Development and Characterization of FLT3-Specific Curcumin-Loaded Polymeric Micelles as a Drug Delivery System for Treating FLT3-Overexpressing Leukemic Cells

Abstract

This study aimed to develop a curcumin (CM) nanoparticle targeted to Feline McDonough Sarcoma (FMS)-like tyrosine kinase 3 (FLT3) protein on the surface of leukemic cells and to evaluate their properties, specificity, cytotoxicity, and inhibitory effect on FLT3 protein level in FLT3 overexpressing leukemic cells, EoL-1 and MV-4-11 cells. FLT3-specific peptides were conjugated onto modified poloxamer 407 (P407) using the copper-catalyzed azide-alkyne cycloaddition reaction (CuAAC). The thin film hydration method was performed for FLT3-specific CM-loaded polymeric micelles (FLT3-CM-micelles) preparation. Flow cytometry and fluorescence microscopy were used to determine rate of cellular uptake. 3-(4,5-dimethythiazol-2-yl)-2,5-diphenyl tetrazolium bromide assay was used to test the cytotoxicity of the micelles on leukemic cells. FLT3-CM-micelles demonstrated a mean particle size less than 50 nm, high entrapment efficiency, and high rate of CM uptake by leukemic cells. The intracellular CM fluorescence is related to FLT3 protein levels on the leukemic cell surfaces. Moreover, FLT3-CM-micelles demonstrated an excellent cytotoxic effect and decreased FLT3 protein expression in the leukemic cells. The FLT3-CM-micelles could enhance both solubility and cytotoxicity of CM on FLT3 overexpressing leukemic cells. These promising nanoparticles may be used for enhancing anti-leukemic activity of CM and developed as a targeted drug delivery system in the future.

Introduction

Leukemia is one of the most common hematological malignancies, occurring with a high incidence and mortality rate worldwide. The overexpression of leukemia-related genes and their protein is useful for diagnosis, monitoring disease progression, and provides a new therapeutic target for leukemia treatment. Feline McDonough Sarcoma (FMS)-like tyrosine kinase 3, or FLT3, is a leukemia-related protein and leukemia marker that overexpresses on the cell surface of leukemic cells, especially acute myelogenous leukemia (AML) and B lineage acute leukemia.^1,2 FLT3 and its ligand (FLT3 ligand; FL) play a pivotal role in the survival or proliferation of leukemic blast cells.^3,4 In the unstimulated state, FLT3 receptor exists in a monomeric form. Upon the binding of FL to the extracellular domain of FLT3 receptor, the receptor undergoes a conformational change, resulting in the unfolding of the receptor and the exposure of the dimerization domain, allowing receptor-receptor dimerization to take place.⁵ After receptor dimerization, tyrosine kinase enzyme is activated leading to phosphorylation of various sites in the intracellular domain. After that, the mechanisms of signal transduction in targeted cells are initiated. This initiation lead to the production of many proteins that involve in cell proliferation, differentiation, and survival.⁶ Then, the dimerized receptors and FL are quickly internalized and degraded.⁷ Another experiment reported that internalization of FLT3 is dependent on both human homolog of murine double minute (Hdm-2) E3 ligase activity and an ubiquitin dependent endocytosis motif (UbE) domain of FLT3.⁸ Thus, endocytosis might play an important role in internalization of FLT3 and FL as well as uptake properties of FLT3 receptor on the leukemic cells. FLT3 mutations, including internal tandem duplication (FLT3-ITD) and a missense point mutation at the D835 residue within a FLT3 tyrosine kinase domain (FLT3-TKD), have been associated with a poor prognosis for overall survival.⁹ Chemotherapy is currently the most effective method for the treatment of leukemia. However, side effects to normal cells and drug resistance of leukemic cells after long periods of treatment are common problems associated with chemotherapy.

Due to the wide range of biological activities and lack of toxicity in human and animal models, natural products have been used as an alternative medicine for malignant tumors as well as leukemia. Curcumin (CM), a yellow pigmented substance, is the most popular medicinal plant found in turmeric (Curcuma longa Linn.). It is well known as the folk and an alternative medicine for many diseases. It is being clinically evaluated as a chemopreventive agent for major cancers, including breast, prostate, lung, stomach, colon, and leukemias.^10,11 Previous studies reported that CM and other curcuminoids exhibited an excellent cytotoxic effect and induced cell death in several types of leukemic cell lines, including K562, HL-60, and Jurkat cells.^12,13 Moreover, the inhibitory effects of CM were associated with a decrease of WT1 gene expression in K562 and Molt4 cell lines.^14,15 In a recent study, CM exhibited an excellent cytotoxic effect, induced cell cycle arrest, and demonstrated the strong inhibitory effect on FLT3 protein in FLT3-overexpressing EoL-1 leukemic cells.¹⁶ Although CM has many biological properties, the low water solubility is a major barrier to its clinical applications.¹⁷

