Biomaterials 35 (2014) 1657e1666

Contents lists available at ScienceDirect

Biomaterials journal homepage: www.elsevier.com/locate/biomaterials

p53 mediated apoptosis by reduction sensitive shielding ternary complexes based on disulfide linked PEI ternary complexes Yiyan He, Yu Nie*, Li Xie, Hongmei Song, Zhongwei Gu** National Engineering Research Center for Biomaterials, Sichuan University, No. 29, Wangjiang Road, Chengdu 610064, Sichuan, PR China

a r t i c l e i n f o

a b s t r a c t

Article history: Received 31 August 2013 Accepted 27 October 2013 Available online 20 November 2013

Reduction-sensitive hyaluronic acid derivatives (HAeSSeCOOH) were shielded on the DNA/polyethylenimine (PEI) to construct ternary complexes (DNA/PEI/HAeSSeCOOH, DPS ternary complexes) with efficient gene transfection. Details studied were conducted to investigation of factors influencing transfection efficiency, including the gene compression by fluorescence resonance energy transfer (FRET) spectrum and the intracellular fate of fluorescent labeled complexes by the confocal laser scanning microscope (CLSM). In the FRET study, DPS complexes were found to enhance condensation of DNA in preparation, while timely loosen gene under exposure to reductive reagent. Similar cellular uptake levels were observed for the designed reduction sensitive complexes and the stable one (DNA/PEI/HA, DPH ternary complexes), but the intracellular process was strikingly different for the two types of complexes. Only DPS showed obvious desired intracellular deshielding and endosomal escape, which contributed to highly efficient gene delivery. After loading with p53 plasmid, DPS complexes achieved significantly upregulated p53 tumor suppressor gene expression at both mRNA and protein levels, as revealed by quantitative polymerase chain reaction (qPCR) and western blot investigations. Transgene induced apoptosis was evaluated by propidium iodide staining and flow cytometry analysis of cell cycle. Tumor cells transfected by DPS complexes containing p53 gene displayed almost 50% higher suppression in proliferation compared to those untreated cells, accompanied with a 46% elevation in the number of cells at sub-G1 phase and remarkable p53 dependent cell cycle perturbations prior to apoptosis. These results demonstrated that targeted delivery of p53 gene via reduction-sensitive DPS ternary complexes enabled up-regulated cellular p53 mRNA level through the exogenous p53 gene, inducing a significant p53dependent anti-proliferative effect on tumor cells, which could be effective means of cancer treatment. Ó 2013 Elsevier Ltd. All rights reserved.

Keywords: Reduction-sensitive Shielding Intracellular p53 Gene therapy Cell apoptosis

1. Introduction With the development in molecular biology, the mechanisms of cancers have been found to be related with the activation of protooncogenes or the deactivation of tumor suppressor genes during DNA replication. Once the genetic errors accumulate to some extent, cells proliferation goes out of control, and the cancer formation occurs [1]. These mutated genes are thought to be good targets for gene therapy. Among them, one of the most important human tumor suppressor genes is p53 and its mutations are present in more than half of all malignancies [2]. It has been reported that enhancing activity of wild-type p53 can mediate a variety of anti-proliferative effects, including suppression of

* Corresponding author. Tel.: þ86 28 85415928. ** Corresponding author. Tel.: þ86 28 85410336; fax: þ86 28 85410653. E-mail addresses: [email protected] (Y. Nie), [email protected] (Z. Gu). 0142-9612/$ e see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.biomaterials.2013.10.073

oncogene-mediated transformation, induction of a growth arrest in the G1 phase of cell cycle, and activation of apoptotic cell death [3]. Thus, induction of wild-type p53 expression would be a promising approach for cancer treatment in gene therapy. However, the lack of safe and effective gene transfection carriers has hindered the application. Based on safety concerns, non-viral vectors are gaining extensive attention as an alternative to viral vectors for gene delivery [4e6], and most of them are composed of cationic lipids and polymers. Polyethyleneimine (PEI) is the most widely used cationic polymers as golden standard, attributing to its protection of DNA from enzymatic degradation in a biologic milieu, facilitation of gene transfer into cells through endocytosis and endosomal escape by proton sponge effect, buffering and membrane lytic capacity [7]. However, the inherent shortcomings of PEI and other cationic vectors lie in their highly positive charge. It could cause adsorption of anionic serum proteins and other extracellular matrices [8], destroy the stability of gene complexes, block the cellular interaction, accelerate the reticuloendothelial system removal and consequently lead to

1658

Y. He et al. / Biomaterials 35 (2014) 1657e1666

Acronyms HA hyaluronic acid HAeSSeCOOH hyaluronic acid with disulfide bonds modification between the carboxyl group and the backbone DP binary complexes of DNA and PEI DPH ternary complexes of DNA, PEI and HA DPS ternary complexes of DNA, PEI and HAeSSeCOOH

