Vinyl Laurate Copolymer

Bioconjugate Chem. 2005, 16, 972−980

972

A Thermoresponsive Chitosan-NIPAAm/Vinyl Laurate Copolymer Vector for Gene Transfection Shujun Sun,† Wenguang Liu,*,† Nan Cheng,† Bingqi Zhang,† Zhiqiang Cao,† Kangde Yao,† Dongchun Liang,‡ Aijun Zuo,‡ Gang Guo,‡ and Jingyu Zhang‡ Research Institute of Polymeric Materials, Tianjin University, Tianjin 300072, P. R. China, and Institute of Endocrinology, Tianjin Medical University, Tianjin 300070, P. R. China. Received March 9, 2005; Revised Manuscript Received June 13, 2005

A carboxyl-terminated N-isopropylacrylamide/vinyl laurate (VL) copolymer was prepared and coupled with chitosan (molecular weight ) 2000) to produce a chitosan-NIPAAm/VL copolymer (PNVLCS) vector. The aqueous solution of PNVLCS displayed an obvious thermoresponsive behavior with a lower critical solution temperature (LCST) about 26 °C. The transmission electron microscopy (TEM) showed that the size of PNVLCS/DNA complexes varied with charge ratios (+/-), and the smaller nanoparticles were formed at higher charge ratios. DLS revealed that the size of complex particles was dependent on temperature. The results of temperature-variable circular dichroism (CD), UV, and electrophoresis retardation indicated that at lower charge ratios, DNA in the complexes assume a B conformation, whereas increasing charge ratios caused B f C type conformation transformation; the dissociationformation of PNVLCS/DNA complexes could be tuned by varying temperature: at 37 °C, the collapse of PNIPAAm in PNVLCS was favorable for the formation of compact complexes, shielding more DNA from exposure; at 20 °C, the hydrated and extended PNIPAAm chains facilitated the unpacking of DNA from PNVLCS, increasing the exposure of DNA. PNVLCS was used to transfer plasmid-encoding β-galactosidase into C2C12 cells. The level of gene expression could be controlled by varying incubation temperature. The transfection efficiency of PNVLCS was well improved by temporarily reducing culture temperature to 20 °C, whereas naked DNA and Lipofectamine 2000 did not demonstrate the characteristics of thermoresponsive gene transfection.

INTRODUCTION

In the past decade, much progress has been made in the study of chitosan (CS)-based nonviral vectors for gene transfection, and many encouraging results have been acquired (1-4). To further improve gene transfection level, a variety of strategies were reportedly employed recently, such as coupling deoxycholic acid (5), galactose (6), poly(vinyl pyrrolidone) (7), and urocanic acid (8) to chitosan. Although there were reports supporting that the modification of chitosan could really increase the transfection efficiency of chitosan-DNA complexes, and there was even in vivo study on using chitosan as a vector for gene therapy (9, 10), the transfection efficiency of chitosan is relatively lower compared with that of liposome (11, 12). Like conventional polycationic vectors, chitosan forms polyelectrolyte complex (PEC) with DNA through electrostatic interaction between primary amino and phosphate groups. The strong electrostatic interaction is favorable for the formation of PEC, but on the other hand, it prevents gene dissociation from its carrier in nucleus, impeding the access of RNA polymerase to DNA so that gene expression level is decreased (13). To weaken the electrostatic attraction of chitosan with DNA, long hydrophobic alkyl side chains were incorpo* To whom correspondence should be addressed: E-mail: [email protected]. † Research Institute of Polymeric Materials, Tianjin University. ‡ Institute of Endocrinology, Tianjin Medical University.

