Transferrin-Conjugated pH-Responsive γ-Cyclodextrin Nanoparticles for Antitumoral Topotecan Delivery

In this study, we developed γ-cyclodextrin-based multifunctional nanoparticles (NPs) for tumor-targeted therapy. The NPs were self-assembled using a γ-cyclodextrin (γCD) coupled with phenylacetic acid (PA), 2,3-dimethylmaleic anhydride (DMA), poly(ethylene glycol) (PEG), and transferrin (Tf), termed γCDP-(DMA/PEG-Tf) NPs. These γCDP-(DMA/PEG-Tf) NPs are effective in entrapping topotecan (TPT, as a model antitumor drug) resulting from the ionic interaction between pH-responsive DMA and TPT or the host–guest interaction between γCDP and TPT. More importantly, the γCDP-(DMA/PEG-Tf) NPs can induce ionic repulsion at an endosomal pH (~6.0) resulting from the chemical detachment of DMA from γCDP, which is followed by extensive TPT release. We demonstrated that γCDP-(DMA/PEG-Tf) NPs led to a significant increase in cellular uptake and MDA-MB-231 tumor cell death. In vivo animal studies using an MDA-MB-231 tumor xenografted mice model supported the finding that γCDP-(DMA/PEG-Tf) NPs are effective carriers of TPT to Tf receptor-positive MDA-MB-231 tumor cells, promoting drug uptake into the tumors through the Tf ligand-mediated endocytic pathway and increasing their toxicity due to DMA-mediated cytosolic TPT delivery.


Introduction
For successful chemotherapy, developing multifunctional drug delivery systems to improve the therapeutic effect while minimizing the side effects of drugs has been widely conducted worldwide [1][2][3]. In particular, recent intensive research on stimulus-responsive drug-carrying systems that induce explosive drug release by reacting sensitively to changes in pH, temperature, and localized enzyme expression in specific tissues has been widely conducted [1,[3][4][5]. Interestingly, among these physical environmental factors, the pH-stimulus factor is remarkable because it is a special feature that appears inside the cellular endosome/lysosome and around solid tumors [1][2][3][4][5]. For example, the weakly acidic weakly acidic pH condition (~6.8) of the tumor-surrounding environment formed by the anaerobic metabolism of tumor tissues can be targeted using pH-responsive polymeric carrier systems [1][2][3][4][5]. Furthermore, if the particles absorbed by tumor cells through endocytosis have an endosomal-escaping ability, more effective cytosolic drug delivery and synergistic therapeutic effects can be expected. However, the particles without an endosomal-escaping ability cannot move to the cytoplasm, are usually exocytosized, or eventually move to the major degradative compartments (lysosomes) of cells [1][2][3][4][5][6][7]. Therefore, recent research has concentrated on exploiting stimulus-responsive multifunctional drug-carrying platforms that effectively recognize site-spatial tumor environments and exhibit excellent endosomal-escaping ability [7][8][9][10][11][12][13][14][15]. Indeed, considering the potential in vitro/in vivo antitumor efficiency of these systems, the design and development of pH-responsive endosomolytic drug-carrying particles may guarantee relatively high antitumor efficiency, even if tumor environments possess an inherently unpredictable complexity [1][2][3][4][5]16].

TPT Loading
The NPs (γCDP-(DMA/PEG-Tf) NPs, γCDP-(DMA/PEG) NPs, γCDP-(PEG-Tf) NPs, and γCDP-(PEG) NPs: 20 mg) were mixed with TPT (10 mg) in deionized water for 1 day. The resulting solution was dialyzed and purified through an ultracentrifugation process at 20,000 rpm for 30 min, and then lyophilized. The amount of TPT entrapped in NPs was calculated after measuring the TPT fluorescence intensity of the supernatant using a microplate reader (Bio-Tek, Winooski, VT, USA) at λ excitation of 400 nm and λ emission of 530 nm. The loading efficiency (%) of TPT in each NP sample was defined as the weight percentage of TPT in the NP sample relative to the initially feeding amount of TPT. The loading content (%) of TPT in each NP sample was calculated as the weight percentage of TPT in the NPs [25,26].