serious inhibition of gene transfection. A typical approach to this problem is to shield the charge by neutral [6,9] or anionic polymers [10e14] (such as PEG, transferrin and alginate). While such shielding may weaken the interaction between gene carriers and targeted cells [15], reduce the endosomal escape ability and incur other adverse consequences [16]. Stimuli-responsive shieldings are designed to overcome these undesired consequences. The strategies of direct cleavage, reduction of negative charge and charge reversion [17] responding to extratumoral or intracellular conditions (such as pH [10,13,18e20], reductive reagents [21e23], and enzymes [24]) could efficiently lead to the shielding detachment (deshielding) at the desired locations. According to this concept, we have devised polyethyleneimine based ternary complexes (DNA/PEI/HAeSSeCOOH, DPS), employing a series of reduction-responsive hyaluronic acid derivatives (HAeSSeCOOH) [25] as polyanion shielding. Disulfide linkages were designed to be cleavable in reductive environment (1e10 mM GSH). Hyaluronic acid was selected as the shielding, for it is a natural anionic polysaccharide present in the extracellular matrix and has been approved by Food and Drug Administration (FDA) for injection. Meanwhile the over expressed HA receptors (CD44 etc.) on various tumor cell surface [26] may contribute to targeted delivery [24,27e30]. In our previous study, even the optimized formula showed several orders of magnitude and 10-fold higher transfection in vitro and in vivo, respectively, the function and process of shielding remained to be defined. More details about the integrity of DNA against DNase, GSH and lyophilization treatment, compression status of loaded gene, cellular uptake and intracellular behaviors of complexes, as well as the intracellular fate of the shielding itself need study. In addition, although significant improvement was obtained with these well-designed carriers in exogenous reporter gene transfection, it is not necessarily the same case for a therapeutic one. Because, unlike the reporter gene, a therapeutic gene exhibits definite biological functions, and the expression is determined not only by the gene delivery system, but also by a series of intracellular signal pathways. To better characterize our DPS formulations in the context of cancer gene therapy, a plasmid containing wild-type p53 tumor suppressor gene was chosen (Scheme 1). The performance of gene transfection was evaluated at mRNA and protein expression levels, and the therapeutic effect was determined by cancer cell apoptosis and cellular cycle analysis. 2. Materials and methods 2.1. Chemicals and reagents Sodium hyaluronate (HA, MW 1.6 MDa) was purchased from Aladdin (China). Ethylenediamine (EDA), Hydroxybenzotriazole (HOBt), 1-ethyl-3-(3dimethylaminopropyl) carbodiimide (EDC), cysteine hydrochloride, branched PEI 25 kDa, OEI 800 Da, propidium iodide (PI) and fluorescein isothiocyanate (FITC) were from SigmaeAldrich (USA). The modified hyaluronic acid derivative (HAeSSeCOOH) with 30% disulfide content (molar ratio) was synthesized according to our recent report [25]. Hoechst 33258 was purchased from Aldrich. Nucleic acid labeling kit Label ITÒ Cy5Ô and Cy3Ô were commercially available from Mirus Bio Corporation

(USA). Alexa FluorÒ 488 carboxylic acid, succinimidyl ester, TRIzolÒ reagent and LysoTracker Blue DND-22 were from Invitrogen (Germany). MTS (3-(4, 5dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium, inner salt) cell proliferation kit was purchased from Promega (USA). RNasefree DNase I and revertAid H Minus First Strand cDNA Synthesis Kit were from MBI-Fermentas (St Leon-Rot, Germany). SYBRÒ Premix Ex TaqÔ II Kit was from TaKaRa. Rabbit polyclonal anti-p53 antibody was obtained from Santa Cruz Biotechnology (USA) and goat anti-rabbit IgG/HRP was from Zhongshan Golden Bridge (Beijing, China). Cell Cycle and Apoptosis Analysis Kit were obtained from Beyotime (China). Dulbecco’s modified Eagle’s medium with high glucose (DMEMHG) and fetal bovine serum (FBS) were from Life Technologies Corporation (GibcoÒ, USA). The human hepatoma cell line HepG2 and murine melanoma cell line B16F10 were obtained from Shanghai Institutes for Biological Sciences (China). BCA Protein Assay Kit was purchased from Pierce (USA). HBG buffer (HEPES 20 mM, pH 7.4, 5% glucose), 2-(N-morpholino) ethanesulfonic acid buffer (MES, 0.1 M, pH 5.0) and other buffers were prepared in MilliQ ultrapure water and filtered (0.22 mm) prior to use, and all the other chemicals were purchased from Aldrich and used as received. 2.2. Assembly of gene complexes DNA/PEI (DP) binary complexes were prepared by gently mixing DNA and PEI solution at N/P ratio of 10 in HBG buffer (HEPES 20 mM, pH 7.4, 5% glucose) and incubated at room temperature for 20 min. HA or HAeSSeCOOH were then added to the DP complexes solution to fabricate ternary complexes of DPH or DPS, respectively. The weight ratio of shielding (HA or HAeSSeCOOH)/DNA ranged from 0.5 to 5, and most DPS complexes used in experiments were prepared at shielding/DNA ratio of 1 unless otherwise specified. 2.3. Stability of complexes in the presence of GSH and DNase The stability of a series DPS ternary complexes in reductive environment and in the presence of deoxyribonuclease (DNase) was evaluated by the agarose gel retardation assay. DPS complexes were incubated with 5 mM glutathione (GSH) at 37  C for 1 h or 166 U/mL DNase for 15 min, followed by addition of EDTA buffer to quench the enzyme activity. The resulted mixture was optionally incubated with 4 mg/mL heparin for 2 h prior to the electrophoresis. After electrophoresis, the gel was stained with ethidium bromide, and the DNA migration patterns were analyzed on the Molecular Imager ChemiDoc XRSþ (Bio-Rad, USA). 2.4. Lyophilization and rehydration of DPS complexes For longer storage and clinical applications, all the DPS complexes were plunged into liquid nitrogen for rapid freeze and lyophilization. The characterization of rehydrated complexes was monitored by inverted fluorescence microscope (LEICA DMI 4000B, Germany) with Hoechst 33258 labeled DNA [31]. 2.5. Compression study of loaded DNA The pGL3 plasmid was dual-labeled with Cy3 and Cy5 fluorescence using the Label ITÒ Kit to monitor FRET spectrum of the gene complexes [14]. According to manufacturer’s specifications, the average density of fluorescent dyes is one dye molecule per 380 DNA base pairs. Various complexes (DPS, DPH and DP complexes) were prepared using 3 mg of Cy3 and Cy5 dual-labeled plasmid in total volume of 200 mL. The complexes were incubation with GSH at final concentration of 5 mM at 37  C for 2 h or kept untreated. The fluorescence emission spectrum was acquired with a Spectrophotometer (F-7000 FL, Hitachi). Excited with 543 nm, the emission signals were collected stepwise from 550 to 750 nm at the step length of 5 nm and at the bandwidth of 5 nm. 2.6. Cell culture Human hepatoma cell line HepG2 and murine melanoma cell line B16F10 were propagated to confluence in DMEM-HG medium supplemented with 100 IU/mL penicillin, 100 mg/mL streptomycin, and 10% fetal bovine serum (FBS). Cells were maintained at 37  C in a humidified atmosphere of 5% CO2 and harvested from flasks with 0.25% trypsin and 0.03% EDTA. 2.7. Cellular uptake of complexes HepG2 cells were seeded at a density of 2  104 cells per well in 35 mm confocal dish (F ¼ 15 mm). After cell attachment reaching 80% confluence, the medium was replaced with fresh culture medium containing 20% Cy5-labeled gene complexes (300 ng DNA per well), followed by 0.5, 1, and 2 h incubation, respectively. The cells were then washed twice with PBS (pH 7.4) and fixed with 4% formaldehyde. The intracellular localization of gene complexes was observed using confocal laser scanning microscope (CLSM, Leica TCS SP5), in which the excitation and emission wavelengths were 633 nm and 670 nm, respectively. To quantify the amount of gene complexes uptake, cells were washed twice with PBS, harvested using trypsin, washed with PBS containing 3% FBS, and re-suspended in PBS for the flow cytometry analysis (Cytomics FC-500, Beckman Coulter). At least 1  104 gated cells were analyzed using Kaluza Flow Cytometry software 1.1. Untreated cells were used in parallel as a negative control.

