through the consumption of glutathione (GSH), may also be inactivated by plumbagin. This modulation is paralleled by ROS formation [66]. A previous analysis of GSTP1 level expression, in our lab revealed that the less plumbagin-sensitive Raji cells do not expressed GSTP1 proteins in contrast to the most sensitive U937 cell model here investigated [67,68]. These differential alterations may potentially provide additional hints to may identify the reason of such differences in ROS generation. Taken all together, our and other findings indicate the ability of plumbagin to may eventually lead to the modulation of important intracellular functions dominated by thiol modulation.