Antileukemic activity and cellular effects of the antimalarial agent artesunate in acute myeloid leukemia

Leukemia Research 59 (2017) 124–135

Contents lists available at ScienceDirect

Leukemia Research journal homepage: www.elsevier.com/locate/leukres

Research paper

Antileukemic activity and cellular effects of the antimalarial agent artesunate in acute myeloid leukemia

MARK

Bijender Kumara,b,c, Arjun Kalvalaa,b,c, Su Chua,b,c, Steven Rosena,c, Stephen J. Formana,c, ⁎ Guido Marcuccia,b,c, Ching-Cheng Chena,b,c, Vinod Pullarkata,c, a b c

Department of Hematology and Hematopoietic Cell Transplantation, City of Hope National Medical Center, Duarte, CA 91010, USA Division of Hematopoietic Stem Cell and Leukemia Research, Beckman Research Institute, Duarte, CA, USA Gehr Family Center for Leukemia Research City of Hope Medical Center, Duarte, CA 91010, USA

A R T I C L E I N F O

A B S T R A C T

Keywords: Artimisinin Artesunate Acute myeloid leukemia BCL-2 Reactive oxygen species Iron Chemotherapy ABT-199 Venetoclax

The artimisinins are a class of antimalarial compounds whose antiparasitic activity is mediated by induction of reactive oxygen species (ROS). Herein, we report that among the artimisinins, artesunate (ARTS), an orally bioavailable compound has the most potent antileukemic activity in AML models and primary patients’ blasts. ARTS was most cytotoxic to the FLT3-ITD+ AML MV4-11 and MOLM-13 cells (IC50 values of 1.1 and 0.82 μM respectively), inhibited colony formation in primary AML and MDS cells and augmented cytotoxicity of chemotherapeutics. ARTS lowered cellular BCL-2 level via ROS induction and increased the cytotoxicity of the BCL2 inhibitor venetoclax (ABT-199). ARTS treatment led to cellular and mitochondrial ROS accumulation, double stranded DNA damage, loss of mitochondrial membrane potential and induction of the intrinsic mitochondrial apoptotic cascade in AML cell lines. The antileukemic activity of ARTS was further confirmed in MV4-11 and FLT3-ITD+ primary AML cell xenografts as well as MLL-AF9 syngeneic murine AML model where ARTS treatment resulted in significant survival prolongation of treated mice compared to control. Our results demonstrate the potent preclinical antileukemic activity of ARTS as well as its potential for a rapid transition to a clinical trial either alone or in combination with conventional chemotherapy or BCL-2 inhibitor, for treatment of AML.

1. Introduction The artemisinins are a family of antimalarial compounds derived from the sweet wormwood Artmisia Annua. Among the artemisinins, artesunate (ARTS) appears to be the most active agent with regard to the antimalarial activity and has optimal water solubility and oral bioavailability [1,2]. The antimalarial activity of artemisinins has been attributed to the generation of reactive oxygen species (ROS) occurring via cleavage of the endoperoxide bond in their structure [3]. Given a relatively high iron content of the cancer cells [4], the iron-catalyzed lysosomal ROS generation appears to be one of the main pro-apoptotic mechanisms mediating the artemisinins’ cytotoxicity [5]. Although artemisinins have shown strong cytotoxic activity against a variety of cancer cell lines in vitro, their mechanisms of action remain to be fully dissected [5–9]. Although up to 80% of younger patients with acute myeloid leukemia (AML) achieve complete remission with conventional chemotherapy, their 5-year survival is only around 40% [10]. Outcome is particularly poor for older patients as well as those with adverse

⁎

cytogenetics and molecular abnormalities like FLT3 internal tandem duplication (FLT3-ITD) [10]. Treatment refractoriness and disease relapse are thought to be due to persistence of quiescent leukemia stem cells (LSC) which remain resistant to conventional chemotherapy [11]. Such functionally defined LSCs have been shown to have low levels of ROS as well as overexpression of BCL-2 [12]. Maintenance of low ROS levels appears to be critical to survival and persistence of LSC. Therefore, inducing ROS generation has the potential to selectively target LSC given the sensitivity of the latter to ROS levels, thereby providing the rationale for testing artimisinins in AML. We therefore examined the antineoplastic activity as well as mechanism of activity of artimisinins against AML cell lines and primary cells both in vitro and in animal models. In this paper, we demonstrate that among the artimisinins tested, the compound ARTS has the most potent antileukemic activity that is synergistic with chemotherapeutic agents. We further show that ARTS treatment also led to decrease in BCL-2 expression and strong synergy with the selective BCL-2 inhibitor venetoclax (ABT-199).

Corresponding author at: Department of Hematology and Hematopoietic Cell Transplantation, City of Hope National Medical Center, 1500 East Duarte Road, Duarte, CA 91010, USA. E-mail address: [email protected] (V. Pullarkat).

http://dx.doi.org/10.1016/j.leukres.2017.05.007 Received 8 November 2016; Received in revised form 4 May 2017; Accepted 8 May 2017 Available online 10 May 2017 0145-2126/ © 2017 Elsevier Ltd. All rights reserved.

Leukemia Research 59 (2017) 124–135

B. Kumar et al.

2. Materials and methods

2.5. Measurement of ROS

2.1. Patient samples

KG1a, MOLM-13 and MV4-11 cells were treated with drugs of interest in IMDM containing 10% Fetal Bovine Serum for 24 h. Following collection, cells were incubated with 5 μM Carboxy-H2DCFDA and 3 μM MitoSox Red at 37 °C for 30 min to stain for cellular and mitochondrial ROS, respectively. Cells were then labeled with Annexin V and analyzed by flow cytometry. Cellular ROS and mitochondrial ROS levels in CD34+ cells were analyzed with BD FlowJo (Ashland,OR) software version 9.6.1.

Bone marrow (BM) samples from health donors, myelodysplastic syndrome (MDS) patients and AML patients were obtained under a specimen banking protocol approved by the Institutional Review Board of City of Hope Medical Center, in accordance with assurances filed with and approved by the Department of Health and Human Services, and meeting all requirements of the Declaration of Helsinki. Mononuclear cells were isolated using Ficoll (Stem Cell Technologies Inc., Vancouver, BC) density gradient separation as previously reported. CD34+ cells were isolated with 95% purity using immunomagnetic column separation per manufacturer's instructions (Miltenyi Biotec, Auburn, CA).