Characterization of NPs
The particle size and zeta potential of each NP sample (0.1 mg/mL) was measured using a Zetasizer 3000 instrument (Malvern Instruments, Malvern, UK) after being stabilized in 150 mM PBS (pH 7.4, 6.5, or 6.0) at 25 • C for 4 h. The morphology of each NP sample was confirmed using a field-emission scanning electron microscopy (FE-SEM, Hitach S-400, Fukuoka, Japan) after being stabilized in 150 mM PBS (pH 7.4, 6.5, or 6.0) at 25 • C for 4 h [32-37].

In Vitro TPT Release Test
Each NP sample (1 mg/mL) was added to a dialysis membrane tube (Spectra/Por MWCO 50 kDa) and immersed in 15 mL of fresh PBS (150 mM, pH 7.4, 6.5, or 6.0). The membrane tubes were incubated in a shaking water bath (100 rpm) at 37 • C for 48 h. At each time point, the external PBS solution of the dialysis tubes was replaced with a fresh PBS solution. The amount of TPT released from the NP sample was determined after measuring the fluorescence intensity of the PBS solution using a microplate reader (Bio-Tek, Winooski, VT, USA) at λ excitation of 400 nm and λ emission of 530 nm [15,[23][24][25].

Hemolysis Test
The endosomolytic ability of NPs was determined by conducting a hemolysis analysis [37]. Red blood cells (RBCs) were isolated using fresh mouse blood from BALB/c mice (7-week old female) and purified through centrifugation at 1500 rpm for 10 min three times. The RBCs in the pellet were washed with fresh PBS, and then dispersed in PBS (150 mM, pH 7.4, 6.5, and 6.0). Each NP sample (100 µg/mL) was then incubated with RBC solution (1 × 10 6 cells/mL) in a shaking water bath at 37 • C for 1 h. The treated RBCs were then centrifuged at 1500 rpm for 10 min, and the light absorbance (LA) of the resulting solution was measured using a microplate reader (Bio-Tek, Winooski, VT, USA) at a 541 nm wavelength. In addition, the RBC solution, which was completely lysed using 2% (w/v) Triton X-100, was used as a positive control, and the PBS-treated RBC solution was used as a negative control. The endosomolytic activity (%) of NPs was determined from the degree of RBC hemolysis [36,37].

In Vitro Cell Cytotoxicity
The MDA-MB-231 or CHO-K1 cells were incubated with each NP sample (equivalent to TPT 10 µg/mL) and free TPT (10 µg/mL) suspended in RPMI-1640 medium (pH 7.4) at 37 • C for 24 h. The cell viabilities of the treated tumor cells were measured using a CCK-8 assay. In addition, the MDA-MB-231 and CHO-K1 cells were incubated with each NP sample (1-100 µg/mL, without TPT) suspended in an RPMI-1640 medium (pH 7.4) at 37 • C for 24 h to verify the original toxicity of each NP. Furthermore, we evaluated the cell apoptosis of both the MDA-MB-231 and CHO-K1 cells treated with NPs (equivalent to TPT 10 µg/mL) or free TPT (10 µg/mL) at 37 • C for 4 h. These tumor cells were stained with Annexin V-FITC and propidium iodide (PI) for 15 min at 25 • C and then analyzed using the FACSCalibur TM flow cytometer (FACS Canto II, Becton Dickinson, Franklin Lakes, NJ, USA) [15,[23][24][25].

