Two-stage pH-sensitive doxorubicin hydrochloride loaded core–shell nanoparticles with dual drug-loading strategies for the potential anti-tumor treatment

Abstract

A novel two-stage pH-sensitive doxorubicin hydrochloride (DOX·HCl) delivery system, programmed to respond to tumor extracellular circumstrances (about pH 6.5) and intercellular endo/lysosome (pH 5.0), was designed with dual drug-loading strategies. Hydrazone linked cationic conjugates poly(ethyleneimine)-C6-succinimidyl 6-hydrazinonicotinate acetone hydrazone-DOX·HCl (PEI-C6-SANH-DOX·HCl, PEI-C-DOX) was synthesized, then condensed the DOX·HCl loaded 21-base (CGA)₇ oligodeoxynucleotides (CGA-ODNs/DOX·HCl, OD) to get the dual methods DOX·HCl loaded nanoparticles PEI-C-DOX/OD (POD) as inner core. Subsequently, O-carboxymethyl-chitosan (CMCS) was coated on POD to construct the core–shell nanoparticles CMCS/POD (CPOD). The average size and zeta potential of CPOD were (165.0 ± 3.3) nm and −(15.6 ± 0.82) mV, respectively. In vitro evaluation showed that CMCS could dissociate from POD at tumor extracellular pH values, and the cellular uptake of CPOD was much higher than that at pH 7.4 (p < 0.005 in B16 cells, p < 0.01 in HepG2 cells). Besides, in vitro release study revealed that the drug release amount from CPOD at endo/lysosome pH was significantly more than that at pH 7.4 (p < 0.005), which was further indicated by nuclear localization test. In vivo NIRF imaging illustrated the accumulation in tumor of CPOD for 24 h. Furthermore, in vivo anti-tumor test indicated that CPOD exhibited more superior anti-tumor efficacy even than double dosage DOX·HCl solution. No visible tissue lesions of main organs in CPOD group were observed, indicating preliminary in vivo safety of CPOD. These results suggest that the two-stage pH-sensitive CPOD would be a promising delivery platform for improving anti-tumor efficacy of DOX·HCl with a fair degree of safety.

---

1. Introduction

Cancer has become the most serious life-threatening disease in the world, with more than 10 million new cases every year.^1,2 As a common treatment (chemotherapy, radiotherapy and surgery), chemotherapy is widely employed in for many cancers. However, its success has been limited by several drawbacks, such as low efficacy and toxic side effects, which are mainly attributed to the non-selectivity biodistribution of drug in the whole body.^3,4 For improving chemotherapy efficacy, achieving tumor target and site-specific drug release in tumor, appropriate drug delivery systems are greatly required.⁵ In recent years, nano-drug delivery systems (NDDSs) have been widely developed with notable advantages including biocompatibility, controlled size and shape, and stimuli-sensitive properties for precise spatiotemporal control.⁶ Furthermore, the suitable size (50–200 nm) of NDDSs enables them to passively reach tumor target through the enhanced permeability and retention (EPR) effect, which make NDDSs preferentially accumulate in tumor tissues following better anti-tumor efficacy than low molecular weight anti-tumor agents.^7–11

Doxorubicin hydrochloride (DOX·HCl), a type of anthracycline broad-spectrum antineoplastic drug, has been used clinically to treat several kinds of tumors, such as hepatocellular carcinoma,¹² lung cancer¹³ and breast cancer.¹⁴ It's clinical application is often coupled with severe side effects, such as cardiotoxicity and renal damage.^15,16 It has been reported that the anti-tumor efficacy of DOX·HCl depends on both the drug-loading concentrations in NDDSs and the composition of tumor matrix.¹⁷ Therefore, in order to enhance anti-tumor efficacy and overcome side effects of DOX·HCl, NDDSs with higher drug-loading and better drug release capability triggered by tumor microenvironment need to be developed.^18–21

In order to obtain high drug-loading, several researchers have made various efforts, namely. Li et al. have built a (3-aminopropyl)triethoxysilane (APTES)-modified mesoporous silica nanoparticles (MSNs) with high drug-loading, DOX has been efficient encapsulated inside MSNs because of the large surface area and pore volume of MSNs.²² However, MSNs are involved in a complicated process and slow biodegradation rate.²³ In addition, Tan et al. have synthesized a D-α-tocopherol polyethylene glycol 1000 succinate (TPGS)-based prodrug conjugated with DOX, which was based on the high drug-loading efficiency of TPGS.²⁴ DOX was loaded by chemical conjugation. In our work, dual drug-loading strategies, in which deoxyribonucleic acid (DNA) loading method combined with chemical modification were used to achieve a higher drug-loading.

Since Trouet et al. reported the first use of DNA as a carrier for anthracycline drugs in 1972,²⁵ DOX·HCl has been known to intercalate within the DNA strand, especially in 5′-GC-3′ or 5′-CG-3′ sequences due to the presence of flat aromatic rings in its molecule structure.²⁶ Jon et al. has established a hybridized aptamer with 21-base (CGA)₇ oligodeoxynucleotides (CGA-ODNs) extended at 3′ end and verified its high drug loading capacity for DOX·HCl, which improved the loaded DOX·HCl molecules from 0.6 to 7.5 per aptamer.^26,27 Subsequently, many researchers have been reported CGA-ODNs as a DOX·HCl carriers with a high and easily controlled drug loading through modifying the length of -CG- repeat sequence.^28–31

Moreover, chemical modification is one of common drug loading methods of NDDSs by conjugation of drugs to hydrophilic or amphiphilic polymers through chemical bonds.^32,33 The benefits of chemical conjugates, include high drug loading and accurate loading efficiency.^34–37 To enable triggered responsiveness inside tumors, drugs are usually conjugated via special linkages which are designed according to the tumor microenvironments, such as pH, redox, enzymes and temperature field.^38,39 Among the stimuli-responsive bonds, pH-sensitive bonds including hydrazone,⁴⁰ cis-acotinyl,⁴¹ and acetal²¹ are the most commonly used due to the distinct pH gradients among normal tissues (pH 7.4), tumor extracellular environments (about pH 6.5)^42,43 as well as endosome (pH 5.0–6.0) and lysosome (pH 4.0–5.0) compartments.^44–46 By taking advantages of the gradual pH decrease in tumor tissues and endo/lysosome, designing a two-stage pH-sensitive nanoparticles, which can respond to the pH changes step by step is necessary for accurate intracellular drug release control.

Herein, intracellular drug release is expected because it can reduce drug loss and increase intracellular drug concentration.⁴⁷ Moreover, DOX·HCl, as a kind of topoisomerase inhibitor,⁴⁸ intracellular release is beneficial to nuclear localization for its anti-tumor efficacy. For effective intracellular drug release, nanoparticles must first traverse the cell membranes. Since cell membranes are negatively charged, positively charged nanoparticles show higher affinity for cell membrane and help bypass the cellular barriers.⁴⁹ However, cationic nanoparticles have strong non-specific cellular uptake in the bloodstream and interact strongly with serum components, which cause severe aggregation and rapid clearance from circulation.^50,51 Therefore, it will be useful to create NDDSs that are negatively charged during blood circulation, but positively charged after accumulated in tumor tissues.^52,53 Core–shell nanoparticles have been considered to be a promising strategy.

Branched poly(ethylenimine) (PEI) with high density of dissociative amine groups has extensive use in biomedical applications, particularly for oligodeoxynucleotides and DNA delivery based on the highly condensing ability of PEI to them.^54,55 Meanwhile, the dissociative amino groups are suitable for chemical modification. Furthermore, compared with some other cationic polymers, such as chitosan and poly-D-lysine, PEI is more efficient in promoting cell attachment and adsorptive endocytosis through proteoglycans in the cell membrane.^56,57 In the present study, DOX·HCl was conjugated to PEI by hydrazone bonds through a heterobifunctional crosslinking agent C6-S-HyNic (C6-succinimidyl 6-hydrazinonicotinate acetone hydrazone, C6-SANH) to form an intracellular pH-sensitive cationic conjugates PEI-C6-SANH-DOX·HCl (PEI-C-DOX). Hydrazone bond appears to be more stable at pH 7.4 than other pH-sensitive bond while showing excellent pH-sensitivity.⁵⁸ On the other hand, DOX·HCl loaded CGA-ODNs (CGA-ODNs/DOX·HCl, OD) was obtained by easily incubation at room temperature. There were no organic reagents incorporated in the preparation process, which was of safety and non-immunogenicity. PEI-C-DOX could further condense negatively charged OD to obtain PEI-C-DOX/OD (POD). Herein, the DOX·HCl loading in POD could be increased through the dual drug-loading strategies, i.e. chemical modification combining with CGA-ODNs loading method. Afterwards, O-carboxymethyl-chitosan (CMCS) is a suitable choice of shell material for coating on POD, which is a biodegradable pH-sensitive amphiprotic polyelectrolyte with good biocompatibility.^59,60 Scheme 1A and B show preparation and functionalization of CPOD, respectively. In blood circulation and normal tissues (pH 7.4), CMCS was expected to retain negative charge and coated on POD, keep CPOD stable and prolong the circulation time, thus relieve side effects of DOX·HCl caused by drug leakage. After CPOD accumulated in tumor tissues via EPR effect, the outer CMCS shell dissociated at the subacid microenvironment (about pH 6.5) and POD was exposed to facilitate the tumor cellular uptake. Upon internalization, DOX·HCl was pH-sensitively released due to both the hydrolysis of hydrazone bonds and the deactivation of CGA-ODNs under acid endo/lysosome environment (about pH 5.0), allowing nuclear accumulation of released DOX·HCl and which resulted in DOX·HCl toxicity in target cells. The effective drug delivery and controlled intracellular drug release made CPOD a potential NDDS to achieve greater anti-tumor efficacy with increased safety.