rated into chitosan, and transfection efficiency was increased with elongating the length of side chains. It was considered that the hydrophobic chains promoted the cell entry and the unpacking of genes from alkylated chitosan vector (14), but the alkylated chitosans were insensitive to environment, so gene expression could not be controlled by varying external stimuli. To fulfill this objective, Hennink’s team explored the feasibility of thermosensitive poly(N-isopropylacrylamide)(PNIPAAm) copolymer as a smart gene delivery vector (15). They merely investigated the influence of temperature on the stability of the complexes, but not on temperaturedependent gene transfection. Recently, Kurisawa et al. reported a thermoreversible terpolymer vector, poly(N-isopropylacrylamide (NIPAAm)-co-2-(dimethylamino)ethyl methacrylate DMAEMA)-co-butyl methacrylate (BMA)) (16). It was found that at a temperature above LCST of this copolymer, a tight copolymer/pCMV-lacZ plasmid DNA complex was formed, which was beneficial for the cellular uptake and protection of DNA from DNase attack. While lowering incubation temperature below LCST, the gene dissociated from its carrier, improving gene transfection even with small doses of DNA. Both branched and linear polyethyleneimines (PEI) have been extensively used to transfect a variety of cells in vitro and in vivo (17, 18). But its significant toxicity has been a bottleneck for its clinical application (19, 20). Tu¨rk et al. prepared temperature-sensitive PNIPAAm-b-PEI copolymers (21). The copolymerization with PNIPAAm reduced the cytotoxicity of branched PEIs with higher molecular weights, and PNIPAAm/PEI25L (linear PEI with molecular weight of 25 kDa) did not show any

10.1021/bc0500701 CCC: $30.25 © 2005 American Chemical Society Published on Web 06/28/2005

Thermoresponsive Copolymer Vector for Gene Transfection

significant toxicity. At body temperature, this copolymer vector could ferry more DNA into Hela cells. In particular, PEI25L obtained the most successful gene expression with low toxicity. Novel smart gene vectors of poly(L-lysine)-graft-poly(N-isopropylacrylamide) were also designed by Oupicky´ et al. (22). It was demonstrated that the structural density and surface charge of the copolymer/ DNA complexes could be adjusted by temperature. Moreover, it increased as the PNIPAM grafts collapsed above their phase transition temperature. In addition, more DNA tended to be released from polyelectrolyte complexes in the presence of PNIPAAm with addition of heparin. To our best knowledge, research work on controllable gene expression of chitosan-based vectors has not been reported yet in the literature to date. In this study, we synthesized carboxyl-terminated NIPAAm/vinyl laurate (VL) random copolymer and coupled PNIPAAm/PVL to chitosan to generate a thermoresponsive CS-NIPAAm/ VL (PNVLCS) copolymer vector. The PNVLCS was investigated as temperature-sensitive transgene vector for the first time. The temperature-dependent physicochemical properties of PNVLCS-DNA complexes were elucidated in detail. This copolymer vector was applied to the mediation of plasmid DNA transfection into C2C12 cells, and the control of gene expression was examined at varied temperatures. EXPERIMENTAL SECTION

Materials. N-Isopropylacrylamide (NIPAAm, Aldrich Chemical. Co.) was purified by recrystallization in hexane and dried in vacuo at 25 °C. Vinyl laurate (VL) was purchased from Fluka. The chain transfer agent, thioglycolic acid (TGA) (Sigma) was utilized to include carboxylic acid ends to the copolymer chains. The initiator, 2,2′-azoisobutyronitrile (AIBN) was recrystallized from methanol. Chitosan (degree of deacetylation ) 80.3%, molecular weight ) 2 kDa) was supplied by Qingdao Haihui Bioengineering Ltd. China. The watersoluble activating agent, 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide hydrochloride (EDAC) was obtained from Sigma. Plasmid pTracer-CMV/Bsd/lacZ and Lipofectamine 2000 reagent were provided by Invitrogen. Ethidium bromide (EB) and N,N,N′,N′-tetramethylethylenediamine (TEMED) were provided by Fluka. All other reagents used were analytical grade and used as received. C2C12 cells were maintained in Dulbecco’s Modified Eagles Medium (DMEM, Sigma) supplemented with L-glutamine, penicillin/streptomycin, and 10% fetal bovine serum. Cells were incubated at 37 °C in a humidified atmosphere containing 5% CO2 and used at appropriate degrees of confluence. Synthesis of Copolymers. First, the carboxylic acidterminated copolymer of NIPAAm and vinyl laurate (PNVLCOOH) was synthesized by a free radical polymerization in THF. N-Isopropylacrylamide(2.0 g), vinyl laurate(0.98 g), TGA (50 µL), and AIBN (20 mg) were dissolved in THF (25 mL). Polymerization was conducted under nitrogen-atmosphere, in a sealed reactor placed into a thermostated oil bath at 65 °C for 8 h. Once cooled to room temperature, the solution was added to excess amount of diethyl ether to precipitate PNVLCOOH. To remove the impurities in precipitated copolymer, dissolution-precipitation procedure was repeated three times, and the resultant precipitate was dried in a vacuum overnight. Second, PNVLCOOH was coupled to chitosan (CS) by the reaction of the carboxylic acid groups activated with

