Effect of monomer sequence of poly(histidine/lysine) catiomers on gene packing capacity and delivery efficiency

Abstract

Polymeric catiomers, which can mimic viruses for gene packing and delivery, have received considerable attention as nonviral vectors for gene therapy. Inspired by the critically important role of the sequence of structural units in various biomolecules, including DNA and protein, in this study, we intend to investigate how the monomer sequence of a polymeric catiomer affects its gene packing capacity and delivery efficiency. The well-documented poly(histidine-co-lysine) was chosen as the scaffold for gene carrier. Four reducible polycations (RPCs) based on sequence-regulated peptides monomers were synthesized. Chemical parameters (namely, composition and molecular weight) of four RPCs were controlled at comparable level except for the monomer sequence. All of the RPCs exhibited low cytotoxicity and effective DNA binding ability. However, these RPCs displayed distinct diversity from each other, especially in their ability of binding to DNA, buffering capacity and transfection efficiency. Using 293T cell as the mode, we found that the regulation of the monomer sequence of polycations could significantly affect their properties for gene delivery, with differences of 100 fold. The sequence effect might be correlated with different chain folding as well as physiochemical properties of RPCs/pDNA complexes, providing new insight for designing gene vector with promising prospects in gene therapy.

---

Introduction

The sequence of structural units in biological macromolecules has been found to play a critically important role in regulation of its biological activity.¹ Proteins, for instance, are generally constituted of chains of amino acids with precise sequence. These precisely ordered amino acids can fold into an exquisite functional tertiary structure to enable a specific biological activity.² Usually, a minor mutation of the sequence (sometimes a single amino acid) will lead to complete disappearance of the activity.³ For example, disulfide-bond formation by a single cysteine mutation in adenovirus protein VI is able to impair capsid release and membrane lysis.⁴

Polymeric catiomers have received considerable attention for their gene packing capacity and delivery capability by mimicking viruses in nature, which is promising for gene therapy. A polymeric catiomer is generally cationic because it is full of amino groups, including primary, secondary, tertiary and quaternary amines, which are able to pack and protect genetic materials during the circulation and transfer into specific human tissues or cells.⁵ The key factor of gene therapy, however, lies in the development of efficient, nontoxic gene delivery vectors that are capable of mediating high and sustained levels of gene expression.⁶ To date, a great number of polymeric catiomers have been designed and developed for gene delivery.⁷ However, more innovative designs are needed because the gene transfection efficiency in this area is far from expectations.⁸

Learning from the nature, there might be plenty of room to design a polymeric catiomer constituted of amino acids by taking advantages of sequence regulations.⁹ This is due to the fact that polymeric catiomers constituted by amino acids structurally resemble biological macromolecules such as proteins.¹⁰ The effort in this direction is also expected to simultaneously maximize the benefit of material structure and chain folding.¹¹ In this study, polymeric catiomers, also constituted of amino acids, might be an ideal platform to explore this idea. For this consideration, poly(histidine-co-lysine) is chosen for respectable gene packing capability of polylysine and buffering capability of polyhistidine, which are two well recognized important aspects for evaluating a gene vector.¹² Polylysine as the nonviral gene vector has been well documented. In past few years, various strategies, including cleavable PEGylation, cross-linking, and histidylation have been developed to engineer polylysine-based gene vectors with high gene delivery efficiency.¹³

In this study, four reducible polycations were synthesized by the oxidative condensation polymerization of regulated monomers.¹⁴ Each monomer of the four RPCs were obtained in a regulated way through solid phase polymerization in strictly same weight ratio and reaction sequence of histidine and lysine, which ensure the acquirement of four copolymers with monomer sequence regulation. The physiochemical properties of the four polycations were characterized by FTIR, ¹H NMR, circular dichroism spectroscopy, agarose gel electrophoresis, and acid–base titration measurements. The particle size, zeta potential, morphology and ability to mediate gene expression were also evaluated (Scheme 1).

Scheme 1 Structure dependent properties of RPCs. Chemical structure of four synthesized RPCs (left), significant physicochemical parameters (e.g., size and zeta potential) (middle) and biological efficacy (gene delivery efficiency) (right).

