The effect of co-treatments of chemotherapeutic drugs and curcumin on cytotoxicity and FLT3 protein expression in leukemic stem cells


 This study aims to enhance efficacy and reduce toxicity of the combination treatment of a drug and curcumin (Cur) on leukemic stem cell and leukemic cell lines, including KG-1a and KG-1 (FLT3+ LSCs), EoL-1 (FLT3+ LCs), and U937 (FLT3− LCs). The cytotoxicity of co-treatments of Dox or Ida at concentrations of the IC10 – IC80 values and each concentration of Cur at the IC20, IC30, IC40, and IC50 values (conditions 1, 2, 3, and 4) was determined by MTT assays. Dox–Cur and Ida-Cur increased cytotoxicity in leukemic cells. Dox–Cur co-treatment showed additive effects in several conditions. The effect of this co-treatment on FLT3 expression in KG-1a, KG-1, and EoL-1 cells was examined by Western blotting. Dox–Cur decreased FLT3 protein levels and total cell numbers in all the cell lines. By contrast, the FLT3 protein levels and total cell number after Cur treatment did not show significant differences as a result of the co-treatments. Dox–Cur decreased FLT3 protein expression in a dose dependent manner. In summary, Cur was the effective compound in inhibiting FLT3 protein expression. Co-treatment with Dox–Cur could enhance the cytotoxicity of Dox by inhibiting the proliferation of AML leukemic stem cells.


Introduction
Leukemia is among the top 10 cancers diagnosed globally. It is a group of cancers of early blood-forming cells, which are characterized by the uncontrolled production and accumulation of blast or immature abnormal blood cells in the peripheral blood and bone marrow. Leukemia can be divided into four major types according to the stage and cell of origin: acute myeloid leukemia (AML), acute lymphoid leukemia (ALL), chronic myeloid leukemia (CML), and chronic lymphocytic leukemia (CLL). AML is the most common type of acute leukemia in adults, with the highest incidence and death rate in both sexes. It can be distinguished by clonal expansion of abnormal myeloid blasts in bone marrow, peripheral blood, or other tissues. According to recent data, 15-25% of AML patients fail to achieve complete remission (CR) due to chemotherapy resistance and may show relapse, with the overall 5-year survival rate of approximately 40% 1,2 . Moreover, between 10 and 40% of newly diagnosed AML patients do not achieve CR with intensive induction therapy, and such patients are categorized as primary refractory or resistant 3 .
Hence, AML is de ned as an aggressive malignant myeloid disorder.
One theory of resistance and relapse in AML patients involves the presence of subpopulations of leukemic stem cells (LSCs) 4 . LSCs have been described as the human AML-initiating cell with a selfrenewal capacity and the ability to give rise to heterogeneous lineages of cancer cells 2,5 . They can be identi ed by the cell surface phenotype CD34 + hematopoietic stem cell and CD38 − subpopulation 6 .
The used traditional chemotherapeutic drugs are incapable of defeating the LSC population due to many reasons. First, these drugs have been designed to eliminate fast-dividing cells by inhibiting cell cycle progression 7 ; thus, they cannot it did not affect LSCs which mostly remain in stage G 0 of the cell cycle 8 .
Second, the expression of P-glycoprotein (MDR1), a multidrug resistance e ux pump protein, in LSCs potentially removes cytotoxic agents from cancer cells 9 . In addition, LSCs may sustain some mutations and epigenetic changes, resulting in conventional drug toxicity reduction and resistance 10,11 . Thus, LSCs are considered to play a fundamental role in AML pathogenesis and have become the main targeted therapies of AML.