Scheme 1 The preparation of two-stage pH-sensitive doxorubicin hydrochloride loaded core–shell nanoparticles CPOD with dual drug-loading strategies (A). Enhanced accumulation of CPOD in tumor tissues by EPR effect, and two-stage pH-sensitive function of CPOD (B).

In this study, cationic intracellular pH-sensitive PEI-C-DOX was synthesized, and ¹H nuclear magnetic resonance (¹H NMR) spectroscopy was used to characterize the successful synthesis. The appropriate molar ratio of CGA-ODNs to DOX·HCl was tested using a fluorescence spectrofluorometer. The proper weight ratios of PEI-C-DOX to OD were evaluated by agarose gel electrophoresis assay, and the optimal formulations of POD and CPOD were chosen after further physical and chemical properties evaluation. The two-stage pH-sensitive of CPOD was confirmed. Firstly, the pH-triggered dissociating of CMCS from CPOD at subacid pH was confirmed in vitro, and pH-sensitive CPOD internalization were tested under different pH conditions using B16 and HepG2 cell lines, respectively. Secondly, the intracellular pH-sensitive DOX·HCl release was proved in vitro, and nuclear localization of released DOX·HCl was tested in B16 and HepG2 cell lines, respectively. The cytotoxicities of CPOD, DOX-free blank nanoparticles and PEI were evaluated in both the two cell lines. Furthermore, tumor accumulation of CPOD was confirmed using a real-time near infrared fluorophore (NIRF) imaging system. The in vivo anti-tumor efficacy of CPOD was evaluated in murine hepatoma cell line H22 tumor-bearing Kunming (KM) mice model. The plasma stability, hemolytic tests and histological analysis were used to investigate the preliminary safety of CPOD.

2. Materials and methods

2.1. Materials, cells and animals

Branched poly(ethyleneimine) (PEI, M_w 25 kDa) was purchased from Sigma-Aldrich (USA). C6-S-HyNic (C6-succinimidyl 6-hydrazinonicotinate acetone hydrazone, C6-SANH) was purchased from Solulink, Inc. (San Diego, CA, USA). Doxorubicin hydrochloride (DOX·HCl) was purchased from Dalian Meilun Biology Technology Co., Ltd (Dalian, China). O-Carboxymethyl-chitosan (CMCS) (average M_w 50 kDa; degree of carboxymethyl substitution = 60%; degree of deacetylation = 85%) was obtained from Jinan Haidebei Biological Engineering Co. (Jinan, China). 21-base (CGA)₇ oligodeoxynucleotides (CGA-ODNs) (sequence: 5′-CGACGACGACGACGACGACGA-3′; complementary sequence: 5′-TCGTCGTCGTCGTCGTCGTCG-3′) were purchased from BGI tech (Shenzhen, China). GoldView was purchased from Beijing Saibaisheng Biological Engineering Co., Ltd (Beijing, China). 3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) was purchased from Sigma-Aldrich (USA). Hoechst 33 342 was obtained from Invitrogen (USA). Sulfo-Cyanine5 carboxylic acid (Cy5) was obtained from LITTLE-PA Sciences Co., Ltd. (Wuhan, China). Cy5 labeled CGA-ODNs (Cy5-CGA-ODNs) (sequence: 5′-Cy5-CGACGACGACGACGACGACGA-3′; complementary sequence: 5′-TCGTCGTCGTCGTCGTCGTCG-3′) were purchased from BGI tech (Shenzhen, China).

Human hepatoma cell line HepG2, murine melanoma cell line B16 and murine hepatoma cell line H22 were kindly provided by Institute of Immunopharmacology and Immunotherapy of Shandong University (Jinan, China). HepG2 cell line was cultured in Dulbecco's Modified Eagle's Medium (DMEM) medium, and B16 cell line was cultured in Roswell Park Memorial Institute (RPMI) 1640 medium supplemented with 10% FBS at 37 °C and 5% CO₂.

Female Kunming mice weighing 18–22 g were supplied by Medical Animal Test Center of the New Drugs Evaluation Center, Shandong University. All experiments were carried out in compliance with the Animal Management Rules of the Ministry of Health of the People's Republic of China (document number 55, 2001) and the Animal Experiment Ethics Review of Shandong University. All procedures were approved by the Animal Care and Use Committee.

2.2. Synthesis and characterization of PEI-C-DOX

As illustrated in Scheme 2, synthesis of PEI-C-DOX was divided into two steps. The synthesis route was shown in Scheme 2.

Scheme 2 Synthesis route of the PEI-C-DOX.

2.2.1. Synthesis of PEI-C6-SANH (PEI-C). C6-SANH was conjugated to PEI by amide bonds. C6-SANH (10 mg) was dissolved in anhydrous DMSO (1 mL), and PEI (40 mg) was dissolved in phosphate buffer solution (PBS, 150 mM sodium chloride, 100 mM sodium phosphate, pH 7.4). The solution of C6-SANH and PEI were mixed in a balloon flask (50 mL) under stirring, and pH 7.4 of PBS was added into the flask up to 20 mL. The reaction solution was stirred for 24 h at room temperature in the environment of nitrogen. The obtained product was purified by dialysis against distilled water (MWCO 3500), lyophilized, the chemical structure was confirmed by ¹H nuclear magnetic resonance (¹H NMR) spectroscopy (AvanceTM DPX-300, Bruker BioSpin GmbH, Rheinstetten, Germany) in D₂O.

2.2.2. Synthesis of PEI-C-DOX. PEI-C-DOX was synthesized by the reaction of hydrazinopyridine groups of PEI-C to C-13 ketone groups of DOX·HCl in order to form pH-sensitive hydrazone bonds. PEI-C (35 mg) and DOX·HCl (20 mg) were dissolved in PBS (150 mM sodium chloride, 100 mM sodium phosphate, pH 6.0), respectively.

The solutions were mixed in a balloon flask (50 mL) under stirring, and pH 6.0 of PBS was added into the flask up to 20 mL. The reaction solution was stirred for 24 h at room temperature in dark and environment of nitrogen. The obtained product was purified by dialysis against distilled water (MWCO 3500), lyophilized, the chemical structure was confirmed by ¹H NMR spectroscopy in D₂O and infrared spectroscopy (IR) (6700 FT-IR NXR FT-RAMAN, Nicolet, USA).

2.2.3. Determination of the drug loading capacity of PEI-C-DOX. The amount of DOX·HCl loaded on PEI-C-DOX was determined by UV-Vis spectrophotometry. PEI-C-DOX was synthesized in triplicate and accurately weighted to prepare 0.1 mg mL⁻¹ solution with PBS (0.01 M, pH 7.4). The absorbance of the solution at 480 nm was detected by a UV-Vis spectrophotometer (UV-2102PCS, UNICOTM, USA). The DOX·HCl concentration was calculated via a calibration curve prepared previously with different concentrations of DOX·HCl in PBS (0.01 M, pH 7.4) (A = 0.0164C − 0.0167, r = 0.9994). The drug loading capacity was calculated using the following equation, and the results were expressed as the mean ± SD (n = 3):

2.3. Preparation and characterization of OD

DOX·HCl was dissolved in double distilled water (ddH₂O), then DOX·HCl and CGA-ODNs were mixed and incubated at room temperature to obtain OD. To confirm the intercalating molar ratio of DOX·HCl to CGA-ODNs, CGA-ODNs (50 μM) was added stepwise to a fixed concentration of DOX·HCl (2.5 μM) in ddH₂O, and the fluorescence intensity of OD was monitored at an excitation wavelength (excitation) of 493 nm and emission wavelength (emission) from 520 nm to 700 nm using a fluorescence spectrofluorometer (F-7000, HITACHI, Japan). The drug loading capacity of OD was calculated using the following equation:

The drug loading capacities of POD and CPOD were calculated using the following equation, respectively, and the results were expressed as the mean ± SD (n = 3):

To observe a possible fluorescence recovery of DOX·HCl, the OD was placed at 4 °C for 24 h. The fluorescence intensity of OD was measured at the same condition.

2.4. Preparation and characterization of CPOD

2.4.1. Preparation of POD. Cationic and pH-sensitive PEI-C-DOX was dissolved and diluted to the corresponding concentrations with ddH₂O. OD was freshly prepared as above. To obtain proper weight ratios of PEI-C-DOX to OD, an equal volume of OD was added to various concentrations of PEI-C-DOX solution under vortex. The mixture system was incubated at room temperature for 30 min to form dual methods DOX·HCl loaded nanoparticles POD.

An agarose gel electrophoresis assay was used to evaluate the weight ratios between PEI-C-DOX and OD. POD prepared at various weight ratios of PEI in PEI-C-DOX to CGA-ODNs (339 : 4096, 339 : 2048, 339 : 1024, 339 : 512, 339 : 256, 339 : 128, 339 : 64, 339 : 32, 339 : 16) was mixed with an appropriate amount of 6× loading buffer and electrophoresed on a 1% (w/v) agarose gel containing an appropriate amount of GoldView (TAE buffer, 90 V, 10 min). The electrophoretic mobility of POD was visualized using a UV transilluminator and digital imaging system (IS-2200, Alpha Innotech, USA). The particle size of POD was determined using dynamic light scattering (DLS), (Zetasizer Nano ZS-90, Malvern, UK) at 25 °C.