To solve this problem, various types of CM nanoparticles have been developed and designed by many research groups. More than 1,500 publications were reported about the effectiveness and advantages of a CM nanocarrier system.^18-22 One of the most popular nanocarrier systems used to overcome the bioavailability of poorly water soluble drugs is polymeric micelles. Advantages of polymeric micelles as a nanodelivery system include small particle size, high stability, high water solubility, low toxicity, and amphiphilic character, have been reviewed previously.^23,24 A tri-block copolymer structure of poloxamer includes two hydrophilic parts of polyethylene oxide (PEO) and a more hydrophobic part of polypropylene oxide (PPO) making this polymer attractive for polymeric micelle preparation. The amphiphilic property of poloxamer plays an important role in micelle formation. In aqueous solution, a hydrophilic part of poloxamer forms the outer shell of the micelle and the hydrophobic drug is incorporated into the inner hydrophobic core.^25-27 During the last two decades, molecularly targeted therapy is very popular in the cancer research field. Small molecules which are specific to cancer cell markers and/or cancer related molecules, including monoclonal antibodies, peptides, synthetic drugs or inhibitors, have been developed and applied for targeted therapy.^28-30 In addition, targeted therapy using nanoparticles and drug nanocarriers is an effective system for cancer treatment with high efficacy, specificity, and toxicity to cancer cells as well as leukemia.^31-34 As a drug nanocarrier, CM nanoformulations enhanced the solubility of CM and increased the rate of CM uptake by tumor cells.^35,36 In addition, tumor-specific CM nanoparticles improved binding and targeting of cancer cells as well as leukemic cell models.^37-39

In the case of FLT3 targeted therapy, FLT3 inhibitors were extensively studied and developed during the last decade. Most of first generation of FLT3 inhibitors was tyrosine kinase inhibitors, such as AG1295, lestaurtinib, sunitinib, sorafinib, and midostaurin. In addition, the second generation FLT3 inhibitors, including quizartinib (AC-220), crenolanib, and PLX3397, demonstrated higher cytotoxic activity and specificity to FLT3 leukemic cells. The information of FLT3 inhibitors is reviewed elsewhere.^40-42 Another approach for FLT3 targeted therapy is an antibody-based approach. IMC-EB10, an anti-FLT3 monoclonal antibody, has been reported to inhibit FLT3 signaling in leukemic cells and improve the survival rate of NOD/SCID mice injected with FLT3 leukemic cells.⁴³ Although there are numerous FLT3 inhibitors and antibodies that have been developed since the last decade, only a small number of FLT3-specific nanoparticles were reported.⁴⁴ According to anti-leukemia properties of CM on FLT3-overexpressing leukemic cells¹⁶ and the effectiveness of drug delivery system (DDS) with target-specific nanoparticles, this study aims to produce FLT3-CM-micelles with surface FLT3-specific peptides, evaluate their properties, and investigate their cytotoxic effect on FLT3-overexpressing leukemic cell lines.

Materials and methods

Materials

A turmeric curcuminoids mixture with 80% of CM, 15% of demethoxycurcumin, and 5% of bisdemethoxycurcumin (Product No. C1386), Poloxamer 407, Dess-Martin periodianine, deuterated chloroform (CDCl₃), dichloromethane anhydrous, and 3-(4,5-dimethythiazol-2-yl)-2,5-diphenyl tetrazolium bromide (MTT) dye were purchased from Sigma-Aldrich (St Louis, MO). FLT3-specific peptides were purchased from Biomatik company (Wilmington, DE). Dichloromethane (CH₂Cl₂), bovine serum albumin (BSA), fetal bovine serum (FBS), tetrahydrofuran (THF), dialysis bag, dimethylsulfoxide (DMSO), and phosphate buffer solution (PBS) were obtained from Fisher Scientific (Pittsburgh, PA). Roswell Park Memorial Institute (RPMI)-1640 medium, and Iscove’s Modified Dulbecco’s Medium were purchased from Life Technologies (Carlsbad, CA). Penicillin/Streptomycin was purchased from MP Biomedicals (Santa Ana, CA). TEM carbon film copper grid and paraformaldehyde were purchased from Electron Microscopy Sciences (Hatfield, PA). The 96-well plate was purchased from Thermo Scientific (Waltham, MA). Rabbit polyclonal anti-FLT3 was purchased from upstate biotechnology (Lake Placid, NY). Rabbit polyclonal anti-GAPDH was purchased from Santa Cruz Bitechnology (Santa Cruz, CA). Horseradish peroxidase (HRP)-conjugated goat anti-rabbit IgG was purchased from Promega (Madison, WI). Luminata™ Forte Western HRP Substrate was purchased from Merck Millipore Corporation (Billerica, MA).

FLT3-Specific Peptide Design and Docking

An extracellular domain of FLT3 protein (PDB code 3QS9 chain E) was used as a target protein receptor and a template for peptide design and docking. The properties of the peptides and the predicted protein-peptide interaction data were analyzed using online available web servers and PyMOL program. The selected peptide sequences were synthesized with N-terminal propargylglycine modification by Biomatik company.