Y. He et al. / Biomaterials 35 (2014) 1657e1666

1659

Scheme 1. Schematic of DNA/PEI/HAeSSeCOOH (DPS) ternary complexes for p53 delivery.

2.8. Intracellular fate of DPS complexes For the intracellular trafficking experiments, the reversible HAeSSeCOOH shielding and stable HA shielding were labeled with a fluorescent dye (Alexa FluorÒ 488). More details in synthesis and structures confirmation were given in the Supporting Information (Figure S1 and S2). The gene complexes were prepared with Cy5-labeled plasmid and Alexa 488 labeled HA derivatives. HepG2 cells were plated in a 35 mm dish with a glass window at a density of 2  104 cells. After overnight culture, the medium was replaced with fresh medium containing DPH and DPS complexes. At desired time points (1, 2 and 3 h), cells were washed twice with PBS and fixed with 4% formaldehyde. Microscopic analysis was performed with a confocal laser scanning microscope (CLSM). Signal of Alexa 488 labeled HA derivatives was excited with a 488 nm laser and the emission was collected using a 505 nm long-pass filter. In addition, the subcellular localization (endosomal/lysosomal escape) of various complexes (DP, DPS and DPH) was also observed. Different from the shielding trafficking, gene complexes were prepared with Cy5-labeled plasmid and FITC-labeled PEI, while the lysosomes in cytoplasm were stained by LysoTracker Blue DND-22 (InvitrogenÔ Molecular ProbesÔ, USA). At pre-arranged incubation time intervals (1 and 4 h), the cells were washed by ice cold PBS buffer thrice and stained with 75 nM LysoTracker at 37  C for 30 min. Subsequently, the cells were washed again and observed immediately by CLSM. Fluorescent probes of LysoTracker Blue, FITC and Cy5 were excited at 373 nm, 495 nm and 633 nm, respectively.

2.9. Transfection and evaluation of p53 tumor suppressor gene at mRNA and protein levels Gene complexes were prepared with the wild-type or mutant p53 tumor suppressor gene according to the same process as pGL3, and incubated with HepG2 cells for 4 h in the 10% serum containing or serum-free medium. After indicated time periods, expression of p53 in the transfected cells was evaluated at the mRNA level by quantitative PCR (qPCR) technique, using b-actin gene as an internal control. Total RNA was extracted from cells using the TRIzolÒ reagent (Invitrogen) and treated with RNase-free DNase I (MBI-Fermentas, St Leon-Rot, Germany). Quantity and quality of RNAs were evaluated by spectrophotometry (Nanodrop 2000, Thermo) and agarose electrophoresis. For reverse transcription (RT), the first strand cDNA was synthesized from 5 mg of total RNA using a commercial revertAid H Minus First Strand cDNA Synthesis Kit (Fermentas) according to the manual’s instructions. Obtained cDNA was used as a template for qPCR amplification and designed according to the MIQE guidelines. qPCR was done in triplicate using SYBRÒ Premix Ex TaqÔ II Kit (TaKaRa) in a 20 mL volume. Amplification conditions are as follows: 39 cycles of denaturation at 95  C for 30 s, annealing at 60  C for 30 s, and extension at 72  C for 30 s with a C1000Ô Thermal Cycler (CFX 96Ô RealTime System, BioRad, Germany). Data were detected and analyzed using CFX Manager Software. The expression of target mRNA was normalized to the b-actin mRNA reference and calculated using the Pfaffl method [32]. Data of relative gene expression were presented as the fold changes compared with the untreated control. The primer sequences were listed as follows: 50 -GGCTCTGACTGTACCACCATCCA-30 (p53, forward), 50 GGCACAAACACGCACCTCAAAG-30 (p53, reverse), 50 -TCTGGCACCACACCTTCTACAATG30 (b-actin, forward), 50 -GGATAGCACAGCCTGGATAGCAA-30 (b-actin, reverse). The