Bioconjugate Chem., Vol. 16, No. 4, 2005 973

EDAC with the amine groups of chitosan. Chitosan (2.2 g) was dissolved in TEMED/HCl buffer (pH4.8, 25 mL) at 10 °C, and PNVLCOOH(1.5 g) was added to the solution. After 0.21 g of EDAC was added, the solution was stirred at 10 °C for 10 h. For purification, the solution of copolymer was raised to 28 °C, and the precipitate was collected by centrifugation (15000 rpm for 5 min) at 28 °C. Other water-soluble impurities in precipitate were removed by dialysis (MWCO 3500) against deionized water for 5 days at 4 °C. The final copolymer (PNVLCS) was recovered by lyophilization of the dialyzed solution. PNVLCS/DNA Complexes Formation. To ensure that chitosan in PNVLCS was positively charged, PNVLCS was dissolved in 0.1 M sodium acetate/0.1 M acetic acid buffer (pH 5.4) to form a solution of 1.0 mg/ mL, and 0.1 mg/mL of DNA solution was prepared in the same way. A series of PNVLCS/DNA complexes at various charge ratios were prepared by mixing PNVLCS solutions with a DNA solution, vortexing for 15 s, and incubating for 30 min at 37 °C prior to characterization use. It was noted that during complex preparation, the volume of DNA was constant and the volume of PNVLCS solution was varied to adjust the theoretical charge ratio (+/-) (molar ratio of amine to phosphate groups). All complexes for the characterization below were prepared in terms of this method unless otherwise statement. FTIR Analysis. The solutions of chitosan, PNVLCOOH, and PNVLCS were poured into freshly cleaned plastic dishes and maintained in a vacuum overnight at 50 °C for film formation. FTIR spectra of the samples were measured over 4000-800 cm-1 on a Bio-Rad FTS 135 spectrophotometer. 1 H NMR Spectroscopy. To estimate the composition ratios of copolymers, 1H NMR spectra of PNVLCOOH and PNVLCS were measured at ambient temperature with a Varian UNITY plus-400 NMR spectrometer using D2O as a solvent. Gel Permeation Chromatography (GPC). The molecular weights and molecular weight distributions of PNVLCOOH and PNVLCS were determined by gel permeation chromatography(GPC Waters 510/M32) using water as a mobile phase. Monodisperse polystyrene (Polymer Laboratories Inc., MA) was used for calibration. GPC analysis showed Mn ) 3079, PDI ) 1.66 (PNVLCOOH) and Mn ) 4715, PDI ) 1.37 (PNVLCS). Noted that the measured molecular weight is lower than theoretical value, 5079. This discrepancy is probably from polystyrene calibration standard and aggregation of NIPAAm-VI chains. Determination of LCST of Polymer Solution. The variation in the turbidity of 1.0 wt % aqueous copolymer solutions was monitored as a function of temperatures at a fixed wavelength of 500 nm on a UV PC2501 UVvis spectrophotometer equipped with a circulating water bath. The LCSTs were taken at the inflection point in the curves of optical density versus temperature. Transmission Electron Microscopy(TEM). The morphology and size of PNVLCS/plasmid DNA complexes at different charge ratios (+/-) were observed using a JEOL JEM-100CXII TEM. First, one drop of the complexes was deposited on carbon-coated grids. After 5 min on the grids, 1.5% phosphotungstic acid (PTA) was added immediately to negatively stain the complexes for additional 5 min, and then the complexes were recorded on films with TEM. Dynamic Light Scattering (DLS). Particle size of PNVLCS/DNA complexes was measured by dynamic