Results and discussions

1. Synthesis and characterization of reducible poly(histidine/lysine) catiomer

In this study, poly(histidine-co-lysine) is chosen for respectable gene packing capability of polylysine (PLL) and buffering capability of polyhistidine. It is well-recognized that PLL can package DNA efficiently via electrostatic interactions, forming discrete polyplexes with diameters in the range of 50–100 nm.¹⁵ However, PLL suffers from undesired transfection efficiency when used independently, eventually compromising the gene delivery efficiency. The underlying reason is associated with its weak ability to escape from endocytic vesicles into the cytoplasm.¹⁶ Thus, the incorporation of endosomolytic agents, such as chloroquine and fusogenic peptides, is desired to significantly improve the transfection performance due to the effect of the endosomal membrane disruption. In this work, histidine(His) groups copolymerized with polylysine (Lys) are employed as a strategy to address this issue, which is also shown to be effective in other studies.¹⁷

Another important consideration is the biocompatibility and bioresponsive release of the packaged gene. It is known that the transfection efficiency of catiomer increases with the molecular weight; however, cytotoxicity increases simultaneously and efficient release of gene inside the cell might also be hindered.¹⁸ Therefore, a strategy to obtain a big-molecular-weight catiomer with reduced toxicity and excellent bio-responsive property is necessary. The method involves the incorporation of disulfide bonds into the main chain of poly(His-co-Lys).¹⁹ Disulfide bond can be selectively cleaved by glutathione that is over-expressed in relevant intracellular reduction conditions, due to which the concentration level of glutathione is a hundred times higher in the intracellular regions than in the extracellular ones. Taking advantage of this bio-responsive property, poly(His-co-Lys) can potentially address the critical issue of both biocompatibility and bio-responsive release of the packaged gene.

To obtain poly(His-co-Lys) with the abovementioned structural requirements, we developed a unique condensation polymerization of peptide oligomers with specific sequence. All of the peptide oligomers carry two cysteine groups at both terminals, which allows their oxidization to form disulfide bonds to obtain poly(His-co-Lys).²⁰ Every peptide oligomer is composed of 12 amino acids (His or Lys) between two cysteine groups. 12 amino acids are designed in one oligomer to minimize the occurrence of intramolecular cyclization. Four oligomers, with the detailed sequence shown in Table 1, were synthesized by the solid phase method and abbreviated as [H₈K₄], [(HHK)₄], [(KKH)₄], and [(HK)₆]. Among them, the oligomers of [H₈K₄] and [(HHK)₄] exhibit exactly the same numbers of His and Lys amino acids, while oligomers of [(KKH)₄] and [(HK)₆] show different contents of His and Lys amino acids. These oligomers are capable of being efficiently copolymerized into poly(His-co-Lys) in a condensation manner under the oxidation procedure of DMSO. The successful co-polymerization can be achieved by decreasing the concentration of thiol group in the reactants, which can be followed by UV spectra. The four reducible polycations (RPCs) were named as poly[H₈K₄], poly[(HHK)₄], poly[(KKH)₄] and poly[(HK)₆]. The molecular weight of RPCs was readily adjusted by controlling the polycondensation reaction time, as shown in Fig. 1. Molecular weights of the four RPCs were calculated according to UV data.

Table 1 Reducible polycation synthesized in this study

Polycation^a	Monomers	Molecular weight^b
a RPCs with sequence regulated monomers linked by disulfide.b Determined by the UV spectra.c H represents histidine; K represents lysine; C represents cysteine.
Poly[(H8K4)]	CHHHHHHHHKKKKC^c	105 600
Poly[(HHK)4]	CHHKHHKHHKHHKC	105 200
Poly[(KKH)4]	CKKHKKHKKHKKHC	105 100
Poly[(HK)6]	CHKHKHKHKHKHKC	104 700

---

Fig. 1 Time course of the oxidative polycondensation of poly[(HHK)₄] monomers. (Molecular weight [] and % residual thiol groups [] vs. time.)

