- Email: [email protected]
Biodegradable gadolinium-chelated cationic poly(urethane amide) copolymers for gene transfection and magnetic resonance imaging
Materials Science and Engineering C 65 (2016) 181–187
Contents lists available at ScienceDirect
Materials Science and Engineering C journal homepage: www.elsevier.com/locate/msec
Biodegradable gadolinium-chelated cationic poly(urethane amide) copolymers for gene transfection and magnetic resonance imaging Xiaolong Gao a, Gangmin Wang c, Ting Shi b, Zhihong Shao a, Peng Zhao b, Donglu Shi b, Jie Ren d, Chao Lin b,⁎, Peijun Wang a,⁎ a
Department of Radiology, Tongji Hospital, Tongji University School of Medicine, Shanghai 200065, PR China The Institute for Translational Nanomedicine, Shanghai East Hospital, Institute for Biomedical Engineering and Nanoscience, Tongji University School of Medicine, Shanghai 200092, PR China Department of Urology, Huashan Hospital, Fudan University, Shanghai 200040, PR China d Institute of Nano and Biopolymeric Materials, School of Materials Science and Engineering, Tongji University, 4800 Caoan Road, Shanghai 201804, PR China b c
a r t i c l e
i n f o
Article history: Received 21 January 2016 Received in revised form 14 March 2016 Accepted 7 April 2016 Available online 13 April 2016 Keywords: Poly(urethane amide) Gene delivery Disulfide Imaging Theranostics
a b s t r a c t Theranostic nano-polyplexes containing gene and imaging agents hold a great promise for tumor diagnosis and therapy. In this work, we develop a group of new gadolinium (Gd)-chelated cationic poly(urethane amide)s for gene delivery and T1-weighted magnetic resonance (MR) imaging. Cationic poly(urethane amide)s (denoted as CPUAs) having multiple disulfide bonds, urethane and amide linkages were synthesized by stepwise polycondensation reaction between 1,4-bis(3-aminopropyl)piperazine and a mixture of di(4-nitrophenyl)-2, 2′dithiodiethanocarbonate (DTDE-PNC) and diethylenetriaminepentaacetic acid (DTPA) dianhydride at varied molar ratios. Then, Gd-chelated CPUAs (denoted as GdCPUAs) were produced by chelating Gd(III) ions with DTPA residues of CPUAs. These GdCPUAs could condense gene into nanosized and positively-charged polyplexes in a physiological condition and, however, liberated gene in an intracellular reductive environment. In vitro transfection experiments revealed that the GdCPUA at a DTDE-PNC/DTPA residue molar ratio of 85/15 induced the highest transfection efficiency in different cancer cells. This efficiency was higher than that yielded with 25 kDa branched polyethylenimine as a positive control. GdCPUAs and their polyplexes exhibited low cytotoxicity when an optimal transfection activity was detected. Moreover, GdCPUAs may serve as contrast agents for T1weighted magnetic resonance imaging. The results of this work indicate that biodegradable Gd-chelated cationic poly(urethane amide) copolymers have high potential for tumor theranostics. © 2016 Elsevier B.V. All rights reserved.
1. Introduction Nucleic acid-based cancer therapy is an alternative strategy to routine chemotherapy [1]. For successful nucleic acid-based cancer therapy, highly efficient delivery of nucleic acids into targeted somatic cells in a bio-safe way is still a challenge. Although engineered viral vectors encoded with nucleic acids can robustly infect human cells, they are associated with a few uncertain bio-safety issues such as uncontrolled mutagenesis, immunogenicity and oncogenicity, thus impeding their clinical translations [2]. Alternatively, non-viral vectors based on cationic polymers reveal high potential for relatively safe gene delivery as a result of their low toxicity, facile synthesis and high nucleic acid-carrying capacity [3]. Accordingly, over the past two decades, a number of polymeric vectors for non-viral gene delivery have been widely studied [4]. A lot of studies have confirmed a massive promise of cationic polymernucleic acid assembled complexes (also called as polyplexes) for cancer gene therapy [5]. ⁎ Corresponding authors. E-mail addresses: [email protected] (C. Lin), [email protected] (P. Wang).
http://dx.doi.org/10.1016/j.msec.2016.04.027 0928-4931/© 2016 Elsevier B.V. All rights reserved.
The development of nano-carriers for tumor imaging and therapy has received increasing interest in recent years [6]. From clinical point of view, such dual-function nano-carriers allow clinical physicians to accurately evaluate anti-cancer efficacy and choose balanced drug dose in time. As such, a variety of theranostic nano-carriers have been designed which combine both therapeutic functions (chemotherapy and gene therapy) and single-/multi-imaging modalities [7]. It should be however noted that most of these theranostic nano-carriers reported are based on inorganic nanoparticles [8]. Only few literatures have focused on theranostic polymers for gene therapy and imaging such as magnetic resonance (MR) imaging. For example, Bryson et al. reported on Gdchelated polycations containing oligoamines for gene delivery and MR imaging [9]. In another work, Wu et al. designed polymethacrylamidebased cationic polymers having europium-labeled residues for MR imaging [10]. Recently, Xue et al. showed a group of Gd-chelated polycations for siRNA delivery [11]. However, these theranostic polymers usually lack biodegradability, thus probably hampering their potential of clinical translations. An ideal theranostic polymer should offer efficient gene transfection and accurate imaging. Importantly, such polymer should be essentially
182
X. Gao et al. / Materials Science and Engineering C 65 (2016) 181–187
low toxic against the cells and the body. In order to generate efficient and low-toxic polymers for gene delivery, disulfide-containing (reducible) cationic polymers for gene delivery have recently received rapid growing interest [12]. A lot of studies confirm that reducible polyethylenimine, polylysine and polyamides may induce superior transfection efficacy to their non-reducible counterparts and, meanwhile, are associated with a minor cytotoxicity [12]. This low toxicity is due to intracellular cleavaging of disulfide linkage by glutathione (0.5–10 mM) [13], leading to intracellular degradation of reducible polymers and thus reduced toxicity. Reducible polymers thus represent suitable non-viral carriers for safe and efficient gene delivery. However, to our best knowledge, no report have presented on reducible cationic polymers for gene delivery and MR imaging. It would be thus meaningful to develop such reducible cationic polymers and evaluate their feasibility for theranostics. We recently reported on reducible cationic polyurethanes synthesized by non-isocyanate chemistry and found that these polyurethanes possessed low toxicity and high ability for gene transfection towards tumor cells [14]. Herein, we aim to design and prepare theranostic reducible cationic polyurethanes for gene delivery and imaging. To this end, a group of linear reducible cationic poly(urethane amide)s (denoted as CPUAs) are designed and prepared that comprise multiple disulfide bonds, tertiary amine groups and diethylenetriaminepentaacetic acid (DTPA) residues. The DTPA residue allows for complexation with Gd(III) ions to generate Gd-chelated CPUAs (denoted as GdCPUAs). We hypothesized that GdCPUAs would be practical as non-viral vectors for gene delivery and MR imaging. We present synthesis and characterization of GdCPUAs and the properties of GdCPUA-based polyplexes in terms of particle size, surface charge, magnetism, and gene release ability. Transfection ability and cytotoxicity of GdCPUA polyplexes are evaluated in vitro against cancer cells. Moreover, T1weighted MR imaging by using GdCPUAs as contrast agents is investigated. 2. Materials and methods 2.1. Materials All chemicals were used directly as received unless otherwise statement. 2,2′-Dithiodiethanol (DTDE), gadolinium chloride (GdCl3), 4-nitrophenyl chloroformate (PNC), dithiothreitol (DTT), 2-[4-(2-hydroxyethyl)piperazin-1-yl] ethane sulfonic acid (HEPES), branched polyethylenimine (BPEI, Mw = 25 kDa), 1,4-diethylenetriaminepentaacetic acid (DTPA) dianhydride, and bis(3-aminopropyl)piperazine (BAP) were ordered from Sigma-Aldrich Co. (USA). Lipofectamine 2000 transfection agent was ordered from Life Technology (USA). The plasmids encoding for green fluorescent protein (GFP) and luciferase (Luc), also termed as pCMV-GFP and pCMV-Luc, respectively, were obtained from plasmid factory (Bielefeld, Germany). shRNA-VEGF plasmid was purchased from InvivoGen (USA). Di(4-nitrophenyl)-2,2′-dithiodiethanocarbonate (DTDE-PNC) was prepared according to our previous report [14]. 2.2. Synthesis and characterization of Gd-chelated polymers Gd-chelated reducible cationic poly(urethane amide) (GdCPUA) polymers was obtained by a two-step procedure. First, a group of cationic poly(urethane amide)s (CPUAs) were obtained by polymerization reaction between BAP and a mixture of DTDE-PNC and DTPA dianhydride at varied molar ratios (i.e. 85/15, 70/30, 50/50). In a typical procedure for the synthesis of CPUA30 at a DTDE-PNC/DTPA anhydride ratio of 70/30, DTDE-PNC (0.5 g, 1.03 mmol), BAP (0.30 g, 1.47 mmol) and DTPA anhydride (0.17 g, 0.44 mmol) were dissolved in anhydrous DMSO (5 mL) as a solvent in a brown-colour round flask. After running for 5 days at 40 °C in nitrogen atmosphere, the reaction was run for another 2 days after adding an excess amount of BAP (0.03 g). Finally,