1660

Y. He et al. / Biomaterials 35 (2014) 1657e1666

efficiency and reproducibility of the amplifications were acquired from the PCR standard curves, while the specificity of amplification was verified by melt-curve analysis and agarose gel electrophoresis. Expression of p53 in HepG2 cells transfected by p53 loaded complexes (DP and DPS) for 24 and 48 h was evaluated at protein level using western blot. The cells were lysed and the protein was extracted and quantified by BCA Protein Assay Kit. Loading samples were prepared using sample buffer (65.5 mM Tris (pH 6.8), 5% mercaptoethanol, 2% SDS, 5% sucrose, 0.002% bromophenol blue), and the equal amount of protein was resolved on 8% SDS-PAGE under denatured conditions, followed by transferring onto nitrocellulose membrane (Millipore). Non-specific binding was blocked with 5% non-fat dry milk in TBST (10 mM Tris (pH 7.5), 100 mM NaCl, 0.1% Tween 20) for 1 h at room temperature. After wash, the membrane was incubated in 1:100 diluted rabbit polyclonal anti-p53 antibody solution overnight at 4  C, and 2 h in 1:5000 diluted goat anti-rabbit IgG/HRP secondary antibody solution at room temperature. After being washed extensively with TBST, bands were visualized using the enhanced chemiluminescence detection kit (Amersham, UK). Meanwhile, cells untreated and transfected with DP complexes containing mutant p53 plasmid (abbreviated as DP (mutant)) were used as control. The p53 expression was normalized with housekeeping gene b-actin. 2.10. Anti-proliferative and apoptosis studies Cells were seeded in 96-well plates (1  104 cells/well for HepG2 and 0.8  104 cells/well for B16F10) one day prior to the experiment. Various

complexes containing wide-type and mutant p53 (DP, DPH and DPS complexes) were incubated with cells for transfection. 48 h later, the MTS reagent (20 mL per well) was added and incubated for an additional 2 h at 37  C for a standard cell proliferation MTS assay. The absorbance was measured at 490 nm using a microplate reader (Model 550, Bio-Rad). Untreated cells were used as respective control. In the apoptosis assay, cells were seeded in 6-well plates at a density of 2  106 cells per well for wide-type and mutant p53 transfection. 24 h later, 10 mg/mL PI was added and incubated for 30 min. The cells were then washed with PBS and cultured in fresh DMEM before observation by inverted fluorescence microscope (Leica CTR 4000, Wetzlar, Germany). 2.11. Cell cycle analysis by flow cytometry Gene complexes (containing 3 mg of p53) treated HepG2 cells (2  106 cells/well for 6-well plate) in 10% serum-containing culture medium were harvested 48 h post-transfection and washed with ice cold PBS. The cells were fixed by ice cold 70% ethanol, vortexed and stored at 4  C for 1 h. After centrifugation at 3000 rpm for 3 min at 4  C and wash with PBS twice, the cells were re-suspended in 500 mL of PBS buffer containing 20 mg/mL DNase-free RNase A and 50 mg/mL PI. After 30 min nuclei staining, all samples were analyzed for DNA-PI fluorescence using Beckman Coulter Cytomics FC-500 and Kaluza Flow Cytometry software 1.1. Resulted cell cycle distributions were analyzed for the proportion of cells in apoptosis and in the sub-G1, G0/G1, S, and G2/M phases of the cell cycle.

Fig. 1. Stability study of complexes and compression status of the loaded DNA. (A) Agarose gel electrophoresis of DPS complexes with or without GSH. (B) Agarose gel electrophoresis of gene complexes after incubation with DNase and/or heparin. (C) Fluorescence microscopic images of lyophilized and rehydrated gene complexes (scale bar ¼ 20 mm). (D) Representative FRET spectra of DP, DPS and DPH complexes acquired at 2 h after treatment with GSH (0 and 5 mM). pGL3 plasmid was dual-labeled with Cy3 and Cy5. (E) Ratio of FRET mediated Cy5 signal to Cy3 signal (I660 nm/I565 nm), which represents different compression status of DNA in the complexes after GSH treatment. Data represent mean  S.D. (n ¼ 6, **: p < 0.01). DP: DNA/PEI complexes. DPS: DNA/PEI/HAeSSeCOOH complexes. DPH: DNA/PEI/HA complexes.

Y. He et al. / Biomaterials 35 (2014) 1657e1666 2.12. Statistical analysis All experiments were repeated at least three times. Data are presented as mean values  S.D. Statistical significance (p < 0.05) was evaluated by using Student t-test when only two groups were compared. While differences between treatments groups (more than two groups) were determined by two-way ANOVA. In all tests, a p value less than 0.05 was considered statistically significant.

3. Results and discussion The particle sizes of the DPS complexes ranged from 110 to 200 nm, while zeta potential was from þ27.5 mV to 23.4 mV, depending on the ratio of HA derivatives to DNA. 3.1. Stability of the complexes and compression status of the loaded DNA Stability of the ternary complexes (Scheme 1) is critical to gene delivery. It directly affects the enzyme degradation of the loaded DNA, the extracellular polyanion exchange, and the elimination of gene from reticuloendothelial system. Therefore, the physical properties of the shielded complexes, as well as interactions between the DNA and carriers were thoroughly investigated in this study. It was reported that addition of polyanions usually led to decomposition of the complexes by competitive dissociation of the DNA molecule from the polycations [33,34], so the agarose gel

1661

electrophoresis assay was performed to determine the stability of complexes (Fig. 1A). All DPS ternary complexes with HAeSSe COOH/DNA weight ratios from 0.5 to 5 were retarded in the loading well, and no DNA release was found after incubation with 5 mM GSH (mimicking intracellular environment where HAeSSeCOOH could detach). It implied that the association between DNA and PEI has not been damaged by the anionic shielding in our systems. Besides, it was obviously that after treatment with DNase and release by heparin, gene in binary or ternary complexes showed well integrity, while naked plasmid DNA was completely degraded (Fig. 1B). However, more DNA in DP complexes changed from supercoil one to open circular and linear one than that in the ternary complexes, similar to the results reported by Moret et al. [35]. It suggested that even an excess of PEI was used to condense polyanionic gene, a fraction of DNA exposed on the surface of vectors or located near the surface is still accessible to serum components. The shielding could prevent this slight degradation by encapsulating DNA in a deeper location of complexes and blocking the interaction with DNase [7]. DNA/polycation binary complexes formed via electrostatic interaction often show easy aggregation during transport and storage. Earlier studies have demonstrated the ability of saccharides to prevent theses aggregation and the disintegration during freeze thawing [11,36]. So whether HAeSSeCOOH could confer a protection effect for lyophilization and rehydration was investigated using fluorescence microscopy. It was showed that, DP