Conversion of Terminal Hydroxyl Groups to Aldehyde Groups on P407

The terminal hydroxyl groups (-OH) were converted to aldehyde groups (-CHO) using an oxidizing reagent according to a previous report.⁴⁵ Briefly, one gram of P407 was dissolved in 30 mL of anhydrous dichloromethane. Subsequently, 58.1 mg of Dess-Martin periodinane (DMP) was added and reacted for 24 h at room temperature. Then, cold petroleum ether was used to precipitate the product and the obtained polymer-CHO was purified by filtration and verified by nuclear magnetic resonance spectroscopy (¹H-NMR) in CDCl₃ and FTIR.

Azide Functionalization of P407-CHO and Peptide Conjugation Reaction

The P407-CHO (1 g) was used as the starting material for cargo azide preparation by dissolving in 100 mL of tetrahydrofuran (THF) to a final concentration of 10 mg/mL. Subsequently, 13.89 mg of azido-polyethylene glycol-amine linker (NH₂-PEG₆-N₃) was added to the solution and stirred at 400 rpm for overnight. Then, THF was evaporated and the obtained cargo azide powder was dried using a vacuum-dryer. The copper-catalyzed azide-alkyne cycloaddition (CuAAC) reaction was performed for the coupling of the cargo azide to alkyne-modified peptides (biomolecule alkyne) as previously described.⁴⁶ Various types of NMR were used to confirm the success of the conjugation reaction. The schematic for the conjugation reaction is shown in Scheme 1.

Scheme 1.

Preparation of cargo azide and conjugation products using the CuAAC reaction. The P407-OH was oxidized to P407-CHO using mild oxidizing agent (DMP). The P407-CHO was used as starting material for cargo azide preparation. Then, the alkyne peptides were conjugated onto the cargo azide using CuAAC reaction to obtain the FLT3-specific peptide containing polymer or conjugation products which were used for FLT3-CM-micelles preparation.

Preparation of CM-Micelles and FLT3-CM-Micelles

CM-micelles and FLT3-CM-micelles were prepared by the thin-film hydration method with a slight modification of the previous procedure described by Khonkarn et al.⁴⁷ Briefly, CM and P407 or cargo azide were prepared in CH₂Cl₂ and mixed together at 1:30 (w/w) of CM:polymer ratio. The solvent was evaporated by rotary evaporator at 40°C for 10 min to obtain CM-containing polymer films. The residual CH₂Cl₂ remaining in the films was removed by air drying in fume hood at room temperature for overnight. After that, the films were rehydrated in PBS, pH 7.5 by extensive mixing and the micelle solutions were filtered through a 0.22 μm filter membrane to obtain CM-micelles and FLT3-CM-micelles. The empty micelles were prepared according to the same procedure in the absence of CM.

Characterization of CM-Micelles and FLT3-CM-Micelles

The properties of CM-micelles and FLT3-CM-micelles, including mean particle size, polydispersity index (PdI), zeta potential (ZP), loading capacity (LC), entrapment efficiency (EE), and micelle morphology were evaluated. Particle size, PdI, and ZP were characterized using photon correlation spectroscopy (PCS; ZetaPALS Particle Sizing; Brookhaven Instrument Corp., Holtsville, NY). The concentration of CM in micelles was measured using a UV-visible spectrophotometer (Agilent 8453, Agilent Technologies, Palo Alto, CA) at 425 nm and calculated using the standard curve of CM in methanol.^35,38 The LC and EE were calculated using the following equations.

$$\mathrm{LC}(\%)=\frac{\mathrm{Amount\; of\; CM\; in\; CM­micelles}}{\mathrm{Total\; amount\; of\; polymer\; used}}\times 100$$ (1)

$$\mathrm{EE}(\%)=\frac{\mathrm{Amount\; of\; CM\; in\; CM­micelles}}{\mathrm{Total\; amount\; of\; CM\; used}}\times 100$$ (2)

The morphology of empty micelles and CM-micelles was observed under a FEI Tecnai™ transmission electron microscope (TEM; Tecnai™, FEI company, Hillsboro, OR). After preparation, the micelle solutions were dropped on a copper grid and dried at room temperature before being observed under the microscope.

Cells and Cell Culture Conditions

Eosinophilic leukemic cell line (EoL-1; RBRC-RCB0641), a model of wild type FLT3-overexpressing leukemic cells, was purchased from RIKEN BRC Cell Bank (Ibaraki, Japan). This cell line was cultured in RPMI-1640 medium containing 10% fetal calf serum, 1 mM L-glutamine, 100 units/mL penicillin, and 100 μg/mL streptomycin.

Biphenotypic myelomonocytic leukemic cell line (MV-4-11; ATCC^® CRL 9591TM), a model of FLT3-ITD overexpressing leukemic cell, was purchased from ATCC (Manassas, VA). This cell line was cultured in IMDM containing 10% fetal calf serum, 100 units/mL penicillin, and 100 μg/mL streptomycin.