resulting solution was purified by ultrafiltration process (3 kDa cutoff) with acidic deionized water (3 × 5 L, pH 4.5). CPUA30 was obtained as a white solid after freeze-drying (yield: 0.3 g, 45%). Chemical composition of CPUAs was tested by 1H NMR spectrum meter (Varian, USA). Next, GdCPUAs were yielded by Gd(III) chelation with DTPA residues in CPUAs. As an example for the preparation of GdCPUA30, CPUA30 (50 mg) was dissolved in 20 mL of deionized water. Next, GdCl3 (~ 4 mmol) in deionized water (5 mL) was dropwise added into CPUA30 solution. During mixing process, the solution became turbid but transparent again after pH was set to pH 6.5 with 4 M HCl. The resulting solution was stirred for 4 h and then transferred in dialysis bag (1 kDa cut-off) for exhaustive dialysis in deionized water at pH ~ 6. GdCPUA30 was obtained as a solid powder after freezing-drying (yield: 0.35 g, 67%). GdCPUAs were analyzed by inductively coupled plasma (ICP) system to determine Gd content after dissolving these polymers in 4 M H2NO3. 2.3. Preparation and characterization of GdCPUA based polyplexes The polyplexes of GdCPUA copolymers were prepared by gently mixing plasmid DNA with GdCPUAs at N/P ratios in HEPES buffer (20 mM, pH 7.4) [15]. Particle size and surface charge of GdCPUA polyplexes were determined by dynamic light scattering (DLS) analysis (Nanosizer NS90, Malvern, UK) at 25 °C. Besides, the polyplex solution was analyzed by agarose gel electrophoresis using the method reported previously [16]. To characterize gene release behavior in a reductive environment, the polyplexes were co-incubated with DTT (5 mM) in the HEPES buffer for 30 min, analyzed by the gel electrophoresis, DNA band was visualized by TanonGel Image system (Shanghai, P.R. China). Relaxation time (T1) of GdCPUAs and their polyplexes was measured using a contrast agent relaxation rate analysis system (PQ100, NIUMAG, Shanghai, P.R. China) according to the manufacture's protocol. A clinical T1-contrast agent, Magnevist (Gd-DTPA complex), was used as a positive control. T1-weighted MR imaging with the polyplexes of GdCPUA copolymers was performed by running Siemens Magnetom Verio 3.0 T MR instrument. In brief, the polyplexes were prepared in HEPES buffer (pH 7.4) in EP tube at different Gd concentrations ranging from 0.1 mM to 0.8 mM. MR imaging was obtained by placing the tubes in an in-house built 10-cm RF coil installed in the instrument according to the manufacturer's instruction. 2.4. Cell culture, transfection test and cytotoxicity assay in vitro SKOV-3 cells (ATCC, USA) were cultured in Mccyo's-5α medium culture medium (GIBCO) having 10% FBS, 100 U/mL penicillin and streptomycin (GIBCO). MCF-7, HepG2 and A549 cells were cultured in DMEM complete medium. The cells were grown at 37 °C in cell incubator (Thermo Scientific, Waltham, MA, USA) supplied with 5% CO2. Transfection activity of GdCPUA copolymers in tumor cells was evaluated according to our previous transfection protocol [16]. In brief, the cells (5–7 × 104 cells/well) were cultured in a 24-well plate for 24 h until 60–70% cell confluence was reached. GdCPUA polyplexes containing pCMV-GFP plasmid (1 μg) in a HEPES buffer (20 mM, pH 7.4) were then prepared at varied N/P ratios. The transfection tests were performed by adding the polyplex solution in each well and coincubating with the cells for 1 h in the absence or presence of 10% FBS. The cells were cultured in complete medium for another 24 h. Besides, a formulation of BPEI polyplexes at an N/P ratio of 10/1 and lipofectamine 2000 were used as positive controls. Those cells without transfection (untreated cells) were considered as a blank (negative) control. The same protocol was applied in the transfection with GdCPUA polyplexes containing shRNA-VEGF plasmid. VEGF protein and mRNA level were analyzed by using ELISA kit (Abcam, USA) and RT-PCR, respectively. Transfection efficiency was determined by testing relative fluorescence intensity of GFP expressed in the cells according to our previous
X. Gao et al. / Materials Science and Engineering C 65 (2016) 181–187
report [16]. The efficiency was presented with the unit of fluorescence intensity (FI) per mg protein. Cytotoxicity of GdCPUAs and their polyplexes was evaluated by determining metabolic activity of the cells with AlamarBlue assay kit (Life Technology). Untreated cells (blank control) were set as 100% cell viability. All results are given as mean ± standard deviation (SD) from triplicate samples (n = 3). 2.5. Statistical analysis
Table 1 Characteristics of cationic poly(urethane amide)s (CPUAs). Sample codea
DTDE-PNC/DTPA anhydride feed molar ratio
DTDE-PNC/DTPA anhydride residue molar ratiob
Mw (kDa)c
PDIc
CPUA15 CPUA30 CPUA50
85/15 70/30 50/50
84/16 68/32 55/45
39.0 42.0 45.0
1.5 1.6 1.8
a
Student's t-test analysis is used in this work. The difference is considered to be statistically significant at P b 0.05. 3. Results 3.1. Synthesis of Gd-loaded cationic poly(urethane amide)s (GdCPUAs) In this work, a group of reducible cationic polyurethanes with diethylenetriaminepentaacetic acid (DTPA) residues for Gd(III) ion chelation were synthesized by stepwise polycondensation reaction between 1,4-bis(3-aminopropyl)piperazine (BAP) and the mixture of DTPA anhydride and di(4-nitrophenyl)-2, 2′-dithiodiethanocarbonate (DTDE-PNC) (Fig. 1a). An equivalent molar ratio between BAP and DTDE-PNC/DTPA anhydride mixture was used to afford resulting polymers with the highest molecular weight. The reaction was terminated with an excess amount of BAP to consume any unreacted 4nitrophenyl group. In the reaction, the molar ratio of DTDE-PNC and DTPA anhydride was set at 85/15, 70/30 and 50/50, respectively. As a result, three reducible copolymers were obtained which comprise disulfide linkages, tertiary amino groups and urethane/amide linkages, and they are termed as cationic poly(urethane amide)s (denoted as CPUA15, CPUA30, and CPUA50, respectively, Table 1). These CPUAs were further characterized. 1H NMR analysis of the CPUAs indicated that their chemical compositions were in good line with expected chemical structures. Fig. 2 shows typical 1H NMR spectrum of CPUA15. Besides, the molar ratio between DTDE residue and DTPA residue could be determined by comparing the integral area of two peaks at δ 4.2 and δ 3.45, attributing to the proton signal of DTDE residue (CH2CH2SS) and DTPA residue (NCH2CH2N), respectively. As revealed in Table 1, the molar ratio of DTDE/DTPA residue was in accordance with feed monomer ratio of DTDE-PNC/DTPA anhydride, suggesting that chemical composition of these CPUAs may be readily adjusted by monomer feed ratio. GPC analysis of these CPUAs showed that they had single-peak molecular weight distribution (Fig. S1) with average weight molecular weight (Mw) ranging from 39 k to 45 kDa and narrow polydispersity index (PDI) of 1.5–1.8. Next, Gd-chelated CPUAs (denoted as GdCPUAs) were prepared by complexation of Gd(III) ions with DTPA residues of three CPUAs
183
b c
These copolymers are coded with feed molar ratio of DTPA anhydride. Determined by 1H NMR. Determined by GPC analysis.
(Fig. 1b) and obtained as solid powder after dialysis and freeze-drying. These GdCPUAs were termed as GdCPUA15, GdCPUA30 and GdCPUA50, respectively. ICP analysis indicated that Gd contents of as-prepared GdCPUAs were in line with theoretical Gd contents (Table 2), implying an efficient complexation of Gd(III) ions with DTPA residues of CPUAs. A solubility test revealed that, at a high polymer concentration of 1 mg/mL, these GdCPUAs were soluble in PBS buffer and completed cell culture medium. Overall, these data indicate successful availability of GdCPUAs. 3.2. Gene condensation and release behavior To ascertain whether GdCPUAs may be used for gene delivery, electrostatic binding between the polymers and plasmid DNA was characterized by electrophoresis gel analysis at different nitrogen/phosphate (N/ P) ratios. It was found that GdCPUAs efficiently retarded the mobility of DNA at and above an N/P ratio of 10/1, implying their binding with DNA (Fig. 3a). Besides, it appeared that retardation ability of these polymers became weaker with increasing amount of Gd-chelated DTPA residues from 15% to 50%. For example, GdCPUA15 could totally hinder DNA mobility at a low N/P ratio of 2.5/1 whereas GdCPUA30 and GdCPUA50 had to afford DNA retardation at a higher N/P ratio of 5/1 and 10/1, respectively. Furthermore, DNA binding ability of GdCPUA polymers was characterized by dynamic light scattering (DLS) analysis. Particle size (in average diameter) of the polyplexes of GdCPUAs was indicated as a function of N/P ratios ranging from 10/1 to 30/1 (Fig. 3b). At a low N/P ratio of 10/1 or 20/1, particle size of GdCPUA polyplexes displayed the following tendency: GdCPUA15 ≈ GdCPUA30 b GdCPUA50, implying that incorporating Gd-chelated DTPA residues into CPUAs may increase the size of the polyplexes. Only at a high N/P ratio of 30/1, this effect was not pronounced due to enhanced gene binding force. At this ratio, the polyplexes of GdCPUAs displayed small size b200 nm. This point was also supported by the results that the particle size of the polyplexes of GdCPUAs markedly reduced with increasing N/P ratios from 10/1 to 30/1. Normally, these polyplexes had a single-peak size distribution
Fig. 1. Schematic synthesis of cationic poly(urethane amide)s with Gd-DTPA complex. a) Synthesis of linear reducible cationic poly(urethane amide)s (CPUAs) by polymerization reaction between BAP and a mixture of DTPA and DTDE-PNC at varied molar ratios; b) Gd(III)-chelating of DTPA residue to yield Gd-chelated CPUAs (GdCPUAs).