Fig. 2. (A) Confocal microscopic images of the cellular uptake of Cy5-labeled gene complexes in HepG2 cells at indicated time points. (B) Flow cytometry analysis of the cellular uptake of Cy5-labeled DPS, DPH and DP complexes in HepG2 cells at 0.5, 1 and 2 h (untreated cells as the control). Gene complexes containing 40% Cy5-labeled DNA (red, excitation 633 nm, emission 670 nm long-pass filter). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

1662

Y. He et al. / Biomaterials 35 (2014) 1657e1666

complexes formed large aggregates after rehydration (Fig. 1C), while the sponge of DPS complexes was soon swollen with good dispersion. In order to probe the tiny changes of the DNA compression status in complexes, FRET spectrum was detected in responses to different GSH levels (0 and 5 mM [37]). DNA was dual-labeled with Cy3 and Cy5, which emit FRET signals only in the distance between donor (Cy3) and acceptor (Cy5) less than 10 nm [38]. Without GSH treatment, DPH and DPS complexes had higher energy transfer ratio of FRET-mediated Cy5 signal to Cy3 signal (acceptor to donor: I660 nm/I565 nm) than DP ones, which represents a tighter condensation of DNA (Fig. 1D and E). These results were contrary to some surmise that negatively shielding would lead to less compression due to charge competition [35], but consistent with the above mentioned better protection of DNA from DNase degradation and the results of the previous study by Yeo et al. [14]. An obvious diversified changes in fluorescent signals appeared after 5 mM GSH treatment. DPS complexes showed relatively lower intensity as compared to DPH complexes, indicating a loosen effect on the DNA condensation. This weaker compression could benefit for timely DNA release in cytoplasm, and consequently contribute to better transfection [11]. According to our previous study, the reduced signal most likely due to the detachment of HAeSSeCOOH [25].

3.2. Cellular uptake of complexes The cellular uptake HAeSSeCOOH shielded complexes was assessed by confocal laser scanning microscopy and flow cytometry, compared with that of non-shielded and stably shielded ones (Fig. 2). After incubation, much more fluorescent signal was observed in the cells treated with DPS and DPH than those treated with DP at 0.5 h time point, and the distinction was more significant after 1 and 2 h. This observation indicated that although the higher positive zeta potential of the DP complexes facilitated the statistic electronic interaction between the nano-scaled carriers and negatively charged cell membrane, the biological recognition and affinity were proved much more powerful for uptake [39]. This fast accumulation was important for the application in vivo, where the contact with target cells is usually transient [40]. Besides optical evaluation by CLSM, flow cytometry analysis was used to confirm the aforementioned observations in a more quantitative manner. It clearly demonstrated the gradual increase of cellular uptake with time (Fig. 2B). And more importantly, no obvious difference was found between cells treated with DPH and DPS ternary complexes, which was consistent with the findings that more than three carboxyl groups (hexasaccharide) in the HA molecule are related to its binding to HA receptors [41].

Fig. 3. Intracellular fate of reduction sensitive DPS complexes and stable DPH complexes, monitored by CLSM on HepG2 cells at indicated time points. DPS and DPH complexes contained 40% Cy5-labeled DNA (red) and Alexa 488 labeled HA derivatives (green). Column 1 and 3 were overlay images of Cy5-labeled DNA and Alexa 488 labeled HA derivatives. Column 2 and 4 were overlay images of Cy5-labeled DNA, Alexa 488 labeled HA derivatives and the bright field. Scale bar ¼ 25 mm.

Y. He et al. / Biomaterials 35 (2014) 1657e1666

1663

Fig. 4. Confocal laser scanning microscope observation of intracellular delivery of DP, DPH and DPS complexes in HepG2 cells at different time (1 and 4 h). The complexes were prepared with Cy5-labeled DNA (red) and FITC-labeled PEI (green). The acidic late endosomes and lysosomes were stained with LysoTracker Blue (blue). 1: Cy5 channel; 2: LysoTracker channel; 3: overlay of 1 (Cy5), 2 (LysoTracker) and FITC channel. White arrows signify the occasions of coincidence between the complexes and the endosomes/lysosomes. Yellow arrows indicate the complexes escaping from the endosomes/lysosomes into the cytoplasm. Blue arrows imply the destruction of endosomes/lysosomes. Scale bar ¼ 10 mm.

3.3. Investigation of intracellular fate

Fig. 5. (A) mRNA levels of p53 in HepG2 cells transfected with DPS and DP complexes in 10% serum containing medium at 24 and 48 h. Data were referenced to the p53 mRNA of untreated HepG2 cells. Data represent mean  S.D. (n ¼ 6, **: p < 0.01 as compared to DP). (B) Western blotting assay of p53 expression in HepG2 cells transfected by DP and DPS complexes for 24 and 48 h a) Transfection in serum-free medium up to 48 h b) Transfection in 10% serum containing medium at 24 and 48 h. Data were referenced to the p53 of untreated cells (control groups). DPS: wild-type p53/PEI/HAe SSeCOOH complexes. DP: wild-type p53/PEI complexes. DP (mutant): mutant p53/PEI complexes. Cells: untreated HepG2 cells.