Human erythroleukemic cell line derived from a chronic myelocytic leukemia (CML) patient (K562) was a generous gift from Dr. Chaisuree Supawilai, Research Institute for Health Sciences, Chiang Mai University, Thailand. Human acute lymphoblastic leukemic cell line (Molt4) was kindly provided by Prof. Dr. Watchara Kasinrerk, Faculty of Associated Medical Sciences, Chiang Mai University, Thailand. Both leukemic cells were cultured in RPMI-1640 medium containing 10% fetal calf serum, 1 mM L-glutamine, 100 units/mL penicillin, and 100 μg/mL streptomycin.

Kinetic Uptake of FLT3-CM-Micelles

The kinetic uptake experiment with continuous monitoring by flow cytometry was performed to evaluate the specificity of FLT3-CM-micelles on FLT3-overexpressing leukemic cells and to compare the internalization rate between FLT3-CM-micelle and CM-micelles. FLT3-overexpressing leukemic cells (EoL-1 and MV-4-11), low FLT3 expressing leukemic cell (Molt4), and very low FLT3 expressing leukemic cell (K562) were used as a model of leukemic cells with different levels of FLT3 protein expression. FLT3-CM-micelle solution and CM-micelle solution with a CM concentration of 6 μM were prepared. Equal volumes of micelle solution and cell suspension were mixed together and the intracellular CM content was measured using Beckman Coulter MoFlo XDP Cell Sorter (Brea, CA) with an excitation at 488 nm and emission at 520-530 nm. The fluorescence emission of CM in sequentially measured cells was monitored for 5 min. The specificity of FLT3-CM-micelles on FLT3 protein of each leukemic cell was expressed as the rate of cellular uptake of FLT3-CM-micelles by each leukemic cell which is show in formula 3. In addition, EoL-1 cells (1 × 10⁶ cells/mL) were incubated with FLT3-CM-micelles, CM-micelle, and free CM solution containing CM of 40 μM for 4 h. Then, the cells were harvested and washed twice with PBS. The intracellular green fluorescence of CM was observed under fluorescent microscope (Olympus BX41, Tokyo, Japan).

$$\mathrm{Rate\; of\; cellular\; uptake}=\frac{\mathrm{MFI\; of\; CM\; from\; FLT3­CM­micelles}}{\mathrm{MFI\; of\; CM\; from\; CM­micelles}}$$ (3)

MTT Cytotoxic Assay

MTT assay was performed to evaluate the cytotoxic effect of FLT3-CM-micelles compared to those of CM micelles and free CM solution. The cells (5.0 × 10⁴ cells/well for EoL-1 and 2.0 × 10⁴ cells/well for MV-4-11) were cultured in a 96-well plate containing 100 μL medium at 37 °C for 24 h before treatment. Then, 100 μL of the fresh medium containing FLT3-CM-micelles, CM-micelles, and free CM solution at various concentrations of CM (0-10 μM) were added into each well and incubated for another 48 h. Their empty micelle formulations were examined in parallel with those of micelles. After that, MTT dye solution (5 mg/mL) was added and then incubated at 37 °C for 4 h. Then, DMSO (200 μL) was added and mixed thoroughly to dissolve the formazan crystals. The absorbance was measured using SpectraMax^® M5 Multi-Mode Microplate Reader (Molecular Devices Crop, CA) at 578 nm with a reference wavelength of 630 nm. The cell viability of each treatment was plotted as a dose-response curve. The 50% inhibitory concentrations (IC₅₀) values of all formulations and free CM solution were determined and used to compare their activities.

Western Blot Analysis

To evaluate the FLT3 protein level, the leukemic cells after treatment were harvested. Total protein extraction was performed using radioimmunoprecipitation assay buffer. The protein concentration was measured by Folin-Lowry method and 80 μg of protein was separated by 7.5% SDS-PAGE. FLT3 protein detection was performed as previously described.¹⁶

Statistical Analysis

All data were expressed as the means ± standard deviation (SD) or means ± standard error of the mean (SEM) from triplicate samples of three independent experiments. The statistical differences between the means were determined using one-way ANOVA. The differences were considered significant when the probability value obtained was found to be less than 0.05 (p < 0.05).