974 Bioconjugate Chem., Vol. 16, No. 4, 2005

light scatting (DLS). DLS measurement was carried out with an argon ion laser system tuned at 514 nm. The complex solutions were filtered through a 0.45 µm filter (Millipore) directly into a freshly cleaned 10 mm-diameter cylindrical cell. The intensity of autocorrelation was measured at a scattering angle (θ) of 90° with a Brookhaven BI-9000AT digital autocorrelator at room temperature. When the difference between the measured and the calculated baselines was less than 0.1%, the correlation function was accepted. The mean diameter was evaluated by the Stokes-Einstein relationship. In this experiment, particle size of complexes was determined at 20 and 37 °C, respectively. Circular Dichroism (CD). Samples were prepared at various PNVLCS/DNA charge ratios with DNA concentration of 40 µg/mL. Then the solutions were diluted with 0.1 M sodium acetate/0.1 M acetic acid buffer to a final DNA concentration of 1.1 × 10-4 M. CD spectra were collected in a 1 cm path length cuvette using a J-7150 spectropolarimeter in a range of wavelengths of 320180 nm at 20 and 37 °C, respectively. Measurements were performed with a speed of 20 nm/min and a resolution of 0.5 nm. The spectra were corrected by subtracting the background of sodium acetate/acetic acid buffer, and three spectra were accumulated and averaged for each sample. The CD signal was converted to molar ellipticity [θ], deg‚cm2/dmol. Temperaure-Variable UV. Temperaure-variable UV spectrophotometer was used to analyze the formation and dissociation of PNVLCS/DNA complex. The complexes at various charge ratios (+/-) were prepared by mixing different amount of PNVLCS solution with 3 µL of plasmid DNA solution. The total volume of the complexes was adjusted to 1 mL before incubation. The absorbance of the complexes at 20 °C and 37 °C was separately recorded at 260 nm on UV-2501PC spectrophotometer. The solvent was used as blank reference. Gel Retardation Assay. PNVLCS/DNA complexes with varied charge ratios were prepared as mentioned above. The complex formation was carried out for 30 min at 20 and 37 °C, respectively. Aliquots of the complex solutions were loaded onto 0.6 wt % agarose gel (100V), and DNA bands were visualized by ethidium bromide staining. These gel assays were performed at temperature below and above LCST of PNVLCS solution for 20 min. In Vitro Gene Expression. Preparation and Purification of Plasmid. Plasmid pTracer-CMV/Bsd/lacZ was amplified in E. coli and then prepared and purified using Wizard Purification Plasmid DNA Purification System (Promega). DNA concentration and purity were assessed via UV optical intensity at 260 nm and 280 nm. The plasmid DNA was stored at -20 °C until transfection experiments. Cell Transfection. C2C12 cells, a mouse myoblast cell line derived from mouse skeletal muscle, were used in this experiment considering that (1) this cell line has been shown to be a suitable host for stable transfection of exogenes (14); (2) C2C12 myoblast has proved to be capable of synthesizing some osteogenesis factors (23) and secreting the hormone or growth factors (24). Thus for our future gene therapeutic application, a target gene will be delivered into muscle cells within the cavum articulare by PNVLCS vector to treat cartilage injury. C2C12 cells were seeded at a concentration of 6 × 104/mL, 0.5 mL/well, in 24-well plate. These wells were then incubated at 37 °C in 5% CO2 for 24 h in DMEM supplemented with fetal bovine serum (FBS) and antibiotics. Before transfection the medium in each well was

Sun et al.

Figure 1. FTIR spectra of PNVLCOOH (a), chitosan (b), and PNVLCS (c).