The molecular weights of poly[H₈K₄], poly[(HHK)₄], poly[(KKH)₄] and poly[(HK)₆] were measured to be 10 600, 105 200, 105 100 and 104 700, respectively. The synthesized RPCs were further analysed by spectroscopy techniques. FTIR spectra show the stretching vibration of N–H at 3267 cm⁻¹, C=N at 2281 cm⁻¹ for the imidazole ring of His and a strong absorbance band at 1728 cm⁻¹ associated with C=O stretching vibration (Fig. 2). The peaks of the FTIR spectra revealed that the four RPCs had been successfully synthesized. The ¹H NMR spectra in DMSO-d6 were obtained to characterize the structure of the final products. As shown in Fig. 3, the peaks at 8.53, 7.83 and 7.12 ppm are assigned to the protons of imidazole ring contained in histidine groups. The peak at 3.12 ppm is assigned to the protons of the –CH₂–S– group in Cys units and those at 2.38, 1.86 and 3.95 ppm are attributable to the protons of the –CHCH₂CH₂– group of histidine. The peaks at 4.13 and 1.32 ppm are assigned to the protons of the –OCCHNH– group. By the integration of the peaks in Fig. 3, the weight ratio of histidine and lysine in each copolymer can be calculated to be 2 : 1, 2 : 1, 1 : 2 and 1 : 1, which is in accordance with theoretical value.

Fig. 2 FTIR spectra of products: (I) poly[(H₈K₄)]; (II) poly[(HHK)₄]; (III) poly[(KKH)₄]; and (IV) poly[(HK)₆].

Fig. 3 ¹H NMR spectra of (A) poly[(H₈K₄)]; (B) poly[(HHK)₄]; (C) poly[(KKH)₄]; (D) poly[(HK)₆].

2. Circular dichroism spectra

The result of circular dichroism (CD) spectroscopy is shown in Fig. 4, where a distinct discrepancy can be identified. For four RPCs, their spectra showed ellipticity bands with slightly different positions and obviously different magnitude. Poly[H₈K₄], poly[(HHK)₄], poly[(KKH)₄], poly[(HK)₆] were located at 193, 195, 196, 197, and 222, 221, 218, 214, respectively. It is remarkable that peaks appeared more and more intense and bands shifted with a narrowing interval between two peak positions in the curve. A possible explanation for the observed differences between the four RPCs circular dichroism spectra could be the following. The four RPCs have different degrees of freedom of motion due to different monomer sequence, which significantly affects intermolecular interaction and the subsequent folding.²¹

Fig. 4 Circular dichroism spectra of four RPC.

Because gene package by catiomers is actually a process of intermolecular interaction, RPCs with different degrees of freedom might show significant differences in gene package and delivery efficiency depending on their specific sequence.

3. Agarose gel retardation assay

The nucleic acids condensation and DNase I protection ability of the copolymers were assessed by agarose gel retardation assay. As shown in Fig. 5(I–II), four RPCs/pDNA complexes exhibit distinctly different performance in DNA packaging capacity.

Fig. 5 Agarose gel electrophoresis retardation assay of (A) poly[(H₈K₄)]/pDNA; (B) poly[(HK)₆]/pDNA; (C and E) poly[(HHK)₄]/pDNA; (D and F) poly[(KKH)₄]/pDNA complexes at weight ratios ranging from 0 to 4. The enzyme in panels (E and F) is DNase I.

In particular, the DNA binding ability of four RPCS with different sequence or composition increases from poly[(HK)₆], poly[(HHK)₄], poly[(KKH)₄] to poly[(H₈K₄)]. The DNA migration of poly[(H₈K₄)] (Fig. 5-IA) is completely retarded at a weight ratio as low as 0.8, which is even comparable with the gold standard polyethylenimine (PEI), showing the highest capacity to bind DNA. In remarkable contrast, poly[(HHK)₄] cannot efficiently bind DNA at a weight ratio even as high as 3. Poly[(HK)₆], with same numbers of K and H, displays the weakest binding capacity. From this point of view, increase of both K and H units appears to augment the binding capacity. As mentioned earlier, gene package is actually a process of intermolecular interaction of relatively rigid DNA and catiomer. Thus, a catiomer with the sufficient freedom of motion is presumably favorable for forming more compact gene complexes. This speculation is in line with data from circular dichroism spectra, which already indicated that poly[(H₈K₄)] has the highest degree of freedom of motion, while poly[(HK)₆] is the most rigid.

DNase I protection ability of the four complexes was estimated to investigate the bio-stability of the complexes. Poly[(HHK)₄]/pDNA and poly[(KKH)₄]/pDNA complexes were subjected to sufficient incubation with DNase I at 37 °C for 10 min (Fig. 5-III). No trace of naked pDNA without the protection of catiomer was observed when incubated with DNase I, suggesting the fragile property of the naked pDNA. However, when the pDNA is protected by the encapsulation of poly[(HHK)₄] and poly[(KKH)₄], it can be kept intact without migration from the well, suggesting the effective protection and condensation. The bio-stability is very important because there are abundant endonucleases in tumor cells.

