Biomacromolecules 2003, 4, 1277-1284
1277
Improved Stability of Polycationic Vector by Dextran-Grafted Branched Polyethylenimine Wen-Chi Tseng* and Chai-Min Jong Department of Chemical Engineering, National Taiwan University of Science and Technology, Taipei, Taiwan Received March 18, 2003; Revised Manuscript Received June 25, 2003
In vivo instability of a polycationic vector limits its efficacy after systemic administration. Conjugation of hydrophilic polymers with neutral charge onto polycationic vectors has been used to improve the stability by reducing the interactions between the vectors and the blood components, such as serum albumin. In this study, dextrans of molecular weight 10 000 (dex-10000) and 1500 (dex-1500) were used to produce various degrees of grafting on linear and branched polyethylenimines (PEI), and the dextran-grafted polymers were used to prepare DNA-polymer complexes. The changes in size and in ζ-potential and the extent of DNA release after the exposure of the complexes to bovine serum albumin (BSA) were used to evaluate the stability of the complexes prepared at various ratios of DNA to polymer. Only the use of dextran-grafted branched PEI was found to be effective to improve the stability of the complexes in the presence of BSA. Dex-10000 was noted to provide a slightly better shielding than dex-1500 against the aggregation caused by BSA and helped maintain the sizes within 200 nm and the ζ-potentials close to neutral. It is thus concluded that the dextran-grafted branched PEI improved the stability of the DNA-polymer complexes and showed potential to conjugate with ligands for in vivo targeted gene delivery. Introduction Gene therapy holds promises for curing the effects of acquired and inherited diseases in a straightforward manner by adding, correcting, or replacing genes. Two major delivery systems are used in current clinic trials-viral and nonviral vectors, of which the former might provoke carcinogenesis and induce systemic immune response despite its high transfection efficiency.1,2 The nonviral protocols generally deliver DNA by physical approaches with less potential toxicity concern. More than 20% of the current clinical protocols utilize the self-assembled complexes the formation of which is through charge interactions between negative DNA and cationic carriers such as cationic liposomes,3-5 polylysine,6,7 chitosan,8 and polyethylenimine (PEI).9-11 Among theses cationic carriers, PEI has been received much attention due to its high transfection efficiency. According to the linkage of its repeating units, PEI can be classified as branched and linear forms.9,12 The linear PEI has secondary amines along its backbone except that the terminal ends are primary amines. The branched PEI has an average of 25% primary, 50% secondary, and 25% tertiary amines.11 The high-density amine groups give PEI several advantages over other cationic carriers, such as to help the formation of tighter and smaller complexes with negative DNA through charge interactions,13 to act as proton sponge to facilitate the release of DNA from the endosomes,9 and to aid the delivered * To whom correspondence should be addressed. Postal address: Department of Chemical Engineering, National Taiwan University of Science and Technology, No. 43, Sec. 4, Keelung Rd., Taipei 106, Taiwan. Fax: 886-2-2462-2586. Tel: 886-2-2730-1078. E-mail: tsengwc@ ch.ntust.edu.tw.