2.6. Flow cytometry for BCL-2 expression MOLM-13 and MV4-11 cells were plated at 2 × 105 cells/mL density and treated with drugs of interest in IMDM containing 10% FBS for 24 h.Cell were washed with PBS and resuspended in 2% paraformaldehyde for 1 h at −4 °C followed by cell washing and permeablized with 70% ethanol at −20 °C overnight. Intracellular BCL2 (BioLegend) staining was performed and analyzed using LSRII flow cytometer.

2.2. Reagents Artemisinin (ARTM), Artesunate (ARTS), dihydroartimisinin (DHA), Deferoxamine, daunorubicin and Cytosine arabinoside were purchased from Sigma (St. Louis, MO). ABT-199 (venetoclax) was purchased from Selleckchem (Houston, TX). Deferasirox was provided by Novartis (Basel, Switzerland). Growth factors for colony forming assay including erythropoietin (EPO), stem cell factor (SCF), granulocyte-macrophage colony stimulating factor (GM-CSF), granulocyte colony stimulating factor (G-CSF) and interleukin-3 (IL-3) were obtained from Peprotech (Rocky Hill, NJ). Carboxyfluorescein succinimidyl ester (CFSE), MitoProbe JC-1 assay kit, Carboxy-H2DCFDA and MitoSox Red used for cellular ROS and mitochondrial ROS detection respectively were obtained from Invitrogen (Carlsbad, CA). Antibodies used for flow cytometry were obtained from BioLegend (San Diego, CA).

2.7. RT-PCR for BCL-2 The RNA expression levels of the BCL2 were analyzed by SYBR green real time PCR. The RNA isolated from control and treated cells were reverse transcribed to cDNA using Superscript ™ II Reverse Transcriptase (Invitrogen). The primers for real time PCR amplification and conditions have been previously described [13]. The primer sequences for human BCL2 were: forward 5′-TTGTGGCCTTCTTTGAGTTCGGTG-3′ and reverse 5′-GGTGCCGGTTCAGGTACTCAGTCA-3′. The SYBR green real time Quantitect PCR kit was purchased from Qiagen (Valencia, CA, USA). GAPDH was used as control in separate reactions.

2.3. Cell culture 2.8. Western blot CD34+ cells from healthy donors or AML patients were cultured in StemSpan serum-free medium (Stem Cell Technologies, Vancouver, BC) supplemented with low concentrations of growth factors (granulocytemacrophage colony stimulating factor (GM-CSF)200 pg/mL, granulocyte colony-stimulating factor(G-CSF) 1 ng/mL, stem cell factor (SCF) 200 pg/mL, leukemia inhibitory factor (LIF)50 pg/mL, macrophage inflammatory protein-1α (MIP-1α) 200 pg/mL, and interleukin-6 (IL-6) 1 ng/mL) at 37 °C with 5% CO2 and high humidity. KG1a, OCI-AML3, MOLM-13 and MV4-11 cells were cultured in IMDM supplemented with 10% FBS and 1% penicillin/streptomycin. CD34+ cells from MDS patients were cultured in IMDM with 30% FBS and GFs [EPO (3 u/mL); SCF (5 ng/mL); GM-CSF (20 ng/mL); G-CSF (20 ng/mL) and IL-3 (5 ng/ mL)] at 37 °C in 5% CO2.

1 × 106 cells were cultured in IMDM containing 10% FBS with or without relevant drugs for 24 h. The cells were washed with PBS and lysed by resuspending in 100 μl RIPA buffer (ThermoFisher Scientific, Waltham, MA) supplemented with protease and phosphatase inhibitor cocktail. Proteins were resolved on 4–12% sodium dodecyl sulfatepolyacrylamide gel electrophoresis (SDS-PAGE) gels and transferred to nitrocellulose membranes. Membranes were sequentially probed with γ-H2AX, cleaved caspase 3, uncleaved Caspase 3, c-Myc, c-Src (Tyr416), pJNK (Tyr183/Tyr185), BCL-2, β-Actin and GAPDH antibodies and horseradish peroxidase-conjugated secondary antibody (Cell Signaling Technology Inc, Danvers, MA). Detection was performed using the SuperSignal West Dura extended duration substrate kit (Thermo Fisher Scientific).

2.4. Proliferation and Apoptosis assays 2.9. Mitochondrial membrane potential (MMP) measurement Cell proliferation was measured by CellTiter-Glo® Luminescent Cell Viability Assay as per manufacturer’s instructions (Promega, Madison, WI). Briefly, human AML cell lines were seeded in 96 well plates at density of 2 × 105 cells/mL. Cells were cultured under different treatment conditions for 48 h. The Cell Titer-Glo substrate and buffer reagents were added into each well and mixed to obtain cell lysis. Plates were read on a Microplate reader (Beckman Coulter DTX880, Brea,CA). Data from three replicates were expressed as percentage of treated cells with respect to untreated controls. To measure apoptosis, cells treated with different drugs for 48–72 h were labeled with Annexin V (BioLegend) and Propidium Iodide (Invitrogen). Fluorescence was measured by flow cytometry in order to determine percentage of apoptotic cells. For proliferation assay, 2 × 105 cells/mL cells were labeled with 5 μM of CFSE and cultured for 24 h under different treatment conditions and analyzed using LSRII flow cytometer (BD Biosciences, San Jose, CA).

The effect of drug treatment on MMP was analyzed using MitoProbe JC-1 Assay kit (Molecular Probes, Eugene, OR). The collapse in the electrochemical gradient across the mitochondrial membrane was measured using a fluorescent cationic dye 5,5′,6,6,-tetrachloro1,1′,3,3′-tetraethyl-benzamidazolo-carbocyanin iodide, also known as JC-1. This dye exhibits potential dependent accumulation in the mitochondrial matrix. 1 × 106 cells were incubated with 1 μM JC-1 for 15 min at room temperature and analyzed on LSRII flow cytometer. 2.10. Colony forming cell (CFC) assays Human CD34 + cells from health donors or AML or MDS patients were treated with drugs of interest for 48 h at 5 μM. Cells were washed and transferred into colony formation culture containing IMDM, 30% FBS, GFs [EPO (3 u/mL); SCF (5 ng/mL); GM-CSF (20 ng/mL); G-CSF 125

Leukemia Research 59 (2017) 124–135

B. Kumar et al.

(20 ng/mL) and IL-3 (5 ng/mL)] and 1.3% methylcellulose (StemCell Technologies, Vancouver, Canada). Burst-forming units-erythroid (BFU-E) and colony forming units-granulocyte and macrophage (CFUGM) were counted after 14 days under light microscope.