4. Cytotoxicity assay

To evaluate the biocompatibility of four reducible biopolycationic catiomers, their cellular toxicity was examined by MTT assays using PEI (25 kDa) as the control. The cytotoxicity of gene vectors strongly depends on the properties of vectors including charge density, structural feature, and molecular weight.²² High positive charge density would generally cause high cytotoxicity. PLL also complies with abovementioned empirical phenomenon.²³ However, in the current study, with disulfide bonds, which could degrade in cell cytoplasm, the cytotoxicity of RPCs was significantly reduced, as shown in Fig. 6(A and B). The results showed that the cell viabilities of the four RPCs were higher than 80% in all concentrations investigated and significantly lower than that of 25 kDa PEI in both cell lines. The reduced cytotoxicity can be attributed to the incorporation of disulfide bonds because RPCs could be degraded into C[H₈K₄]C, C[(HHK)₄]C, C[(KKH)₄]C and C[(HK)₆]C in cytoplasm that are easily metabolized.

Fig. 6 Cytotoxicity of four RPCs in 293T cells (A) and Hela cells (B) incubation at various concentrations for 24 h.

5. Buffering capacity

The buffering capacities of the RPCs and PEI (25 kDa) were determined by an acid–base titration method. The polymer with a high buffering ability would undergo a small change in pH when the same amount of HCl is added into the polymer solution during titration. As shown in Fig. 7B, PEI (25 kDa) has a strong buffer capacity because of its multi-amine structure (i.e. primary, secondary, and tertiary amines groups).²⁴ However, in contrast, the four RPCs showed relatively lower buffering capacity. In addition, the buffering capacity of catiomers has a slight increase along with increased histidine content. It has been reported that the imidazole group of histidine has the proton sponge effect in the endolysosomal pH range and could enhance the transfection efficiency for impelling gene escaping.²⁵ Presumably, the conjugated bonds of the imidazole groups would allow them to be protonated at acidic pH, which results in a broader buffering pH region of the catiomers containing higher histidine segments.

Fig. 7 Size and ζ-potential distribution of four RPCs/pDNA. The dynamic hydration diameter distribution of poly[(H₈K₄)] is shown in panel (A) with the inserted TEM image (bar: 0.5 μm). Buffering capacity of four RPCs and PEI are displayed in panel (B). Mean particles size (C) and ζ-potential (D) of the RPCs/pDNA complexes are measured at weight ratios from 10 to 50 and 0.2 to 20.

6. Particle size and zeta potential

According to a previous study, the ability of lysine to bind DNA can be affected by its charge density and the length, which could perfectly explain the results. As shown in Fig. 7A, the average dynamic hydration diameter of the complex is around 200 nm. The TEM image in the inset shows the homogeneous spherical shape of the representative complex with a diameter around 80 nm. The difference between the two results is associated with the shrinkage effect of the TEM sample. Comparatively, sizes of the RPCs/pDNA complexes formed at different weight ratios show a dependence on sequence and composition (Fig. 7C). It is seen that poly[(H₈K₄)]/pDNA and poly[(HHK)₄]/pDNA maintain an almost constant size regardless of weight ratio, while poly[(HK)₆]/pDNA only forms a stable complex at weight ratio over 30 and poly[(KKH)₄] has a looser structure when the size approaches 400 or goes above it.

The zeta potential exhibits similar trend with size, as shown in Fig. 7D. Zeta potential of poly[(H₈K₄)]/pDNA increases to approx. 20 mV at a relatively low weight ratio of 5, which was recognized to be within the optimum range of surface charge for gene delivery. Poly[(KKH)₄]/pDNA and poly[(HK)₆]/pDNA, however, are still at negative charge at this weight ratio, indicating that they have lower gene loading capacity. The discrepancy on size and zeta potential might be attributed to different freedoms of chain motion, as shown in previously in the CD spectra. More the freedom of motion is, more compactness as well as more positive surface charge of the RPCs are expected during/after the gene complex formation process.

7. In vitro transfection

Abovementioned results have showed that both sequence and composition significantly affect the chain folding and physicochemical properties of RPCs. This further motivates us to investigate whether these differences can be translated into in vitro transfection efficiency. For this aim, 293T cells as a regular cell line for gene transfection study were selected as model cells and accessed by luciferase assay and the BCA method. The reporter plasmid pEGFP complexed with PEI at optimal weight ratio of 1.3 was used as a positive control.