3.2.
Characterization of γCDP-(DMA/PEG-Tf) NPs Figure 2a,b show that the average particle sizes of the NP samples ranged from 120 to 134 nm at pH 7.4 and 6.0, indicating that there was no change in particle size according to pH. However, the zeta potential of the γCDP-(DMA/PEG-Tf) NPs increased from −27.6 mV at pH 7.4 to −11.6 mV at pH 6.0 (Figure 2c), most likely due to the cleavage of DMA (backing into cationic primary amine, by hydrolysis) [16,22,23] at pH 6.0, as described in our previous reports. Similarly, the zeta potential of the γCDP-(DMA/PEG) NPs also increased from −24.6 mV at pH 7.4 to −8.4 mV at pH 6.0. However, γCDP-(PEG) NPs and γCDP-(PEG-Tf) NPs without DMA moieties indicated no significant difference in zeta potential when the pH of the solution was reduced to pH 6.0. Consequently, the cleavage of DMA in γCDP-(DMA/PEG-Tf) NPs and γCDP-(DMA/PEG) NPs at pH 6.0-6.5 caused extensive ionic repulsion between γCDP-ADH (without DMA, cationic primary amine) [22,23] and TPT, resulting in the activation of TPT release at pH 6.0-6.5 (Figure 3). In particular, the drug release behaviors of all NPs at pH 7.4 were due to the passive release of TPT from NPs; however, at pH 6.0, the drug release amount of γCDP-(DMA/PEG-Tf) NPs or γCDP-(DMA/PEG) NPs was approximately twofold higher than that of γCDP-(PEG-Tf) NPs or γCDP-(PEG) NPs. Interestingly, within 12 h, the drug release behavior of NPs reached plateaus, and their drug release behaviors up to 12 h showed almost zero-order kinetics. In addition, the  Figure 2a,b show that the average particle sizes of the NP samples ranged from 120 to 134 nm at pH 7.4 and 6.0, indicating that there was no change in particle size according to pH. However, the zeta potential of the γCDP-(DMA/PEG-Tf) NPs increased from −27.6 mV at pH 7.4 to −11.6 mV at pH 6.0 (Figure 2c), most likely due to the cleavage of DMA (backing into cationic primary amine, by hydrolysis) [16,22,23] at pH 6.0, as described in our previous reports. Similarly, the zeta potential of the γCDP-(DMA/PEG) NPs also increased from −24.6 mV at pH 7.4 to −8.4 mV at pH 6.0. However, γCDP-(PEG) NPs and γCDP-(PEG-Tf) NPs without DMA moieties indicated no significant difference in zeta potential when the pH of the solution was reduced to pH 6.0. Consequently, the cleavage of DMA in γCDP-(DMA/PEG-Tf) NPs and γCDP-(DMA/PEG) NPs at pH 6.0-6.5 caused extensive ionic repulsion between γCDP-ADH (without DMA, cationic primary amine) [22,23] and TPT, resulting in the activation of TPT release at pH 6.0-6.5 ( Figure 3). In particular, the drug release behaviors of all NPs at pH 7.4 were due to the passive release of TPT from NPs; however, at pH 6.0, the drug release amount of γCDP-(DMA/PEG-Tf) NPs or γCDP-(DMA/PEG) NPs was approximately twofold higher than that of γCDP-(PEG-Tf) NPs or γCDP-(PEG) NPs. Interestingly, within 12 h, the drug release behavior of NPs reached plateaus, and their drug release behaviors up to 12 h showed almost zero-order kinetics. In addition, the

In Vitro/In Vivo Tumoral Uptake and Tumor Inhibition
We evaluated the tumor-specific cellular internalization of γCDP-(DMA/PEG-Tf) NPs using confocal microscopy and flow cytometry. Here, MDA-MB-231 cells (Tf receptor-positive) and CHO-K1 cells (as a control, Tf receptor-negative) were treated with fluorescent TPT drug-loaded NP samples ( Figure 4). First, the quantitative cellular uptake of TPT-loaded NP samples was measured