removed, and rinsed with fresh DMEM (without FBS and antibiotics). Then the transfection solutions (containing PNVLCS/DNA complexes described above) were placed in each well and left in the CO2 incubator. Three hours later, the medium containing complex solution was removed and then incubated in fresh DMEM medium supplemented with FBS and antibiotics at different temperatures and times: (1) 45 h at 37 °C, and (2) 18 h at 37 °C + 3 h at 20 °C + 24 h at 37 °C. After the incubation was finished, the activity of β-galactosidase in transfected cells was determined using β-Galactosidase Enzyme Assay System with Reporter Lysis Buffer (Promega, Madison, WI). The cells were washed twice with PBS and harvested with reporter lysis buffer. The β-galactosidase activity in transfected cells was measured at wavelength 450 nm by plate-reader of Labsystem using o-nitrophenyl-β-D-galactopyranoside (ONPG) as substrate. Cell lysates and a series of protein standard solutions prepared with bovine serum albumin (BSA) were subjected to BCA Protein Assay (Pierce, Rockford, IL) to quantify the total amount of protein in the transfected cells. The transfection efficiency was evaluated as the β-galactosidase expression level per microgram of cellular protein harvested (milliunit/mg protein). For comparison, the naked DNA and Lipofectamine 2000 were used as controls. The ratio of Lipofectamine 2000 to DNA was 2 µL/1 µg according to the recommended protocol. Each transfection experiment was carried out in triplicate. Statistical Analysis. Both Student’s t-test and oneway ANOVA with post hoc tests were used to evaluate the thermoresponsive transfection efficiency with P < 0.05. The transfection efficiency at constant 37 °C was treated as a control. RESULTS AND DISCUSSION

Characteristics of PNVLCS Copolymer. The FTIR spectra of the PNVLCOOH, chitosan, and PNVLCS are given in Figure 1. As shown in Figures 1a and 1c, the amide peaks of NIPAAm units appear at 1650-1660 cm-1 (CdO stretching, amide I), 1530-1540 cm-1 (N-H bending, amide II), and 3420-3550 cm-1 (N-H stretching). A CdO stretching peak (related to carboxylic acid unit and ester unit of vinyl laurate) appears at 17301740 cm-1 in Figure 1a. This peak decreases when the copolymerization between PNVLCOOH and CS occurred (in Figure 1c). The bands at 1365 and 1386 cm-1 are attributed to the deformation of the two methyl groups in isopropyl groups (25). In the chitosan spectrum (Figure



Figure 2. 1H NMR spectrum of PNVLCOOH (in D2O). The composition ratio(NIPAAm/VL, mol/mol) of PNVLCOOH is determined by comparing the peak areas of d and a in the spectrum.

1b), the characteristic absorbance of amide I is around 1657 cm-1, and C-O stretching peak of pyranose ring is at 1029-1071 cm-1 (26), whose appearance in Figure 1c indicates the introduction of the chitosan unit into the copolymer chain. The 1H NMR spectrum of PNVLCOOH is depicted in Figure 2. It is evident that the peaks of NIPAAm and VL are well resolved (27, 28), so the composition ratio (NIPAAm/VL, mol/mol) of PNVLCOOH is determined by comparing the peak areas of d and a in 1H NMR spectrum, which represent the proton of the isopropyl CH in NIPAAm and the CH3 of laurate (28), respectively. The calculated NIPAAm/VL composition ratio is 25:1, suggesting that only partial VL monomer reacted with NIPAAm supposedly due to its low reactivity caused by its steric hindrance of long alkyl chain (the actual feed ratio is 4:1). Noted that in Figure 2, the structure of PNVLCOOH does not necessarily reflect the single case of chain transfer; statistically, thioglycolic acid is likely to attach to the NIPAAm end. Herein, only one case is presented. In our study, hydrophobic vinyl laurate was used to lower the LCST of the copolymer so that the formation and dissociation of PNVLCS/DNA complexes can be easily modulated by controlling temperature. It was also expected that incorporation of a small quantity of laurate vinyl could increase the cellular entry of PNVLCS vector. In addition, we will focus on the effect of charge ratios of vector to DNA on the properties of the complexes. In 1H NMR spectrum of PNVLCS (Figure 3), there still appear some feature peaks of NIPAAm (29), but the signal peaks of VL cannot be differentiated. The peaks at 2.702, 3.449, and 4.434 ppm are assigned to the H-2, H-3, and H-1 of chitosan (30). The signals of H-4, H-5, H-6 of chitosan cannot be resolved due to overlapping with those of NIPAAm. Combining the results of FTIR and NMR, it is reasonable to consider that chitosan has been coupled to PNVLCOOH. In designing chitosan-NIPAAm/vinyl laurate, we first considered the LCST of the resultant copolymer solution.

Figure 3.

1H

NMR spectrum of PNVLCS (in D2O).

Figure 4. Absorbance versus temperature plots obtained for 1.0 wt % aqueous solutions of PNVLCOOH and PNVLCS. The LCSTs were taken at the inflection point in the curves of optical density versus temperature.