Four RPCs carrying pEGFP at weight ratios from 10 to 50 (genes being completely bounded within this window) were investigated by quantitatively evaluating their gene delivery efficiency. Fig. 8 shows the quantitative assays of the four complexes, which clearly demonstrates the correlation of gene transfection efficiency with sequence, composition and weight ratio. The protein expressed by poly[(H₈K₄)] is the highest among four RPCs/pDNA complexes, though it is still relatively lower than the positive control PEI. The discrepancy of data between poly[(H8K4)] and poly[(HK)6] is as high as 100 times at several weight ratios, which is a significant difference according to the Student's t-test. Poly[(HHK)₄] and poly[(KKH)₄] RPCs show comparable efficiency at all of the five investigated weight ratios. In addition, the overall transfection efficiency of all the four RPCs/pDNA complexes increases as a function of weight ratio. Thus, the optimal weight ratios of each copolymer/EGFP for transfection were chosen in the subsequent confocal observation. The optimal weight ratios of four RPCs were 40 for poly[(H₈K₄)] and 50 for poly[(HHK)₄], poly[(KKH)₄], and poly[(HK)₆]. pEGFP expression in 293T cells following exposure to four RPCs/pDNA complexes fabricated at optimal weight ratios was qualitatively evaluated by fluorescence confocal microscopy. As shown in Fig. 9, numbers of fluorescent cells visible under the microscope after transfection obviously decrease from poly[(H₈K₄)], poly[(HHK)₄], poly[(KKH)₄] to poly[(HK)₆] in accordance with the quantitative assay. Collectively, the results of in vitro transfection definitely demonstrate the effect of monomer sequence on the transfection efficiency, which is possibly regulated by the freedom of chain folding of the polymer catiomer.

Fig. 8 Quantitative transfection efficiency of four RPCs/DNA complexes at various w/w ratios ranging from 10 to 50 in 293T cells by BCA method.

Fig. 9 GFP detection of the transfected 293T cells, the images are the fluorescent images. The w/w ratios of PEI/pDNA and RPCs/pDNA complexes were 1.33, 40, 50, 40, 50 : 1 (A, B, C, D, and E) observed in 293T cells at a magnification of 20.

Conclusion

In this study, we have developed a method of condensation polymerization to synthesize reducible polycations with predetermined sequence ratio of K and H monomers. The molecular weights of these polycations were controlled at comparable level by tuning polymerization time. These polycations are reducible because a reduction responsive disulfide bond is designed into the main chain. It was observed that sequence and composition are two major parameters regulating chain folding and physicochemical properties. Importantly, we found that the monomer sequence encoded in the polymeric catiomer determines the ability of gene packing capacity and delivery efficiency. The differences in gene delivery efficiencies can be as high as 100 fold.

Experimental section

1. Materials

Four monomer peptides (CH₈K₄C, C(HHK)₄C, C(KKH)₄C and C(HK)₆C) were purchased from GL Biochem (China). PEI (25 kDa) was obtained from Sigma-Aldrich (Natick, MA). Tetrahydrofuran (THF, 99.85%), 1,4-dioxane (99.8%), dimethylformamide (DMF), and hydrogen bromide (HBr) 33 wt% solution in glacial acetic acid were purchased from Acros Organics. Dimethyl sulfoxide (DMSO) was supplied from Aladdin and used as received. Dulbecco's modified Eagle's medium (DMEM), fetal bovine serum (FBS), 1% penicillin–streptomycin (PS), 0.25% trypsin–EDTA, phosphate buffered saline (PBS) and BODIPY 650/665-X were purchased from Gibco Invitrogen. Deoxyribonuclease I (DNase I), 3-(4,5-dimethylthiazol-2-yl)-2,5-dipenyltetrazoium bromide (MTT) and bicinchoninic acid (BCA) protein assay kit were purchased from Beyotime Institute of Biotechnology (China). The reporter gene plasmid, pEGFP-C1 was supplied by Invitrogen.