Table 1 IC50 values of ARTS for AML cell lines.

2.11. In vivo studies C57BL/Ka (B6), C57BL/Ka-CD45.1 (B6-45.1), NOD-scid IL2Rgammanull(NSG) and B6-Rag2-/-γc−/− (DKO) mice were maintained by the Animal Resource Center of City of Hope. NSG and MLLAF9 knock-in mice were obtained from Jackson laboratory (Sacramento, CA). Mouse care and experimental procedures were performed in accordance with federal guidelines and protocols approved by the Institutional Animal Care and Use Committee at the City of Hope. For in-vivo treatment studies in the MLL-AF9 knock in (KI) transplantation model, 1 × 104 viable CD45.2 MLL-AF9 KI leukemic GMP cells along with 2 × 105 CD45.1 helper bone marrow (BM) cells were transplanted into 8–10 weeks old sublethally irradiated (7.0 Gy) recipient mice by retro orbital injection. Sublethal irradiation and helper BM were used to achieve a more uniform disease onset in recipient animals. Recipient mice were maintained on Sulfatrim diet. Treatment with either DMSO or ARTS (25 mg/kg) intraperitoneally (i.p.) was started 10 days after transplantation and mice were observed for survival endpoints. BM was analyzed for leukemic cell chimerism when mice were moribund. For the cell line xenograft models, 3–4 day-old B6-Rag-2-/-γc−/− mice were intrahepatically injected with 5 × 106 MV4-11 cells after irradiation (1.5 Gy). Mice were treated with twice weekly dose of vehicle (DMSO) or ARTS (25 mg/kg) i.p. starting after 10 days posttransplantation and their survival was observed. In order to test combination of ARTS with chemotherapy, sublethally irradiated (3.5 Gy) NSG mice were injected with 5 × 106 MV4-11cells and treated with ARTS (50 mg/kg) alone or in combination with Ara-C (50 mg/kg) for 4 days starting at day 6 after injection. For the primary AML model, sub lethally irradiated(3.0 Gy) NSG mice were injected with 1 × 107 patient derived FLT3-ITD+ AML cells by retro orbital injection. Recipient mice were maintained on Sulfatrim diet. Treatment with either vehicle (DMSO) or ARTS (50 mg/kg) i.p. was started 2 weeks after transplantation and mice were observed for survival.

Cell line

IC50 (μM)

MV4-11 MOLM-13 OCI/AML3 NB4 KG1A

1.1 0.82 5.1 4.9 7.2

values of 0.82 and 1.1 μM respectively (Table 1, Supplementary Table 1). Hence MV4-11 and MOLM-13 cell lines were selected for further cytotoxicity experiments. For some experiments, less sensitive cells such as KG-1a and OCI-AML3 were used in order to obtain adequate numbers of viable cells for functional analysis. Among the three artimisinins tested, ARTS also showed the highest cytotoxicity (Fig. 1A and B). After 72 h of exposure to ARTS at 2 μM concentration (∼ 2-fold IC50 value for these cell lines), the apoptosis rate of MV4-11 cells was 55.9 ± 1.5% compared to 15.5 ± 1.5% (P = 0.0003) and 14.75 ± 1.5 (P = 0.0002) for ARTM and DHA respectively. The apoptosis rate for MOLM-13 cells was 81.5 ± 2.5% for ARTS compared to 16.3 ± 1.8% (P = 0.003) for ARTM and 25.6 ± 1.5, (P = 0.0063) for DHA (Fig. 1A and B). Hence, we selected ARTS for further evaluation. We next examined the potential synergy of ARTS with the chemotherapeutic agents commonly utilized for treatment of AML. Anthracyclines and the nucleoside analog cytosine arabinoside (cytarabine, Ara-C) represent the backbone of induction remission regimens for AML patients. The MV4-11 cells were treated for 48 h with 2 μM ARTS alone or in combination with 100 nM of doxorubicin or cytarabine, following which apoptosis was analyzed. The combination of ARTS with doxorubicin or cytarabine induced apoptosis rates of 88.6 ± 2% and 84 ± 5.5% respectively compared to 39.8 ± 1.6% with daunorubicin (P = 0.0005) or 49.5 ± 4.5% with cytarabine alone (P = 0.017) (Fig. 1C). Similar experiment was performed in using the FLT3-ITD+ AML cell line MOLM-13. Since IC50 values for ARTS are lower in MOLM-13 cells compared to MV4-11, we used lesser amount of ARTS for synergy experiments. The MOLM-13 cells were treated for 48 h with 500 nM ARTS alone or in combination with 100 nM of either doxorubicin or cytarabine. The combination of ARTS with daunorubicin or cytarabine induced apoptosis in 47.3 ± 2% and 55 ± 1.3% of cells respectively compared to 36.8 ± 1.2% for doxorubicin and 43.6 ± 1.1% cytarabine alone (P-values 0.038 and 0.0136 respectively) (Fig. 1D). We observed ARTS to be cytotoxic to CD34+ primary human AML blasts as well. Data from 2 representative FLT3-ITD + patients is shown in Fig. 1E and F. Data from 2 FLT3-ITD negative patients is shown in Fig. 3G and H. ARTS significantly augmented the cytotoxicity of doxorubicin or cytarabine in primary CD34+ patient blasts as well. (Fig. 1E and F). These results demonstrate the potential usefulness of ARTS in combination with chemotherapy for AML.

2.12. Statistical analysis All statistical analyses were performed using GraphPad Prism software 6.0d. Data are presented as mean ± standard deviation (SD), and two group comparisons were done with a two-tailed Student’s t-test. Three group comparisons was performed using one way ANOVA. For Kaplan Meier survival curve analysis, the comparisons were done using Log rank test. A P-value of < 0.05 was considered statistically significant. Combination Index values were calculated using Compusyn software and a CI value less than 1 indicates synergy between drugs tested.