2.4.2. Preparation of CPOD. A selected optimal POD was used to prepare the two-stage pH-sensitive nanoparticles CPOD. CMCS was dissolved and diluted to the corresponding concentrations with PBS (0.01 M, pH 7.4). POD was freshly prepared and then dropped into equal volume various concentrations of CMCS solution under vortex. The mixture system was incubated at room temperature for 30 min to form CPOD.

Similarly, the mixture binding conditions for CPOD with different weight ratios of CMCS to CGA-ODNs (75 : 128, 75 : 64, 75 : 32, 75 : 16, 75 : 8, 75 : 4, 75 : 2, 75 : 1) were evaluated by agarose gel electrophoresis assay with the same method. The study was conducted to examine whether the addition of CMCS led to the decomposition of the complex. The zeta potentials of CPOD were determined using DLS at 25 °C.

2.4.3. Characterization of POD and CPOD. The optimal POD and CPOD were carried out in triplicate. The size distributions and zeta potentials were determined using DLS at 25 °C, respectively. The results were expressed as the mean ± SD (n = 3). The micromorphologies of POD and CPOD were visualized by a transmission electronic microscopy (TEM) (JEM-1200EX, Japan) after negative staining with one drop of 3% aqueous solution of sodium phosphotungstate for contrast enhancement.

2.5. In vitro evaluation of firstly/extracellular pH-sensitive

2.5.1. In vitro pH-sensitive dissociation of CMCS. To confirm the pH-sensitive dissociation of CMCS, 200 μL CPOD was incubated with 800 μL of PBS (0.01 M, pH 7.4, 7.0, 6.5, 6.0, 5.5, 5.0) at 37 °C for 2 h. The experiment was carried out in triplicate. After incubation, zeta potentials were determined using DLS at 25 °C, respectively.

2.5.2. In vitro pH-sensitive cellular uptake. To evaluate the pH-sensitive cellular uptake of CPOD, DMEM or RPMI 1640 media with different pH were prepared to mimic normal tissues pH (pH 7.4) and tumor tissues pH (pH 6.5 and pH 6.0), respectively. B16 or HepG2 cells were seeded in a twelve-well plate at a density of 1.5 × 10⁵ cells per well. After cultured overnight, when the cells reached about 80% confluence, fresh media of pH 7.4, pH 6.5 and pH 6.0 including free DOX·HCl, POD or CPOD were added into each well, respectively. The final concentration of DOX·HCl was 2.0 μg mL⁻¹. After incubation for 3 h, cells were thoroughly washed three times with cold PBS and then fixed with 4% paraformaldehyde for 15 min at room temperature. Excess paraformaldehyde was removed by rinsing with sterile PBS, then visualized using a reverse fluorescence microscope (ZX71, Olympus, Japan).

To quantify the cellular uptake, the cells were treated as above. After washing three times with cold PBS, all the cells were trypsinized and washed with cold PBS, then the cells were resuspended in 100 μL of PBS after centrifugated. The quantified cellular uptake was carried out in triplicate and detected using a flow cytometer (FACSCaliber, Becton Dickinson, USA). The results were expressed as the mean ± SD (n = 3).

2.6. In vitro evaluation of secondly/intracellular pH-sensitive

2.6.1. In vitro pH-sensitive DOX·HCl release. In vitro pH-sensitive DOX·HCl release was performed at two different pH conditions to mimic normal physiological pH (pH 7.4) and endo/lysosome acidic pH (pH 5.0) using a dialysis bag diffusion technique, respectively. Furthermore, in order to confirm the respective pH-sensitive release of DOX·HCl from PEI-C-DOX or OD, nanoparticles based on PEI-C-DOX conjugation (CMCS/PEI-C-DOX/CGA-ODNs, CPO) and nanoparticles based on OD complex (CMCS/PEI/OD, COD) were prepared as the same method of CPOD, respectively. 200 μL of freshly prepared CPOD, CPO or COD was placed into pre-swelled dialysis bag (MWCO 3500) that was incubated in 5 mL of pre-warmed PBS buffer at 37 ± 0.5 °C under horizontal shaking (100 rpm), respectively. At desired time intervals (0.5 h, 1 h, 2 h, 4 h, 6 h, 8 h, 12 h, 24 h, 48 h), the release media was removed and replaced with 5 mL fresh PBS. The amount of released DOX·HCl was determined by a fluorescence spectrophotometer with the excitation set at 493 nm and emission set at 560 nm. The release experiment was carried out in triplicate. The DOX·HCl concentration was calculated via the calibration curves prepared previously with different concentrations of DOX·HCl in PBS of pH 5.0 and 7.4, respectively (A = 0.519C + 11.2, r = 0.9993, pH 5.0, A = 0.491C + 15.4, r = 0.9995, pH 7.4). The release rate was calculated using the following equation, and the results were expressed as the mean ± SD (n = 3):Where W_n was the accumulated DOX·HCl release mass; V was the volume of the release media; C_n was the DOX·HCl concentration in the release media at each time point; and W was the total DOX·HCl content of the release sample.

2.6.2. Nuclear localization. The intracellular pH-sensitive DOX·HCl release was further confirmed by nuclear localization study. The B16 or HepG2 cells were treated as described in Section 2.5.2. After incubation for 3 h (with DOX·HCl, POD and CPOD) or 4 h (with POD and CPOD), cells were thoroughly washed three times with cold PBS and then fixed with 4% paraformaldehyde. The cell nuclei were then stained with Hoechst 33342 (2.5 μg mL⁻¹) during a 15 min incubation following standard protocols. The incubated cells were washed three times to remove excess Hoechst 33342, then visualized using a reverse fluorescence microscope.

2.7. Cytotoxicity assay

In vitro antitumor efficiency of CPOD was evaluated by MTT assay using B16 and HepG2 cells, respectively. B16 or HepG2 cells were seeded into 96-well plates at a density of 5 × 10³ and cultured for 24 h at 37 °C before the assay, respectively. Different concentrations of free DOX·HCl, CPOD, PEI and blank nanoparticles (CMCS/PEI/CGA-ODNs) prepared as the same method of CPOD were added to five wells each concentration and incubated for 48 h, respectively. Following that, 20 μL MTT (5 mg mL⁻¹) was added to each well and the cells were incubated for another 4 h at 37 °C. The media were removed and then 150 μL DMSO was added to each well to dissolve the formazan crystals formed by living cells. Cells without treatment were used as control. Cell culture media (RPMI 1640 or DMEM) without cells was used as blank. The absorbance at a test wavelength of 570 nm and a reference wavelength of 630 nm in each well was recorded using a Microplate Reader (Elx800, BioTek, USA). The MTT assay was carried out in triplicate. Cell viability was calculated according to the following equation, and the results were expressed as the mean ± SD (n = 3):Where abs (sample), abs (blank) and abs (control) referred to the absorbance of the samples, blank and control, respectively.

2.8. Plasma stability and hemolytic tests

In order to investigate the stability and safety of CPOD in blood circulation, and prepared for the following in vivo experiments. Plasma stability and hemolytic tests were conducted.

2.8.1. Plasma stability test. The plasma was obtained by centrifuged at 3000 rpm for 10 min of fresh rat blood after heparin added. 10% plasma was selected for the test. POD or CPOD were freshly prepared, and then 500 μL POD or CPOD were added into isovolumetric 20% plasma and stored at 37 °C. At desired time intervals (0 h, 1 h, 2 h, 4 h, 6 h, 8 h, 24 h, 48 h), the size distributions were determined using DLS at 25 °C. The test was conducted in triplicate and the results were expressed as the mean ± SD (n = 3).

2.8.2. Hemolytic test. The fresh rat blood was stirred to remove fibrinogen. Red blood cells (RBCs) were obtained after centrifugation and washed with normal saline. The RBCs suspension (2%, v/v) was then mixed with CPOD according to Table 1. Deionized water was added as a positive control, and saline solution was added for the negative control group. After incubated at 37 ± 0.5 °C for 3 h, the mixture was centrifuged at 3000 rpm for 10 min to remove erythrocytes, and the supernatant was collected and analyzed for released hemoglobin by a UV-Vis spectrophotometer at 577 nm as an indication of RBC lysis. The test was processed in triplicate. The haemolysis ratio (HR%) was calculated as follows, and the results were expressed as the mean ± SD (n = 3):Where abs (sample), abs (−) and abs (+) referred to the absorbance of the samples, negative control and positive control, respectively.

Table 1 The adding methods of hemolytic test

	Number
	1	2	3	4	5	6 (−)	7 (+)
CPOD (mL)	0.1	0.2	0.3	0.4	0.5	—	—
Normal saline (mL)	2.4	2.3	2.2	2.1	2.0	2.5	—
Distilled water (mL)	—	—	—	—	—	—	2.5
2% RBCs solution (mL)	2.5	2.5	2.5	2.5	2.5	2.5	2.5
Concentration of DOX·HCl (μg mL^−1)	2	4	6	8	10

---

2.9. In vivo biodistribution of CPOD

The healthy female Kunming (KM) mice (age of six weeks; body weight of 18–22 g) were obtained from Medical Animal Test Center of the New Drugs Evaluation Center, Shandong University.

