Eltrombopag modulates reactive oxygen species and decreases acute myeloid leukemia cell survival

Abstract

Previous studies have demonstrated that the small molecule thrombopoietin (TPO) mimetic, eltrombopag (E), induces apoptosis in acute myeloid leukemia (AML) cells. Here, we sought to define the mechanism of the anti-leukemic effect of eltrombopag. Our studies demonstrate that, at a concentration of 5 μM E in 2% serum, E induces apoptosis in leukemia cells by triggering PARP cleavage and activation of caspase cascades within 2-6 hours. The induction of apoptotic enzymes is critically dependent on drug concentration and the concentration of serum. This effect is not associated with an alteration in mitochondrial potential but is associated with a rapid decrease in a reactive oxygen species (ROS) in particular hydrogen peroxide (H2O2). Interestingly, E also decreases mitochondrial maximal and spare respiratory capacities suggesting an induced mitochondrial dysfunction that may not be readily apparent under basal conditions but becomes manifest only under stress. Co-treatment of MOLM14 AML cells with E plus Tempol or H2O2 provides a partial rescue of cell toxicity. Ferric ammonioum citrate (FAC) also antagonized the E induced toxicity, by inducing notable increase in ROS level. Overall, we propose that E dramatically decreases ROS levels leading to a disruption of AML intracellular metabolism and rapid cell death.

Fig 1. Cytotoxic effect of Eltrombopag on AML cells is dependent on dose and serum concentration.

A) Dose response of primary AML cells and MOLM14 AML cell line (left panel) to E in culture medium containing 2% FBS and response to 5 μM E at the increasing concentration of serum in culture medium (2–10%) (right panel). Graphs represent cell counts at Day 4 for primary AML sample and Day 2 for MOLM 14 cells. Total number of viable cells in treated samples is presented as a % of total number of viable cells in control samples (untreated cells). Data are presented as an average of 3 independent experiments with a standard error of the mean (SEM). B) Time course of the response to 5 μM E in representative primary AML sample and MOLM14 cell line in 2% serum. C) Summary of the E’s effect on human primary AML cells n = 9 compared to peripheral blood mononuclear cells (PB-MNCs) from normal donors (ND) n = 3 (left panel) and AML cell lines n = 4 (right panel). Graphs demonstrate decrease in a total cell number after exposure to 5 μM E at 2% serum, presented as a percent of control at day 2 for cell lines and day 4 for primary cells.

Fig 2. Activation of apoptotic pathway in AML cells by Eltrombopag.

A) Graph represents time-course response of MOLM14 cells to E presented as an average of total number of viable cells from 3 independent experiments and SEM. B) Representative blots of Western blot analysis of activation of the apoptotic pathway at early time points (2–6 hours). Top panel—activation of Caspase 9, bottom panel—activation of Caspase 8. β- actin was used as a loading control. Positive CTR – cells incubated with etoposide 25 μM for 5 h. C) Western blot analysis of the activation of PARP protein. D) Representative flow cytometry plots of Annexin V and Propidium iodide staining of MOLM14 cells treated without or with E for 12 h. E) Western blot analysis of cleaved PARP and cleaved caspase 9 expression in additional AML cell lines MV4-11 (left panel) and KG1a (middle panel) and in primary AML cells (right panel). β-actin was used as a loading control. CTR—untreated control cells, E—cells treated with 5 μM E in the presence of 2% serum.

Fig 3. Eltrombopag completely inhibits S-phase DNA replication as measures by BrdU/7AAD staining in AML cells lines.

Cell cycle analysis of A) MOLM14 B) KG1a, and C) MV4-11 AML cells treated without or with E assessed by BrdU/7AAD staining. Gates R1-R4 represent different cell cycle phases: R1—apoptotic cells, R2—G0/G1, R3—S phase and R4—G2+M. Control (CTR)—cells cultured in medium only, E- cells cultured in the presence of E (5 μM).

Fig 4. MOLM14 cells incubated with Eltrombopag for 6 hours show higher expression of stress response genes.

Quantitative real time PCR (Q-PCR) measurement of mRNA level expression of five stress-response related genes in untreated cells (CTR), cells treated with rhTPO (100 ng/mL) and E (5 μM) for 6h (top panels) and earlier time point 2–6 h (bottom panel). Expression level of each gene is normalized to β-actin expression and presented as a percent of expression in untreated cells.

Fig 5. Eltrombopag Does Not Significantly Affect Mitochondria Function and ATP Production.

A) Graph shows results of TMRE staining as a measurement of the mitochondria membrane potential in MOLM14 cells after exposure to E at various time points. CCCP, a known mitochondrial membrane potential disrupter was used as a control for the assay. Bars represent percent of control (untreated cells) mean fluorescence intensity (MFI) with SEM. B) Effect of E on mitochondrial membrane potential in AML cell lines (left panel) and primary AML cells (right panel) measured by a TMRE staining. C) Measurements of oxygen consumption rate in control and cells treated with E (5 μM) for 24 hours. The oxygen consumption rates are expressed as nanomoles O₂ consumed/ min/ 10⁶ cells. The differences in maximal and spare respiratory capacity (SRC) between control and E treated cells are statistically significant with p value from 3 independent experiments equal 0.04 and 0.01 respectively. D) Graph represents an ATP production in MOLM14 cells cultured with or without (CTR) E for 20 hours, then exposed either to IAA, AA+O or remained untreated. Dark gray bars (untreated) represent total ATP, black bars (IAA) represent ATP produced by mitochondria and light gray bars (AA+O) represent ATP produced during glycolysis.