3.2. Cytotoxicity of ARTS in AML cells is mediated by induction of ROS 3. Results Artemisinins mediate their antimalarial activity by generation of ROS in the parasite facilitated by intraparasitic heme iron. These ROS in turn induces DNA damage and death in the malaria parasite (3). Since ROS induction is critical to activity of artimisinins we examined ROS level in AML cell lines after exposure to ARTS. Both cellular and mitochondrial ROS were induced by ARTS. Induction of cellular ROS after ARTS exposure was attenuated by pretreatment with the free radical scavenger free radical scavenger N-acetyl cysteine (NAC). (Figs. 2 A and B, 4 C and D). ARTS treatment augmented cellular ROS levels in primary AML cells as well (Fig. 2C and D). In order to assess the role of ROS in mediating cytotoxicity of ARTS,

3.1. ARTS exhibits antiproliferative activity and cytotoxicity against AML and augments cytotoxicity of anthracyclines and cytarabine We examined the antiproliferative activity and cytotoxicity of ARTM, ARTS and DHA against the AML cell lines, MV4-11, MOLM-13, OCI-AML3, NB4 and KG1A. These cells lines were selected as representative of relatively frequent cytogenetic/molecular subsets of AML. Among the artimisinins tested ARTS showed the most antiproliferative activity (Supplementary Fig. S1). The FLT3-ITD+ cell lines MOLM-13 and MV4-11 were the most sensitive to ARTS with IC50 126

Leukemia Research 59 (2017) 124–135

B. Kumar et al.

(caption on next page)

the anti leukemic activity of ARTS. Apoptosis in MV4-11 cells treated with 2 μM ARTS for 48 h was decreased from 48.5 ± 2.5% to 34 ± 1.1% by NAC treatment (P = 0.03) (Fig. 2E). Attenuation of

we examined the effects of NAC treatment on cytotoxicity of ARTS. NAC treatment resulted in attenuation of cytotoxicity of ARTS against MV4-11 cells supporting ROS generation as an important component of 127

Leukemia Research 59 (2017) 124–135

B. Kumar et al.

Fig. 1. Cytotoxicity and antiproliferative effect of artimisinin (ARTM), dihydroartimisinin (DHA) and artesunate (ARTS) in AML cell lines and combination of ARTS with chemotherapy (A, B,). MV4-11 and MOLM-13 cells were cultured with 1 μM, 2 μM and 5 μM ARTS, ARTM and DHA for 72 h. ARTS treated MV4-11 and MOLM-13 AML cells showed higher apoptosis at 1 μM (ANOVA P < 0.0001), 2 μM (P < 0.0001) and 5 μM (P < 0.0001) drug concentrations compared to ARTM and DHA treatment. (C) MV4-11 cells were cultured with 2 μM ARTS alone or with 100 nM doxorubicin, or cytosine arabinoside (Ara-C) for 48 h and analyzed for apoptosis. Combination of ARTS with doxorubicin or Ara-C was more cytotoxic than doxorubicin or Ara-C alone. (D) MOLM-13 cells were cultured with 500 nM ARTS alone or with 100 nM doxorubicin, or Ara-C for 48 h and analyzed for apoptosis. (E, F) ARTS showed cytotoxicity against primary AML cells. Combination with ARTS with daunorubicin or Ara-C significantly augmented cytotoxicity of either agent compared to either agent alone against primary AML blasts. In COH 2742 patient sample, the apoptosis rates were 45.8 ± 2.4% at 500 nM daunorubicin vs. 65.6 ± 0.9% for daunorubicin with 5 μM ARTS combination (P = 0.016) and 52.1 ± 0.8% at 500 nM cytarabine alone vs. 69.6 ± 1.7% for cytarabine with 5 μM ARTS combination (P = 0.004). Similarly, for COH 2782 patient sample, the apoptosis rates was 51.7 ± 2.1% for daunorubicin alone vs. 69.4 ± 1.4% daunorubicin with 5 μM ARTS combination (P = 0.001) while apoptosis rate was 53.3 ± 1.5% for cytarabine alone vs. 4.8 ± 1.25% for cytarabine with ARTS combination (P = 0.008) Data are from triplicate experiments. Mean ± SD values are shown.

For MV4-11 cells, the combination of ARTS with 5 nM ABT-199 induced apoptosis in 85.5 ± 0.8% of cells compared to 36.1 ± 0.8% for ABT-199 alone (P = 0.006). The combination of ARTS and 5 nM ABT199 induced apoptosis in 47.4 ± 2.1% of MOLM-13 cells compared to 21.5 ± 1.2% for 5 nM ABT-199 alone (P = 0.007) (Fig. 3E and F). The addition of ABT-199 to ARTS significantly augmented the cytotoxicity of ARTS in CD34+ primary patient blasts as well (Fig. 3G and H). These results show the potential to use ARTS in combination with ABT-199.

cytotoxicty of ARTS by ROS scavenging with NAC was observed in primary AML cells as well (Fig. 2F and G). Since intracellular generation of ROS by artemisinins is an iron catalyzed reaction, we also examined the effect of the iron chelators deferasirox (DFX) and deferoxamine (DFO) on cytotoxicity of artemisinins. Iron chelation was expected to attenuate artesunate’s cytotoxicity by lowering iron-mediated ROS production. Surprisingly, the cytotoxicity of ARTS was instead significantly augmented by DFX but not by DFO (Supplementary Fig. S2). DFX itself has been shown to have some antileukemic activity against AML and has been shown to induce ROS in AML cells [14]. Indeed, we observed augmented ROS induction by the combination of ARTS and DFX. (Supplementary Fig. S2) which likely explains the increased cytotoxicity observed with combination of ARTS and DFX. Consistent with these results, we also showed an increase of the phosphorylated histone H2AX (γ-H2AX), a sensitive marker for double stranded DNA damage induced by genotoxic agents in ARTS-treated MV4-11 and MOLM-13 AML cell lines by Western blot. C-Jun N-terminal kinase (JNK), a member of the mitogen-activated protein kinase family is activated by ROS and mediates apoptosis induced by excessive ROS (15). As expected, induction of signaling via the ROS-activated JNK pathway was also observed in ARTS-treated AML cell lines (Fig. 2H and I).

3.4. ARTS treatment leads to loss of MMP, triggers mitochondrial pathway of apoptosis and downregulates phosphorylation of Src family kinases Since loss of mitochondrial membrane potential (MMP, ΔYm) is an early event in the mitochondrial pathway of apoptosis, we examined the effect of ARTS treatment on MMP and on generation of mitochondrial ROS. ARTS treatment resulted in loss of MMP as measured by flow cytometry using the cationic dye JC-1 (Fig. 4A and B). (Relative drop in ΔΨm% ARTS vs Control P = 0.0067 for MV4-11 and 0.018 for KG1a respectively). ARTS also induced mitochondrial ROS in KG1a and MV4-11 cells (Fig. 4C and D) (ARTS vs Control P = 0.001 and 0.003 respectively) and triggered mitochondria-dependent apoptosis as evidenced by increased level of cleaved caspase-3 and decrease in uncleaved caspase-3 (Fig. 4E and F). Caspase-3 cleavage was further enhanced by the combination of ARTS and the anthracycline daunorubicin which supports the observed synergy between anthracyclines and ARTS (Fig. 4E and F). Artimisinins have been shown to have multiple effects on regulators of cell proliferation and survival. The Src family of kinases are overexpressed and constitutively activated in AML and have been shown to be important for AML cell survival. We observed decreased phosphorylation at Tyr416 of Src family of kinases thereby demonstrating that ARTS has multiple effects on mediators of cell proliferation and survival (Fig. 4E and F). Total Src kinase levels were unchanged (data not shown).