The tumor-bearing mice model was established by inoculating subcutaneously of H22 hepatic cancer cells (2 × 10⁷) in the right flank of each mouse. Water-soluble dye free Cy5 was used to mimic the biodistribution of free DOX·HCl as a positive control, CPOD prepared by Cy5-CGA-ODNs was used as experimental group. When tumor volume reached about 100 mm³, the tumor-bearing mice were administrated with free Cy5 and CPOD prepared with Cy5-CGA-ODNs through tail vein at a dosage of 62.5 nmol kg⁻¹, respectively. At desired time intervals (1 h, 2 h, 4 h, 8 h, 24 h), the mice were imaged using a real-time near infrared fluorophore (NIRF) imaging system (IVIS Kinetic, Cailper Life Science, USA) at appropriate wavelength (Ex 640 nm, Em 694 nm) after narcosis. After 24 h, the mice were sacrificed, tumors and main organs including heart, liver, spleen, lung, and kidney were excised and imaged using a NIRF imaging system. The fluorescence intensities of tumors and main organs were then quantitated and analyzed by Living Image 3.1 software (Cailper Life Science, USA).

2.10. In vivo anti-tumor efficacy

The tumor-bearing KM mice model was established as the same method in Section 2.9. Then the tumor-bearing mice (n = 24) were randomly and equally divided into four groups: normal saline group, free DOX·HCl group (1 mg kg⁻¹ and double dosage 2 mg kg⁻¹) and CPOD group (1 mg kg⁻¹). When the tumor volumes reached about 100 mm³, mice in the four groups were administrated with the formulations through tail vein everyday (total for six times), respectively. The tumor dimensions were measured with a vernier caliper on day 1, 3, 5, 7, 9, 11, and the tumor volume was calculated using the formula:Where width and length referred to the widest dimension and the longest dimension, respectively.

In addition, body weights of each mice were recorded everyday. On day 11, each group of mice were sacrificed, and the tumors were excised, following weighting.

2.11. Histological analysis

The healthy female KM mice were administrated with the formulations through tail vein at the same dosage and frequency as mentioned in Section 2.10. On day 11, each group of mice was sacrificed, the organs including heart, liver, spleen, lung, and kidney were excised and fixed with formalin, embedded in paraffin, and cut into 5 μm thick sections for hematoxylin and eosin (H&E) staining. Images were observed using a light microscope.

3. Results and discussion

3.1. Synthesis and characterization of PEI-C-DOX

In order to obtain the feature of dual drug-loading, branched PEI was selected as the material both for chemical conjugation and OD condensing. A heterobifunctional crosslinking agent, C6-SANH, was used to prepare the PEI-C-DOX with intracellular pH-sensitive hydrazone bonds. C6-SANH helped to avoid the complex synthesis steps, harsh reaction conditions and high amounts of organic solvents in hydrazone bond synthesis process. Thus, the synthesis of PEI-C-DOX took advantages of simple operation and mild reaction conditions, which was divided into two steps, and synthesis route was as shown in Scheme 2.

3.1.1. Synthesis of PEI-C. PEI-C was synthesized and characterized by ¹H NMR. The ¹H NMR spectra with attributed peaks were as shown in Fig. 1A and B. The peaks at δ 2.7–2.5 ppm were assigned to protons of PEI (Fig. 1A). As shown in Fig. 1B, after synthesis of PEI-C, the three peaks at δ 8.4–6.7 ppm (f, h, g) were assigned to protons of the pyridine. The peaks at δ 1.6–0.8 ppm (b, c, d) were attributed to protons of the methylenes in C6-SANH. The methylenes peaks of (a and e) were sheltered by peaks of PEI. It determined that PEI-C was successful synthesized.

Fig. 1 ¹H NMR spectra of PEI (A), PEI-C (B), DOX·HCl (C) and PEI-C-DOX (D) (in D₂O). IR spectra of PEI (E), PEI-C (F), DOX·HCl (G) and PEI-C-DOX (H).

3.1.2. Synthesis of PEI-C-DOX. In order to characterize the structure of PEI-C-DOX, the ¹H NMR spectra with attributed peaks were as shown in Fig. 1C and D. The peaks at δ 7.6–7.2 ppm (b′, c′, d′), δ 5.382 ppm (h′), δ 4.2–4.1 ppm (l′), δ 3.808 ppm (a′), δ 3.8–3.7 ppm (g′), δ 3.7–3.5 ppm (j′), δ 2.9–2.4 ppm (e′), δ 2.3–1.9 ppm (f′), δ 1.915 ppm (i′) and δ 1.3–1.2 ppm (m′) were assigned to protons of DOX·HCl (Fig. 1C). After synthesis of PEI-C-DOX (Fig. 1D), the peaks at δ 7.5–7.4 ppm (b′, c′, d′), δ 4.1–3.9 ppm (l′), δ 3.9–3.8 ppm (g′), δ 3.8–3.7 ppm (j′), δ 2.2–2.1 ppm (f′) and δ 1.3–1.2 ppm (m′) were assigned to protons of DOX·HCl in PEI-C-DOX. Meanwhile, the peaks of pyridine protons and methylene protons in C6-SANH were also appeared at δ 8.2–7.7 ppm (f, h) and δ 1.6–0.8 ppm (b, c, d). The structure of PEI-C-DOX was confirmed.

PEI-C-DOX was further verified by IR spectroscopy. As shown in Fig. 1G and H, the carbonyl peak of DOX·HCl at 1730 cm⁻¹ was disappeared after linked to PEI-C. In addition, although the peaks of several functional groups appeared around 1600 cm⁻¹, such as amide, carbon–carbon double bond, benzene and hydrazone bond, the peak at 1652 cm⁻¹ (Fig. 1H) was most likely attributed to the absorption of hydrazone bond. Because as shown in Fig. 1F and H, the intensity of absorption peak at around 1600 cm⁻¹ (1633 cm⁻¹ in Fig. 1F and 1652 cm⁻¹ in Fig. 1H) was obviously increased after conjugation of DOX·HCl (Fig. 1H). It indicated that PEI-C conjugated to the carbonyl of DOX·HCl via hydrazone bond, and confirmed the successful synthesis of intracellular pH-sensitive PEI-C-DOX.

3.1.3. Determination of the drug loading capacity of PEI-C-DOX. The amount of DOX·HCl loaded in PEI-C-DOX conjugates determined by a UV-Vis spectrophotometry (λ = 480 nm) was 9.15 ± 0.265% (weight ratio of DOX·HCl to PEI-C-DOX).

3.2. Preparation and characterization of OD

The preparation of OD was easily operated by incubation at room temperature, and did not induce any other modifications of the drug or CGA-ODNs in the process. It has been reported that the fluorescence of DOX·HCl will quench after intercalated into CGA-ODNs.^61,62 As shown in Fig. 2A, the fluorescence intensity of DOX·HCl decreased progressively along with the increase of CGA-ODNs after incubation with CGA-ODNs, suggesting the successful intercalation. When the molar ratio of CGA-ODNs to DOX·HCl was higher than 0.12 : 1, there was no significant change of fluorescence intensity. It demonstrated that DOX·HCl was totally intercalated into CGA-ODNs with a high loading capacity of 8.33 DOX·HCl molecules per CGA-ODNs. Therefore, in the present work, the molar ratio was fixed at 0.12 : 1 for the following experiment. The drug loading capacity of OD was 27.3%, and the drug loading capacities of POD and CPOD were 16.2 ± 0.977% and 4.72 ± 0.09%, respectively. To investigate the store stability of OD, OD at molar ratio of 0.12 : 1 was placed at 4 °C for 24 h and the emission spectrum was as shown in Fig. 2B, and fluorescence of DOX·HCl was hardly recovered indicating the stability of the intercalation.

Fig. 2 Fluorescence emission spectra of OD with different molar ratios of CGA-ODNs to DOX·HCl (A). Fluorescence emission spectra of OD stored at 4 °C for 24 h (B).

3.3. Preparation and characterization of CPOD

3.3.1. Preparation of POD. POD was prepared based on the electrostatic interaction between PEI-C-DOX to OD, thus the change in charge density of PEI-C-DOX after drug conjugation was considered, which affected its ability to condense OD. To eliminate the tiny disparities of charge density caused by the differences of DOX·HCl modification rate, the weight ratios of PEI-C-DOX to CGA-ODNs were converted into PEI to CGA-ODNs. Therefore, POD was prepared at various weight ratios of PEI to CGA-ODNs from 339 : 4096 to 339 : 16 and agarose gel electrophoresis assay was carried out to determine the OD binding ability of PEI-C-DOX. As shown in Fig. S1A,† at low weight ratios of PEI to CGA-ODNs, a fraction of migration-free OD could still be visualized. When the weight ratio of PEI to CGA-ODNs was up to 339 : 256, free OD could not be observed, indicating that OD was completely condensed by PEI-C-DOX. Subsequently, the size distributions of POD with weight ratios between 339 : 256 to 339 : 16 were measured as shown in Fig. S1B.† For mass ratios ≤339 : 512, POD could not be accurately analyzed due to incomplete complex formation and too broad complex size ranges. When the weight ratio of PEI to CGA-ODNs increased from 339 : 256 to 339 : 64, the average sizes of POD were 140–150 nm. In this phase, the mean size of POD was hardly changed, the binding ability between PEI-C-DOX to already formed POD may be described as an exponential function.⁶³ Then, an increase size of POD was observed in the weight ratio from 339 : 32 to 339 : 16, indicating that the POD suspension became unstable. At the weight ratio of 339 : 64, a more symmetrical size distribution with a small polydispersity index (PDI) was observed, and some excessive positive charges provided by PEI-C-DOX was necessary for electrostatic crosslinking with CMCS. In summary, 339 : 64 was chosen as the optimal weight ratio for the following experiments.