184
X. Gao et al. / Materials Science and Engineering C 65 (2016) 181–187
Fig. 2. 1H NMR spectrum of CPUA15 polymer (D2O, 300 MHz).
(Fig. 3c) with a low polydispersity index b0.25 (data not shown). TEM observation exhibited a spherical morphology of GdCPUA15 polyplexes (Fig. S2). The surface charges of GdCPUA polyplexes were moderate (b+15 mV) at N/P ratios of 10/1 and 20/1 but markedly increased above +20 mV at a high N/P ratio of 30/1 (Fig. 3d). Gene release ability of GdCPUA polyplexes at different N/P ratios was analyzed by gel electrophoresis after they were treated with DTT (5 mM), mimicking an intracellular reductive environment (Fig. S3). In this case, efficient gene release from GdCPUA polyplexes could be observed, implying that they can actively mediate an efficient intracellular gene release through disulfide cleavaging and the presence of Gd-DTPA residues in these GdCPUAs has a negligible effect on gene liberation. Because GdCPUA polyplexes exhibit a few favorable gene delivery features such as nanoscale particle size, positive surface charge and efficient intracellular gene release, they are suitable candidates for non-viral gene transfection.
3.3. GdCPUAs induce efficient gene transfection in vitro In vitro transfection activity of the GdCPUAs was evaluated in SKOV3 cells using pCMV-GFP plasmid encoding for GFP gene. Fig. 4a exhibits transfection efficacy of GdCPUA15 as a function of N/P ratios in the absence or presence of 10% serum. Under serum-free condition, increasing N/P ratios from 10/1 to 30/1 led to marked increment tendency in transfection efficacy. An optimal efficiency of GdCPUA15 was obtained at an N/P ratio of 20/1. Moreover, compared to the polyplexes of 25 kDa BPEI as a positive control, those of GdCPUA15 at the same N/P ratio of 10/1 afforded higher transfection efficiency. However, the transfection ability of GdCPUA15 is compromised by the serum. As exhibited in Fig. S4, 10% serum badly decreased the amount of GFP-expressing cells induced by GdCPUA15 and BPEI. In this case, transfection efficiencies of GdCPUAs showed ~2–4-folded lower compared to those obtained under serumfree condition. Again, optimal transfection efficiency of GdCPUA15 was found at an optimal N/P ratio of 20/1, that is, ca. 6.8-folded higher than that of 25 kDa BPEI. Transfection ability of GdCPUAs (GdCPUA15, GdCPUA30, and GdCPUA50) was compared in SKOV-3 cells at the same N/P ratio of 10/1 (Fig. 4b). In general, transfection efficiency of these polymers reduced with increasing compositions of Gd-chelated DTPA residues from 15 mol% to 50 mol%, showing that integrating Gd-chelated DTPA
Table 2 Gd content and relaxivity (r1) analysis of GdCPUAs. Theoretical Gd content (wt%)a
Tested Gd content (wt%)b
r1
GdCPUA15 GdCPUA30 GdCPUA50
5% 10% 16.7%
4.8 9.2% 15.5%
11.8 11.6 11.4
b
Calculated by chemical structure. Determined by ICP assay.
3.4. T1-weighted MR imaging with GdCPUAs Since GdCPUA polymers comprise Gd-chelated residues, they were potential as T1-contrast agents for MR imaging. Thus, T1-weighted MR imaging in vitro using GdCPUA15 was conducted by preparing the solution of its polyplexes at an N/P ratio of 10/1. Fig. 6a exhibits visual contrast phantom of the solution at varied Gd concentrations. As expected, significant T1-contrast enhancement (signal brightening) was seen with increasing Gd concentrations from 0.2 to 0.8 mM. Moreover, the relaxivity of GdCPUAs, determining by the slope of 1/T1 (relaxation time) vs. Gd concentrations (Fig. 6b), was determined using a group of GdCPUA aqueous solution at varied Gd concentrations. These GdCPUAs generally displayed similar relaxivity value (Table 2) which is higher than that of DTPA-Gd complex (Magnevist), a small molecular contrast agent. The relaxivity of GdCPUA15 was ca. 2.6 times higher than that of DTPA-Gd. However, when the relaxivity was measured with GdCPUA15 polyplexes, it was lower than that obtained with GdCPUA15 (9.1 vs. 11.8 mM− 1·s− 1, Fig. 6b). Overall, GdCPUAs may serve as contrast agents for T1-weighted MR imaging. 4. Discussion
Sample code
a
residue into cationic polymers may adversely influence their transfection activity. A rational interpretation could be ascribed to the decrease in the surface charge of the polyplexes which are not favorable for efficient cellular uptake. Again, the serum could decrease transfection efficiency of these GdCPUAs by ~3–6 times as compared to that obtained under serum-free condition. Moreover, transfection activity of the GdCPUA polyplexes at an N/P ratio of 10/1 was evaluated against MCF-7, HepG2, and A549 cancer cells. Again, GdCPUA15 induced superior transfection activity to the other two polymers (Fig. 4c). Transfection efficiency of GdCPUA15 against these cells was ~2–3 times higher than that of 25 kDa BPEI. GdCPUA15 was thus used for further study. To ascertain the feasibility of GdCPUAs for gene therapy, a therapeutic small hairpin RNA (shRNA) plasmid for silencing vascular endothelial growth factor (shRNA-VEGF) was applied to prepare the polyplexes of GdCPUA15 at an N/P ratio of 10/1. ELISA analysis showed that, after in vitro transfection of SKOV-3 cells with the polyplexes, they could afford significant down-regulation of VEGF expression in the cells (Fig. 4d), that is, ca. 58.3 ± 5.6% of residual VEGF protein expression relative to the blank. Beside, this silencing level was more pronounced when compared to that induced with polyplexes of BPEI (72.6 ± 6.5%) and Lipofectamine 2000 (83.3 ± 1.5%). By contrast, no marked reduction of VEGF expression was detected with GdCPUA15 polyplexes containing pCMV-Luc (Luc NC) as a negative control. Support for VEGF silencing was also found by real time RT-PCR analysis of mRNA level of VEGF after transfection. It was found that the polyplexes afforded markedly higher mRNA silencing efficiency compared to those of BPEI and Lipofectamine 2000 (i.e. 36.3 ± 4.5% vs. 76 ± 10.5% vs. 57.3 ± 13.9% of residual mRNA level compared to the blank, Fig. S5). Cytotoxicity evaluation showed that none of the GdCPUA polyplexes containing pCMV-GFP exerted an adverse effect on metabolic activity of SKOV-3 cells after transfection (Fig. 5). Besides, GdCPUA15 itself was of low cytotoxicity in different cells (with N90% cell viability) at a high polymer concentration of 200 μg/mL (Fig. S6). By contrast, BPEI based polyplexes at an N/P ratio of 20/1 showed detectable cytotoxicity (~ 80% cell viability) and BPEI itself had high cytotoxicity (IC50 ≈ 15– 30 μg/mL) at a high polymer concentration from 25 to 100 μg/mL. These results support good cyto-compatibility of GdCPUAs.
Theranostic nano-systems can offer simultaneous imaging and therapy of cancer. As such, cancer theranostics will simplify traditional clinical operation procedure and provide physician with accurate therapy status in time. Although those inorganic nano-systems have high potential for cancer theranostics [8], their clinical translations are seriously impeded due to insufficient bio-safety evaluations. Nano-polyplexes
X. Gao et al. / Materials Science and Engineering C 65 (2016) 181–187
185
Fig. 3. Gene binding analysis of GdCPUA polymers. a) Agarose gel retardation analysis of the mixture of GdCPUA polymers and DNA as a function of N/P ratios; b) particle size and of the polyplexes of GdCPUA polymers at an N/P ratio of 10/1, 20/1 and 30/1; c) size distribution of the polyplexes of GdCPUA15 at an N/P ratio of 10/1; d) zeta-potential of the polyplexes of GdCPUA polymers at the N/P ratios.
based on degradable polymers and nucleic acids represent innovative nano-systems for cancer therapy. Therefore, in this work, we propose for theranostic cationic polymers which can condense genes into nano-polyplexes for gene delivery and chelate Gd(III) ions to form Gd complex for MR imaging. Although we and other groups present many reducible polymers for gene delivery [13,17], those polymers reported are not applicable as MR contrast agents. To our best knowledge, this is the first report on reducible polymeric gene carriers for MR imaging. Those previous studies have focused on the design of functionalized cationic polymers for improved transfection [5,12]. We recently find that 1,4-bis(3-aminopropyl)piperazine (BAP) residue is a hit chemical structure to design cationic polymers for efficient gene transfection [16]. Accordingly, in this study, we design this theranostic reducible
cationic polymer containing BAP residues (Fig. 1). Besides, we recently developed reducible cationic polyurethanes with BAP residues for cancer therapy and showed that these polyurethanes are of low cytotoxicity in vitro and in vivo [14]. Thus, in this work, we design reducible cationic poly(urethane amide)s comprising multiple urethane linkages for gene delivery. It is not well known whether Gd-chelated DPTA residue in cationic polymers has an effect on gene delivery properties. Herein, we find that increasing the amounts of Gd-chelated DTPA residue in GdCPUAs leads to large particle size and low surface charge of resulting polyplexes. The results are probably attributed to the reason that Gd-chelated DTPA complex is a neutral group which cannot contribute to gene binding. Also, steric hindrance of the complex might hinder the condensation between positively-charged GdCPUAs and negatively-charged gene. These
Fig. 4. Transfection activity study of GdCPUAs. a) Transfection efficiency of GdCPUA15 polyplexes against SKOV-3 cells as a function of N/P ratios in the absence (w/o) or presence (w/) of 10% serum; b) comparing transfection ability of three GdCPUAs against SKOV-3 cells at the same N/P ratio of 10/1 in the absence (w/o) or presence (w/) of 10% serum; c) transfection efficiency of GdCPUAs against different cell lines in the absence of serum; GdCPUA15 vs. BPEI:*P b 0.05, ***P b 0.001; d) VEGF-shRNA-containing polyplexes of GdCPUA15 afford the silencing of VEGF protein expression in SKOV-3 cells. BPEI polyplexes (N/P = 10/1) and Lipofectamine 2000 (Lipo2000) formulation were used as positive controls. ***P b 0.001.