3.3. ARTS lowers BCL-2 expression and increases cytotoxicity of the BCL-2 inhibitor venetoclax (ABT-199) In order to investigate the effect of artimisinins on programmed cell death, we then examined effect of drug treatment on the antiapoptotic BCL-2 protein. BCL-2 not only functions as a prosurvival protein by suppressing the pro-apoptotic cascade, but also has been shown to have antioxidant properties. [16,17] An inverse correlation between the levels of BCL-2 and ROS has been reported and increased ROS generation reduces BCL-2 expression [18]. ARTS treatment resulted in lowering of BCL-2 level in MV4-11 and MOLM-13 cells as demonstrated by flow cytometry. For MOLM-13 cells, the relative mean fluorescence intensity (MFI) for BCL-2 was significantly reduced in ARTS treated cells compared to control (relative MFI 1 vs. 0.66 ± 0.03, P = 0.0019). Similarly, for MV4-11 cells, relative BCL-2 MFI values were 1 vs. 0.65 ± 0.017 for ARTS and control respectively, (P = 0.0013) (Fig. 3A and B). These results were further confirmed in MV4-11, MOLM-13, KG-1a and OCI-AML3 cell lines by Western blotting. Cellular BCL-2 levels decreased in these cell lines after 24 h of exposure to ARTS. Treatment with NAC attenuated the decrease in BCL-2 induced by ARTS alone suggesting that this effect is ROS-mediated (Fig. 3C). Significant decrease in BCL-2 RNA expression after exposure to ARTS for 24 h was seen in AML cells (Fig. 3D) showing that the lowering of BCL-2 after ARTS treatment is at least partly mediated by inhibition of BCL-2 transcription. This decrease in BCL-2 mRNA transcription by ARTS was attenuated by NAC treatment suggesting that this effect is ROS-mediated (Supplementary Fig. S3). Given the inhibition of BCL-2 observed in AML cell lines treated with ARTS, we then examined the cytotoxicity of the combination of ARTS with ABT-199 (venetoclax), a BH3 protein–protein interaction inhibitor that has already shown promising results in clinical studies of AML patients [19]. MV4-11 and MOLM-13 cell lines were treated with 1 μM ARTS in combination with 5–15 nM concentrations of ABT-199.

3.5. ARTS treatment inhibited colony formation by AML and MDS cells In order to determine the activity of artimisinins against clonigenic primary AML and MDS cells, we examined the effect of treatment with ARTM, ARTS and DHA on colony forming cells (CFC) in bone marrow CD34+ cells from AML and MDS patients as well as normal controls (n = 3) (Fig. 5). ARTS was the most effective of the artimisinins in suppressing colony formation. ARTS treatment for 48 h prior to plating in methylcellulose resulted in significant inhibition CFU-GM from AML and MDS patients (AML CFU-GM numbers for ARTS vs Untreated 21.3 ± 4.1 vs 39.2 ± 11.3 P = 0.04 and MDS CFU-GM numbers for ARTS vs Untreated 41 ± 6.2 vs 93.5 ± 23.5 P = 0.005 respectively). CFC from normal bone marrow progenitors were also significantly inhibited. This was mainly due to suppression of CFU-GM and no effect on BFU-E colony numbers was observed (Fig. 5). These data show the activity of ARTS against clonigenic AML and MDS progenitors. 3.6. ARTS shows potent antileukemic activity in murine models of AML As ARTS also showed marked in vitro cytotoxicity towards leukemia cells from the MLL-AF9 knock in mouse [(Supplementary Fig. S4) 128

Leukemia Research 59 (2017) 124–135

B. Kumar et al.

Fig. 2. Induction of ROS mediates cytotoxicity of artimisinins against AML. (A,B) Treatment of AML cell lines KG1a and MV4-11 with artesunate (ARTS) resulted in induction of cellular ROS as detected by flow cytometry. Pretreatment with NAC attenuated ROS induction by ARTS. Induction of cellular ROS by ARTS was observed in primary AML cells as well (Fig. 2C and D)(E) Treatment with the free radical scavenger N-acetyl cysteine (NAC) attenuated the cytotoxicity of ARTS in MV4-11 cells. (F,G) Cytotoxicity of ARTS was attenuated by NAC treatment in primary AML cells Data are from triplicate experiments. Mean ± SD values are shown. (H,I) Induction of double-stranded DNA damage and JNK pathway activation in MOLM-13 and MV4-11 cell lines by artemisinins. The effect of ARTS, ARTM and DHA treatment on DNA damage was determined by Western blot analysis for the phosphorylated histone H2AX (γ-H2AX). ARTS treatment showed marked induction of γ-H2AX. Artesunate treatment also induced signaling via the ROS-activated JNK pathway in MV4-11 and MOLM-13 cells as detected by Western blotting using the pJNK (Tyr183/Tyr185) antibody. Data shown are representative of repeated experiments.