3.3.2. Preparation of CPOD. CMCS was chosen to shelter the positive charge of POD and achieve the tumor extracellular pH-sensitivity. To investigate whether the addition of CMCS can decompose POD by competitively releasing OD from PEI-C-DOX, an agarose gel electrophoresis assay was utilized to measure the stability of POD. The agarose gel electrophoresis results with various amounts of CMCS (weight ratios of CMCS to CGA-ODNs from 75 : 128 to 75 : 1) were shown in Fig. S2A.† At the test weight ratios of CMCS to CGA-ODNs, no free OD was observed, indicating that CMCS could not affect the stability of POD. Additionally, as shown in Fig. S2B,† the addition of negatively charged CMCS could result in surface charge (zeta potential) variations. Along with the weight ratios increasing, the zeta potentials of CPOD decreased from (20.4 ± 1.9) mV and finally reached a plateau of about −15 mV. The charges overturn from positive to negative indicating that anionic CMCS was coated on cationic POD at pH 7.4. The plateau reflection point of 75 : 4 (CMCS to CGA-ODNs) was selected as the ideal weight ratio for the following experiments.

3.3.3. Characterization of POD and CPOD. The mean particle sizes, polydispersity index (PDI) and zeta potentials of three batches of POD and CPOD were measured using DLS at 25 °C, and the results were as summarized in Table 2. In the preparation of CPOD, CMCS was dissolved in pH 7.4 of PBS, which could keep CMCS negatively charged in order to coat on cationic POD. The average size of CPOD was increased owing to the coating of CMCS. The zeta potential of POD was (20.4 ± 1.9) mV, conforming to that of reported PEI/DNA complex which was >20 mV,²¹ while, because the shielding effect of CMCS, the zeta potential of CPOD decreased to −(15.6 ± 0.82) mV. The increased size and reversed surface charge confirmed the fabrication of CPOD. The suitable particle size of CPOD [(165.0 ± 3.3) nm] allowed for a more efficient translocation through the tumor angiogenesis and accumulation in tumor tissues due to the EPR effect.^7–10 The negatively charged CPOD could avoid non-specific binding to serum components in blood circulation and prolong the circulation time. Meanwhile, the high absolute value of the zeta potential [−(15.6 ± 0.82) mV] tended to stabilize particle suspension, because the electrostatic repulsion between particles with the same electric charges prevented the aggregation of the spheres.

Table 2 Summary of average sizes, polydispersity index and zeta potentials of POD and CPOD (n = 3)

	Average size (nm)	PDI	Zeta potential (mV)
POD (w/w = 339 : 64)	140.2 ± 5.2	0.144 ± 0.0182	20.4 ± 1.9
CPOD (w/w = 75 : 4)	165.0 ± 3.3	0.191 ± 0.0097	−15.6 ± 0.82

---

The TEM picture and size distributions were as shown in Fig. 3A and B. Both POD and CPOD displayed a spherical morphology with a narrow distribution, and the results of PDI (<0.2 both for POD and CPOD) showed in Table 2 was further indication of the characteristic colloidal suspension reliable disparity.

Fig. 3 Morphology and size distribution of POD (w/w = 339 : 64) (A) and CPOD (w/w = 75 : 4) (B).

3.4. In vitro evaluation of firstly/extracellular pH-sensitivity

3.4.1. In vitro pH-sensitive dissociation of CMCS. In order to evaluate the pH-sensitive dissociation of CMCS, CPOD was incubated with different pH of PBS (0.01 M) at 37 °C for 2 h. After incubation, zeta potentials were determined as depicted in Fig. 4. As shown, the zeta potentials were nearly unchanged in simulative normal physiological pH (pH 7.4 and pH 7.0). However, in the simulative tumor subacid extracellular pH (pH 6.5 and pH 6.0), the zeta potentials of CPOD increased gradually. In the more acid pH of PBS (pH 5.5 and pH 5.0), zeta potentials reversed to positive. It's attributed to the fact that CMCS is a kind of amphiprotic polyelectrolyte. When pH was higher than its isoelectric point (pI, about pH 6.5), CMCS was negatively charged and remained coating on POD to keep CPOD stable. When pH reduced below pI, the amino groups of CMCS would be protonated resulting in the charges of negative CMCS reversed to positive. Finally CMCS dissociated from POD because of the electrostatic repulsion.^64,65 In summary, CMCS shielding could keep CPOD stable in normal physiological pH, while, in the tumor extracellular pH, exposed positive POD which was benificial for cellular internalization.

Fig. 4 Zeta potentials of CPOD after incubation with different pH of PBS (0.01 M) at 37 °C for 2 h.

3.4.2. In vitro pH-sensitive cellular uptake. To further prove the firstly pH-sensitive function of CPOD, which is exposed POD internalized by tumor cells, in vitro pH-sensitive cellular uptake experiments were evaluated in B16 and HepG2 cell lines, respectively. The micrographs took by a reverse fluorescence microscope were as shown in Fig. 5A and B. The red fluorescence (DOX·HCl) confirmed the DOX·HCl internalization in both B16 and HepG2 cell lines. As visualized, the DOX·HCl red fluorescence of CPOD group was enhanced along with the decrease of culture media pH, while, at pH 6.5 and pH 6.0, it was similar to that of POD and DOX·HCl solution, and the fluorescence intensity of POD group was similar to that of DOX·HCl solution group at each pH. To quantify the cellular uptake, the results detected by flow cytometer were shown in Fig. 5C and D. The uptake of CPOD in B16 at pH 7.4 was 36.3 ± 2.71%, significantly less than that at pH 6.5 (83.5 ± 1.21%) and pH 6.0 (94.8 ± 0.15%), (p < 0.005). Similarly, in HepG2 cell lines, the uptake of CPOD at pH 7.4 was 59.5 ± 4.96%, much lower than that at pH 6.5 (99.7 ± 0.29%) and pH 6.0 (99.7 ± 0.43%), (p < 0.01).

Fig. 5 Fluorescence micrographs of B16 cell line (A) and HepG2 cell line (B) treated with DOX·HCl solution, POD and CPOD at pH 7.4, pH 6.5 and pH 6.0 for 3 h, respectively. Red fluorescence: DOX·HCl, the scale bars represent 50 μm (200×). Flow cytometric analysis of B16 cell line (C) and HepG2 cell line (D) treated with DOX·HCl solution, POD and CPOD at pH 7.4, pH 6.5 and pH 6.0 for 3 h, respectively. **p < 0.01, ***p < 0.005.

At pH 7.4, in both B16 and HepG2 cell lines, POD displayed similar internalization to DOX·HCl solution, which were all close to 100%, indicating the strong cellular uptake ability of POD. However, less uptake of CPOD than POD at pH 7.4 was observed, while, at pH 6.5 and pH 6.0 the cellular uptake of CPOD was no significant differences to that of POD and DOX·HCl solution at pH 7.4. It may be because in normal physiological environment (pH 7.4), the negatively charged CMCS coating made CPOD difficult to approach the same electronegative cytomembranes and caused little internalization of CPOD.^66–68 Meanwhile, in tumor subacid extracellular pH (pH 6.5 and pH 6.0), the outermost anionic layer, CMCS, may completely fall off and then exposed to the cationic POD. The exposed positive POD enhanced the cellular uptake of the payloads.

3.5. In vitro evaluation of secondly/intracellular pH-sensitivity

3.5.1. In vitro pH-sensitive DOX·HCl release. In vitro drug release experiments were carried out in pH 7.4 (pH of normal physiological environment) and pH 5.0 (pH of endo/lysosome), respectively. To confirm the pH-sensitive release of DOX·HCl from PEI-C-DOX and OD, the results of free drug release and DOX·HCl release from CPO and COD were shown in Fig. 6A. The release of DOX·HCl solution at pH 5.0 had no significant difference to that at pH 7.4. The cumulative release percentage was up to 90% after 4 h at both pH 5.0 and pH 7.4. As for CPO, after 48 h, 92.84 ± 2.98% DOX·HCl was released at pH 5.0, which was up to a 60% enhancement release at acidic pH (30.43 ± 3.69% of DOX·HCl was released from CPO at pH 7.4 under the same conditions, p < 0.005). The pH-sensitive DOX·HCl release from CPO was mainly caused by the cleavage of hydrazone bonds under acidic condition. Meanwhile, similar result was obtained in COD, DOX·HCl release at pH 5.0 (70.39 ± 0.97%, 48 h) was more than that at pH 7.4 (48.36 ± 0.86%, 48 h), (p < 0.05). It has been reported that, CGA-ODNs was not stable and would be degraded in endo/lysosome, which promoted DOX·HCl release in acidic environment.⁶⁹ Besides, under acidic conditions, the primary amine group in DOX·HCl would protonate and the solubility of DOX·HCl would be increased, thereby, further enhancing the release of DOX·HCl.²⁷ It is known that deactivation of CGA-ODNs is due to the pH-sensitive DOX·HCl release from COD. However, at beginning, DOX·HCl release from CPO and COD at pH 5.0 was consistent and synchronous. After 2 h, the release of DOX·HCl from COD retarded, while DOX·HCl in CPO was sustained release, which in turn was sufficient to prolong the action time of DOX·HCl and enhance anti-tumor activity. It was shown that, CPO exhibited more pH-sensitive than COD. The reason may be that the hydrazone bonds in CPO were completely cleaved at pH 5.0 and almost all DOX was released. Although the deactivation of CGA-ODNs in COD may cause longer time, some DOX·HCl still retained in CGA-ODNs which was not degraded.