Figure 4 presents the variation in the turbidity of PVLCOOH and PNVLCS solutions as a function of temperature. From the figure, it is seen that there exist sharp changes of absorbance in both cases of PNVLCOOH and PNVLCS solutions. Obviously, introducing hydrophobic laurate moieties to the PNIPAAm chains causes a significant decrease in LCST from 32 °C to 20 °C. While the influence of hydrophilic carboxyl-capping on LCST can be neglected. In comparison, the LCST of PNVLCS solution is increased to 26 °C when PNVLCOOH is coupled with hydrophilic chitosan chains. Physicochemical Properties of PNVLCS/DNA Complexes. The representative TEM images of PNVLCS vector and PNVLCS/DNA complexes with varied charge ratios are exhibited in Figure 5. The complexes at the selected charge ratios display globular shape, and the average sizes of globules in TEM images are summarized in Table 1. Somewhat surprisingly, PNVLCS also appears micelle-like morphology, and its average diameter is as high as 118 nm. In our previous work on thermoresponsive transition behavior of vinyl laurate-


Sun et al.

Figure 5. TEM images of PNVLCS (a) and PNVLCS/DNA complexes at charge ratios of 1:2 (b), 1:1 (c), and 5:1 (d). The scale bars ) 200 nm. 1.5% phosphotungstic acid (PTA) was used as a negative staining agent. Table 1. Particle Size Obtained from TEM Images complex charge ratio (+/-)

diameter (nm)

PNVLCS 1:8 1:2 1:1 2:1 5:1 15:1

118 ( 10 100 ( 9 121 ( 15 148 ( 29 87 ( 23 83 ( 13 67 ( 14

Table 2. Mean Diameters of PNVLCS/DNA Complexes at 20 and 37 °Ca mean diameter (nm) complex charge ratio (+/-)

at 20 °C

at 37 °C

1:8 1:2 1:1

87 ( 11 158 ( 24 146 ( 22

92 ( 15 202 ( 23 141 ( 17

a Values are the results of three measurements and are denoted as mean ( SD.

modified PNIPAAm copolymer solution (28), we found that incorporation of VL moieties could facilitate hydrophobic aggregation of chains far below the LCST. Therefore, NIPAAm/VL-coupled hydrophilic chitosan copolymer is inclined to form a larger micelle-like structure even at room temperature. With the increment of charge ratio, the size of complexes increases and reaches a maximum at 1:1. Interestingly, with a further increasing charge ratio up to 15:1, the size of complex shows a decreasing trend. The size of complex is dependent on the interplay of hydrophobic and electrostatic interactions between DNA and PNVLCS. At lower charge ratio, the self-aggregation from hydrophobic alkyl chains of VL in PNVLCS causes a slight increase in the size of complex. Whereas at higher charge ratio, more compact complexes may be resulted from the strong electrostatic attraction. To study the effects of temperature on the size of PNVLCS/DNA complexes, DLS experiment was performed at 20 °C (lower than LCST) and 37 °C (higher than LCST). In our experiment, we only selected three charge ratios, 1:8, 1:2, and 1:1 (Table 2). In the cases of 1:8 and 1:2, the mean diameter of complexes is increased

Figure 6. CD spectra of DNA and PNVLCS/DNA complexes at different charge ratios (at 20 °C). In each complex solution, the concentration of DNA is 1.1 × 10-4 M. The spectra were corrected by subtracting the background of sodium acetate/acetic acid buffer (pH 5.4), and three spectra were accumulated and averaged for each sample.