Although drug resistance in AML patients usually occurs, the traditional chemotherapy remains a popular method for leukemia treatment due to its high ability to destroy cancer cells that can spread throughout the whole body. Anthracycline antibiotics, such as doxorubicin (Dox (14-hydroxydaunorubicin)) and idarubicin (Ida (4-demethoxydaunorubicin)), are generally used as standard chemotherapeutic agents for AML treatment 12 . These drugs function by inhibiting topoisomerase II activity in DNA transcription and also trigger apoptosis or autophagy in cells 13 . The combination of anthracyclines and cytarabine in the initial treatment is capable of inducing complete remission (CR) in approximately 45-70% of patients 14 ; however, more than 40% of CR cases eventually experience relapse within 2 years 15 . The previous studies on AML leukemic stem cells demonstrated that anthracycline is less effective in killing LSCs (CD34 + /CD38 − cells) than committed leukemic cells (CD34 + /CD38 + cells) 16 , and the co-treatment of cytarabine and anthracyclines is less effective against primitive AML cells than against leukemia blasts 17,18 . Furthermore, with high dose administration, anthracyclines are able to cause the development of side effects in patients in relation to its chemical structure, including nausea, vomiting, hair loss, and myelosuppression 19 . Several reports expressed their concern about the presence of cardiac, renal, and liver toxicity in patients treated with Dox 20,21 . Thus, combination therapy with natural substances with chemosensitizing and chemoprotective activities may be a promising strategy to overcome LSCs and reduce the side effects of anthracyclines.
Curcumin (Cur) is a natural polyphenol constituent of turmeric (Curcuma longa Linn.). It exhibits a wide range of pharmacological activities, such as antioxidant, anti-cancer, anti-in ammatory, and antimicrobial effects [22][23][24] . Previous studies reported that Cur exhibited an excellent cytotoxic effect; induced cell death in several types of leukemic cell lines 25,26 ; and showed inhibitory effects on WT1 and FLT3 protein expression, which are associated with cell proliferation 26,27 . Moreover, it inhibited the activity of Pglycoprotein (MDR1) 28 and exhibited cancer chemopreventive properties, especially in myocardial protection 29 by inhibiting ROS generation 30 . Consequently, it may be possible to manipulate the combination of Cur and anthracyclines for reduction in anthracycline toxicity and to overcome drug e ux via Pgp-mediated MDR in leukemia on AML leukemic cells and LSCs. Although Dox and Cur exhibit synergistic cytotoxic effects on cancer cell models, the combination of free Dox and free Cur shows a slightly obvious synergistic effect on the animal model 31 .
The aims of this study were to study the cytotoxicity of co-treatment with anthracycline drugs and curcumin for FLT3-overexpressing leukemic stem cells (KG-1a and KG1), FLT3-overexpressing leukemic cells (EoL-1), and non FLT3-expressing leukemic cells (U937). Moreover, the effect of co-treatments on FLT3 protein expression and total cell numbers were determined.
The IC 50 values of Dox in leukemic stem cells (KG-1a and KG-1) were found to be signi cantly higher than for leukemic cells, EoL-1 and U937 cells. However, in the group of the leukemic stem cell line, KG-1 cells were substantially more responsive to Dox and Ida than KG-1a cells were, indicating that a high number of LSCs affected the chemotherapeutic treatment's sensitivity. Furthermore, the IC 50 values of Cur in KG-1a cells were considerably higher than those in the other cells. These ndings demonstrated the drug resistance in LSCs against LCs and suggest that it might be possible to use to improve the potency of AML treatment.