Fig. 6 In vitro DOX·HCl release from DOX·HCl solution, CPO and COD (A); DOX·HCl solution and CPOD (B) at pH 5.0 and pH 7.4 of PBS buffers (0.01 M), respectively, *p < 0.05, ***p < 0.005.

As shown in Fig. 6B, the amount of DOX·HCl released from CPOD was both time- and pH-dependent. After 48 h, 73.89 ± 2.91% DOX·HCl was released from CPOD at pH 5.0, which was significantly more than that at pH 7.4 (p < 0.005, about 40.44 ± 1.27% of DOX·HCl was released from CPOD at pH 7.4 under the same conditions). These results revealed that intracellular pH-sensitive CPOD would rapidly release DOX·HCl in response to endo/lysosomal pH. It confirmed that CPOD was an ideal platform for constructing intracellular pH-responsive system.

3.5.2. Nuclear localization. DOX·HCl is a topoisomerase inhibitor which results in damage to DNA replication, RNA transcription, and protein synthesis.⁴⁸ In order to obtain effective anti-tumor activity, it is necessary for CPOD to achieve intracellular drug release and nuclear localization of released DOX·HCl. The nuclear localization was carried out for 3 h and 4 h at various pH using a reverse fluorescence microscope, respectively. As shown in Fig. 7A and B, the red fluorescence (DOX·HCl) confirmed that DOX·HCl was internalized, the blue fluorescence were cell nuclei stained with Hoechst 33342, and the merged purple fluorescence confirmed that DOX·HCl located in cell nuclei. After 3 h incubation, both in B16 and HepG2 cell lines, for free DOX·HCl (pH 7.4), significant purple fluorescence was observed in the nuclei, indicating the nuclear loacalization of free DOX·HCl.

Fig. 7 Fluorescence micrographs of B16 cell line (A) and HepG2 cell line (B) treated with DOX·HCl solution, POD and CPOD at pH 7.4, pH 6.5 and pH 6.0 for 3 h and 4 h, respectively. Red fluorescence: DOX·HCl, blue fluorescence: nuclei stained with Hoechst 33342, the scale bars represent 50 μm (200×).

Although, most of red fluorescence accumulated in cytoplasm of both cell lines treated with POD (at pH 7.4) and CPOD (at pH 6.5 and pH 6.0) and much weaker purple fluorescence was showed in the nuclei. It may be attributed to that internalization of free DOX·HCl was due to efficient passive diffusion,⁷⁰ while the cellular uptake of POD (at pH 7.4) and CPOD (at pH 6.5 and 6.0) as well as intracellular pH-sensitive DOX·HCl release needed longer time. On the other hand, B16 and HepG2 cell lines were treated with POD and CPOD for a longer incubation time (4 h), respectively. After 4 h, there was stronger purple fluorescence in nuclei of both cell lines treated with POD (at pH 7.4) and CPOD (at pH 6.5 and 6.0) than that at 3 h, indicating much more DOX·HCl was intracellular released and located in nuclei. For CPOD, no purple fluorescence was observed at pH 7.4 of both cell lines, while at pH 6.5 (of both cell lines) and pH 6.0 (of B16 cell line) the intensity of purple fluorescence was similar to that of POD. At pH 6.0 of HepG2 cell line, the intensity of purple fluorescence was much stronger than others, further indicating the extracellular pH-sensitive function. The effective nuclear localization of CPOD at tumor extracellular pH (pH 6.5 and 6.0) after 4 h demonstrated two-stage pH-sensitive function, which had first high cellular uptake at tumor extracellular pH and second intracellular pH-sensitive DOX·HCl release.

3.6. Cytotoxicity assay

The in vitro therapeutic efficacy of CPOD was tested using B16 and HepG2 cell lines and compared with DOX·HCl solution as a positive control. Both DOX·HCl and CPOD exhibited a dosage-dependent cytotoxic effect in B16 and HepG2 cells (Fig. 8A and B, respectively). In B16 cell line, the (half maximal inhibitory concentration) IC50 of DOX·HCl solution was 0.277 ± 0.031 μg mL⁻¹ while IC50 of CPOD was 1.11 ± 0.085 μg mL⁻¹. DOX·HCl solution showed higher cytotoxicity than CPOD (p < 0.05). Meanwhile, in HepG2 cell lines, the cytotoxicity of CPOD was similar to DOX·HCl solution (IC50 of DOX·HCl solution and CPOD was 3.82 ± 0.51 μg mL⁻¹ and 3.66 ± 0.14 μg mL⁻¹, respectively). The results can be explain by different sensibility of various cell lines to DOX·HCl. In addition, the internalization of CPOD and DOX·HCl release from CPOD needed a longer time than passive diffusion of DOX·HCl solution, thus the in vitro cytotoxicity of CPOD was reduced at the test concentration range.

Fig. 8 Cell viability of DOX·HCl solution and CPOD (A and B), blank nanoparticles (C and D) and PEI (E and F), the (A), (C) and (E) were carried out in B16 cell line and (B), (D) and (F) were carried out in HepG2 cell line.

The cytotoxicity of nanocarriers is an important factor that should be taken into consideration. As shown in Fig. 8C and D, after incubation with blank nanoparticles for 48 h, the cell viabilities were higher than 90% at all experimental dosage for both B16 and HepG2 cells, indicating the low toxicities of blank nanoparticles. Meanwhile as shown in Fig. 8E and F, high cytotoxicity of PEI was observed in both the two cells (the cell viabilities were lower than 40% at all experiment dosages). This may be attributed to the fact that the positive charges of PEI was shielded by the anionic shell, CMCS, which was able to improve the viability of the cells, thus preliminary proving the in vitro safety of the bland nanoparticles.

3.7. Plasma stability and hemolytic tests

3.7.1. Plasma stability test. As shown in Fig. 9A, after incubation with 10% plasma, CPOD showed stable size within 24 h (p > 0.05), and to 48 h, the increase of size was observed (p < 0.05).

Fig. 9 The average size of POD and CPOD after incubation with 10% plasma (A). The image of hemolysis test and hemolysis percentage of CPOD (B), *p < 0.05.

However, the size of POD had abrupt increase (p < 0.05), and after 2 h, macroscopic sediments were observed. It may be caused by the existence of substances in plasma, such as proteins. The plasma proteins were easy to absorb on the surface of positive POD, which caused the change of particle size and formation of coagulations.^71,72 CMCS coating could relieve the adsorption and make CPOD stable in blood circulation.

3.7.2. Hemolytic test. As shown in Fig. 9B, no hemolysis of negative control tube (number 6) was observed, and the positive control tube (number 7) showed hemolysis phenomenon. For CPOD, in the range of setting concentrations (2–10 μg mL⁻¹), no hemolysis or aggregation phenomena were observed in the tubes of number 1 to 5. In addition, the calculated HRs of CPOD were all less than 1%. Generally, a hemolysis ratio of < 5% is regarded as nontoxic and safe.⁷³ The results indicated that CPOD had good hemocompatibility under preliminary experimental conditions, and CPOD were suitable for intravenous injection administration.

3.8. In vivo biodistribution of CPOD

The tumor target and tissue biodistribution of CPOD was evaluated by the in vivo biodistribution study using H22 tumor-bearing KM mice model. As shown in Fig. 10A, after 1 h of administration through tail vein, fluorescence signal was observed in both free Cy5 and CPOD groups. The distribution of free Cy5 was non-selective, at 24 h, no fluorescence signal was observed indicating the elimination of free Cy5. However, CPOD accumulated in tumor within 1 h and were distinctly observed up to 24 h post-injection at the tumor site. At the same time, the live accumulation of CPOD was observed from 1 h to 8 h, and CPOD was also shown in spleen at 8 h. The liver or spleen accumulation pattern of CPOD was highly representative of charged nanoparticles, widely known to undergo a sequestration in the major organs of the reticuloendothelial system (RES), mainly in the liver and spleen.⁷⁴ The effective tumor target of CPOD was based on EPR effect, which was due to the suitable particle size of CPOD [(165.0 ± 3.3) nm]. This caused an efficient translocation through the tumor angiogenesis and accumulation in tumor tissues. The fluorescence signal of CPOD in bladder indicated the renal excretion of CPOD. At 24 h, free Cy5 was already eliminated, but CPOD still accumulated in tumor and no fluorescence signal in other tissues was observed. Subsequently, the tumor and major organs including heart, liver, spleen, lung, and kidney were collected and imaged after the mice were sacrificed at 24 h post-injection. The image is shown in Fig. 10B. After quantitation, the fluorescence intensity of tumor in CPOD group was 2.261 × 10⁷, which was 3.46 times stronger than that of free Cy5 group (6.533 × 10⁶). It further demonstrated the effective tumor accumulation of CPOD. Simultaneously, CPOD showed a high liver accumulation. The fluorescence intensity of liver (3.676 × 10⁸) was 7.95 times stronger than that of free Cy5 group (4.624 × 10⁷), which was consistent with the in vivo imaging results. The fluorescence intensity of kidney in CPOD group (3.211 × 10⁷) was 4.03 times stronger than that of free Cy5 group (7.964 × 10⁶), further indicating the renal excretion of CPOD.