plasmid to enter the nucleus.14,15 The use of PEI as a gene carrier has also gained more and more in vivo applications, for example, treating tumors in animal models.16 Although nonviral vectors have the advantages of low immunogenicity, low acute toxicity, simplicity for preparation, and feasibility to be produced on a large scale, there exist two major problems associated with the clinic applications of nonviral vectors, namely, the instability after systemic administration and the lack of specificity for gene targeting.17 After intravenous injection, the cationic vectors tend to form aggregates and are subsequently dismantled because of the charge interactions between the vector and the serum components.18 The vectors also interact with other blood components and result in thrombocytopenia and leucopenia.19 Reducing the overall surface charges of polycationic vectors can minimize such inactivation through grafting biocompatible polymers of neutral charge onto the cationic carriers. The vectors can be then prepared by selfassembly between the modified carrier and DNA. Several studies have demonstrated that the circulation time of cationic vectors after intravenous injection could be improved by grafting biocompatible polymers onto the cationic carriers, such as the grafting of poly(ethylene glycol) (PEG) onto polyethylenimine,20,21 the grafting of PEG onto polylysine,22,23 and the grafting of N-(2-hydroxypropyl)methacrylamide onto polylysine.24 Dextran, a branched polymer consisting of repeating units of glucose, is a commonly used biopolymer in drug delivery systems to enhance the circulation time of drug. Its branched structure might not only provide better shielding effect than those linear polymers, such as PEG, to minimize the charge
10.1021/bm034083y CCC: $25.00 © 2003 American Chemical Society Published on Web 08/12/2003
1278
Biomacromolecules, Vol. 4, No. 5, 2003
interactions with serum proteins but also allow the conjugation of multiple ligands on each dextran molecule to increase the valance of the modified vector. The conjugation of dextran onto PEI has been shown to improve the stability of the DNA-polymer complex in the presence of serum.20,22 Nevertheless it is unclear how dextran molecular weight and the degree of dextran grafting affect the stability of DNApolymer complexes. Serum albumin is a major component in the blood and has been identified as a key element that destabilizes the vectors after intravenous injection.24,25 In this study, bovine serum albumin (BSA) was used to simulate the serum proteins of in vivo environment that the DNA-polymer complexes encountered after systemic administration. The changes in physicochemical properties, mainly size and ζ-potential, were indicators for the stability of the complexes prepared with either dextran-grafted linear PEI or dextrangrafted branched PEI after the exposure of the complexes to BSA. The use of dextran in grafting had two purposes: one was to evaluate the shielding effect of dextran for enhancing the stability of the complexes; the other was to mimic the conjugation of a targeting ligand. Four factors that might affect the stability of the complexes were investigated: the types of PEI, the molecular weights of dextran, the degrees of grafting, and the ratios of the amine groups on PEI to the phosphate groups on DNA. Materials and Methods Plasmid Preparation. The pEGFP-C1 vector (Clontech, Palo Alto, CA) contains a humanized and mutated green fluorescent protein gene driven by a cytomegalovirus promoter. It was amplified in E. coli DH5R (Life Technologies, Gaithersburg, MD) by a standard procedure and purified using a commercial kit (Qiagen, Hilden, German). Modification of Polyethylenimine by Dextran. Dextran was grafted onto polyethylenimine (PEI) by a reductive reaction between the aldehyde group of dextran and the amine group of PEI using sodium cyanoborohydride.20 Dextrans of two different molecular weights were useds 1500 (Fluka, Buchs, Switzerland), denoted as dex-1500, and 10 000 (Amersham-Pharmacia, Piscataway, NJ), denoted as dex-10000. Because previous studies have shown that the PEIs of molecular weight 25 000 are more effective reagents for transfection than other PEIs,9,10 this study used linear and branched PEIs (Aldrich, St. Louis, MO) of approximately the same molecular weight of 25 000. PEI (10 mg) was mixed with the desired amount of dextran plus 15 mg of sodium cyanoborohydride in 5 mL of borate buffer (100 mM, pH 8.0). The reaction mixture was incubated at 45 °C for 48 h. The dextran-grafted PEI, denoted as dex-g-PEI, was lyophilized after the removal of sodium cyanoborohydride and the unreacted dextran by dialysis. Element analysis performed on Heraeus Vario EL-III (Elementar Analyzensysteme, Hanau, Germany) was employed to determine the carbon and nitrogen percentages in the synthesized dex-gPEI. Preparation of DNA-Cationic Polymer Complexes. PEI was dissolved at a concentration of 40 mg/mL in water