2. The synthesis of four reducible polycations

Through the oxidation of terminal cysteinyl thiol groups of CH₈K₄C, C(HHK)₄C, C(HK)₆C and C(KKH)₄C, four copolymers (poly[H₈K₄], poly[(HHK)₄], poly[(HK)₆], and poly[(KKH)₄]) were obtained, which were connected via disulfide bonds. The synthesis involved the following procedure: (1) oxidative polymerization from peptides in 2 mL PBS containing 30% DMSO at room temperature for different time periods; (2) four RPCs were purified using centrifugal membrane filters (MWCO: 10 000 Da).

3. Measurements

Using Ellman's reagent [5,5′-dithiobis(2-nitrobenzoic acid), DTNB], we followed the routine procedure to estimate free thiol concentration during or after the polymerization. Before the assay, a standard curve to correlate UV data with thiol concentration was obtained by dissolving cysteine sample in DTNB solution with concentration ranging from 0 to 0.2 mmol L⁻¹. During the polymerization, the reaction was monitored by withdrawing the sample at pre-determined times and subjecting to UV measurements. The thiol concentration was calculated according to the standard curve obtained via UV data. The molecular weight of the catiomer could be further derived from thiol concentration according to end group analysis. The equation: molecular weight of the synthesized catiomer = 2 × m/[thiol group], where m and [thiol group] refer to the weight of catiomer, concentration of thiol group, respectively, was used.

Proton nuclear magnetic resonance ¹H NMR spectra were recorded on an Avance 500 MHz spectrometer (Switzerland) using DMSO-d6 as solvent, TMS (tetramethylsilane) as standard. The ratio of histidine to lysine was calculated according the equation: [Histidine]/[Lysine] = 2S_a/S_i, where S_a and S_i refer to signal integral of the corresponding hydrogen in histidine and lysine structures indicated in Fig. 2. The morphology of RPCs/pDNA complexes was observed by transmission electron microscopy (TEM). The representative TEM micrographs of RPCs/pDNA complexes at w/w ratio of 30 are shown.

4. Preparation of RPCs/pDNA complexes

A plasmid DNA stock solution (664 ng mL⁻¹) was prepared in 40 mM Tris–HCl buffer solution. Four RPCs were dissolved in NaCl solution (150 mM) with a concentration of 2 mg mL⁻¹ and the solution was filtered using a 0.22 μm filter. Then, the RPCs solution and DNA stock solution were mixed to prepare complexes at various weight ratios with gentle vortexing, and they were incubated at 37 °C for 30 min before use.

5. Buffering capacity

The buffering capacity of the copolymers was measured by acid–base titration. In brief, copolymers (6 mg) were dissolved in 150 mM NaCl (30 mL) and were adjusted to pH 10.48 with 0.1 M NaOH. 0.1 M HCl (5 μL each time) was added stepwise into the abovementioned solution and the pH values measured by a microprocessor pH meter were recorded.

6. Agarose gel retardation assay and DNase I protection assay

Comparison of the ability of monomers to form stable complexes with pDNA is necessary to evaluate their properties as gene carriers. In this study, agarose gel electrophoresis assay was conducted. We fabricated complexes with polycations/pDNA weight ratios ranging from 0.2 : 1 to 4 : 1 for the comparison using agarose gel electrophoresis. Polycations/pDNA complex suspensions containing 0.1 μg of pDNA were loaded onto a 1% (w/v) agarose gel containing ethidium bromide. Then, electrophoretic separation was carried out for 40 min under a voltage of 120 V in Tris–acetate(TAE) running buffer. DNA bands were visualized through a UV lamp using an Imago GelDoc system. The DNase I protection assay was similar to the abovementioned assay with some modifications. In brief, before being loaded on the agarose gel, the pDNA complexes were incubated with DNase I at 37 °C for 10 min, and the reaction was terminated by 25 mM EDTA (1 μL) at 65 °C after 10 min.

7. Measurement of particle size and zeta potential

To conduct relevant morphometric analysis of polycations/DNA complexes, the particle size and zeta potential of these complexes were measured. Several complexes with different weight ratios, ranging from 10 to 50, were obtained according to the conditions enlisted before. For more accuracy, the acquired complexes were diluted with 150 mM NaCl solution to 1.0 mL volume prior to the measurement.