129

Leukemia Research 59 (2017) 124–135

B. Kumar et al.

Fig. 3. ARTS lower BCL-2 expression and augments the cytotoxicity of the BCL-2 inhibitor venetoclax (ABT-199) (A,B) MV4-11 and MOLM-13 cells were treated with 2 μM artemisinin (ARTM), dihydroartimisinin (DHA) and artesunate (ARTS) for 24 h and analyzed by flow cytometry. MFI for BCL-2 was significantly lower in ARTS treated cells compared to control. Mean ± SD of MFI values relative to control of 3 independent experiments is shown. (C) Decrease in BCL-2 protein expression by ARTS exposure for 24 h was confirmed by Western blot analysis in 4 AML cell lines. ARTS treatment lowered BCL-2 protein level at 24 h and this effect was attenuated by treatment with NAC suggesting that the lowering of BCL-2 is mediated by induction of ROS by ARTS. (D) Real-time RT-PCR showed decreased BCL-2 transcription after exposure to ARTS for 24 h (E,F) ARTS augmented apoptosis induced by the BCL-2 inhibitor ABT-199 in MV4-11 and MOLM-13 cell lines. (G, H) Cytotoxicity of ARTS (5 μM) in primary patient blasts is significantly augmented by combination with 10–50 nM of ABT-199. The apoptosis observed in COH 2516 at 10 nM ABT-199 alone was 34.6 ± 2.6% which increased at 44.1 ± 1.5% with combination of 10 nM ABT-199 and 5 μM ARTS. (P = 0.03). Similarly, the apoptosis rates in COH 2573 at 10 nM ABT-199 concentration was 26.8 ± 1.4% which increased to 37.8 ± 1.4% with 10 nM ABT-199 and 5 μM ARTS combination (P = 0.018 and P = 0.002 respectively). Data are from triplicate experiments. Mean ± SD values are shown.

27.7 ± 2.3% vs. 96.5 ± 1.5%(P = 0.0002)], we utilized this model for in vivo experiments (Fig. 6). First, wild type B6 mice were injected with MLL-AF9 knock-in cells along with “helper” whole bone marrow

cells. ARTS treatment (25 mg/kg) administered intraperitoneally on alternate days starting at day 10 after AML cell injection resulted in significant reduction in leukemic cells engraftment in the recipient mice 130

Leukemia Research 59 (2017) 124–135

B. Kumar et al.

Fig. 3. (continued)

alternate days from days 12–18 after AML cell injection resulted in significant survival prolongation (median survival 20 vs. 23days, P = 0. 0.001) (Fig. 6B). Since ARTS had shown highest activity against FLT3-ITD expressing MV4-11 and MOLM-13 cell lines, we then tested the antileukemia activity of this compound in FLT3-ITD human AML xenograft. Sublethally irradiated NSG mice were injected intraperitoneally with patient

(90.2 ± 3.6% vs. 68.5 ± 7.1%, P = 0.022), reduced splenomegaly (median weight 362.5 ± 18.9 vs. 267.7 ± 21.9 mg, P = 0.021) and significant prolongation of survival duration (median survival 26 vs. 30 days, P = 0.0001 Log rank test) (Fig. 6A). In a second experiment, we utilized MV4-11 cells engrafted in 3 day-old Rag–2 gamma chain knockout (DKO) mice followed by treatment with ARTS 25 mg/kg. Again, ARTS treatment (25 mg/kg) on 131

Leukemia Research 59 (2017) 124–135

B. Kumar et al.

Fig. 4. Effects of Artemisinin treatment on MMP, mitochondrial ROS and mitochondrial apoptosis pathway in AML cell lines. (A, B) Artimisinin treatment of MV4-11 and KG-1a cell lines with artimisinins for 24 h shows loss of MMP (ΔΨm) which was most pronounced with ARTS. (C, D) Flow cytometry for mitochondrial ROS in MV4-11 and KG1a cells detected using Mitosox Red. Among artimisinins tested, ARTS was most potent in inducing mitochondrial ROS. Data are from triplicate experiments. Mean ± SD values are shown. (E,F) MV4-11 and MOLM-13 cells treated with 2 μM ARTS alone or in combination with 100 nM daunorubicin for 24 h showed reduced pSrc (Tyr 416) tyrosine phosphorylation, uncleaved caspase-3 as well as increased cleaved caspase-3.

kg) and Ara-C (50 mg/kg) resulted in prolongation of survival of mice receiving the combination compared to treatment with ARTS alone (median survival 45 vs. 38 days. P = 0.0023) (Fig. 6D). Thus, utilizing multiple murine models, we consistently showed a

derived FLT3-ITD+ AML cells. Treatment with ARTS 50 mg/kg resulted in significant survival prolongation (median survival 35 vs.39 days, P = 0.012) (Fig. 6C). In a model of sublethally irradiated NSG mice injected with MV4-11 cells, combination treatment with ARTS (50 mg/ 132

Leukemia Research 59 (2017) 124–135

B. Kumar et al.

Fig. 5. Effect of treatment with artimisinins on colony forming cells (CFC) derived from normal donors, patients with AML and MDS. CD34+ cells obtained by immunomagnetic separation from bone marrow of normal donors, AML and MDS patients (n = 3 for each group) were treated with 5 μM ARTM, ARTS and DHA and plated on semisolid methylcellulose based media. Compared to untreated cells (UT), ARTS treated AML and MDS progenitors showed significantly reduced total number of CFU-GM colony forming cells. ARTS treatment in normal CD34+ progenitors (NL) had reduced CFU-GM suggesting myelosuppressive properties of the drug. (* denotes P-value < 0.05, **denotes P-value < 0.005). Mean ± SD values are shown.

ROS, loss of MMP and caspase-3 activation consistent with induction of mitochondrial pathway of apoptosis triggered by intracellular ROS are consistent with studies of ARTS in other cancers [8]. Iron-catalyzed ROS production could provide selectivity to malignant cells given their higher expression of transferrin receptors and higher iron content compared to normal cells4 and may explain the low toxicity of ARTS to normal tissues. The robust ROS induction observed with ARTS in AML could provide an opportunity to target the quiescent LSC population. LSC are low in ROS and induction of ROS could theoretically induce their differentiation and enhance sensitivity to chemotherapy [12]. Particularly interesting in the context of ARTS treatment is the interplay between ROS and BCL-2. Induction of ROS has been shown to have proapoptotic effects on BCL-2 family proteins including suppression of BCL-2 by regulation of its phosphorylation and ubiquitination [20]. Furthermore, ROS-induced activation of JNK can lead to mitochondrial translocation of JNK and inhibition of BCL-2 function [15]. Consistent with this data, we show that ARTS treatment decreases cellular BCL-2 levels and synergizes with the BCL-2 inhibitor ABT-199. Aberrantly high BCL-2 expression has been observed in functionally characterized LSC that are low in ROS and BCL-2 inhibition has been shown to induce cell death in LSC [12]. The ability of ARTS to induce ROS and thereby lower BCL-2 likely explains the strong synergy between the two drugs that we observed in AML. Early results of a trial of ABT-199 (venetoclax) in combination with hypomethylating agents has shown impressive results in AML [21]. Our data showing suppression of BCL-2 levels by ARTS as well as marked synergy between ABT-199 and ARTS provides a basis for testing this combination in clinical trial. Given the dependence of LSC on low ROS and high BCL-2 levels, this could provide an effective strategy to specifically target LSC. The effect of ARTS on a variety of cellular pathways have been postulated as mediators its antineoplastic activity [1,7]. We observed

significant survival prolongation after in vivo treatment with relatively low doses of ARTS. The sensitivity of FLT3-ITD+ AML cells to ARTS is particularly interesting, given the relatively poor prognosis of these patients treated with conventional chemotherapy regimens.