Tseng and Jong
from Milli-Q Ultrapure Water System (Bedford, MA) and titrated to pH ) 7.0 with 1 N HCl. The desired amounts of cationic polymers were diluted in 200 µL of dilution buffer (20 mM Hepes, 5.2% glucose, pH 7.0) and added into an equal volume of dilution buffer containing 40 µg of DNA. Equilibrium was reached by allowing the mixture to stand at room temperature for 20 min. The charge ratio of cationic polymer to DNA was expressed as N/P ratio in which N represented the number of amine in the polymer and P represented the number of phosphate in the DNA. Measurement of the Interactions between DNA and Cationic Polymer. DNA intercalating dye, TO-PRO-1 (Molecular Probes, Eugene, OR), was used to examine the association of DNA with cationic polymers. The fluorescence intensity of TO-PRO-1 was measured at an excitation wavelength 515 nm and emission wavelength 531 nm with a 5 nm slit by a Hitachi fluorimeter FP-2500 (Tokyo, Japan). After the complex formation, a sample containing 20 µg of DNA was diluted to 3 mL by using dilution buffer for fluorescence measurement. Measurements of Particle Size and ζ-Potential. The surface charges and particle sizes of the complexes were analyzed by using a Zetamaster system (Malvern, Malvern, UK). The dilution buffer was filtered through stacked 0.2 µm filters to reduce the particulate interferences from the buffer. Before measurement, the samples containing 40 µg of DNA were diluted to obtain a suitable count rate. Each sample was measured 10 times for 120 s, and the distribution was analyzed by automatic mode. Stability of DNA-Polymer Complex in the Presence of BSA. One volume of the complex containing 15 µg of DNA was incubated with three volumes of bovine serum albumin (BSA) (Sigma, St. Louis, MO) solution (240 µg/ mL) at 37 °C for 30 min. After incubation, the changes in particle size and in ζ-potential were analyzed by a Zetamaster system, and the release of DNA was analyzed by TO-PRO-1 fluorescence measurement. Cell Culture and Transfection. MDA-MB-231 cells (a human breast carcinoma cell line, ATCC-HTB-26) were maintained at 37 °C, 100% humidity, in Dulbecco’s modified Eagle medium (DMEM; Sigma, St. Louis, MO) supplemented with 10% fetal bovine serum (Hyclone, Logan, Utah) and antibiotics of penicillin-streptomycin-amphotericin (Hyclone, Logan, UT). Cells were maintained in T-75 flasks with regular medium changes over 3-4 days and were seeded onto 12-well plates at 20 000 cells/cm2 24 h before transfection. The transfection reagent for each well was prepared by mixing 2 µg of plasmid with the desired polymer at N/P ratio of 9 in 200 µL of DMEM. The mixture was allowed to stand at room temperature for 15-30 min followed by addition of 0.8 mL of DMEM. After the cells were cultured with this medium for 6 h, fresh cell culture medium, containing serum but without the transfection reagent, replaced the DMEM at 6 h and was maintained for an additional 18 h. The transfected cells were visualized under a fluorescence microscope (Olympus IX 70, Tokyo, Japan), and the photographs of the cells were taken by a digital camera (Olympus C-4040, Tokyo, Japan) under a magnifica-
Stabilization of Polycationic Vector by dex-g-PEI
Biomacromolecules, Vol. 4, No. 5, 2003 1279
Table 1. Various Degrees of Dextran-Grafting Generated on PEIa type of PEI
dextran MW 10000
branched
linear
a
amount of dextran added (mg)
ratio of weight percentage of carbon/nitrogen (%)
estimated dextran molecules per PEI molecule
percentage of reacted amine of PEI (%)
175.0 87.5 17.5
24.09 ( 2.62 12.89 ( 0.78 4.16 ( 0.22
44.50 ( 5.20 22.22 ( 1.54 4.87 ( 0.43
7.83 ( 0.92 3.91 ( 0.28 0.86 ( 0.07
1500
52.5 26.5 5.25
7.42 ( 0.12 4.85 ( 0.05 2.49 ( 0.03
75.64 ( 1.61 41.60 ( 0.73 10.29 ( 0.54
13.31 ( 0.28 7.32 ( 0.12 1.81 ( 0.09
10000
25.5 12.5
2.76 ( 0.05 2.27 ( 0.07
2.08 ( 0.10 1.11 ( 0.14
0.37 ( 0.02 0.20 ( 0.02
1500
25.5 12.5
2.36 ( 0.21 2.04 ( 0.02
10.06 ( 0.07 4.33 ( 0.05
1.77 ( 0.02 0.76 ( 0.02
Mean ( SD; n ) three independent reactions.