Results and Discussion

FLT3 Specific Peptide Sequences and Their Properties

Two of FLT3-specific peptide sequences (CKR and EVQ peptides) were designed and some of their important properties are shown in Table 1. The sequence of each peptide and FLT3 protein (PDB code 3QS9 chain E) were sent to PEP-SiteFinder web server to identify candidate regions for protein-peptide interactions.⁴⁸ Normally, the PEP-SiteFinder generated 200 models of candidate protein-peptide binding site per each peptide. The predicted protein-peptide models are sorted according to their ATTRACT2 scores which corresponded to their binding energy.⁴⁹ In this study, model 1 which is the best ATTRACT2 scores of each peptide was selected for data analysis using PyMOL program. The best predicted protein-peptide interaction of each peptide is shown in Figure 1. From the docking data, the CKR peptide linked to FLT3 protein template with 5 hydrogen bonds at amino acid threonine (T167), 2 bonds to leucine (L168), arginine (R169), and glutamic acid (E266). On the other hand, the EVQ peptide interacted to the protein template with 8 hydrogen bonds at amino acid arginine (R243), phenylalanine (F245), glycine (G396), leucine (L397), 2 bonds to aspartic acid (D398), glycine (G400), and 2 bonds to tyrosine (Y401). According to the model of FLT3-FL complex reported by Verstraete K and colleagues.⁵⁰ the binding site of FL on FLT3 receptor is located mainly on domain 3 (amino acid residues 245-346) and some residues on domain 2 (amino acid residues 162-244) of the extracellular part of FLT3. This area is expected to be a high-affinity ligand binding site. Compared to our docking result, the binding sites of both peptides are related to domain 2 and 3 of the FLT3 extracellular domain. Thus, the binding of our peptides to the extracellular domain of FLT3 protein around the high-affinity binding site makes our peptides more attractive for targeting the FLT3 protein on leukemic cell membrane.

Table 1.

FLT3-specific peptide sequences and their properties

Peptide name	Sequence	No. of residues	Atomic compositiona (Total No. of atoms)	MWa (g/mol)	Instability indexa	p-value of binding prediction to FLT3 proteinb	Water solubilityc
CKR	CKRFQNSHL	9	C48H77N17O13S1 (156)	1132.2	Unstable	0.053	Good
EVQ	EVQTCISHLL	10	C49H83N13O16S1 (162)	1142.3	Stable	0.107	Poor

The data were obtained from

^(a)ProtParam web server at web.expasy.org/protparam/,

^(b)PepSite web server at pepsite2.russelllab.org/webserver, and

^(c)Innovagen web server at www.innovagen.com

Figure 1.

Structure of FLT3 protein (PDB 3QS9 chain E) in complex with (a) CKR peptide and (b) EVQ peptide were analyzed by the PyMOL program. The FLT3 protein template is shown in ribbon diagram (cartoon representation) and the peptide sequences are illustrated in stick representation. The interaction between FLT3 protein and each peptide with hydrogen bonds was shown as yellow dashed lines.

Determination of Modified Poloxamer 407, Cargo Azide, and Conjugation Products

¹H-NMR and FTIR spectroscopy were used to confirm the success of the conversion reaction of the terminal hydroxyl groups to aldehyde groups on P407. It was found that, P407-CHO demonstrated the aldehyde peak at δ = 9.75 ppm in ¹H-NMR spectrum (Figure 2a) and the C=O peak of aldehyde was observed at 1726.7 cm^-1 in the FTIR spectrum (Figure 2b). After that, the P407-CHO was used as starting material for cargo azide preparation. ¹H-NMR, ¹³C-NMR, and HSQC-NMR were performed to evaluate the cargo azide formation and conjugation products. The absence of aldehyde resonance and the presence of additional peaks in ¹H-NMR spectrum of the cargo azide were used to confirm the success of cargo azide synthesis (Figure 3). After peptide conjugation, the alkyne group in peptide disappeared in both ¹H-NMR and ¹³C-NMR spectra (Figure 4a and 4b). Moreover, the new expected peaks of triazole ring at δ 8.16 ppm in ¹H-NMR spectrum and 126.55 ppm in the ¹³C-NMR spectrum were found in conjugation products (Figure 4c). All of the results confirm the success of the conjugation process.

Figure 2.

¹H-NMR and FTIR spectra of P407-OH and P407-CHO. (a) The ¹H-NMR spectrum and (b) FTIR spectra of P407-CHO (upper) were compared with P407-OH (lower). The arrow indicates the presence of aldehyde (CHO) peak and the C=O absorption peak of aldehyde in P407-CHO sample after conversion reaction.

Figure 3.

NMR characterization of cargo azide, including (a) ¹H-NMR, (b) ¹³C-NMR. The absence of a CHO resonance (arrow) after cargo azide synthesis in ¹H-NMR and ¹³C-NMR were used to confirm the success of cargo azide synthesis.

Figure 4.

NMR characterization of conjugation products, including (a, b) ¹H-NMR, ¹³C-NMR, and ¹H-¹³C heteronuclear single quantum coherence-NMR spectra demonstrated the presence and absence of the alkyne resonance (red arrow and red circle) at 2.56 and 74.42 ppm in the peptide and conjugation product, respectively. (c) ¹H-¹³C HSQC-NMR spectra of the conjugation product demonstrated a new triazole peak (red circle).