Fig 6. Eltrombopag markedly decreases levels of ROS in MOLM14 cells.

Flow cytometry analysis of ROS in MOLM14 cells cultured with or without E. A) Representative flow charts show gating strategy (left panel) based on size (Forward Scatter) and granularity (Side Scatter) of the cells, and (central and right panel) Carboxy-H₂DCFDA signal. B) Graphic presentation of ROS levels measured by flow cytometry in cells treated with E (5 μM) or DPI (25 μM), an inhibitor of NADPH oxidase known to decrease an intracellular level of ROS, or untreated (CTR) cells. Results are presented as an average of percent of control (untreated cells) mean fluorescence intensity with SEM. Data represent 4 independent experiments and the differences are statistically significant (*) p = 0.0000018. C) Relative level of ROS in AML cell lines: MOLM14, HL-60, MV4-11 and KG1a (left panel) and primary AML samples (right panel) after exposure to E. Differences between CTR and E treated cells are statistically significant (*) p = 0.0026 for cell lines and p = 0.001 for primary cells. D) Levels of O₂ ^- in MOLM14 cells exposed to E measures at various time points using Dihydroethidium (DHE) stain. Control represents untreated cells, E—cells exposed to E (5 μM), DPI—cells exposed to DPI (25 μM). Data are presented as an average percent of control calculated based on mean fluorescence from 3 independent experiments and standard error of the mean.

Fig 7. Amplex Red measurement of H₂O₂ released from MOLM14 cells correlates with the decrease in intracellular decrease of H₂O₂ measured using carboxy- H₂DCFDA dye.

A) H₂O₂ released from MOLM14 calculated based on the standard curve and presented in μM concentration. CTR—untreated cells, E—cells treated with 5 μM E, neg CTR—culture medium without cells, neg CTR+E—culture medium without cells + 5 μM E. B) Effect of E on H₂O₂ in the presence of hydrogen peroxidase and in the absence of cells. * p value < 0.05.

Fig 8. Correlation between eltrombopag’s cytotoxic effect and ROS level at various drug doses and serum concentrations.

A) Additive effect of E combined with antioxidant (NAC) on ROS decrease and survival of MOLM14 cells. Left panel—levels of H₂O₂ measured using carboxy- H₂DCFDA in MOLM14 cells exposed to E and NAC for 1 hour. Right panel—Proliferation of MOLM14 cells exposed to E (2–5 μM) alone NAC (15 mM) alone or combination. Results are presented as an average of total number of cells counted at days: 1 and 2 and SEM from 2 independent experiments. B) Increasing serum concentration decreases effect of E on ROS level (top panels) and cell survival (bottom panels) in AML cell lines. Results are presented as an average of percent of control of total number of cells counted at day 2 and SEM from 2 independent experiments.

Fig 9. Effect of E is partially diminished by addition of H₂O₂ or SOD mimetic to augment ROS level.

A) Proliferation of MOLM14 cells in the presence of E with and with or without addition of 4 μM H₂O₂. Top panel includes untreated cells (CTR) and cells treated with 4 μM H₂O₂ only, bottom panel is a magnification of bars representing cells treated with E alone and E+ H₂O₂. B) Proliferation of MOLM14 cells in the presence of E with or without SOD mimetic tempol (4 mM). Top panel includes untreated cells (CTR) and cells treated with 4 μM H₂O₂ only, bottom panel is a magnification of bars representing cells treated with E alone and E+Tempol. C) ROS levels in MOLM 14 cells cultured in the presence of E alone, H₂O₂ alone and E+ H₂O₂ (left panel) and E alone, tempol alone and E+ tempol (right panel) for 1h. ROS level is presented as a mean fluorescence intensity (MFI).

Fig 10. Modification of intracellular ROS level abrogates eltrombopag’s effect on AML cells.

A) E blocks completely increase of ROS caused by BSO and rescues cells from BSO induced death. Left panel—ROS level in MOLM14 cells untreated (CTR) or pre-treated with BSO for 72 hours and then exposed to E or left untreated. Right panel—proliferation assay—results are presented as an average and SEM of a total number of cells from 3 independent experiments. At day 0 cells cultured in the presence of BSO (50 or 100 μM) for 72hours or in plain medium (CTR) were plated at concentration 5x10^4 per 250 μL of culture medium and exposed to E or left untreated. B) Pre-loading with ferric ammonium citrate (FAC) rescues MOLM14 cells from E cytotoxic effect by inducing ROS level. Levels of H₂O₂ measured using carboxy- H₂DCFDA in MOLM14 cells treated as described above (left panel). Proliferation of MOLM14 cells un-manipulated or pre-loaded with 500 μg/mL of FAC for 24 h and then exposed to E or left untreated (CTR) (right panel).