tion of 10 × 10. Photographs of five different fields in total were taken for each well to analyze transfection efficiency, which was expressed as the percentage of the cells expressing GFP and calculated by dividing the number of the green cells by the number of the total cells in the five photographs. Results Degrees of Dextran Grafting on PEI. The degrees of dextran grafting on PEI were modulated by the amounts of dextran in the reaction as shown in Table 1. We estimated the average number of dextran molecules on each PEI molecule on the basis of the ratio of carbon percentage to nitrogen percentage in the dex-g-PEI by assuming that the polydispersities of all of these polymers are equal to one. Branched PEI had three levels of grafting for each molecular weight of dextran. The reaction grafted 76, 42, and 10 molecules of dex-1500 and 45, 22, and 5 molecules of dex10000 onto one molecule of branched PEI. Two levels of grafting were produced on linear PEI for each molecular weight of dextran. One and two molecules of dex-10000 and four and ten molecules of dex-1500 were grafted onto one molecule of linear PEI. After grafting, most of the positive charge groups on PEI were still retained because the percentages of the reacted amine of PEI were below 14%. More than 80% of the added dextran reacted with branched PEI irrespective of the molecular weight of dextran, indicating that the conjugation was very efficient for branched PEI. On the other hand, linear PEI was less reactive with dextran because less than 30% of the added dex-10000 or dex-1500 reacted with the amine groups of linear PEI. Interactions between Dextran-Grafted PEIs and DNA. Figure 1A,B showed the extent of DNA being condensed by the dex-g-PEI. When DNA was condensed after the addition of polymer, the intercalating dye was excluded out of the DNA double helix, causing a reduction in fluorescence intensity.18 The percentage of uncondensed DNA was calculated as the fluorescence intensity after the addition of polymer divided by that before the addition of polymer. When N/P ratio exceeded 6, the percentages of uncondensed DNA approached different asymptotes irrespective of the forms of polymer. The extent of condensation, however, varied with different forms of polymers. Branched PEI
Figure 1. Percentage (mean ( SD; n ) three independent preparations of complex) of uncondensed DNA within the complexes prepared with dextran-grafted and unmodified PEIs at different charge ratios. Panel A shows the amount of uncondensed DNA for using linear PEIs. The estimated degrees of grafting on each linear PEI are 4 (4) and 10 (0) molecules of dex-1500, and one (2) and two (9) molecules of dex-10000, respectively. Panel B shows the amount of uncondensed DNA for using branched PEIs. The estimated degree of grafting on each branched PEI are 5 (2), 22 (9), and 45 ([) molecules of dex10000, and 10 (4), 42 (0), and 76 (]) molecules of dex-1500, respectively. In both panels the solid and dashed lines represent the unmodified and dextran-grafted PEIs, respectively.
condensed DNA more efficiently than linear PEI presumably because of the strong interactions between DNA and the tertiary amines on branched PEI. The grafting of dextran promoted the ability of DNA condensation of linear PEI, for example, a reduction in the percentage of uncondensed DNA from 30% to 15% at N/P ratio of 6 (Figure 1A). For dex-g-linear PEI, the various degrees of grafting and the different molecular weights of dextran used in this study showed roughly the same extent of DNA condensation. For branched PEI, the grafting of dextran slightly decreased the percentages of uncondensed DNA when N/P ratios were below 2 but had little effect on
1280
Biomacromolecules, Vol. 4, No. 5, 2003
Tseng and Jong
Table 2. Sizes (nm) and ζ-Potentials (mV) of the Complexes Prepared by Unmodified and dex-g-PEIs before the Exposure to BSAa linear PEI N/P ratio
dextran MW
branched PEI
molecules of grafted dextran
size
ζ-potential
molecules of grafted dextran
size
ζ-potential
U
1810 ( 199
-12.8 ( 1.2
U
5509 ( 719
-30.8 ( 0.9
10000
1 2
173 ( 2 174 ( 16
-24.7 ( 4.2 -16.2 ( 2.4
5 22 45
308 ( 59 251 ( 17 223 ( 10
-11.3 ( 1.2 -3.5 ( 2.4 -6.2 ( 1.7
1500
4 10
305 ( 35 283 ( 14
-26.2 ( 4.5 -30.3 ( 4.5
10 42 76
1901 ( 388 427 ( 43 576 ( 204
-6.2 ( 2.1 -4.6 ( 0.7 -14.5 ( 0.7
U
456 ( 38
22.2 ( 4.5
U
109 ( 5
12.9 ( 0.2
10000
1 2
224 ( 23 236 ( 7
5.2 ( 1.0 4.7 ( 0.9
5 22 45
133 ( 10 155 ( 10 131 ( 43
3.9 ( 1.6 3.4 ( 0.9 -0.4 ( 1.0
1500
4 10
481 ( 109 440 ( 10
17.0 ( 2.1 17.8 ( 3.3
10 42 76
105 ( 7 163 ( 10 189 ( 21
5.5 ( 0.1 3.1 ( 0.7 2.8 ( 0.3
U
329 ( 137
22.1 ( 4.8
U
77 ( 21
16.4 ( 0.5
10000
1 2
227 ( 9 173 ( 26
6.7 ( 0.3 5.5 ( 1.0
5 22 45
146 ( 3 126 ( 17 129 ( 12
3.4 ( 1.7 2.8 ( 0.2 2.9 ( 0.5
1500
4 10
268 ( 38 294 ( 29
19.2 ( 1.4 19.2 ( 3.3
10 42 76
115 ( 17 166 ( 28 159 ( 21
5.7 ( 0.2 5.4 ( 0.2 3.6 ( 1.7
2
6
9
a
U represents either unmodified linear PEI or unmodified branched PEI. Mean ( SD; n ) three independent preparations of complex.