Properties of CM-Micelles and FLT3-CM-Micelles

After micelle preparation, both CM-micelles and FLT3-CM-micelles demonstrated a clear yellowish solution and the empty micelles were clear and colorless (Figure 5a). The micelle properties were evaluated and the summarized data are presented in Table 2. All micelle formulations demonstrated high curcumin entrapment with %EE more than 75%. This present study demonstrated the effectiveness of the FLT3-CM-micelles in term of increasing CM solubility. The CKR and EVQ-CM-micelles increased the solubility of CM up to 64±2 μg/mL (174±6 μM) and 63±1 μg/mL (171±5 μM), respectively. The previous study reported that only 0.6 μg/mL of CM was dissolved in water.⁵¹ Thus, the FLT3-CM-micelles obtained in this study increased CM solubility with 105 times higher than the solubility of CM in water. A slightly larger mean particle size of FLT3-CM-micelles was found when compared to their empty micelle and CM-micelle. The increasing of micelle size after using conjugation product to form the micelle confirmed the successful of conjugation reaction. However, there were no statistically significant differences between micelle size of FLT3-CM-micelles and CM-micelle. This may dues to the small number of residues of peptides used in this study (only 9 and 10 residues). So, it is not long enough to make a micelle size bigger than CM-micelle. In addition, the TEM micrograph confirmed that all micelle formulations showed the nanoscale particle size (Figure 5b) with the mean micellar size less than 50 nm.

Figure 5.

Photographic image of CM-micelles and FLT3-CM-micelles. (a) Physical appearance of CM-micelles, FLT3-CM-micelles (CKR-CM-micelles and EVQ-CM-micelles), and their empty micelle. (b) TEM micrographs of all micelle formulations demonstrated a nanoscale particle size. The scale bar is 100 nm.

Table 2.

Physical properties of CM-micelles and FLT3-CM-micelles

Micelle formulations	Mean particle size (nm)	PdI	ZP (mV)	LC (%)	EE (%)
Empty micelles	28.1 ± 4.9	0.318 ± 0.014	-5.61 ± 1.14	ND	ND
CM-micelles	42.2 ± 7.4	0.273 ± 0.038	-5.04 ± 1.69	2.94 ± 0.14	88.4 ± 4.1
Empty FLT3-micelles
- Empty CKR-micelles	41.3 ± 5.1	0.344 ± 0.010	-8.26 ± 1.55	ND	ND
- Empty EVQ-micelles	45.8 ± 2.7	0.322 ± 0.009	-6.80 ± 1.78	ND	ND
FLT3-CM-micelles
- CKR-CM-micelles	46.6 ± 4.2	0.280 ± 0.007	-7.89 ± 2.13	2.57 ± 0.07	77.7 ± 2.1
- EVQ-CM-micelles	48.6 ± 4.6	0.274 ± 0.020	-8.95 ± 2.16	2.50 ± 0.06	76.0 ± 1.7

The data are represented as mean ± SEM of three independent experiments.

Kinetic Uptake of CM in FLT3-CM-Micelles

For kinetic uptake study, FLT3-overexpressing leukemic cells (EoL-1 and MV-4-11) were mixed together with the solution of FLT3-CM-micelles (CKR and EVQ-CM-micelles) and CM-micelles. The intracellular CM intensity was continuously measured using flow cytometry for 5 min. A very rapid accumulation of intracellular CM content from FLT3-CM-micelles was found and higher than those of intracellular CM contents from CM-micelles (Figure 6a and 6b). Both FLT3-CM-micelles demonstrated about 4.13 to 7.84 times higher CM accumulation than CM-micelles. In addition, CKR-CM-micelles exhibited about 1.45 folds higher CM accumulation than EVQ-CM-micelles in EoL-1 and MV-4-11 leukemic cells. To evaluate the specificity of FLT3-CM-micelles on FLT3-overexpressing leukemic cells, four leukemic cell types with different levels of FLT3 expression on their cell surface were used. FLT3-CM-micelle solution was added into the cell suspension and the rate of cellular uptake was evaluated using flow cytometry. The intracellular CM content in each cell type was continuously monitored for 5 min and the specificity of FLT3-CM-micelles on FLT3-overexpressing leukemic cells was evaluated. As shown in Figure 6c–6e, rate of cellular uptake of FLT3-CM-micelles is correlated to the level of FLT3 protein on leukemic cell surface. Both FLT3-overexpressing leukemic cells (EoL-1 and MV-4-11) demonstrated higher rate of cellular uptake of FLT3-CM-micelles than Molt4 and K562 cells. To confirm the binding of the peptide on FLT3 protein in EoL-1 cells, the CKR peptide was tagged with 3-azido-7-hydroxycoumarin using click chemistry as previously described.⁵² Then, equal volumes of cell suspension and the fluorophore-conjugated CKR peptide solution at 10, 100, and 200 μg/mL were mixed together and the fluorescence intensity was continuous measured for 5 min using Beckman Coulter MoFlo XDP Cell Sorter (Brea, CA) with an excitation at 405 nm and emission at 457 nm. The results showed that mean fluorescence intensity (MFI) of the fluorophore increases in proportion with the concentration of fluorophore-conjugated CKR peptide (Figure 6f). However, a competitive inhibition experiment should be considered to perform as another confirmation experiment for testing the specificity of the peptides on FLT3 protein in the cell surface. The fluorescence microscopy images confirmed that CKR-CM-micelles demonstrated a stronger intracellular green fluorescence of CM than those of CM-micelles and free CM solution (Figure 7). The expressions of FLT3 mRNA and/or FLT3 receptor levels on leukemic cell surface were previously reported that depend on their cell type.^53-55 The levels of FLT3 mRNA in myeloid cell lines such as EoL-1 and MV-4-11 was higher than the levels in T-lymphoid cell lines and erythroid cell lines. Moreover, the percentage of FLT3-positive cells of EoL-1 was 93% when compared to MV-4-11 (10%).⁵⁵ These previous findings help to support the results that cellular uptake rate of FLT3-CM-micelles relates to expression level of FLT3 on leukemic cell surface that specific to FLT3-CM-micelles on FLT3-overexpressing leukemic cells. Moreover, the different level of intracellular MFI of CM in FLT3-overexpressing leukemic cells among free CM solution, CM micelles, and FLT3-CM-micelles indicated that FLT3-CM-micelle formulation obtained in this study is an effective targeted drug delivery system for delivery of CM to FLT3-overexpressing leukemic cells.