4. Discussion Herein, we first report potent antileukemic activity of ARTS in vitro as well as in three different murine models including a xenograft model using primary AML cells. We also observed in vitro activity against primary MDS cells. We demonstrate that ARTS antileukemia activity depended on generation of ROS and suppression of BCL-2 as also supported by its synergistic activity with the emerging BCL-2 inhibitor ABT-199 (venetoclax). Also interesting to us was the marked synergy between ARTS and conventional chemotherapy used in AML. These data therefore support testing the combination of ARTS with chemotherapy and/or with BCL-2 inhibitors. Induction of ROS is critical to the antimalarial activity of ARTS. Similar to previous studies examining antineoplastic activity of ARTS, we show that its cytotoxicity against AML is also mediated by induction of ROS. ARTS has shown antineoplastic activity against various cancer cell lines [5–9]. In a recent study using AML cell lines, ARTS and its dimer ART-838 inhibited a variety of cell lines with IC50 values that were achievable in vivo in mice by oral therapy [6]. We have further characterized the activity of ARTS in AML including its activity in in vivo models, synergy with AML therapies and cellular effects that mediate its antileukemic activity, particularly its effect on BCL-2. ARTSinduced mitochondrial apoptosis was shown to be mediated by ironcatalyzed ROS production in the lysosomal compartment in breast cancer cells [5]. The lysosomal accumulation of ARTS and degradation of lysosomal ferritin is implicated in its ability to induce mitochondrial ROS [9]. Our data showing elevation of cellular and mitochondrial 133

Leukemia Research 59 (2017) 124–135

B. Kumar et al.

(caption on next page)

their inhibition may allow targeting of AML stem cells [23]. Therefore, ARTS is likely to mediate its activity by has multiple cellular effects that are detrimental to AML cell survival including mechanisms that could

down regulation of Src kinase activvation after ARTS treatment. Src kinases mediate STAT phosphorylation and is important in AML cell survival [22]. Src kinases are also overexpressed in AML stem cells and 134

Leukemia Research 59 (2017) 124–135

B. Kumar et al.

Fig. 6. ARTS has potent in vivo activity in multiple murine models of AML. (A) ARTS treatment in-vivo delays MLL-AF9 KI cell induced leukemogenesis in C57/B6 mice. (i) Schematic diagram showing transplantation of 1 × 104 MLL-AF9 KI BM cells along with 2 × 105 helper BM cells (day 0) followed by control (DMSO) or ARTS (25 mg/kg) i.p injections on days 10,12,14,16 and 18 (ii) ARTS treatment significantly prolonged survival of treated mice (P = 0.0001, n = 8 in each group). (iii–v) ARTS treated mice showed reduced leukemic engraftment (P = 0.0217) and reduced spleen weight (P = 0.0210) and size. (B) ARTS treatment in vivo delays MV4-11 induced leukemogenesis in Rag2γc −/− mice (n = 10 in each group) (i) Schematic diagram showing intrahepatic transplantation of 5 × 106 MV4-11 cells followed by Control (DMSO) or ARTS (25 mg/kg) i.p. injections on days12,14,16 and 18. (ii) ARTS treatment significantly prolonged survival showing that ARTS has antileukemic activity in vivo (P = 0.0097). (C) ARTS treatment in-vivo delays FLT3-ITD + AML patient-derived leukemia development in NSG mice. 10 × 106 leukemic blasts from a FLT3-ITD + AML patient were transplanted by retro-orbital injection in sub-lethally radiated (3 Gy) NSG mice followed by treatment with ARTS (50 mg/kg) intraperitoneally (n = 4-5 mice). ARTS treatment significantly delayed leukemia development and prolonged survival (P = 0.0120). (D) Injection of sub lethally irradiated NSG mice with MV4-11 cells followd by combination treatment with ARTS (50 mg/kg) and Ara-C (50 mg/kg) resulted in prolongation of survival of mice receiving the combination compared to treatment with ARTS alone (median survival 45 vs. 38 days. P = 0.0023).

particularly target LSC. Despite the multitude of cellular effects in neoplastic cells, ARTS appears surprising nontoxic to normal tissues. Its toxicity profile is well studied in the setting of antimalarial therapy. In a study of intravenous ARTS for treatment of severe malaria, the drug was extremely well tolerated, even in critically ill patients including those with hepatic dysfunction. All deaths and most of the adverse events were attributed to the malaria itself [24]. In a trial of ARTS in combination with breast cancer therapy, doses up to 200 mg/kg of ARTS was also well tolerated [25]. However, our data showing suppression of CFU-GM from normal bone marrow suggest that some degree of myelosuppression can be expected. However, anemia has been the only significant cytopenia observed during artimisinin based therapy for malaria [26]. In conclusion, ARTS, a readily available antimalarial agent offers the potential for rapid translation to clinical trials of AML. Its oral bioavailability, excellent safety profile as well as synergy with conventional chemotherapy as well as BCL-2 inhibitor offer promising opportunities for testing it in de novo as well as relapsed/refractory AML.

[8]

[9]

[10] [11] [12]

[13]

[14]

[15]

[16]

Acknowledgements

[17] [18]

The authors also acknowledge support of the Gehr Family Center for Leukemia Research at City of Hope Medical Center. Partly supported by Grant P30CA033572 to City of Hope Comprehensive Cancer Center from National Cancer Institute of the National Institutes of Health.