186
X. Gao et al. / Materials Science and Engineering C 65 (2016) 181–187
Fig. 5. AlamarBlue assay showing cytotoxicity of GdCPUA based polyplexes against SKOV3 cells. a) cytotoxicity of GdCPUA15 polyplexes as a function of N/P ratios; b) cytotoxicity of the polyplexes of GdCPUAs at the same N/P ratio of 10/1. BPEI based polyplexes at N/ P = 10/1 or 20/1 were used as a control.
reasons may interpret that GdCPUA50 polyplexes possess larger size and lower surface charge than GdCPUA15 polyplexes at the same N/P ratios (Fig. 3b&d). Because large size and low surface charge are unfavorable for efficient gene transfection, GdCPUA50 polyplexes reveal inferior transfection ability in vitro to GdCPUA15 polyplexes (Fig. 4). More detailed mechanism underlying the transfection of GdCPUA polyplexes should be elucidated in future work. Notably, active gene transfection induced by GdCPUA15 is not associated with high cytotoxicity (Fig. 5), implying that Gd-DTPA chelate has a minor effect on cell viability. This low cytotoxicity may be due to the fact that the concentration of Gd-DTPA complex used for transfection is relatively low (below ~0.04 μmol/mL). Besides, intracellular degradation of GdCPUAs through disulfide cleavaging may contribute to low cytotoxicity of the polymers (Fig. S6). These results again reflect that multiple disulfide-containing cationic polymers are low cytotoxic, as confirmed by previous studies [13]. We do not perform in vivo transfection of GdCPUA-based polyplexes since their positive surface charge is unfavorable for systemic (e.g. intravenous) gene delivery. In general, cationic polyplexes are prone to aggregation in the blood, which may cause a low transfection. This may interpret that the serum seriously decreases transfection efficacy of GdCPUA15 (Fig. 4a). Further studies need to be conducted on functionalized GdCPUA system with stealth polymers such as poly(ethylene glycol) and targeting ligands for tumor-targeted therapy in vivo. Angiogenesis is a pivotal pathway to supply oxygen and nutrients for tumor growth, where vascular endothelial growth factor (VEGF) is one of the most pivotal angiogenic regulation factors. Many types of tumor cells highly express VEGF. Thus, down-regulation of VEGF expression is an efficient approach for cancer therapy. Because of their high transfection ability, VEGF-shRNA-containing polyplexes of GdCPUA15 affords more marked silencing level of VEGF expression in SKOV-3 cells as compared to those of BPEI and lipofectamine 2000. It can thus be deduced that GdCPUA15 based polyplexes provide a high possibility for tumor gene therapy. In general, Gd-chelated polymer conjugates display higher T1 relaxivity (r1) values than a low molecular weight Gd-DTPA complex because of limited motion (i.e. increased rotational correlation lifetime) of the conjugates. This principle may rationally explain that GdPCUA15 has higher r1 value than Gd-DTPA (Magnevist) (11.8 vs. 4.6 mM−1·s−1, Fig. 6b). Furthermore, upon complexation of GdPCUA15 with
Fig. 6. In vitro MR imaging with the polyplexes of GdCPUA15 as a contrast agent. a) In vitro T1-weighted imaging of GdCPUA15 polyplexes as a function of Gd concentration measured at 3.0 T and room temperature; b) reciprocal relaxation time (1/T1) of GdCPUA15, GdCPUA15/DNA, and DTPA-Gd (Magnevist) as a function of Gd concentrations. Relaxivity (r1) is calculated as the slope of the line. Gd-DTPA complex (Magnevist) is used as a positive control.
negatively-charged DNA into nano-polyplexes, this r1 value is reduced to 9.1 mM−1·s−1. This outcome is likely due to the formation of hydrophobic inner core, which hampers efficient interaction of water molecule with chelated Gd ions located in the core. It should be noted that an excess amount of GdCPUA15 is employed to prepare polyplexes. As such, those GdPCUA15 polymers loosely surrounding polyplexes can still contribute enhanced relaxivity. Thus, r1 value of GdCPUA15 polyplexes is higher than that of Gd-DTPA complex. A similar phenomenon was reported by Nakamura et al., who found that Gd-chelated poly(ethylene glycol)-b-poly(aspartic acid) polymers had higher r1 value than Gd-DTPA complex [18]. This value however reduced slightly after self-assembly of the polymer with cationic polyallylamine or protamine to form nano-complexes. 5. Conclusions In summary, we have revealed that biodegradable cationic poly(urethane amide)s (CPUAs) can be designed and prepared which comprise multiple disulfide, urethane and amide linkages as well as diethylenetriaminepentaacetic acid (DTPA) residues. Furthermore, gadolinium-chelated CPUAs (GdCPUAs) may be achieved by complexation of gadolinium (III) ions with DTPA residues of CPUAs. These GdCPUAs may serve as non-viral carriers for highly efficient gene transfection in cancer cells and as contrast agents for T1-weighted magnetic resonance imaging. This GdCPUA copolymer system represents a new platform for use in tumor theranostics. Acknowledgements This article is supported by the grants from the Shanghai Municipal Natural Science Foundation (13ZR1443600), National High-Tech R&D Program of China (2013AA032202), Fundamental Research Funds for the Central Universities, and partially by National Natural Science Funds of China (20904041, 81571655). The authors claim no confliction of interest.
X. Gao et al. / Materials Science and Engineering C 65 (2016) 181–187
Appendix A. Supplementary data Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.msec.2016.04.027. References [1] W. Walther, P.M. Schlag, Current status of gene therapy for cancer, Curr. Opin. Oncol. 25 (2013) 659–664. [2] M. Giacca, S. Zacchigna, Virus-mediated gene delivery for human gene therapy, J. Control. Release 161 (2012) 377–388. [3] H. Yin, R.L. Kanasty, A.A. Eltoukhy, A.J. Vegas, J.R. Dorkin, D.G. Anderson, Non-viral vectors for gene-based therapy, Nat. Rev. Genet. 15 (2014) 541–555. [4] J. Luten, C.F. van Nostruin, S.C. De Smedt, W.E. Hennink, Biodegradable polymers as non-viral carriers for plasmid DNA delivery, J. Control. Release 126 (2008) 97–110. [5] M.A. Mintzer, E.E. Simanek, Nonviral vectors for gene delivery, Chem. Rev. 109 (2009) 259–302. [6] E.-K. Lim, T. Kim, S. Paik, S. Haam, Y.-M. Huh, K. Lee, Nanomaterials for theranostics: recent advances and future challenges, Chem. Rev. 115 (2015) 327–394. [7] S.S. Kelkar, T.M. Reineke, Theranostics: combining imaging and therapy, Bioconjug. Chem. 22 (2011) 1879–1903. [8] J.M. Knipe, J.T. Peters, N.A. Peppas, Theranostic agents for intracellular gene delivery with spatiotemporal imaging, Nano Today 8 (2013) 21–38. [9] J.M. Bryson, K.M. Fichter, W.-J. Chu, J.-H. Lee, J. Li, L.A. Madsen, P.M. McLendon, T.M. Reineke, Polymer beacons for luminescence and magnetic resonance imaging of DNA delivery, Proc. Natl. Acad. Sci. U. S. A. 106 (2009) 16913–16918.
187
[10] Y. Wu, C.E. Carney, M. Denton, E. Hart, P. Zhao, D.N. Streblow, A.D. Sherry, M. Woods, Polymeric PARACEST MRI contrast agents as potential reporters for gene therapy, Org. Biomol. Chem. 8 (2010) 5333–5338. [11] L. Xue, S.S. Kelkar, X. Wang, J. Ma, L.A. Madsen, T.M. Reineke, A theranostic polycation containing trehalose and lanthanide chelate domains for siRNA delivery and monitoring, RSC Adv. 5 (2015) 74102–74106. [12] S. Son, R. Namgung, J. Kim, K. Singha, W.J. Kim, Bioreducible polymers for gene silencing and delivery, Acc. Chem. Res. 45 (2012) 1100–1112. [13] C. Lin, J.F.J. Engbersen, The role of the disulfide group in disulfide-based polymeric gene carriers, Expert Opin. Drug Deliv. 6 (2009) 421–439. [14] J. Cheng, X. Tang, J. Zhao, T. Shi, P. Zhao, C. Lin, Multifunctional cationic polyurethanes designed for non-viral cancer gene therapy, Acta Biomater. 30 (2016) 155–167. [15] C. Lin, Y. Song, B. Lou, P. Zhao, Dextranation of bioreducible cationic polyamide for systemic gene delivery, Bio-Med. Mater. Eng. 24 (2014) 673–682. [16] C. Zhang, R. Jin, P. Zhao, C. Lin, A family of cationic polyamides for in vitro and in vivo gene transfection, Acta Biomater. 22 (2015) 120–130. [17] Y. Song, B. Lou, P. Zhao, C. Lin, Multifunctional disulfide-based cationic dextran conjugates for intravenous gene delivery targeting ovarian cancer cells, Mol. Pharm. 11 (2014) 2250–2261. [18] E. Nakamura, K. Makino, T. Okano, T. Yamamoto, M. Yokoyama, A polymeric micelle MRI contrast agent with changeable relaxivity, J. Control. Release 114 (2006) 325–333.