while temperature is raised from 20 to 37 °C. For charge ratio of 1:1, the diameter is diminished to a measurable extent while temperature is increased. The temperaturedependent change in the size of complex could be ascribed to two opposite factors. On one hand, at temperature above LCST, PNIPAAm chains collapse to form highly compacted structure, but meanwhile the hydrophobic chains are prone to self-aggregate. In some range of charge ratios, the aggregation plays a dominant role, rendering the size of complexes larger; at other charge ratios, the role of aggregation is marginal, and instead the collapse of PNIPAAm chains in combination with electrostatic attraction facilitates the formation of dense complexes, that is, a smaller size of particles. The CD spectra of PNVLCS/DNA complexes with various charge ratios determined at 20 °C are illustrated in Figure 6. The positive and negative ellipticities centered around 275 and 245 nm arise from the DNA itself, revealing a type of double-stranded structure of DNA. The free DNA shows a typical CD spectrum for B-type DNA conformation. At 1:8 charge ratio, the peak intensity remains nearly unchanged compared to that of free DNA; with the increase of charge ratio, the molar ellipticities of the DNA are suppressed, and the positive band at 275 nm is redshifted. Interestingly, one can see two distinct variation trends of DNA conformation in the selected range of charge ratios. From 1:8 to 2:1, DNA in these complexes assumes a B conformation. A similar result was reported in our previous work on chitosan/DNA complexes (31), but the Cotton patterns observed with charge ratios from 5:1 to 15:1 demonstrate that the DNA changes from B to C conformation. The B-C type transformation was also reported in PLL-g-dextran/DNA complex system (32). We also recorded the CD spectra for the above series of PNVLCS/DNA complexes at 37 °C. The variation trend of CD pattern is identical to that at 20 °C. Though at the temperature above LCST, a heavy syneresis of PNIPAAm occurs, we see no B-A type transformation. Figure 7a displays the CD spectra of complexes with varied charge ratios at 20 and 37 °C. It should be pointed out that at 37 °C, the intensity of free DNA is slightly decreased. To isolate the thermoresponsive phase transition, the difference in the CD intensity caused by temperature has been subtracted in the data treatment. It



Figure 7. Effect of temperature-induced phase transition of PNVLCS on CD spectra of complexes. (a) Representative CD spectra of complexes at various charge ratios. (Solid line: 20 °C; dash line: 37 °C). The difference in the intensity of CD caused by temperature has been subtracted to eliminate the influence of the decrease in the intensity of free DNA at 37 °C. (b) Dependence of [θ]275 on charge ratios determined at 20 and 37 °C. (b): 20 °C; (1): 37 °C.

is clearly seen from the figures that the CD intensity decreases to a certain degree when the temperature is increased from 20 °C to 37 °C. To give a clear observation of the effect of temperature-induced hydrophobicity on the CD signal, we plotted [θ] values of the CD signal at 275 nm determined at 20 and 37 °C versus charge ratios (Figure 7b). Obviously, the values of [θ]275 at 20 °C are greater than at 37 °C. It has been confirmed that at temperatures below LCST, PNIPAAm chains are hydrated; thus a loose structure of PNVLCS/DNA complex is formed in this case, rendering more DNA exposure, while at 37 °C, the PNIPAAm chains in PNVLCS collapse due to heavy dehydration. The collapsed chains tightly cover the surface of complex, protecting DNA from exposure. Therefore, the CD signals are suppressed at an elevated temperature. To further verify the influence of thermoresponsive phase transition of PNVLCS on the packing-unpacking

of DNA in the complexes, temperature-variable UV was performed. Figure 8 shows the change in OD260 of free DNA and the complex solutions at 20 and 37 °C versus charge ratios. As shown in the figure, OD260 of free DNA is constant at these two temperatures. At the selected temperature, DNA is not denatured, so the direct influence of temperature on the variation in the absorbance of naked DNA can be rationally neglected. We can see that after complexing with PNVLCS vectors, the OD260 values decrease relative to that of free DNA. It is not hard to figure out that with the addition of PNVLCS, DNA is condensed in polyelectrolyte complexes, which reduces the exposure of base-pairs; hence the OD260 decreases. Figure 8 also provides other useful information that all the OD260 values at 37 °C are lower than those at 20 °C. As aforementioned, while the temperature is raised to 37 °C, the PNIPAAm chains in PNVLCS collapse to tightly bind to the surface of complexes, resulting in more


Sun et al.

Figure 8. Change in OD260 of free DNA and PNVLCS/DNA complex solutions at 20 and 37 °C versus charge ratios. (b): 20 °C; (1): 37 °C. Figure 10. Thermoresponsive transfection efficiency of PNVLCS/DNA complexes at different charge ratios. The transfection level at constant 37 °C was used as a control. Data are shown as mean ( SD (n ) 3). P*

Vinyl Laurate Copolymer

Recommend Documents