[19]

[20]

Appendix A. Supplementary data [21]

Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.leukres.2017.05.007. [22]

References [23] [1] W.E. Ho, H.Y. Peh, T.K. Chan, W.S. Wong, Artemisinins: pharmacologic actions beyond anti-malarial, Pharmacol. Ther. 142 (1) (2014) 126–139. [2] L. Cui, X.-Z. Su, Discovery: mechanisms of action and combination therapy of artemisinin, Expert Rev. Anti Infect. Ther. 7 (8) (2009) 999–1013. [3] A.M. Gopalakrishnan, N. Kumar, Antimalarialmaction of artesunate involves DNA damage mediated by reactive oxygen species, Antimicrobial. Agents Chemother. 59 (1) (2015) 317–325. [4] S.V. Torti, F.M. Torti, Iron and cancer: more ore to be mined, Nat. Rev. Cancer 13 (5) (2013) 342–355. [5] A. Hamacher-Brady, H.A. Stein, S. Turschner, I. Toegel, R. Mora, N. Jennewein, et al., Artesunate activates mitochondrial apoptosis in breastcancer cells via ironcatalyzed lysosomal reactive oxygen species production, J. Biol. Chem. 286 (8) (2011) 6587–6601. [6] J.M. Fox, J.R. Moynihan, B.T. Mott, J.R. Mazzone, N.M. Anders, P.A. Brown, et al., Artemisinin-derived dimer ART-838 potently inhibited human acute leukemias: persisted in vivo and synergized with antileukemic drugs, Oncotarget 7 (6) (2016) 7268–7279. [7] C. Kim, J.H. Lee, S.H. Kim, G. Sethi, K.S. Ahn, Artesunate suppresses tumor growth

[24]

[25]

[26]

135

and induces apoptosis through the modulation of multiple oncogenic cascades in a chronic myeloid leukemia xenograft model, Oncotarget 6 (6) (2016) 4020–4035. L. Liu, L.F. Zuo, J. Zuo, J. Wang, Artesunate induces apoptosis and inhibits growth of Eca109 and Ec9706 human esophageal cancer cell lines in vitro and in vivo, Mol. Med. Rep. 12 (1) (2015) 1465–1472. N.D. Yang, S.H. Tan, S. Ng, Y. Shi, J. Zhou, K.S. Tan, et al., Artesunate induces cell death in human cancer cells via enhancing lysosomal function and lysosomal degradation of ferritin, J. Biol. Chem. 289 (48) (2014) 33425–33441. H. Dombret, C. Gardin, An update of current treatments for adult acute myeloid leukemia, Blood 127 (1) (2016) 53–61. J. Felipe Rico, D.C. Hassane, M.L. Guzman, AML stem cells: from bench to bedside, Cancer Lett. 338 (1) (2013) 4–9. E. Lagadinou, A. Sach, et al., BCL-2 inhibition targets oxidative phosphorylation and selectively eradicates quiescent human leukemia stem cells, Cell Stem Cell 12 (3) (2013) 329–341. M. Paul-Samojedny, D. Kokocińska, A. Samojedny, U. Mazurek, R. Partyka, Z. Lorenz, T. Wilczok, Expression of cell survival/death genes: cl-2 and Bax at the rate of colon cancer prognosis, Biochim. Biophys. Acta 1751 (1–2) (2005) 25–29. C. Callens, S. Coulon, J. Naudin, I. Radford-Weiss, N. Boissel, E. Raffoux, et al., Targeting iron homeostasis induces cellular differentiation and synergizes with differentiating agents in acute myeloid leukemia, J. Exp. Med. 207 (4) (2010) 731–750. H.-M. Shen, Z.-G. Liu, JNK signaling pathway is a key modulator in cell death mediated by reactive oxygen and nitrogen species, Free Radic. Biol. Med. 40 (6) (2006) 928–939. S. Mazel, D. Burtrum, H. Petrie, Regulation of cell division cycle progression by Bcl2 expression: a potential mechanism for inhibition of programmed cell death, J. Exp. Med. 183 (5) (1996) 2219–2226. D.H. Hockenbery, Z.N. Oltvai, X.M. Yin, C.L. Milliman, S.J. Korsmeyer, Bcl2 functions in an antioxidant pathway to prevent apoptosis, Cell 75 (2) (1993) 241–251. D.A. Hildeman, T. Mitchell, B. Aronow, S. Wojciechowski, J. Kappler, P. Marrack, Control of Bcl-2 expression by reactive oxygen species, Proc. Natl. Acad Sci. U. S. A. 100 (25) (2003) 15035–15040. R. Pan, L.J. Hogdal, J.M. Benito, D. Bucci, L. Han, G. Borthakur, et al., selective BCL-2 inhibition by ABT-199 causes on target cell death in acute myeloid leukemia, Cancer Discov. 4 (3) (2014) 362–375. E. Li Dechao Ueta, T. Kimura, T. Yamamoto, T. Osaki, Reactive oxygen species (ROS) control the expression of bcl-2 family proteins by regulating their phosphorylation and ubiquitination, Cancer Sci. 95 (8) (2004) 644–650. C. DiNardo, D. Pollyea, K. Pratz, M.J. Thirman, A. Letai, M. Frattini, et al., A Phase 1b study of venetoclax (ABT-) in combination with decitabine or azacytidine in treatment-naїve patients with acute myelogeneous leukemia who are ≥ 65 years and not eligible for standard induction therapy, Blood 126 (23) (2015) Abstract. Y. Ozawa, A.H. Williams, M.L. Estes, N. Matsushita, F. Boschelli, R. Jove, et al., Src family kinases promote AMl cell survival through activation of signal transducers and activators of transcription (STAT), Leuk. Res. 32 (6) (2008) 893–903. C. Dos Santos, T. McDonald, Y.W. Ho, H. Liu, A. Lin, S.J. Forman, et al., The Src and c-Kit kinase inhibitor dasatanib enhances p53-mediated targeting of human acute myeloid leukemia stem cells by chemotherapeutic agents, Blood 122 (11) (2013) 1900–1913. P.S. Twomey, B.L. Smith, C. McDermott, A. Novitt-Moreno, W. McCarthy, Sp. Kachur, et al., Intravenous artesunate for the treatment of severe and complicated malaria in the United States: clinical use under an investigational new drug protocol, Ann. Intern. Med. 163 (7) (2015) 498–506. M. Konig, C. von Hagens, S. Hoth, I. Baumann, I. Walter-Sack, L. Edler, et al., Investigation of ototoxicity of artesunate as add-on therapy in patients with metastatic or locally advanced breast cancer: new audiological results fro a prospective, open, uncontrolled monocentric phase I study, Cancer Chemother. Pharmacol. 77 (2) (2016) 413–427. J. Zwang, J.L. Ndiaye, A. Djimde, G. Dorsey, A. Martensson, C. Karema, et al., Comparing changes in haematologic parameters occurring in patients included in randomized controlled trials of artesunate-amodiaquine vs single and combination treatments of uncomplicated falciparum in sub-Saharan Africa, Malar. J. 13 (2014) 114.