8. Cell viability assay

293T and Hela cells, provided by the cell center of Tumor Hospital, Tongji University, were cultured in the DMEM medium supplemented with 10% fetal bovine serum and 1% penicillin–streptomycin. Cells were maintained at 37 °C in a humidified 5% CO₂ atmosphere. Evaluation of the cytotoxicity of RPCs was performed by the MTT method. In brief, the cells were seeded into a 96-well plate at a density of 5 × 10³ cells per well. Following an overnight attachment period, cells were then exposed for 24 h to various concentrations of four reducible polycations ranging from 1.5 × 10⁻² to 0.5 mg mL⁻¹ that were prepared in cell culture media. Then, the medium was replaced by 200 μL of fresh DMEM and 20 μL of MTT (5 mg mL⁻¹). The plate was incubated for an additional 4 h at 37 °C. After the addition of 150 μL of DMSO to each well, the plates were incubated for 10 min at 37 °C before the optical density (OD) was measured at 492 nm using Multiscan MK3 (Thermo Fisher Scientific, Waltham, MA, USA). The relative cell viability was calculated by the following equation: the viability (%) = (OD_sample/OD_control) × 10, where the OD_control was obtained in the absence of polycations, whereas the OD_sample was obtained in the presence of polycations. Each value was averaged from six independent experiments.

9. In vitro transfection

293T cells were used to perform the in vitro transfection, which was evaluated by BCA protein assay kit and flow cytometry method (FCM). The reporter plasmid, pEGFP, was used for the qualitative and quantitative determination. PEI was used for positive control at a weight ratio of 1.3. Briefly, 293T cells (8 × 10⁴ cells per well) were seeded into 24-well plate in full medium. After 24 h, the copolymers/pDNA complexes (0.5 μg pDNA per well) were prepared with weight ratios ranging from 10 to 50 in pure DMEM and added in each well containing 400 μL fresh pure DMEM. After 4 h, the mixed medium was substituted for full medium and the cells were incubated for additional 44 h before the determination. For qualitative observation, the transfected cells were imaged under an inverted fluorescence microscope (Nikon Ti–S invert, ECLIPSE 80i). For the GFP relative quantitative determination, cells were washed with PBS and lysed by reporter lysis buffer (200 μL, Promega, U.S.A.). Then, 100 μL lysates were used for determining the fluorescence intensity using a fluorospectrophotometer (Fluoroskan Ascent FL, Thermo Scientific) at excitation of 488 nm and emission of 509 nm. Total protein content was analysed by BCA protein assay kit according to the manufacturer's instruction. For the quantitative determination of GFP, the cells were prepared in fixative (PBS with 2% paraformaldehyde) and measured by a flow cytometer (FACScan, Becton, and Dickinson). The transfection efficiency was described as the percentage of the GFP-positive cells. Student's t-test was used to determine the statistical difference among groups at a significance level p < 0.05. Data are presented as mean ± standard errors.

Acknowledgements

This work was financially supported by 973 program (2013CB967500) and National Natural Science Foundation of China (51473124, 51173136 and 21104059), and “Chen Guang” project (12CG17) founded by Shanghai Educational Development Foundation.