Contents lists available at ScienceDirect
Materials Science and Engineering C journal homepage: www.elsevier.com/locate/msec
Biodegradable gadolinium-chelated cationic poly(urethane amide) copolymers for gene transfection and magnetic resonance imaging Xiaolong Gao a, Gangmin Wang c, Ting Shi b, Zhihong Shao a, Peng Zhao b, Donglu Shi b, Jie Ren d, Chao Lin b,⁎, Peijun Wang a,⁎ a
Department of Radiology, Tongji Hospital, Tongji University School of Medicine, Shanghai 200065, PR China The Institute for Translational Nanomedicine, Shanghai East Hospital, Institute for Biomedical Engineering and Nanoscience, Tongji University School of Medicine, Shanghai 200092, PR China Department of Urology, Huashan Hospital, Fudan University, Shanghai 200040, PR China d Institute of Nano and Biopolymeric Materials, School of Materials Science and Engineering, Tongji University, 4800 Caoan Road, Shanghai 201804, PR China b c
a r t i c l e
i n f o
Article history: Received 21 January 2016 Received in revised form 14 March 2016 Accepted 7 April 2016 Available online 13 April 2016 Keywords: Poly(urethane amide) Gene delivery Disulfide Imaging Theranostics
a b s t r a c t Theranostic nano-polyplexes containing gene and imaging agents hold a great promise for tumor diagnosis and therapy. In this work, we develop a group of new gadolinium (Gd)-chelated cationic poly(urethane amide)s for gene delivery and T1-weighted magnetic resonance (MR) imaging. Cationic poly(urethane amide)s (denoted as CPUAs) having multiple disulfide bonds, urethane and amide linkages were synthesized by stepwise polycondensation reaction between 1,4-bis(3-aminopropyl)piperazine and a mixture of di(4-nitrophenyl)-2, 2′dithiodiethanocarbonate (DTDE-PNC) and diethylenetriaminepentaacetic acid (DTPA) dianhydride at varied molar ratios. Then, Gd-chelated CPUAs (denoted as GdCPUAs) were produced by chelating Gd(III) ions with DTPA residues of CPUAs. These GdCPUAs could condense gene into nanosized and positively-charged polyplexes in a physiological condition and, however, liberated gene in an intracellular reductive environment. In vitro transfection experiments revealed that the GdCPUA at a DTDE-PNC/DTPA residue molar ratio of 85/15 induced the highest transfection efficiency in different cancer cells. This efficiency was higher than that yielded with 25 kDa branched polyethylenimine as a positive control. GdCPUAs and their polyplexes exhibited low cytotoxicity when an optimal transfection activity was detected. Moreover, GdCPUAs may serve as contrast agents for T1weighted magnetic resonance imaging. The results of this work indicate that biodegradable Gd-chelated cationic poly(urethane amide) copolymers have high potential for tumor theranostics. © 2016 Elsevier B.V. All rights reserved.
1. Introduction Nucleic acid-based cancer therapy is an alternative strategy to routine chemotherapy [1]. For successful nucleic acid-based cancer therapy, highly efficient delivery of nucleic acids into targeted somatic cells in a bio-safe way is still a challenge. Although engineered viral vectors encoded with nucleic acids can robustly infect human cells, they are associated with a few uncertain bio-safety issues such as uncontrolled mutagenesis, immunogenicity and oncogenicity, thus impeding their clinical translations [2]. Alternatively, non-viral vectors based on cationic polymers reveal high potential for relatively safe gene delivery as a result of their low toxicity, facile synthesis and high nucleic acid-carrying capacity [3]. Accordingly, over the past two decades, a number of polymeric vectors for non-viral gene delivery have been widely studied [4]. A lot of studies have confirmed a massive promise of cationic polymernucleic acid assembled complexes (also called as polyplexes) for cancer gene therapy [5]. ⁎ Corresponding authors. E-mail addresses: [email protected] (C. Lin), [email protected] (P. Wang).
http://dx.doi.org/10.1016/j.msec.2016.04.027 0928-4931/© 2016 Elsevier B.V. All rights reserved.
The development of nano-carriers for tumor imaging and therapy has received increasing interest in recent years [6]. From clinical point of view, such dual-function nano-carriers allow clinical physicians to accurately evaluate anti-cancer efficacy and choose balanced drug dose in time. As such, a variety of theranostic nano-carriers have been designed which combine both therapeutic functions (chemotherapy and gene therapy) and single-/multi-imaging modalities [7]. It should be however noted that most of these theranostic nano-carriers reported are based on inorganic nanoparticles [8]. Only few literatures have focused on theranostic polymers for gene therapy and imaging such as magnetic resonance (MR) imaging. For example, Bryson et al. reported on Gdchelated polycations containing oligoamines for gene delivery and MR imaging [9]. In another work, Wu et al. designed polymethacrylamidebased cationic polymers having europium-labeled residues for MR imaging [10]. Recently, Xue et al. showed a group of Gd-chelated polycations for siRNA delivery [11]. However, these theranostic polymers usually lack biodegradability, thus probably hampering their potential of clinical translations. An ideal theranostic polymer should offer efficient gene transfection and accurate imaging. Importantly, such polymer should be essentially
182
X. Gao et al. / Materials Science and Engineering C 65 (2016) 181–187
low toxic against the cells and the body. In order to generate efficient and low-toxic polymers for gene delivery, disulfide-containing (reducible) cationic polymers for gene delivery have recently received rapid growing interest [12]. A lot of studies confirm that reducible polyethylenimine, polylysine and polyamides may induce superior transfection efficacy to their non-reducible counterparts and, meanwhile, are associated with a minor cytotoxicity [12]. This low toxicity is due to intracellular cleavaging of disulfide linkage by glutathione (0.5–10 mM) [13], leading to intracellular degradation of reducible polymers and thus reduced toxicity. Reducible polymers thus represent suitable non-viral carriers for safe and efficient gene delivery. However, to our best knowledge, no report have presented on reducible cationic polymers for gene delivery and MR imaging. It would be thus meaningful to develop such reducible cationic polymers and evaluate their feasibility for theranostics. We recently reported on reducible cationic polyurethanes synthesized by non-isocyanate chemistry and found that these polyurethanes possessed low toxicity and high ability for gene transfection towards tumor cells [14]. Herein, we aim to design and prepare theranostic reducible cationic polyurethanes for gene delivery and imaging. To this end, a group of linear reducible cationic poly(urethane amide)s (denoted as CPUAs) are designed and prepared that comprise multiple disulfide bonds, tertiary amine groups and diethylenetriaminepentaacetic acid (DTPA) residues. The DTPA residue allows for complexation with Gd(III) ions to generate Gd-chelated CPUAs (denoted as GdCPUAs). We hypothesized that GdCPUAs would be practical as non-viral vectors for gene delivery and MR imaging. We present synthesis and characterization of GdCPUAs and the properties of GdCPUA-based polyplexes in terms of particle size, surface charge, magnetism, and gene release ability. Transfection ability and cytotoxicity of GdCPUA polyplexes are evaluated in vitro against cancer cells. Moreover, T1weighted MR imaging by using GdCPUAs as contrast agents is investigated. 2. Materials and methods 2.1. Materials All chemicals were used directly as received unless otherwise statement. 2,2′-Dithiodiethanol (DTDE), gadolinium chloride (GdCl3), 4-nitrophenyl chloroformate (PNC), dithiothreitol (DTT), 2-[4-(2-hydroxyethyl)piperazin-1-yl] ethane sulfonic acid (HEPES), branched polyethylenimine (BPEI, Mw = 25 kDa), 1,4-diethylenetriaminepentaacetic acid (DTPA) dianhydride, and bis(3-aminopropyl)piperazine (BAP) were ordered from Sigma-Aldrich Co. (USA). Lipofectamine 2000 transfection agent was ordered from Life Technology (USA). The plasmids encoding for green fluorescent protein (GFP) and luciferase (Luc), also termed as pCMV-GFP and pCMV-Luc, respectively, were obtained from plasmid factory (Bielefeld, Germany). shRNA-VEGF plasmid was purchased from InvivoGen (USA). Di(4-nitrophenyl)-2,2′-dithiodiethanocarbonate (DTDE-PNC) was prepared according to our previous report [14]. 2.2. Synthesis and characterization of Gd-chelated polymers Gd-chelated reducible cationic poly(urethane amide) (GdCPUA) polymers was obtained by a two-step procedure. First, a group of cationic poly(urethane amide)s (CPUAs) were obtained by polymerization reaction between BAP and a mixture of DTDE-PNC and DTPA dianhydride at varied molar ratios (i.e. 85/15, 70/30, 50/50). In a typical procedure for the synthesis of CPUA30 at a DTDE-PNC/DTPA anhydride ratio of 70/30, DTDE-PNC (0.5 g, 1.03 mmol), BAP (0.30 g, 1.47 mmol) and DTPA anhydride (0.17 g, 0.44 mmol) were dissolved in anhydrous DMSO (5 mL) as a solvent in a brown-colour round flask. After running for 5 days at 40 °C in nitrogen atmosphere, the reaction was run for another 2 days after adding an excess amount of BAP (0.03 g). Finally,