Effects of combination treatments of various concentrations of Cur and a xed concentration of Dox on cell number and viability in leukemic stem cells and leukemic cells. According to the results presented in previous section, Cur and Dox-Cur treatments could inhibit AML LSC and LC cell proliferation more effectively than Dox treatment alone. Three non-toxic concentrations within the range of Cur's IC 20 value and a xed concentration of Dox from Dox-Cur condition 1 were tested with KG-1a, KG-1, and EoL-1 cells for 48 h to con rm the impact of Cur on cell proliferation inhibition of Dox. The results demonstrate that the co-treatments of Dox-Cur signi cantly decreased the cell number of both cell lines in a dose dependent manner when compared to a single Dox treatment and control (Fig. 5). The cell number of KG-1a cells in the control group was 3.44 ⋅ 10 5 cells/mL, and decreased to 2.96 × 10 5 , 2.21 × 10 5 , 1.91 × 10 5 , and 1.46 × 10 5 cells/mL in response to Dox, Dox + Cur at 4 µg/mL, Dox + Cur at 4.5 µg/mL, and Dox + Cur at 5 µg/mL, respectively (Fig. 5A). In addition, the cell number of KG-1 cells decreased from 4.18 × 10 5 cells/mL in the control group to 3.48 × 10 5 , 2.93 × 10 5 , 2.47 × 10 5 , and 2.22 × 10 5 cells/mL in response to Dox, Dox + Cur at 3 µg/mL, Dox + Cur at 3.5 µg/mL, and Dox + Cur at 4 µg/mL, respectively (Fig. 5C).

Discussion
In this experiment, doxorubicin (Dox) and idarubicin (Ida), standard chemotherapy for AML patients, were chosen as chemotherapeutic substance models to be studied. They can destroy leukemic cells by becoming incorporated into the DNA single strand, inhibiting topoisomerase II activity in DNA transcription, and triggering apoptosis or autophagy 13,34,35 . The cytotoxic activity of these anthracyclines was determined in each leukemic cell line by MTT assays. Both drugs showed the greatest cytotoxicity for EoL-1 cells, followed by U937, KG-1, and KG-1a cells. The inhibitory concentrations at cell growth (IC 50 ) values of 50 of idarubicin on KG-1a and KG-1 cells were 19.82 ± 1.80 and 5.45 ± 0.89 ng/mL, respectively. In contrast, doxorubicin showed lower cytotoxicity than idarubicin with IC 50 values of 0.65 ± 0.13 and 0.21 ± 0.02 µg/mL, respectively. This may have resulted from the absence of the methoxyl group at position 4 of idarubicin's structure which increased the lipophilicity and rate of cellular uptake, leading to greater toxicity than that of daunorubicin or doxorubicin 36 .
In addition, the cytotoxicity of curcumin (Cur), a natural substance with chemosensitizing and chemoprotective activities 23 , was also examined with four leukemic cell lines by MTT assay. The results showed that Cur demonstrated the highest cytotoxic effect on EoL-1 cells, followed by U937, KG-1, and KG-1a cells. Thus, Cur was selected as a supplementary substance for enhancing the e ciency and decreasing the toxicity of anthracycline drugs in this study. According to the cell viability curve, the co-treatment of Dox-Cur and Ida-Cur tended to increase the cytotoxicity for KG-1a, KG-1, EoL-1, and U937 cells in dose-dependent manners as compared to single drug treatment. Moreover, Cur also enhanced the cytotoxic e cacy on both chemotherapeutic drugs in dose dependent manners due to the lower IC 50 values of anthracyclines used in co-treatment in each cell line.
Dox and Ida are usually ineffective due to an increase in LSCs, drug resistance, and relapse in AML patients. In this study, Cur which is a natural supplementary substance, was found to improve the cytotoxicity of Dox and Ida in all the cell lines due to its anti-leukemic (apoptotic induction) 37 and chemosensitizing (decreasing MDR-1 gene expression) 28 activities. For these reasons, it could decrease the toxicity of both chemotherapies, resulting in lower IC 50 values for drugs in co-treatments when compared with single drug treatments.
It is notable that effective doses of the co-treatments used to treat KG-1a cells were higher than those for KG-1, EoL-1, and U937 cells. KG-1a and KG-1 cells are leukemic stem cell lines with a high percentage of leukemic stem cells (∼95% and ∼55%, respectively). These cells are well-known for their chemotherapy resistance, which includes primary rest in the stage G 0 of the cell cycle and high expression of the drug e ux pump. Since the EoL-1 and U937 cells lack a stem cell population, they were more vulnerable to the co-treatments.