The designed reduction-sensitive DPS complexes were expected to keep stable and shielded in the extracellular environment, while detach timely via disulfide cleavage under intracellular reducing conditions [42]. Confocal microscopy was adopted to gain intuitive insight into the behavior of the cleavable part in DPS complexes, together with the DPH complexes with stable shielding. After 1 h incubation with complexes containing Alexa 488 labeled HA derivatives and Cy5 labeled plasmid, the fluorescent signals were well colocalized on the cell surface (Fig. 3). The yellow pixels indicated the integrity of both DPS and DPH complexes. It remained yellow for longer incubation (2 h) in DPH group, while that of DPS complexes gradually changed over time. At 3 h time point, more fluorescent dots appeared inside the cells. And importantly, most intracellular fluorescent dots in the DPS group displayed only red color, while the dots in DPH group were yellow. This result was consistent with our anticipation that the cleavage of disulfide bond between carboxyl groups and backbone occurred in the cytoplasm, contributing to the detachment of the shielding and exposure of PEI/DNA complexes [9,25]. As the cleavage of disulfide bond in shielding was confirmed inside cells, the reduced endosomal escape ability was expected recovered [19,20]. So the subcellular localization of the complexes was observed by CLSM, using Cy5-labeled DNA (red), FITC-labeled PEI (green) and LysoTracker (blue) marked late endosomes/lysosomes (Fig. 4). As shown in the image after 1 h incubation, a majority of DP, DPH and DPS complexes were located in the blue fluorescent endosomes/lysosomes, as reflected by the white fluorescence from the colocalization of the yellow complexes with the blue organelle dyes [43]. Four hours later, when lots of the DPH complexes were still sequestered in the endosomes/lysosomes, most DPS complexes were separated from the blue LysoTracker as the DP ones. It indicated that the reduction sensitive DPS complexes successfully resumed the endosomal escape ability of PEI, corresponding to the deshielding profile in Fig. 3. Furthermore, the

1664

Y. He et al. / Biomaterials 35 (2014) 1657e1666

staining of endosomes/lysosomes by the tracker is known to depend upon the acidity of the endosomal/lysosomal compartments. In the case of DP and DPS, the blue fluorescence of endosomes/lysosomes became attenuated at 4 h time point, suggesting the destruction of the acid compartment environment as a result of the proton sponge effect of PEI [43]. Above results (Figs. 2e4) may well explain the better transfection performance of DPS compared with the DPH (with similar cellular uptake) and DP (with similar endosomal escape capacity) [25] (Supporting Figure S3), that gene delivery is a sophisticated process, each step (including cellular uptake, endosome escape and DNA release) is important, neglecting anyone of them may result in lose. 3.4. Transfection of p53-encoding plasmid Therapeutic gene transfection by DPS complexes was analyzed using p53 tumor suppressor gene. After transfection of p53 containing DPS complexes into HepG2 cells, the mRNA expression was analyzed by qPCR at 24 and 48 h. As summarized in Fig. 5A, the p53 mRNA level was extraordinary high after transfection by DPS formulation, reaching w30-fold higher than that expressed in DP groups at 48 h (p < 0.05). Endogenous wild-type p53 gene is known non-existent in HepG2 cells [3], and its accumulation in other cells

is relatively low under normal conditions [44]. Thus, the high level of p53 mRNA indicated an efficient delivery and expression of exogenous p53 gene. The promotion of p53 expression was also confirmed at protein level in serum-free or 10% serum-containing medium (Fig. 5B). It was obviously that the expression of p53 protein was always higher than that in DP group, using b-actin as reference (Fig. 5B a and b). And the protein expression was up-regulated with the time, showing higher superiority at 48 h time point. 3.5. Apoptosis and anti-proliferation studies p53 induced apoptosis of carcinoma cells was assessed on HepG2 and B16F10 cells by fluorescence microscopy with PI staining. 24 h post-transfection, images of PI staining (Fig. 6A and Supporting Figure S4A) clearly showed that the number of late apoptotic cells and dead cells (PI positive, red (in web version)) were different in each groups. Rare PI-positive cells were observed in DP (mutant) group and the untreated cells, while DPS group showed the maximum number, followed by DPH and DP groups. These phenomena indicated that a serious p53-dependent apoptosis was achieved by our designed stimuli-deshielding complexes, in good agreement with the results from the luciferaseencoding plasmids [25]. The number of cells expressing the

Fig. 6. Cell apoptosis and cycle assay after transfection of various p53 loaded complexes. (A) Detection of apoptosis induced by different gene complexes (DP, DPH and DPS) in HepG2 cells for 24 h. Apoptosis was assessed by fluorescence microscopy with PI labeling. (B) Anti-proliferative activity of the different gene complexes (DP, DPH and DPS) in HepG2 cells for 48 h (C) and (D) Histograms and graph bar of HepG2 cells cycle distribution after various treatments. Data represent mean  S.D. (n ¼ 6, **: p < 0.01 as compared to DP or DPH).