resulting solution was purified by ultrafiltration process (3 kDa cutoff) with acidic deionized water (3 × 5 L, pH 4.5). CPUA30 was obtained as a white solid after freeze-drying (yield: 0.3 g, 45%). Chemical composition of CPUAs was tested by 1H NMR spectrum meter (Varian, USA). Next, GdCPUAs were yielded by Gd(III) chelation with DTPA residues in CPUAs. As an example for the preparation of GdCPUA30, CPUA30 (50 mg) was dissolved in 20 mL of deionized water. Next, GdCl3 (~ 4 mmol) in deionized water (5 mL) was dropwise added into CPUA30 solution. During mixing process, the solution became turbid but transparent again after pH was set to pH 6.5 with 4 M HCl. The resulting solution was stirred for 4 h and then transferred in dialysis bag (1 kDa cut-off) for exhaustive dialysis in deionized water at pH ~ 6. GdCPUA30 was obtained as a solid powder after freezing-drying (yield: 0.35 g, 67%). GdCPUAs were analyzed by inductively coupled plasma (ICP) system to determine Gd content after dissolving these polymers in 4 M H2NO3. 2.3. Preparation and characterization of GdCPUA based polyplexes The polyplexes of GdCPUA copolymers were prepared by gently mixing plasmid DNA with GdCPUAs at N/P ratios in HEPES buffer (20 mM, pH 7.4) [15]. Particle size and surface charge of GdCPUA polyplexes were determined by dynamic light scattering (DLS) analysis (Nanosizer NS90, Malvern, UK) at 25 °C. Besides, the polyplex solution was analyzed by agarose gel electrophoresis using the method reported previously [16]. To characterize gene release behavior in a reductive environment, the polyplexes were co-incubated with DTT (5 mM) in the HEPES buffer for 30 min, analyzed by the gel electrophoresis, DNA band was visualized by TanonGel Image system (Shanghai, P.R. China). Relaxation time (T1) of GdCPUAs and their polyplexes was measured using a contrast agent relaxation rate analysis system (PQ100, NIUMAG, Shanghai, P.R. China) according to the manufacture's protocol. A clinical T1-contrast agent, Magnevist (Gd-DTPA complex), was used as a positive control. T1-weighted MR imaging with the polyplexes of GdCPUA copolymers was performed by running Siemens Magnetom Verio 3.0 T MR instrument. In brief, the polyplexes were prepared in HEPES buffer (pH 7.4) in EP tube at different Gd concentrations ranging from 0.1 mM to 0.8 mM. MR imaging was obtained by placing the tubes in an in-house built 10-cm RF coil installed in the instrument according to the manufacturer's instruction. 2.4. Cell culture, transfection test and cytotoxicity assay in vitro SKOV-3 cells (ATCC, USA) were cultured in Mccyo's-5α medium culture medium (GIBCO) having 10% FBS, 100 U/mL penicillin and streptomycin (GIBCO). MCF-7, HepG2 and A549 cells were cultured in DMEM complete medium. The cells were grown at 37 °C in cell incubator (Thermo Scientific, Waltham, MA, USA) supplied with 5% CO2. Transfection activity of GdCPUA copolymers in tumor cells was evaluated according to our previous transfection protocol [16]. In brief, the cells (5–7 × 104 cells/well) were cultured in a 24-well plate for 24 h until 60–70% cell confluence was reached. GdCPUA polyplexes containing pCMV-GFP plasmid (1 μg) in a HEPES buffer (20 mM, pH 7.4) were then prepared at varied N/P ratios. The transfection tests were performed by adding the polyplex solution in each well and coincubating with the cells for 1 h in the absence or presence of 10% FBS. The cells were cultured in complete medium for another 24 h. Besides, a formulation of BPEI polyplexes at an N/P ratio of 10/1 and lipofectamine 2000 were used as positive controls. Those cells without transfection (untreated cells) were considered as a blank (negative) control. The same protocol was applied in the transfection with GdCPUA polyplexes containing shRNA-VEGF plasmid. VEGF protein and mRNA level were analyzed by using ELISA kit (Abcam, USA) and RT-PCR, respectively. Transfection efficiency was determined by testing relative fluorescence intensity of GFP expressed in the cells according to our previous
X. Gao et al. / Materials Science and Engineering C 65 (2016) 181–187
report [16]. The efficiency was presented with the unit of fluorescence intensity (FI) per mg protein. Cytotoxicity of GdCPUAs and their polyplexes was evaluated by determining metabolic activity of the cells with AlamarBlue assay kit (Life Technology). Untreated cells (blank control) were set as 100% cell viability. All results are given as mean ± standard deviation (SD) from triplicate samples (n = 3). 2.5. Statistical analysis
Table 1 Characteristics of cationic poly(urethane amide)s (CPUAs). Sample codea
DTDE-PNC/DTPA anhydride feed molar ratio
DTDE-PNC/DTPA anhydride residue molar ratiob
Mw (kDa)c
PDIc
CPUA15 CPUA30 CPUA50
85/15 70/30 50/50
84/16 68/32 55/45
39.0 42.0 45.0
1.5 1.6 1.8
a
Student's t-test analysis is used in this work. The difference is considered to be statistically significant at P b 0.05. 3. Results 3.1. Synthesis of Gd-loaded cationic poly(urethane amide)s (GdCPUAs) In this work, a group of reducible cationic polyurethanes with diethylenetriaminepentaacetic acid (DTPA) residues for Gd(III) ion chelation were synthesized by stepwise polycondensation reaction between 1,4-bis(3-aminopropyl)piperazine (BAP) and the mixture of DTPA anhydride and di(4-nitrophenyl)-2, 2′-dithiodiethanocarbonate (DTDE-PNC) (Fig. 1a). An equivalent molar ratio between BAP and DTDE-PNC/DTPA anhydride mixture was used to afford resulting polymers with the highest molecular weight. The reaction was terminated with an excess amount of BAP to consume any unreacted 4nitrophenyl group. In the reaction, the molar ratio of DTDE-PNC and DTPA anhydride was set at 85/15, 70/30 and 50/50, respectively. As a result, three reducible copolymers were obtained which comprise disulfide linkages, tertiary amino groups and urethane/amide linkages, and they are termed as cationic poly(urethane amide)s (denoted as CPUA15, CPUA30, and CPUA50, respectively, Table 1). These CPUAs were further characterized. 1H NMR analysis of the CPUAs indicated that their chemical compositions were in good line with expected chemical structures. Fig. 2 shows typical 1H NMR spectrum of CPUA15. Besides, the molar ratio between DTDE residue and DTPA residue could be determined by comparing the integral area of two peaks at δ 4.2 and δ 3.45, attributing to the proton signal of DTDE residue (CH2CH2SS) and DTPA residue (NCH2CH2N), respectively. As revealed in Table 1, the molar ratio of DTDE/DTPA residue was in accordance with feed monomer ratio of DTDE-PNC/DTPA anhydride, suggesting that chemical composition of these CPUAs may be readily adjusted by monomer feed ratio. GPC analysis of these CPUAs showed that they had single-peak molecular weight distribution (Fig. S1) with average weight molecular weight (Mw) ranging from 39 k to 45 kDa and narrow polydispersity index (PDI) of 1.5–1.8. Next, Gd-chelated CPUAs (denoted as GdCPUAs) were prepared by complexation of Gd(III) ions with DTPA residues of three CPUAs
183
b c
These copolymers are coded with feed molar ratio of DTPA anhydride. Determined by 1H NMR. Determined by GPC analysis.
(Fig. 1b) and obtained as solid powder after dialysis and freeze-drying. These GdCPUAs were termed as GdCPUA15, GdCPUA30 and GdCPUA50, respectively. ICP analysis indicated that Gd contents of as-prepared GdCPUAs were in line with theoretical Gd contents (Table 2), implying an efficient complexation of Gd(III) ions with DTPA residues of CPUAs. A solubility test revealed that, at a high polymer concentration of 1 mg/mL, these GdCPUAs were soluble in PBS buffer and completed cell culture medium. Overall, these data indicate successful availability of GdCPUAs. 3.2. Gene condensation and release behavior To ascertain whether GdCPUAs may be used for gene delivery, electrostatic binding between the polymers and plasmid DNA was characterized by electrophoresis gel analysis at different nitrogen/phosphate (N/ P) ratios. It was found that GdCPUAs efficiently retarded the mobility of DNA at and above an N/P ratio of 10/1, implying their binding with DNA (Fig. 3a). Besides, it appeared that retardation ability of these polymers became weaker with increasing amount of Gd-chelated DTPA residues from 15% to 50%. For example, GdCPUA15 could totally hinder DNA mobility at a low N/P ratio of 2.5/1 whereas GdCPUA30 and GdCPUA50 had to afford DNA retardation at a higher N/P ratio of 5/1 and 10/1, respectively. Furthermore, DNA binding ability of GdCPUA polymers was characterized by dynamic light scattering (DLS) analysis. Particle size (in average diameter) of the polyplexes of GdCPUAs was indicated as a function of N/P ratios ranging from 10/1 to 30/1 (Fig. 3b). At a low N/P ratio of 10/1 or 20/1, particle size of GdCPUA polyplexes displayed the following tendency: GdCPUA15 ≈ GdCPUA30 b GdCPUA50, implying that incorporating Gd-chelated DTPA residues into CPUAs may increase the size of the polyplexes. Only at a high N/P ratio of 30/1, this effect was not pronounced due to enhanced gene binding force. At this ratio, the polyplexes of GdCPUAs displayed small size b200 nm. This point was also supported by the results that the particle size of the polyplexes of GdCPUAs markedly reduced with increasing N/P ratios from 10/1 to 30/1. Normally, these polyplexes had a single-peak size distribution
Fig. 1. Schematic synthesis of cationic poly(urethane amide)s with Gd-DTPA complex. a) Synthesis of linear reducible cationic poly(urethane amide)s (CPUAs) by polymerization reaction between BAP and a mixture of DTPA and DTDE-PNC at varied molar ratios; b) Gd(III)-chelating of DTPA residue to yield Gd-chelated CPUAs (GdCPUAs).