In Vitro/In Vivo Tumoral Uptake and Tumor Inhibition
We evaluated the tumor-specific cellular internalization of γCDP-(DMA/PEG-Tf) NPs using confocal microscopy and flow cytometry. Here, MDA-MB-231 cells (Tf receptor-positive) and CHO-K1 cells (as a control, Tf receptor-negative) were treated with fluorescent TPT drug-loaded NP samples ( Figure 4). First, the quantitative cellular uptake of TPT-loaded NP samples was measured using a  (~67 and~78, respectively). These results indicate that the NPs with Tf ligands were efficiently endocytosed to MDA-MB-231 tumor cells [16,[32][33][34][35][36][37], which was also readily apparent in the confocal image results.
In addition, to verify the degree to which the multifunctionality of γCDP-(DMA/PEG-Tf) NPs affected in vivo tumor ablation, we investigated the in vivo pharmaceutical potential of γCDP-(DMA/PEG-Tf) NPs using MDA-MB-231 tumor-bearing BALB/c nude mice [27][28][29][30][31][32]37]. Here, the NP samples were intravenously administered to MDA-MB-231 tumor-bearing BALB/c nude mice, and their fluorescence images and in vivo antitumor efficacy were obtained. Figure 7a,b show that γCDP-(DMA/PEG-Tf) NPs and γCDP-(PEG-Tf) NPs were highly accumulated in the local tumor site, thus supporting their efficient Tf ligand-mediating tumor targeting ability [16,[23][24][25]27]. Although the accumulation of γCDP-(DMA/PEG-Tf) NPs and γCDP-(PEG-Tf) NPs in the liver was significant, most likely due to the extensive NP uptake of reticuloendothelial system in the liver [28,37], it was significant that γCDP-(DMA/PEG-Tf) NPs enabled immediate tumor inhibition with the help of the Tf ligands and DMA moieties. The relative tumor volumes (at 7 days post injection) in the nude mice injected with the γCDP-(DMA/PEG-Tf) NPs were approximately 2.8-, 2.1-, 4.6-, and 6.7-fold smaller than those of the nude mice injected with the γCDP-(DMA/PEG) NPs, γCDP-(PEG-Tf) NPs, free TPT, and saline (control), respectively (Figure 7c,d). These results reveal that γCDP-(DMA/PEG-Tf) NPs can preferentially bind to in vivo tumor cells and improve their antitumor activity.

Conclusions
In this study, γCDP-(DMA/PEG-Tf) NPs was successfully fabricated for highly efficient MDA-MB-231 tumor treatment. In vitro/in vivo results demonstrated that the multifunctionality of γCDP-(DMA/PEG-Tf) NPs enabled increased tumor cell binding affinity and led to significant tumor cell death in vivo. In particular, the destabilization of γCDP pores (resulting from the detachment of DMA moieties at endosomal pH) influenced cytosolic drug release and enhanced cell cytotoxicity. On the basis of the results of this study, we believe that these properties of γCDP-(DMA/PEG-Tf) NPs, prepared using biocompatible and functional materials, can be effective in selectively killing in vivo tumor cells, and there is a high possibility that they will be developed into tumor-targeting nanomedicine with high potential for application to tumor treatment in the future.
Supplementary Materials: The following are available online at www.mdpi.com/xxx/s1: Figure S1

Conclusions
In this study, γCDP-(DMA/PEG-Tf) NPs was successfully fabricated for highly efficient MDA-MB-231 tumor treatment. In vitro/in vivo results demonstrated that the multifunctionality of γCDP-(DMA/PEG-Tf) NPs enabled increased tumor cell binding affinity and led to significant tumor cell death in vivo. In particular, the destabilization of γCDP pores (resulting from the detachment of DMA moieties at endosomal pH) influenced cytosolic drug release and enhanced cell cytotoxicity. On the basis of the results of this study, we believe that these properties of γCDP-(DMA/PEG-Tf) NPs, prepared using biocompatible and functional materials, can be effective in selectively killing in vivo tumor cells, and there is a high possibility that they will be developed into tumor-targeting nanomedicine with high potential for application to tumor treatment in the future.