Notes and references

1. C. M. Thomas and J. F. Lutz, Angew. Chem., Int. Ed. Engl., 2011, 50, 9244–9246.
2. (a) N. Badi and J. F. Lutz, Chem. Soc. Rev., 2009, 38, 3383–3390; (b) Y. Brudno and D. R. Liu, Chem. Biol., 2009, 16, 265–276.
3. J. Li, S. N. Rothstein, S. R. Little, H. M. Edenborn and T. Y. Meyer, J. Am. Chem. Soc., 2012, 134, 16352–16359.
4. C. L. Moyer and G. R. Nemerow, Virology, 2012, 428, 41–47.
5. S. C. De Smedt, J. Demeester and W. E. Hennink, Pharm. Res., 2000, 17, 113–126.
6. (a) B. Lu, C. F. Wang, D. Q. Wu, C. Li, X. Z. Zhang and R. X. Zhuo, J. Controlled Release, 2009, 137, 54–62; (b) T. Kurosaki, T. Kitahara, S. Fumoto, K. Nishida, J. Nakamura, T. Niidome, Y. Kodama, H. Nakagawa, H. To and H. Sasaki, Biomaterials, 2009, 30, 2846–2853; (c) T. Niidome, K. Takaji, M. Urakawa, N. Ohmori, A. Wada, T. Hirayama and H. Aoyagi, Bioconjugate Chem., 1999, 10, 773–780.
7. H. F. Zhou, W. T. Yu, X. Guo, X. D. Liu, N. Li, Y. Zhang and X. J. Ma, Biomacromolecules, 2010, 11, 3480–3486.
8. D. Luo and W. M. Saltzman, Nat. Biotechnol., 2000, 18, 33–37.
9. (a) C. Troiber and E. Wagner, Bioconjugate Chem., 2011, 22, 1737–1752; (b) J. F. Lutz, M. Ouchi, D. R. Liu and M. Sawamoto, Science, 2013, 341, 628–636; (c) D. Fischer, T. Bieber, Y. X. Li, H. P. Elsasser and T. Kissel, Pharm. Res., 1999, 16, 1273–1279.
10. (a) J. F. Lutz, Polym. Chem., 2010, 1, 55; (b) Y. Hibi, M. Ouchi and M. Sawamoto, Angew. Chem., Int. Ed. Engl., 2011, 50, 7434–7437.
11. (a) Y. Lee, K. Miyata, M. Oba, T. Ishii, S. Fukushima, M. Han, H. Koyama, N. Nishiyama and K. Kataoka, Angew. Chem., Int. Ed., 2008, 47, 5163–5166; (b) D. L. McKenzie, E. Smiley, K. Y. Kwok and K. G. Rice, Bioconjugate Chem., 2000, 11, 901–909; (c) S. Park and K. E. Healy, Bioconjugate Chem., 2003, 14, 311–319.
12. (a) C. M. Ward, M. L. Read and L. W. Seymour, Blood, 2001, 97, 2221–2229; (b) W. T. Godbey and A. G. Mikos, J. Controlled Release, 2001, 72, 115–125.
13. (a) J. Gu, X. Wang, X. Jiang, Y. Chen, L. Chen, X. Fang and X. Sha, Biomaterials, 2012, 33, 644–658; (b) C. M. Ward, M. Pechar, D. Oupicky, K. Ulbrich and L. W. Seymour, J. Gene Med., 2002, 4, 536–547; (c) H. Zhu, C. Dong, H. Dong, T. Ren, X. Wen, J. Su and Y. Li, ACS Appl. Mater. Interfaces, 2014, 6, 10393–10407.
14. D. Oupicky, A. L. Parker and L. W. Seymour, J. Am. Chem. Soc., 2002, 124, 8–9.
15. (a) S. Danielsen, K. M. Varum and B. T. Stokke, Biomacromolecules, 2004, 5, 928–936; (b) Y. M. Liu and T. M. Reineke, J. Am. Chem. Soc., 2005, 127, 3004–3015.
16. J. Yang, H. Y. Wang, W. J. Yi, Y. H. Gong, X. Zhou, R. X. Zhuo and X. Z. Zhang, Adv. Healthcare Mater., 2013, 2, 481–489.
17. C. Pichon, C. Goncalves and P. Midoux, Adv. Drug Delivery Rev., 2001, 53, 75–94.
18. M. L. Read, S. Singh, Z. Ahmed, M. Stevenson, S. S. Briggs, D. Oupicky, L. B. Barrett, R. Spice, M. Kendall, M. Berry, J. A. Preece, A. Logan and L. W. Seymour, Nucleic Acids Res., 2005, 33, e86.
19. K. Miyata, Y. Kakizawa, N. Nishiyama, A. Harada, Y. Yamasaki, H. Koyama and K. Kataoka, J. Am. Chem. Soc., 2004, 126, 2355–2361.
20. D. L. McKenzie, K. Y. Kwok and K. G. Rice, J. Biol. Chem., 2000, 275, 9970–9977.
21. (a) A. Spector, J. Am. Chem. Soc., 1969, 91, 220–222; (b) G. D. Fasman, H. Hoving and S. N. Timasheff, Biochemistry, 1970, 9, 3316–3324.
22. L. Yin, H. Tang, K. H. Kim, N. Zheng, Z. Song, N. P. Gabrielson, H. Lu and J. Cheng, Angew. Chem., Int. Ed. Engl., 2013, 52, 9182–9186.
23. S. Asayama, M. Sudo, S. Nagaoka and H. Kawakami, Mol. Pharm., 2008, 5, 898–901.
24. W. T. Godbey, K. K. Wu and A. G. Mikos, J. Controlled Release, 1999, 60, 149–160.
25. W. C. Tseng, T. Y. Fang, L. Y. Su and C. H. Tang, Mol. Pharm., 2005, 2, 224–232.

---