184
X. Gao et al. / Materials Science and Engineering C 65 (2016) 181–187
Fig. 2. 1H NMR spectrum of CPUA15 polymer (D2O, 300 MHz).
(Fig. 3c) with a low polydispersity index b0.25 (data not shown). TEM observation exhibited a spherical morphology of GdCPUA15 polyplexes (Fig. S2). The surface charges of GdCPUA polyplexes were moderate (b+15 mV) at N/P ratios of 10/1 and 20/1 but markedly increased above +20 mV at a high N/P ratio of 30/1 (Fig. 3d). Gene release ability of GdCPUA polyplexes at different N/P ratios was analyzed by gel electrophoresis after they were treated with DTT (5 mM), mimicking an intracellular reductive environment (Fig. S3). In this case, efficient gene release from GdCPUA polyplexes could be observed, implying that they can actively mediate an efficient intracellular gene release through disulfide cleavaging and the presence of Gd-DTPA residues in these GdCPUAs has a negligible effect on gene liberation. Because GdCPUA polyplexes exhibit a few favorable gene delivery features such as nanoscale particle size, positive surface charge and efficient intracellular gene release, they are suitable candidates for non-viral gene transfection.
3.3. GdCPUAs induce efficient gene transfection in vitro In vitro transfection activity of the GdCPUAs was evaluated in SKOV3 cells using pCMV-GFP plasmid encoding for GFP gene. Fig. 4a exhibits transfection efficacy of GdCPUA15 as a function of N/P ratios in the absence or presence of 10% serum. Under serum-free condition, increasing N/P ratios from 10/1 to 30/1 led to marked increment tendency in transfection efficacy. An optimal efficiency of GdCPUA15 was obtained at an N/P ratio of 20/1. Moreover, compared to the polyplexes of 25 kDa BPEI as a positive control, those of GdCPUA15 at the same N/P ratio of 10/1 afforded higher transfection efficiency. However, the transfection ability of GdCPUA15 is compromised by the serum. As exhibited in Fig. S4, 10% serum badly decreased the amount of GFP-expressing cells induced by GdCPUA15 and BPEI. In this case, transfection efficiencies of GdCPUAs showed ~2–4-folded lower compared to those obtained under serumfree condition. Again, optimal transfection efficiency of GdCPUA15 was found at an optimal N/P ratio of 20/1, that is, ca. 6.8-folded higher than that of 25 kDa BPEI. Transfection ability of GdCPUAs (GdCPUA15, GdCPUA30, and GdCPUA50) was compared in SKOV-3 cells at the same N/P ratio of 10/1 (Fig. 4b). In general, transfection efficiency of these polymers reduced with increasing compositions of Gd-chelated DTPA residues from 15 mol% to 50 mol%, showing that integrating Gd-chelated DTPA
Table 2 Gd content and relaxivity (r1) analysis of GdCPUAs. Theoretical Gd content (wt%)a
Tested Gd content (wt%)b
r1
GdCPUA15 GdCPUA30 GdCPUA50
5% 10% 16.7%
4.8 9.2% 15.5%
11.8 11.6 11.4
b
Calculated by chemical structure. Determined by ICP assay.
3.4. T1-weighted MR imaging with GdCPUAs Since GdCPUA polymers comprise Gd-chelated residues, they were potential as T1-contrast agents for MR imaging. Thus, T1-weighted MR imaging in vitro using GdCPUA15 was conducted by preparing the solution of its polyplexes at an N/P ratio of 10/1. Fig. 6a exhibits visual contrast phantom of the solution at varied Gd concentrations. As expected, significant T1-contrast enhancement (signal brightening) was seen with increasing Gd concentrations from 0.2 to 0.8 mM. Moreover, the relaxivity of GdCPUAs, determining by the slope of 1/T1 (relaxation time) vs. Gd concentrations (Fig. 6b), was determined using a group of GdCPUA aqueous solution at varied Gd concentrations. These GdCPUAs generally displayed similar relaxivity value (Table 2) which is higher than that of DTPA-Gd complex (Magnevist), a small molecular contrast agent. The relaxivity of GdCPUA15 was ca. 2.6 times higher than that of DTPA-Gd. However, when the relaxivity was measured with GdCPUA15 polyplexes, it was lower than that obtained with GdCPUA15 (9.1 vs. 11.8 mM− 1·s− 1, Fig. 6b). Overall, GdCPUAs may serve as contrast agents for T1-weighted MR imaging. 4. Discussion
Sample code
a
residue into cationic polymers may adversely influence their transfection activity. A rational interpretation could be ascribed to the decrease in the surface charge of the polyplexes which are not favorable for efficient cellular uptake. Again, the serum could decrease transfection efficiency of these GdCPUAs by ~3–6 times as compared to that obtained under serum-free condition. Moreover, transfection activity of the GdCPUA polyplexes at an N/P ratio of 10/1 was evaluated against MCF-7, HepG2, and A549 cancer cells. Again, GdCPUA15 induced superior transfection activity to the other two polymers (Fig. 4c). Transfection efficiency of GdCPUA15 against these cells was ~2–3 times higher than that of 25 kDa BPEI. GdCPUA15 was thus used for further study. To ascertain the feasibility of GdCPUAs for gene therapy, a therapeutic small hairpin RNA (shRNA) plasmid for silencing vascular endothelial growth factor (shRNA-VEGF) was applied to prepare the polyplexes of GdCPUA15 at an N/P ratio of 10/1. ELISA analysis showed that, after in vitro transfection of SKOV-3 cells with the polyplexes, they could afford significant down-regulation of VEGF expression in the cells (Fig. 4d), that is, ca. 58.3 ± 5.6% of residual VEGF protein expression relative to the blank. Beside, this silencing level was more pronounced when compared to that induced with polyplexes of BPEI (72.6 ± 6.5%) and Lipofectamine 2000 (83.3 ± 1.5%). By contrast, no marked reduction of VEGF expression was detected with GdCPUA15 polyplexes containing pCMV-Luc (Luc NC) as a negative control. Support for VEGF silencing was also found by real time RT-PCR analysis of mRNA level of VEGF after transfection. It was found that the polyplexes afforded markedly higher mRNA silencing efficiency compared to those of BPEI and Lipofectamine 2000 (i.e. 36.3 ± 4.5% vs. 76 ± 10.5% vs. 57.3 ± 13.9% of residual mRNA level compared to the blank, Fig. S5). Cytotoxicity evaluation showed that none of the GdCPUA polyplexes containing pCMV-GFP exerted an adverse effect on metabolic activity of SKOV-3 cells after transfection (Fig. 5). Besides, GdCPUA15 itself was of low cytotoxicity in different cells (with N90% cell viability) at a high polymer concentration of 200 μg/mL (Fig. S6). By contrast, BPEI based polyplexes at an N/P ratio of 20/1 showed detectable cytotoxicity (~ 80% cell viability) and BPEI itself had high cytotoxicity (IC50 ≈ 15– 30 μg/mL) at a high polymer concentration from 25 to 100 μg/mL. These results support good cyto-compatibility of GdCPUAs.
Theranostic nano-systems can offer simultaneous imaging and therapy of cancer. As such, cancer theranostics will simplify traditional clinical operation procedure and provide physician with accurate therapy status in time. Although those inorganic nano-systems have high potential for cancer theranostics [8], their clinical translations are seriously impeded due to insufficient bio-safety evaluations. Nano-polyplexes
X. Gao et al. / Materials Science and Engineering C 65 (2016) 181–187
185
Fig. 3. Gene binding analysis of GdCPUA polymers. a) Agarose gel retardation analysis of the mixture of GdCPUA polymers and DNA as a function of N/P ratios; b) particle size and of the polyplexes of GdCPUA polymers at an N/P ratio of 10/1, 20/1 and 30/1; c) size distribution of the polyplexes of GdCPUA15 at an N/P ratio of 10/1; d) zeta-potential of the polyplexes of GdCPUA polymers at the N/P ratios.
based on degradable polymers and nucleic acids represent innovative nano-systems for cancer therapy. Therefore, in this work, we propose for theranostic cationic polymers which can condense genes into nano-polyplexes for gene delivery and chelate Gd(III) ions to form Gd complex for MR imaging. Although we and other groups present many reducible polymers for gene delivery [13,17], those polymers reported are not applicable as MR contrast agents. To our best knowledge, this is the first report on reducible polymeric gene carriers for MR imaging. Those previous studies have focused on the design of functionalized cationic polymers for improved transfection [5,12]. We recently find that 1,4-bis(3-aminopropyl)piperazine (BAP) residue is a hit chemical structure to design cationic polymers for efficient gene transfection [16]. Accordingly, in this study, we design this theranostic reducible
cationic polymer containing BAP residues (Fig. 1). Besides, we recently developed reducible cationic polyurethanes with BAP residues for cancer therapy and showed that these polyurethanes are of low cytotoxicity in vitro and in vivo [14]. Thus, in this work, we design reducible cationic poly(urethane amide)s comprising multiple urethane linkages for gene delivery. It is not well known whether Gd-chelated DPTA residue in cationic polymers has an effect on gene delivery properties. Herein, we find that increasing the amounts of Gd-chelated DTPA residue in GdCPUAs leads to large particle size and low surface charge of resulting polyplexes. The results are probably attributed to the reason that Gd-chelated DTPA complex is a neutral group which cannot contribute to gene binding. Also, steric hindrance of the complex might hinder the condensation between positively-charged GdCPUAs and negatively-charged gene. These
Fig. 4. Transfection activity study of GdCPUAs. a) Transfection efficiency of GdCPUA15 polyplexes against SKOV-3 cells as a function of N/P ratios in the absence (w/o) or presence (w/) of 10% serum; b) comparing transfection ability of three GdCPUAs against SKOV-3 cells at the same N/P ratio of 10/1 in the absence (w/o) or presence (w/) of 10% serum; c) transfection efficiency of GdCPUAs against different cell lines in the absence of serum; GdCPUA15 vs. BPEI:*P b 0.05, ***P b 0.001; d) VEGF-shRNA-containing polyplexes of GdCPUA15 afford the silencing of VEGF protein expression in SKOV-3 cells. BPEI polyplexes (N/P = 10/1) and Lipofectamine 2000 (Lipo2000) formulation were used as positive controls. ***P b 0.001.