Fig. 10 NIRF images of H22 tumor-bearing KM mice after administrated through tail vein of free Cy5 and CPOD prepared by Cy5-CGA-ODNs at desired time intervals, respectively. The tumor sites were marked by the black cycle (A). NIRF images of tumors and main organs including heart, liver, spleen, lung, and kidney of free Cy5 and CPOD prepared by Cy5-CGA-ODNs at 24 h, respectively (B).

In profile, the retention of CPOD in tumor tissues observed in both the in vivo and ex vivo imaging was beneficial for anti-tumor efficacy. It was also helpful in elimination in other tissues, which improved medicinal safety of CPOD.

3.9. In vivo anti-tumor efficacy

Anti-tumor efficacy of CPOD was assessed on H22 tumor-bearing KM mice. The changes of tumor volumes, tumor weights and body weights of the mice were as shown in Fig. 11. After 11 days of treatment, the average tumor volume of mice treated with normal saline reached about 600 mm³ (Fig. 11A). For CPOD group, the average tumor volume was about 100 mm³, which was significantly smaller than that of 1 mg kg⁻¹ DOX·HCl group (p < 0.005) and double dosage 2 mg kg⁻¹ DOX·HCl group (p < 0.005). Thereafter, the mice were sacrificed, and then the solid tumors were excised. As seen in Fig. 11B, the tumors of CPOD group were smaller than tumors of other groups. Upon weighing of the tumors (Fig. 11C), the average tumor weight of mice treated with CPOD was 146.8 ± 76.8 mg, which was significantly lighter than that of 1 mg kg⁻¹ DOX·HCl group (p < 0.005) and 2 mg kg⁻¹ DOX·HCl group (p < 0.05). Overall, in contrast to the lower cytotoxicity in vitro, CPOD showed preliminarily greater in vivo anti-tumor efficacy than free DOX·HCl (even better than double dosage of DOX·HCl solution) in H22 tumor-bearing KM mice model. The result could be explained as follows. Firstly, the in vivo biodistribution of DOX·HCl was changed after loaded into CPOD, CPOD was demonstrated to target tumor tissue and accumulate in it for a longer time in Section 3.8. Secondly, in the Sections 3.4 and 3.5, the internalization in tumor subacid microenvironment and controlled intracellular pH-sensitive drug release of CPOD was confirmed, and it was beneficial to improve the anti-tumor efficacy of CPOD.

Fig. 11 Antitumor efficacy of evaluated using H22 tumor-bearing KM mice. Tumor volumes (A), tumor image (B), tumor weights (C) and body weights (D) of normal saline, 1 mg kg⁻¹ DOX·HCl, 2 mg kg⁻¹ DOX·HCl and 1 mg kg⁻¹ CPOD, respectively. *p < 0.05, ***p < 0.005.

The body weight changes were monitored for reflecting the adverse effects of various DOX·HCl formulations. As shown in Fig. 11D, all groups exhibited similar smooth fluctuations of body weight (p > 0.05), suggesting weight loss caused by systemic toxicities wasn't revealed in the test.

3.10. Histological analysis

To further evaluate the safety of CPOD on main organs, histological evaluation effect was conducted, and administration was the same with in vivo anti-tumor efficacy test. The H&E staining histological sections of main organs including heart, liver, spleen, lung and kidney were shown in Fig. 12. The heart, liver, spleen, lung and kidney of CPOD group showed normal histological morphology, and no visible difference was observed in comparison with the negative control (normal saline) group. However, for 1 mg kg⁻¹ and 2 mg kg⁻¹ DOX·HCl solution groups, there was vacuolation degeneration in the cells of hearts, livers, and inflammatory cells infiltration in the lungs. Moreover, the heart, liver and lung tissues damage and inflammation of 2 mg kg⁻¹ DOX·HCl solution group was more serious than that of 1 mg kg⁻¹ DOX·HCl solution group, indicating that DOX·HCl solution may have a dosage-dependent toxicity on main organs caused by non-specific in vivo biodistribution. In the in vivo biodistribution test, CPOD showed liver accumulation, however lesions were not observed in the liver section. It may be that CPOD could keep stable in normal tissues and seldom DOX·HCl was released. It preliminarily indicated that CPOD was a potential platform for DOX·HCl loading to improve medicinal safety.

Fig. 12 H&E staining histological sections of main organs from KM mice (100×).

4. Conclusion

In conclusion, a two-stage pH-sensitive DOX·HCl loaded core–shell nanoparticles with dual drug-loading strategies aiming at improving antitumor efficacy and medicinal safety were simultaneously designed and investigated. DOX·HCl was loaded in inner core POD by dual drug-loading strategies of chemical modification combining with CGA-ODNs loading method, and thereafter extracellular pH-sensitive CMCS was coated on POD to construct the core–shell nanoparticles CPOD. Both POD and CPOD showed spherical and uniform size distribution. CMCS could shelter the positively charged POD, stailize and prolong blood circulation time of CPOD. CPOD was observed to accumulate in tumor tissues, which lasted for 24 h. Meanwhile, at 24 h the ex vivo fluorescence intensity of tumor in CPOD group was 3.46 times stronger than that of free drug group. After enriched in tumor tissues, CPOD played the two-stage pH-sensitive functions that were firstly pH-sensitive internalization and secondly intracellular pH-sensitive DOX·HCl release, which finally achieved the cytotoxicity of nuclear localizated DOX·HCl. More importantly, CPOD displayed a good curative efficacy on the inhibition of tumor growth with a fair degree of safety. This approach provides a new platform with effective DOX·HCl delivery and controlled drug release to improve anti-tumor efficacy and improve medicinal safety.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (No. 81402867). A special thanks to Dr Livesey Olerile for his support of English grammar corrections.