Figure 6.

Cellular uptake of FLT3-CM-micelles and CM-micelles. Intracellular mean fluorescence intensity (MFI) of CM in (a) EoL-1 and (b) MV-4-11 compared to both FLT3-CM-micelles and CM micelles. (c-e) Rate of cellular uptake of both FLT3-CM-micelles in leukemic cells with different levels of FLT3 protein expression on their cell surface. (f) Cellular uptake of different concentrations of fluorescent-conjugated CKR peptide by EoL-1 cells. The data are represented as mean ± SEM of three independent experiments.

Figure 7.

Intracellular green fluorescence of CM from CKR-CM-micelles, CM-micelles, and free CM solution compared with their empty micelles and cell control.

Cytotoxic Activity of FLT3-CM-Micelles on FLT3-Overexpressing Leukemic Cells

Both EVQ and CKR-CM-micelles were studied their cytotoxic effect on EoL-1 and MV-4-11 cells compared to the activity of CM-micelles and free CM solution. FLT3-CM-micelle solution, CM-micelles, and free CM solution with various concentrations of CM at 0-10 μM were added into the cell suspension and incubated for 48 h. MTT assay was employed in this study. The results showed that, both FLT3-CM-micelles demonstrated a cytotoxic effect on EoL-1 and MV-4-11 cells with IC₅₀ values lower than those of CM-micelles and free CM solution treatments (Figure 8a, 8b, and Table 3). Furthermore, MV-4-11 cells were more sensitive to CM treatment than EoL-1 cells. A statistically significant difference of IC₅₀ values among the treatment of FLT3-CM-micelles, CM-micelles, and free CM solution was found, especially in EoL-1 treatment. However, there were no statistically significant difference of IC₅₀ values between the FLT3-CM-micelles groups and CM-micelles group in MV-4-11 treatment. It might be due to the characteristics of MV-4-11 cells that are highly sensitive to CM treatment when compared to EoL-1 cells. It can be suggested that FLT3-CM-micelles demonstrated a strong cytotoxic activity on both FLT3-overexpressing leukemic cells without the effects of their empty micelle formulation. From our previous report, EoL-1 cells were arrested at G₀/G₁ phase of cell cycle after CM treatment. Moreover, CM exhibited a strong inhibitory effect on rate of proliferation as well as FLT3 and STAT5A protein expressions in EoL-1 cells.¹⁶ Thus, the mechanism of CM action in EoL-1 and MV-4-11 cells might involve in the inhibition of STAT5A-induced cell proliferation. On the other hand, FLT3 causes the activation of phosphatidylinositol-3-kinase (PI3K) which stimulates protein kinase B or AKT protein that regulates many downstream proteins involve in cell survival and anti-apoptosis.⁵⁶ Thus, the inhibition of FLT3 protein expression by CM in the cell lines used in this study might inhibit the activity of PI3K and its downstream effectors to promote cell survival and anti-apoptosis. Previous studies also reported that CM directly inhibited PI3K and anti-apoptotic protein (B-cell lymphoma 2 or Bcl-2) expression.^57,58 These findings might support the cytotoxic activity of CM against both leukemic cell lines used in this study. However, EoL-1 and MV-4-11 cells have different phenotypes. Thus, the cytotoxic mechanisms were also different in response to CM treatments. Cell line with FLT3-ITD (MV-4-11 cells) was reported that it have a more rapid cycling of receptors on its cell surface than that of wild type FLT3 (EoL-1 cells).⁸ This might increases long term accumulation of CM in the cells and enhances cytotoxic effect on MV-4-11 than EoL-1 cells. However, the exact cytotoxic mechanistic actions of CM on the FLT3-overexpressing leukemic cells need further investigation.

Figure 8.