the promotion of DNA condensation at other N/P ratios (Figure 1B). For those branched PEIs that grafted more than 40 molecules of either dex-1500 or dex-10000, the extents of DNA condensation decreased in comparisons with the use of unmodified branched PEI. Sizes and ζ-Potentials of the Complexes Prepared by the Unmodified and Dextran-Grafted PEI. For unmodified linear PEI, the average sizes of the complexes were about 1800, 450, and 330 nm at N/P ratios of 2, 6, and 9, respectively (Table 2). The use of dex-10000-g-linear PEI reduced the average sizes to be below 250 nm, whereas the use of dex-1500-g-linear PEI hardly affected the average sizes except at N/P ratio of 2. For linear PEI with or without grafted dextran, the ζ-potentials of the complexes were negative at N/P ratio of 2 and turned out to be positive at N/P ratios of 6 and 9. The use of dex-10000-g-linear PEI reduced the ζ-potentials from 22 mV to be below 7 mV at N/P ratios of 6 and 9, but the use of dex-1500-g-linear PEI had little effect on the ζ-potentials at these N/P ratios. On the other hand, the degree of grafting showed very minimal effect on the sizes and ζ-potentials of the complexes prepared with dex-g-linear PEI. For the use of unmodified branched PEI, the average sizes of the complexes were around 5500, 110, and 80 nm and the average ζ-potentials of the corresponding complexes were -31, 13, and 17 mV at N/P ratios of 2, 6, and 9, respectively (Table 1). At N/P ratios of 6 and 9, the use of dex-10000-g and dex-1500-g branched PEIs increased the average sizes to be around 140 and 170 nm, respectively. At N/P ratio of 2, the use of dex-1500-g-branched PEI reduced the sizes to be around 2000 and 500 nm for those branched PEIs that
grafted 10 molecules and more than 42 molecules of dextran1500, respectively, whereas the use of dex-10000-g-branched PEI reduced the sizes to be in the range of 200-300 nm for all the degrees of grafting. The average ζ-potentials of the complexes were increased to be in the range of -4 to -14 mV at N/P ratio of 2 and were reduced to be below 5 mV at N/P ratios of 6 and 9 by using the dex-g-branched PEI. The degree of grafting had various effects on the ζ-potential, depending on the N/P ratio of the complexes. At N/P ratio of 2, there existed a maximum ζ-potential for those PEIs with a medium degree of dextran grafting. At N/P ratios of 6 and 9, the ζ-potentials slightly decreased with the degree of grafting. Stability of the DNA-Polymer Complexes in the Presence of BSA. The stability of the DNA-polymer complexes was evaluated by three measures: the change in size, the change in ζ-potential, and the percentage of DNA release. The change in size was calculated as the absolute difference between the average size of the complex before exposure to BSA and that after exposure to BSA. The same differences of uncondensed DNA and change in ζ-potential were taken as the percentage of DNA release and the change in ζ-potential, respectively. Enhanced stability was indicated by the small change in size, the small change in ζ-potential, and the low percentage of DNA release. After exposure to BSA most of the complexes exhibited increases in size and decreases in ζ-potentials, except for those complexes at N/P ratio of 2 (Table 2, Figures 2 and 3). For those prepared with either unmodified linear or unmodified branched PEI, the complexes tended to be disintegrated at N/P ratio of 2 and to form aggregates at N/P
Stabilization of Polycationic Vector by dex-g-PEI
Figure 2. The stability of DNA-polymer complex in the presence of BSA by the use of dextran-grafted linear PEI. The change in size (panel A), the change in ζ-potential (panel B), and the percentage of DNA release (panel C) (mean ( SD for each; n ) three independent preparations of complex) were used to indicate the stability of the complexes prepared by using the dextran-grafted and unmodified linear PEIs at different N/P ratios after the exposure of the complexes to BSA. Most of the complex sizes increased except those labeled an asterisk (/) above the bar, and most of the complex ζ-potentials decreased except those labeled with a triangle above the bar. U represents the unmodified linear PEI.