Notes and references

1. J. Chen, W. Zhang, M. Zhang, Z. Guo, H. Wang, M. He, P. Xu, J. Zhou, Z. Liu and Q. Chen, Nanoscale, 2015, 7, 12542–12551.
2. Q. Cheng, L. Du, L. Meng, S. Han, T. Wei, X. Wang, Y. Wu, X. Song, J. Zhou, S. Zheng, Y. Huang, X. J. Liang, H. Cao, A. Dong and Z. Liang, ACS Appl. Mater. Interfaces, 2016, 8, 4347–4356.
3. H. Mcilwain, Nature, 1943, 151, 270–273.
4. Z. Khatun, M. Nurunnabi, M. Nafiujjaman, G. R. Reeck, H. A. Khan, K. J. Cho and Y. K. Lee, Nanoscale, 2015, 7, 10680–10689.
5. J. Liu, B. Zhang, Z. Luo, X. Ding, J. Li, L. Dai, J. Zhou, X. Zhao, J. Ye and K. Cai, Nanoscale, 2015, 7, 3614–3626.
6. D. B. Pacardo, F. S. Ligler and Z. Gu, Nanoscale, 2015, 7, 3381–3391.
7. V. Torchilin, Adv. Drug Delivery Rev., 2011, 63, 131–135.
8. D. Chao, Z. Zhong, Y. Jiang, C. Ru and F. Meng, Nano Today, 2012, 7, 467–480.
9. F. Danhier, O. Feron and V. Preat, J. Controlled Release, 2010, 148, 135–146.
10. N. Bertrand, J. Wu, X. Xu, N. Kamaly and O. C. Farokhzad, Adv. Drug Delivery Rev., 2014, 66, 2–25.
11. Y. J. Son, J. S. Jang, Y. W. Cho, H. Chung, R. W. Park, I. C. Kwon, I. S. Kim, J. Y. Park, S. B. Seo, C. R. Park and S. Y. Jeong, J. Controlled Release, 2003, 91, 135–145.
12. H. Li, Y. Cui, J. Sui, S. Bian, Y. Sun, J. Liang, Y. Fan and X. Zhang, ACS Appl. Mater. Interfaces, 2015, 7, 15855–15865.
13. S. Lv, Z. Tang, M. Li, J. Lin, W. Song, H. Liu, Y. Huang, Y. Zhang and X. Chen, Biomaterials, 2014, 35, 6118–6129.
14. T. Wei, C. Chen, J. Liu, C. Liu, P. Posocco, X. Liu, Q. Cheng, S. Huo, Z. Liang, M. Fermeglia, S. Pricl, X. J. Liang, P. Rocchi and L. Peng, Proc. Natl. Acad. Sci. U. S. A., 2015, 112, 2978–2983.
15. J. Lu, W. Zhao, Y. Huang, H. Liu, R. Marquez, R. B. Gibbs, J. Li, R. Venkataramanan, L. Xu, S. Li and S. Li, Mol. Pharm., 2014, 11, 4164–4178.
16. Y. F. Song, D. Z. Liu, Y. Cheng, M. Liu, W. L. Ye, B. L. Zhang, X. Y. Liu and S. Y. Zhou, Sci. Rep., 2015, 5, 16125.
17. C. Fang, F. M. Kievit, O. Veiseh, Z. R. Stephen, T. Wang, D. Lee, R. G. Ellenbogen and M. Zhang, J. Controlled Release, 2012, 162, 233–241.
18. J. Yu, Y. Zhang, W. Sun, C. Wang, D. Ranson, Y. Ye, Y. Weng and Z. Gu, Nanoscale, 2016, 8, 9178–9184.
19. J. Xu, Q. Zhao, Y. Jin and L. Qiu, Nanomedicine, 2014, 10, 349–358.
20. J. Guo, H. Hong, G. Chen, S. Shi, Q. Zheng, Y. Zhang, C. P. Theuer, T. E. Barnhart, W. Cai and S. Gong, Biomaterials, 2013, 34, 8323–8332.
21. C. Liu, F. Liu, L. Feng, M. Li, J. Zhang and N. Zhang, Biomaterials, 2013, 34, 2547–2564.
22. Y. Wang, Y. Sun, J. Wang, Y. Yang, Y. Li, Y. Yuan and C. Liu, ACS Appl. Mater. Interfaces, 2016, 8, 17166–17175.
23. Y. Chen, H. Chen and J. Shi, Adv. Mater., 2013, 25, 3144–3176.
24. Y. Bao, M. Yin, X. Hu, X. Zhuang, Y. Sun, Y. Guo, S. Tan and Z. Zhang, J. Controlled Release, 2016, 235, 182–194.
25. A. Trouet, D. D. Campeneere and C. De Duve, Nature, 1972, 239, 110–112.
26. V. Bagalkot, O. C. Farokhzad, R. Langer and S. Jon, Angew. Chem., Int. Ed. Engl., 2006, 45, 8149–8152.
27. D. Kim, Y. Y. Jeong and S. Jon, ACS Nano, 2010, 4, 3689–3696.
28. M. Chang, C. S. Yang and D. M. Huang, ACS Nano, 2011, 5, 6156–6163.
29. I. H. Lee, M. K. Yu, I. H. Kim, J. H. Lee, T. G. Park and S. Jon, J. Controlled Release, 2011, 155, 88–95.
30. L. Meng, L. Yang, X. Zhao, L. Zhang, H. Zhu, C. Liu and W. Tan, PLoS One, 2012, 7, e33434.
31. T. Liu, M. Wang, T. Wang, Y. Yao and N. Zhang, Colloids Surf., B, 2015, 126, 531–540.
32. J. Kopecek, Adv. Drug Delivery Rev., 2013, 65, 49–59.
33. M. Li, Y. Liu, L. Feng, F. Liu, L. Zhang and N. Zhang, Colloids Surf., B, 2015, 131, 191–201.
34. L. Zhang and N. Zhang, Int. J. Nanomed., 2013, 8, 2927–2941.
35. A. J. Ferdous, N. Y. Stembridge and M. Singh, J. Controlled Release, 1998, 50, 71–78.
36. M. Elsabahy and K. L. Wooley, Chem. Soc. Rev., 2012, 41, 2545–2561.
37. Z. Su, Y. Liang, Y. Yao, T. Wang and N. Zhang, J. Mater. Chem. B, 2016, 4, 1122–1133.
38. Y. Bae, S. Fukushima, A. Harada and K. Kataoka, Angew. Chem., Int. Ed., 2003, 42, 4640–4643.
39. S. Yoon, W. J. Kim and H. S. Yoo, Small, 2013, 9, 284–293.
40. R. Tang, W. Ji, D. Panus, R. N. Palumbo and C. Wang, J. Controlled Release, 2010, 151, 18–27.
41. Z. Y. Qiao, F. S. Du, R. Zhang, D. H. Liang and Z. C. Li, Macromolecules, 2010, 43, 6485–6494.
42. D. M. Prescott, H. C. Charles, J. M. Poulson, R. L. Page, D. E. Thrall, Z. Vujaskovic and M. W. Dewhirst, Clin. Cancer Res., 2000, 6, 2501–2505.
43. C. J. Gulledge and M. W. Dewhirst, Anticancer Res., 1996, 16, 741–749.
44. E. S. Lee, Z. Gao and Y. H. Bae, J. Controlled Release, 2008, 132, 164–170.
45. M. Murakami, H. Cabral, Y. Matsumoto, S. Wu, M. R. Kano, T. Yamori, N. Nishiyama and K. Kataoka, Sci. Transl. Med., 2011, 3, 599–612.
46. S. Ganta, H. Devalapally, A. Shahiwala and M. Amiji, J. Controlled Release, 2008, 126, 187–204.
47. H. Maeda, T. Sawa and T. Konno, J. Controlled Release, 2001, 74, 47–61.
48. T. Li, M. Zhang, J. Wang, T. Wang, Y. Yao, X. Zhang, C. Zhang and N. Zhang, AAPS J., 2016, 18, 146–155.
49. S. E. Gratton, P. A. Ropp, P. D. Pohlhaus, J. C. Luft, V. J. Madden, M. E. Napier and J. M. DeSimone, Proc. Natl. Acad. Sci. U. S. A., 2008, 105, 11613–11618.
50. J. Z. Du, T. M. Sun, W. J. Song, J. Wu and J. Wang, Angew. Chem., Int. Ed., 2010, 49, 3703–3708.
51. Y. Huang, Z. Tang, X. Zhang, H. Yu, H. Sun, X. Pang and X. Chen, Biomacromolecules, 2013, 14, 2023–2032.
52. H. Deng, X. Zhao, J. Liu, J. Zhang, L. Deng, J. Liu and A. Dong, Nanoscale, 2016, 8, 1437–1450.
53. Y. Dong, J. Yang, H. Liu, T. Wang, S. Tang, J. Zhang and X. Zhang, Theranostics, 2015, 5, 890–904.
54. W. T. Godbey, K. K. Wu and A. G. Mikos, J. Controlled Release, 1999, 60, 149–160.
55. S. Kumar, S. Raj, K. Sarkar and K. Chatterjee, Nanoscale, 2016, 8, 6820–6836.
56. A. R. Vancha, S. Govindaraju, K. V. Parsa, M. Jasti, M. González-García and R. P. Ballestero, BMC Biotechnol., 2004, 4, 663–664.
57. F. Chen, J. Zhang, L. Wang, Y. Wang and M. Chen, Nanoscale, 2015, 7, 15763–15779.
58. Y. Bae and K. Kataoka, Adv. Drug Delivery Rev., 2009, 61, 768–784.
59. L. Upadhyaya, J. Singh, V. Agarwal and R. P. Tewari, J. Controlled Release, 2014, 186, 54–87.
60. N. N. Inamdar and V. K. Mourya, Adv. Mater. Lett., 2010, 1, 11–33.
61. H. T. Haj, M. Salerno, W. Priebe, H. Kozlowski and A. Garnier-Suillerot, Chem.-Biol. Interact., 2003, 145, 349–358.
62. L. Valentini, V. Nicolella, E. Vannini, M. Menozzi, S. Penco and F. Arcamone, Farmaco, Ed. Sci., 1985, 40, 377–390.
63. I. Y. Perevyazko, M. Bauer, G. M. Pavlov, S. Hoeppener, S. Schubert, D. Fischer and U. S. Schubert, Langmuir, 2012, 28, 16167–16176.
64. Y. Yao, Z. Su, Y. Liang and N. Zhang, Int. J. Nanomed., 2015, 10, 6185–6198.
65. F. Liu, M. Li, C. Liu, Y. Liu and N. Zhang, J. Biomed. Nanotechnol., 2014, 10, 3397–3406.
66. P. L. Rodriguez, T. Harada, D. A. Christian, D. A. Pantano, R. K. Tsai and D. E. Discher, Science, 2013, 339, 971–975.
67. T. Stylianopoulos, K. Soteriou, F. Dai and R. K. Jain, Ann. Biomed. Eng., 2013, 41, 68–77.
68. T. Stylianopoulos, M. Z. Poh, N. Insin, M. G. Bawendi, F. Dai, L. Munn and R. K. Jain, Biophys. J., 2010, 99, 1342–1349.
69. E. Kim, Y. Jung, H. Choi, J. Yang, J. S. Suh, Y. M. Huh, K. Kim and S. Haam, Biomaterials, 2010, 31, 4592–4599.
70. C. Fang, F. M. Kievit, Y. C. Cho, H. Mok, O. W. Press and M. Zhang, Nanoscale, 2012, 4, 7012–7020.
71. K. Fukukawa, R. Rossin, A. Hagooly, E. D. Pressly, J. N. Hunt, B. W. Messmore, K. L. Wooley, M. J. Welch and C. J. Hawker, Biomacromolecules, 2008, 9, 1329–1339.
72. E. D. Pressly, R. Rossin, A. Hagooly, K. Fukukawa, B. W. Messmore, M. J. Welch, K. L. Wooley, M. S. Lamm, R. A. Hule, D. J. Pochan and C. J. Hawker, Biomacromolecules, 2007, 8, 3126–3134.
73. T. Ramasamy, Z. S. Haidar, T. H. Tran, J. Y. Choi, J. H. Jeong, B. S. Shin, H. G. Choi, C. S. Yong and J. O. Kim, Acta Biomater., 2014, 10, 5116–5127.
74. T. Maldiney, B. Ballet, M. Bessodes, D. Scherman and C. Richard, Nanoscale, 2014, 6, 13970–13976.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra19242d

---