186
X. Gao et al. / Materials Science and Engineering C 65 (2016) 181–187
Fig. 5. AlamarBlue assay showing cytotoxicity of GdCPUA based polyplexes against SKOV3 cells. a) cytotoxicity of GdCPUA15 polyplexes as a function of N/P ratios; b) cytotoxicity of the polyplexes of GdCPUAs at the same N/P ratio of 10/1. BPEI based polyplexes at N/ P = 10/1 or 20/1 were used as a control.
reasons may interpret that GdCPUA50 polyplexes possess larger size and lower surface charge than GdCPUA15 polyplexes at the same N/P ratios (Fig. 3b&d). Because large size and low surface charge are unfavorable for efficient gene transfection, GdCPUA50 polyplexes reveal inferior transfection ability in vitro to GdCPUA15 polyplexes (Fig. 4). More detailed mechanism underlying the transfection of GdCPUA polyplexes should be elucidated in future work. Notably, active gene transfection induced by GdCPUA15 is not associated with high cytotoxicity (Fig. 5), implying that Gd-DTPA chelate has a minor effect on cell viability. This low cytotoxicity may be due to the fact that the concentration of Gd-DTPA complex used for transfection is relatively low (below ~0.04 μmol/mL). Besides, intracellular degradation of GdCPUAs through disulfide cleavaging may contribute to low cytotoxicity of the polymers (Fig. S6). These results again reflect that multiple disulfide-containing cationic polymers are low cytotoxic, as confirmed by previous studies [13]. We do not perform in vivo transfection of GdCPUA-based polyplexes since their positive surface charge is unfavorable for systemic (e.g. intravenous) gene delivery. In general, cationic polyplexes are prone to aggregation in the blood, which may cause a low transfection. This may interpret that the serum seriously decreases transfection efficacy of GdCPUA15 (Fig. 4a). Further studies need to be conducted on functionalized GdCPUA system with stealth polymers such as poly(ethylene glycol) and targeting ligands for tumor-targeted therapy in vivo. Angiogenesis is a pivotal pathway to supply oxygen and nutrients for tumor growth, where vascular endothelial growth factor (VEGF) is one of the most pivotal angiogenic regulation factors. Many types of tumor cells highly express VEGF. Thus, down-regulation of VEGF expression is an efficient approach for cancer therapy. Because of their high transfection ability, VEGF-shRNA-containing polyplexes of GdCPUA15 affords more marked silencing level of VEGF expression in SKOV-3 cells as compared to those of BPEI and lipofectamine 2000. It can thus be deduced that GdCPUA15 based polyplexes provide a high possibility for tumor gene therapy. In general, Gd-chelated polymer conjugates display higher T1 relaxivity (r1) values than a low molecular weight Gd-DTPA complex because of limited motion (i.e. increased rotational correlation lifetime) of the conjugates. This principle may rationally explain that GdPCUA15 has higher r1 value than Gd-DTPA (Magnevist) (11.8 vs. 4.6 mM−1·s−1, Fig. 6b). Furthermore, upon complexation of GdPCUA15 with
Fig. 6. In vitro MR imaging with the polyplexes of GdCPUA15 as a contrast agent. a) In vitro T1-weighted imaging of GdCPUA15 polyplexes as a function of Gd concentration measured at 3.0 T and room temperature; b) reciprocal relaxation time (1/T1) of GdCPUA15, GdCPUA15/DNA, and DTPA-Gd (Magnevist) as a function of Gd concentrations. Relaxivity (r1) is calculated as the slope of the line. Gd-DTPA complex (Magnevist) is used as a positive control.
negatively-charged DNA into nano-polyplexes, this r1 value is reduced to 9.1 mM−1·s−1. This outcome is likely due to the formation of hydrophobic inner core, which hampers efficient interaction of water molecule with chelated Gd ions located in the core. It should be noted that an excess amount of GdCPUA15 is employed to prepare polyplexes. As such, those GdPCUA15 polymers loosely surrounding polyplexes can still contribute enhanced relaxivity. Thus, r1 value of GdCPUA15 polyplexes is higher than that of Gd-DTPA complex. A similar phenomenon was reported by Nakamura et al., who found that Gd-chelated poly(ethylene glycol)-b-poly(aspartic acid) polymers had higher r1 value than Gd-DTPA complex [18]. This value however reduced slightly after self-assembly of the polymer with cationic polyallylamine or protamine to form nano-complexes. 5. Conclusions In summary, we have revealed that biodegradable cationic poly(urethane amide)s (CPUAs) can be designed and prepared which comprise multiple disulfide, urethane and amide linkages as well as diethylenetriaminepentaacetic acid (DTPA) residues. Furthermore, gadolinium-chelated CPUAs (GdCPUAs) may be achieved by complexation of gadolinium (III) ions with DTPA residues of CPUAs. These GdCPUAs may serve as non-viral carriers for highly efficient gene transfection in cancer cells and as contrast agents for T1-weighted magnetic resonance imaging. This GdCPUA copolymer system represents a new platform for use in tumor theranostics. Acknowledgements This article is supported by the grants from the Shanghai Municipal Natural Science Foundation (13ZR1443600), National High-Tech R&D Program of China (2013AA032202), Fundamental Research Funds for the Central Universities, and partially by National Natural Science Funds of China (20904041, 81571655). The authors claim no confliction of interest.
X. Gao et al. / Materials Science and Engineering C 65 (2016) 181–187
Appendix A. Supplementary data Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.msec.2016.04.027. References [1] W. Walther, P.M. Schlag, Current status of gene therapy for cancer, Curr. Opin. Oncol. 25 (2013) 659–664. [2] M. Giacca, S. Zacchigna, Virus-mediated gene delivery for human gene therapy, J. Control. Release 161 (2012) 377–388. [3] H. Yin, R.L. Kanasty, A.A. Eltoukhy, A.J. Vegas, J.R. Dorkin, D.G. Anderson, Non-viral vectors for gene-based therapy, Nat. Rev. Genet. 15 (2014) 541–555. [4] J. Luten, C.F. van Nostruin, S.C. De Smedt, W.E. Hennink, Biodegradable polymers as non-viral carriers for plasmid DNA delivery, J. Control. Release 126 (2008) 97–110. [5] M.A. Mintzer, E.E. Simanek, Nonviral vectors for gene delivery, Chem. Rev. 109 (2009) 259–302. [6] E.-K. Lim, T. Kim, S. Paik, S. Haam, Y.-M. Huh, K. Lee, Nanomaterials for theranostics: recent advances and future challenges, Chem. Rev. 115 (2015) 327–394. [7] S.S. Kelkar, T.M. Reineke, Theranostics: combining imaging and therapy, Bioconjug. Chem. 22 (2011) 1879–1903. [8] J.M. Knipe, J.T. Peters, N.A. Peppas, Theranostic agents for intracellular gene delivery with spatiotemporal imaging, Nano Today 8 (2013) 21–38. [9] J.M. Bryson, K.M. Fichter, W.-J. Chu, J.-H. Lee, J. Li, L.A. Madsen, P.M. McLendon, T.M. Reineke, Polymer beacons for luminescence and magnetic resonance imaging of DNA delivery, Proc. Natl. Acad. Sci. U. S. A. 106 (2009) 16913–16918.
187
[10] Y. Wu, C.E. Carney, M. Denton, E. Hart, P. Zhao, D.N. Streblow, A.D. Sherry, M. Woods, Polymeric PARACEST MRI contrast agents as potential reporters for gene therapy, Org. Biomol. Chem. 8 (2010) 5333–5338. [11] L. Xue, S.S. Kelkar, X. Wang, J. Ma, L.A. Madsen, T.M. Reineke, A theranostic polycation containing trehalose and lanthanide chelate domains for siRNA delivery and monitoring, RSC Adv. 5 (2015) 74102–74106. [12] S. Son, R. Namgung, J. Kim, K. Singha, W.J. Kim, Bioreducible polymers for gene silencing and delivery, Acc. Chem. Res. 45 (2012) 1100–1112. [13] C. Lin, J.F.J. Engbersen, The role of the disulfide group in disulfide-based polymeric gene carriers, Expert Opin. Drug Deliv. 6 (2009) 421–439. [14] J. Cheng, X. Tang, J. Zhao, T. Shi, P. Zhao, C. Lin, Multifunctional cationic polyurethanes designed for non-viral cancer gene therapy, Acta Biomater. 30 (2016) 155–167. [15] C. Lin, Y. Song, B. Lou, P. Zhao, Dextranation of bioreducible cationic polyamide for systemic gene delivery, Bio-Med. Mater. Eng. 24 (2014) 673–682. [16] C. Zhang, R. Jin, P. Zhao, C. Lin, A family of cationic polyamides for in vitro and in vivo gene transfection, Acta Biomater. 22 (2015) 120–130. [17] Y. Song, B. Lou, P. Zhao, C. Lin, Multifunctional disulfide-based cationic dextran conjugates for intravenous gene delivery targeting ovarian cancer cells, Mol. Pharm. 11 (2014) 2250–2261. [18] E. Nakamura, K. Makino, T. Okano, T. Yamamoto, M. Yokoyama, A polymeric micelle MRI contrast agent with changeable relaxivity, J. Control. Release 114 (2006) 325–333.