Cytotoxic activity of both FLT3-CM-micelles, CM-micelles, free CM solution, and their empty micelle formulations on FLT3-overexpressing (a) EoL-1 and (b) MV-4-11 leukemic cells. (c) Inhibitory effect of both FLT3-CM-micelles comparing with CM-micelles, free CM solution, and their empty micelles formulations on FLT3 protein level (upper band) in EoL-1 cells. The GAPDH protein (lower band) was used as normalized control of the system. The data are represented as mean ± SEM of three independent experiments. Asterisks (*) denote significant difference at p < 0.05).

Table 3.

Inhibitory concentration at 50% cell viability (IC₅₀) values of FLT3-CM-micelles, CM-micelles, and free CM solution on FLT3-overexpressing leukemic cells.

Micelle formulations	IC50 values (μM)
	EoL-1	MV-4-11
Empty micelles	>10	>10
Empty FLT3-micelles
- Empty CKR-micelles	>10	>10
- Empty EVQ-micelles	>10	>10
Free CM solution	9.70 ± 0.16	2.03 ± 0.02
CM-micelles	6.56 ± 0.42	1.08 ± 0.03
FLT3-CM-micelles
- CKR-CM-micelles	2.01 ± 0.05	1.06 ± 0.09
- EVQ-CM-micelles	1.96 ± 0.09	1.05 ± 0.02

The data are represented as mean ± SEM of three independent experiments.

Inhibitory Effect of FLT3-CM-Micelles on FLT3 Protein Expression in EoL-1 Leukemic Cells

To determine the inhibitory effect of both FLT3-CM-micelles on FLT3 protein level in EoL-1 leukemic cell model, the 20% inhibitory concentration (IC₂₀) values of all formulations obtained from MTT experiment were used. The leukemic cells were treated with free CM-solution, CM-micelles, both FLT3-CM-micelles formulations and their empty micelles formulations. After 12 h of treatment, EVQ-CM-micelles demonstrated an excellent inhibitory effect on FLT3 protein expression when compared to free CM solution, CM-micelles and CKR-CM-micelles. FLT3 protein levels were decreased by 85%, 71%, 68%, and 49% in response to treatments with EVQ-CM-micelles, CKR-CM-micelles, CM micelles, and free CM solution, respectively (Figure 8c). In addition, treatment with EVQ-CM-micelles demonstrated statistically significant difference in FLT3 protein level when compared to free CM-solution treatment. These findings confirm that the FLT3-CM-micelle formulations performed in this study can enhance rate of CM uptake and increase CM accumulation in EoL-1 cells. The previous studies reported that the cytotoxic activity of chemotherapeutic drugs to the cancer cells may be related with their accumulation rate, and the reduced drug accumulation may play a role in the mechanism of resistance of cancer to the drug.^59,60 Although there were no significant difference between the mean values of FLT3 protein expression in EoL-1 cells after treatment with CKR-CM-micelles and CM-micelles or free CM solution, the concentration of CM at IC₂₀ value used in CKR-CM-micelle formulation (0.5 μM) is very low when compared to 2.0 μM for CM-micelle formulation and 6.0 μM for free CM solution. All of these results indicate that FLT3-CM-micelle formulations are more effective than those of CM-micelles and free CM solution in term of cellular uptake, cytotoxicity, specificity to targeted cells, and inhibitory effect on FLT3 protein expression.

Conclusion

FLT3 has been recognized as a leukemia marker for acute myelogenous leukemia and B lineage acute lymphoblastic leukemia. There are many inhibitors that have been developed for treating FLT3-overexpressing leukemic patients. Moreover, specific targeted therapy and nanotechnology approaches are more attractive for leukemia treatment. In this study, FLT3-CM-micelles with FLT3-specific peptide conjugation were formulated and developed as FLT3-specific CM nanocarrier system. The obtained FLT3-CM-micelles improved the solubility of CM and demonstrated a nanoscale particle size with high CM entrapment. In addition, the cellular uptake of FLT3-CM-micelles represented a strong intracellular fluorescence intensity of CM when compared to those of CM-micelles and free CM solution. The rate of cellular uptake of FLT3-CM-micelles is related to FLT3 protein level on leukemic cell surfaces and confirmed the specificity of the peptides on FLT3 protein. Moreover, both FLT3-CM-micelles exhibited an excellent cytotoxic effects and inhibitory effect on FLT3 protein expression in FLT3-overexpressing leukemic cells due to the specificity of the peptides to the target cells and the effectiveness of CM and micelle formulations. The highlight of this experiment showed the decrease of FLT3 levels after FLT3-CM-micelles treatment with non-cytotoxic dose. These recent findings suggested that the CKR and EVQ peptides improve targeting of the FLT3-CM-micelles to FLT3-overexpressing leukemic cells. Thus, it might be useful to develop these formulations as hydrophobic drug nanocarrier for clinical application and leukemia treatment, especially FLT3-overexpressing leukemic patients.

Abbreviations used

CM

curcumin

FLT3

Feline McDonough Sarcoma (FMS)-like tyrosine kinase 3

P407

poloxamer 407

PDB

protein data bank