Y. He et al. / Biomaterials 35 (2014) 1657e1666

reporter gene after gene transfer was comparable to that of cells undergoing apoptosis in our study. Additionally, MTS assay was carried out for quantitative evaluation of p53-dependent anti-proliferation on HepG2 and B16F10 cells (Fig. 6B and Supporting Figure S4B). The cellular inhibition of reduction-sensitive DPS was 18% higher than that of stable DPH group (p < 0.05), and even up to 34% than that in DP complexes (p < 0.05). Overall, these results showed that the successful transfection of wild-type p53 by DPS formulation may not only inhibit the carcinoma cells growth but also induce cell apoptosis. 3.6. Cell cycle analysis by flow cytometry p53 is known to arrest cell growth by holding cell cycle at the G1/ S phase. Cell cycle analysis via flow cytometry was performed to gain more insights into the anti-proliferative effect of the complexes. After transfection of wide-type p53, cell cycle of HepG2 revealed a significant increase in the sub-G1 phase, with a concomitant decrease in G0/G1 phase and a minor decrease in the G2/M population (Fig. 6C, D). Cells transfected with DPS showed the strongest increased sub-G1 population (from 1.4% in the control group to 45.9%), while those in the DP and DPH treated cells was 18.6% and 30.1%, respectively. Since no significant apoptosis (subG1) was detected in untreated cells and cells treated by DP (mutant), the apoptosis observed can be reasonably attributed to the wildtype p53 protein expression by the DPS delivery system. These results indicated that DPS resulted in significant cell-cycle perturbations prior to apoptosis and increased apoptosis in HepG2 cells. Further in vivo studies using mouse transplanted tumor models are in progress to evaluate the therapeutic effect of targeted and reduction-sensitive ternary complexes carrying p53 gene. 4. Conclusions The reduction sensitive shielding ternary complexes (DPS) based on disulfide linked PEI ternary complexes were developed, and the present study focused on the function of shielding and mechanism of delivery. Both reduction sensitive DPS and stable DPH ternary complexes displayed better protection against DNase, stronger compress status of DNA and higher level of cellular uptakes, compared with binary complexes (DP). But the intracellular fate study revealed that only DPS could effectively detach the HA shielding due to the disulfide reduction and facilitate endosomal escape. Application of DPS complexes for tumor suppressor gene p53 delivery could up-regulate exogenous p53 expression at both mRNA and protein levels, and consequently induce better antiproliferation and higher apoptosis population. In sum, DPS system would contribute to the establishment of design criteria for smart gene vectors towards efficient gene therapy. Acknowledgments This study was supported by National Basic Research Program of China (National 973 programs, No. 2011CB606206), National Natural Science Foundation of China (NSFC, No. 51133004, 81361140343, 81000657, 31271020), Fund from Sino-German Center for Research Promotion (GZ 756), Research Fund for the Doctoral Program of Higher Education of China (Grant No. 20100181120075), and the Excellent Young Scholar Program of Sichuan University (2012SCU04A06). Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.biomaterials.2013.10.073.

1665

References [1] Gaspar HB, Parsley KL, Howe S, King D, Gilmour KC, Sinclair J, et al. Gene therapy of X-linked severe combined immunodeficiency by use of a pseudotyped gammaretroviral vector. Lancet 2004;364:2181e7. [2] Hollstein M, Sidransky D, Vogelstein B, Harris CC. p53 mutations in human cancers. Science 1991;253:49e53. [3] Bressac B, Galvin KM, Liang TJ, Isselbacher KJ, Wands JR, Ozturk M. Abnormal structure and expression of p53 gene in human hepatocellular carcinoma. Proc Natl Acad Sci U S A 1990;87:1973e7. [4] Chang J, Xu X, Li H, Jian Y, Wang G, He B, et al. Components simulation of viral envelope via amino acid modified chitosans for efficient nucleic acid delivery: in vitro and in vivo study. Adv Funct Mater 2013;23:2691e9. [5] Cheng G, He Y, Xie L, Nie Y, He B, Zhang Z, et al. Development of a reductionsensitive diselenide-conjugated oligoethylenimine nanoparticulate system as a gene carrier. Int J Nanomed 2012;7:3991e4006. [6] Dai J, Zou S, Pei Y, Cheng D, Ai H, Shuai X. Polyethylenimine-grafted copolymer of poly(L-lysine) and poly(ethylene glycol) for gene delivery. Biomaterials 2011;32:1694e705. [7] Burke RS, Pun SH. Extracellular barriers to in vivo PEI and PEGylated PEI polyplexmediated gene delivery to the liver. Bioconjug Chem 2008;19:693e704. [8] Wang Y, Zheng M, Meng F, Zhang J, Peng R, Zhong Z. Branched polyethylenimine derivatives with reductively cleavable periphery for safe and efficient in vitro gene transfer. Biomacromolecules 2011;12:1032e40. [9] Nie Y, Gunther M, Gu Z, Wagner E. Pyridylhydrazone-based PEGylation for pH-reversible lipopolyplex shielding. Biomaterials 2011;32:858e69. [10] Cho SK, Kwon YJ. Polyamine/DNA polyplexes with acid-degradable polymeric shell as structurally and functionally virus-mimicking nonviral vectors. J Control Release 2011;150:287e97. [11] Ito T, Iida-Tanaka N, Koyama Y. Efficient in vivo gene transfection by stable DNA/PEI complexes coated by hyaluronic acid. J Drug Target 2008;16:276e81. [12] Kim EJ, Cho HJ, Park D, Kim JY, Kim YB, Park TG, et al. Antifibrotic effect of MMP13-encoding plasmid DNA delivered using polyethylenimine shielded with hyaluronic acid. Mol Ther 2011;19:355e61. [13] Soliman M, Nasanit R, Abulateefeh SR, Allen S, Davies MC, Briggs SS, et al. Multicomponent synthetic polymers with viral-mimetic chemistry for nucleic acid delivery. Mol Pharm 2012;9:1e13. [14] Yeo Y, Xu PS, Quick GK. Gene delivery through the use of a hyaluronateassociated intracellularly degradable crosslinked polyethyleneimine. Biomaterials 2009;30:5834e43. [15] Tian H, Lin L, Chen J, Chen X, Park TG, Maruyama A. RGD targeting hyaluronic acid coating system for PEI-PBLG polycation gene carriers. J Control Release 2011;155:47e53. [16] Knorr V, Allmendinger L, Walker GF, Paintner FF, Wagner E. An acetal-based PEGylation reagent for pH-sensitive shielding of DNA polyplexes. Bioconjug Chem 2007;18:1218e25. [17] Sethuraman V, Lee M, Bae Y. A biodegradable pH-sensitive micelle system for targeting acidic solid tumors. Pharm Res 2008;25:657e66. [18] Yu H, Zou Y, Wang Y, Huang X, Huang G, Sumer BD, et al. Overcoming endosomal barrier by amphotericin B-loaded dual pH-responsive PDMA-bPDPA micelleplexes for siRNA delivery. ACS Nano 2011;5:9246e55. [19] Sanjoh M, Hiki S, Lee Y, Oba M, Miyata K, Ishii T, et al. pDNA/poly(L-lysine) polyplexes functionalized with a pH-sensitive charge-conversional poly(aspartamide) derivative for controlled gene delivery to human umbilical vein endothelial cells. Macromol Rapid Commun 2010;31:1181e6. [20] Takae S, Miyata K, Oba M, Ishii T, Nishiyama N, Itaka K, et al. PEG-detachable polyplex micelles based on disulfide-linked block catiomers as bioresponsive nonviral gene vectors. J Am Chem Soc 2008;130:6001e9. [21] Park K, Lee MY, Kim KS, Hahn SK. Target specific tumor treatment by VEGF siRNA complexed with reducible polyethyleneimine-hyaluronic acid conjugate. Biomaterials 2010;31:5258e65. [22] Sun H, Guo B, Cheng R, Meng F, Liu H, Zhong Z. Biodegradable micelles with sheddable poly (ethylene glycol) shells for triggered intracellular release of doxorubicin. Biomaterials 2009;30:6358e66. [23] Hatakeyama H, Akita H, Harashima H. A multifunctional envelope type nano device (MEND) for gene delivery to tumours based on the EPR effect: a strategy for overcoming the PEG dilemma. Adv Drug Deliv Rev 2011;63:152e60. [24] Choi KY, Yoon HY, Kim J-H, Bae SM, Park R-W, Kang YM, et al. Smart nanocarrier based on PEGylated hyaluronic acid for cancer therapy. ACS Nano 2011;5:8591e9. [25] He Y, Cheng G, Xie L, Nie Y, He B, Gu Z. Polyethyleneimine/DNA polyplexes with reduction-sensitive hyaluronic acid derivatives shielding for targeted gene delivery. Biomaterials 2013;34:1235e45. [26] Abetamann V, Kern HF, Elsasser HP. Differential expression of the hyaluronan receptors CD44 and RHAMM in human pancreatic cancer cells. Clin Cancer Res 1996;2:1607e18. [27] Jiang G, Park K, Kim J, Kim KS, Oh EJ, Kang H, et al. Hyaluronic acidpolyethyleneimine conjugate for target specific intracellular delivery of siRNA. Biopolymers 2008;89:635e42. [28] Luo Y, Ziebell MR, Prestwich GD. A hyaluronic acid-taxol antitumor bioconjugate targeted to cancer cells. Biomacromolecules 2000;1:208e18. [29] Oh EJ, Park K, Kim KS, Kim J, Yang J-A, Kong J-H, et al. Target specific and longacting delivery of protein, peptide, and nucleotide therapeutics using hyaluronic acid derivatives. J Control Release 2010;141:2e12.