The combination treatment of Dox-Cur showed synergistic and additive cytotoxic effects on both AML leukemic stem cell lines (KG-1a and KG-1 cells) and AML leukemic cell lines (EoL-1 and U937 cells). Despite the fact that only Dox-Cur condition 3 showed synergism on KG-1a and EoL-1 cells, Cur was able to be used as a supplement to lower chemotherapeutic agent doses. The combination treatment also reduced the concentration at the IC 50 value of Dox in each cell line which could be a useful formulation to decrease the cytotoxicity of Dox on normal cells. However, the poor solubility and short biological half-life of Cur, as well as the non-speci c activity of Dox, resulted in low absorption and cytotoxicity of these drugs in tumor cells 13,24 .
FLT3 is a key driver of AML, and its mutations are associated with the development of high risk of relapse in patients. Previous studies demonstrated that Cur has an inhibitory effect on FLT3 protein expression in leukemic cells 27 . Thus, the combination of Dox and Cur for AML treatment may lead to FLT3 protein expression reduction, which is involved in the proliferation process of leukemic cells.
In this study, the non-toxic doses at IC 20  In the future, the effects of Dox-Cur on cell cycle progression and apoptosis induction will be assessed to validate the mechanism of co-treatment's effect on cell proliferation inhibition and cell death.

Conclusion
Overall, anthracyclines (Dox and Ida) and Cur, a natural phenolic compound with anti-tumor activity, were shown to be effective AML chemotherapeutic agents. Our results show that the combination of Dox and Cur had a synergistic effect and could improve Dox anti-tumor activity in AML cells, particularly leukemic stem cells, by inhibiting cell proliferation through FLT-3 protein suppression. This nding presents an alternative choice that may be useful in the development of a promising regimen for the treatment of AML relapse in the future. (monoblastic leukemic cell line) was purchased from ATCC ® . These were cultured in RPMI-1640 medium containing 10% fetal calf serum, 1 mM L-glutamine, 100 units/mL penicillin, and 100 mg/mL streptomycin. All the leukemic cell lines were cultured at 37°C in a humidi ed incubator with 5% CO 2 .

Methods
Cytotoxicity of single doxorubicin, idarubicin, and curcumin (curcuminoid mixture) on leukemic stem cell and leukemic cell viability by MTT assay. KG-1a and KG-1 cell lines were adjusted to 1.5 ´ 10 4 cells, while EoL-1 and U937 cells were adjusted to 3.0 ´ 10 4 and 1.0 ´ 10 4 cells in 100 µL of complete medium, and then seeded into at-bottom 96-well plate and incubated at 37°C under 5% CO 2 atmosphere for 24 h. Western Blotting. KG-1a, KG-1, and EoL-1 cells were prepared and treated with Dox, Cur, and the cotreatment. After that, the cells were harvested after 48 h of incubation, and the whole proteins were extracted using RIPA buffer. The protein concentra-tion was measured with the Folin-Lowry method. The protein lysates were separated through 12% SDS-PAGE and then transferred to PVDF membranes. The membranes were blocked in 5% skim milk and probed by rabbit polyclonal anti-FLT3 and rabbit polyclonal anti-GAPDH antibody at a dilution of 1:1,000. The reaction was followed by HRP-conjugated goat anti-rabbit IgG at 1:15,000 dilution. The proteins were visualized using LuminataTM Forte Western HRP substrate. Finally, the protein band signal was quanti ed using a scan densitometer (Bio-Rad, CA, USA) or Fluorchem E Western blot and gel imager (ProteinSimple, CA, USA).
Statistical Analysis. The average of triplicate experiments and standard derivation (SD) were used for quanti cation. The levels of target protein expressions were compared to those of the vehicle control in each experiment. The results are shown as mean ± SD. The differences between the means of each sample were analyzed by one-way analysis of variance (one-way ANOVA). Statistical signi cance was considered at p < 0.05, p < 0.01, and p < 0.001. Figure 1 Cytotoxicity of (A) doxorubicin, (B) idarubicin, and (C) curcumin on KG-1a, KG-1, EoL-1, and U937 cells.