1666

Y. He et al. / Biomaterials 35 (2014) 1657e1666

[30] Surace C, Arpicco S, Dufaÿ-Wojcicki Al, Marsaud Vr, Bouclier Cl, Clay D, et al. Lipoplexes targeting the CD44 hyaluronic acid receptor for efficient transfection of breast cancer cells. Mol Pharm 2009;6:1062e73. [31] Ito T, Yoshihara C, Hamada K, Koyama Y. DNA/polyethyleneimine/hyaluronic acid small complex particles and tumor suppression in mice. Biomaterials 2010;31:2912e8. [32] Pfaffl MW, Horgan GW, Dempfle L. Relative expression software tool (RESTÓ) for group-wise comparison and statistical analysis of relative expression results in real-time PCR. Nucleic Acids Res 2002;30:36. [33] Arigita C, Zuidam NJ, Crommelin DJ, Hennink WE. Association and dissociation characteristics of polymer/DNA complexes used for gene delivery. Pharm Res 1999;16:1534e41. [34] Koyama Y, Yamada E, Ito T, Mizutani Y, Yamaoka T. Sugar-containing polyanions as a self-assembled coating of plasmid/polycation complexes for receptor-mediated gene delivery. Macromol Biosci 2002;2:251e6. [35] Moret I, Esteban Peris J, Guillem VM, Benet M, Revert F, Dasi F, et al. Stability of PEI-DNA and DOTAP-DNA complexes: effect of alkaline pH, heparin and serum. J Control Release 2001;76:169e81. [36] Anchordoquy TJ, Girouard LG, Carpenter JF, Kroll DJ. Stability of lipid/DNA complexes during agitation and freeze-thawing. J Pharm Sci 1998;87:1046e51. [37] Saito G, Swanson JA, Lee K-D. Drug delivery strategy utilizing conjugation via reversible disulfide linkages: role and site of cellular reducing activities. Adv Drug Deliv Rev 2003;55:199e215.

[38] Roy R, Hohng S, Ha T. A practical guide to single-molecule FRET. Nat Meth 2008;5:507e16. [39] Conner SD, Schmid SL. Regulated portals of entry into the cell. Nature 2003;422:37e44. [40] Huth S, Lausier J, Gersting SW, Rudolph C, Plank C, Welsch U, et al. Insights into the mechanism of magnetofection using PEI-based magnetofectins for gene transfer. J Gene Med 2004;6:923e36. [41] Jiang G, Park K, Kim J, Kim KS, Hahn SK. Target specific intracellular delivery of siRNA/PEI-HA complex by receptor mediated endocytosis. Mol Pharm 2009;6: 727e37. [42] Kuai R, Yuan W, Qin Y, Chen H, Tang J, Yuan M, et al. Efficient delivery of payload into tumor cells in a controlled manner by TAT and thiolytic cleavable PEG co-modified liposomes. Mol Pharm 2010;7:1816e26. [43] Jiang T, Zhang Z, Zhang Y, Lv H, Zhou J, Li C, et al. Dual-functional liposomes based on pH-responsive cell-penetrating peptide and hyaluronic acid for tumor-targeted anticancer drug delivery. Biomaterials 2012;33: 9246e58. [44] Rotter V, Abutbul H, Ben-Ze’ev A. P53 transformation-related protein accumulates in the nucleus of transformed fibroblasts in association with the chromatin and is found in the cytoplasm of non-transformed fibroblasts. EMBO J 1983;2:1041e7.