ratios of 6 and 9 in the presence of BSA (Table 2, Figures 2 and 3). Accompanying the change in size, the ζ-potential increased at N/P ratio of 2 and decreased at N/P ratios of 6 and 9 (Figures 2 and 3). At N/P ratio of 2, the complex formed aggregates instead of being disintegrated when dextran was grafted onto linear PEI, and the use of dex-1500 led to less changes in size than the use of dex-10000 (Figure 2A). At N/P ratio of 9, the use of dex-10000, however, resulted in less changes of size than the use of dex-1500. Regardless of the molecular weight of dextran, the use of dex-g-linear PEI reduced the change in size after exposure to BSA except at N/P ratio of 6 under which condition the changes in size were higher than for the use of unmodified linear PEI (Figure 2A). For the same molecular weight of dextran, the dependency between the degree of grafting and the change in size seemed not so obvious. At N/P ratios of 6 and 9, the use of dex-10000-glinear PEI reduced the change in ζ-potential, and the reductions continued as the molecules of grafted dextran on PEI increased. On the other hand, the use of dex-1500-glinear PEI was unable to reduce the changes in ζ-potential at all tested N/P ratios (Figure 2B).
Biomacromolecules, Vol. 4, No. 5, 2003 1281
Figure 3. The stability of DNA-polymer complex in the presence of BSA by the use of dextran-grafted branched PEI. Using dextrangrafted branched PEI resulted in less changes in size (panel A), ζ-potential (panel B), and DNA release (panel C) (mean ( SD for each; n ) three independent preparations of complex) after the exposure of the complexes to BSA (Figure 2). Most of the complex sizes increased except those labeled an asterisk (*) above the bar, and most of the complex ζ-potentials decreased except those labeled with a triangle above the bar. U represents the unmodified branched PEI.
After exposure to BSA, most changes in size were within 50 nm at all tested N/P ratios for the use of dex-10000-gbranched PEI and were smaller than those for the use of dex-1500-g-branched PEI (Figure 3A). The effect of grafting degree on the change in size seemed to depend on the molecular weight of dextran. For example, at least 42 molecules of dex-1500 were required to keep the change in size within 100 nm, while 5 molecules of dex-10000 were sufficient to obtain the change in size within 50 nm. The use of dex-10000-g-branched PEI resulted in larger changes in ζ-potential under the same conditions than the use of dex1500-g-branched PEI did (Figure 3B). A minimal degree of grafting could reduce the changes in ζ-potential below 10 mV, but the dependency between the degree of grafting and the change in ζ-potential was not so obvious. DNA intercalating dye was used to examine whether DNA was still associated with the polymer upon exposure to BSA. The results revealed that the grafting of dextran was able to reduce the amount of released DNA for both linear and branched PEI (Figures 2C and 3C). But at N/P ratio of 9, the use of dex-g-PEI resulted in higher amounts of released DNA than the use of unmodified PEI did. For the complexes prepared with dex-g-branched PEI at the tested N/P ratios, the amount of released DNA seemed to be independent of
1282
Biomacromolecules, Vol. 4, No. 5, 2003
Figure 4. Dependency of the dextran grafting and the transfection efficiency of the DNA-polymer complexes. Panel A shows the transfection efficiency of unmodified linear PEI (U) and dex-g-linear PEI. Panel B shows the transfection efficiency of unmodified branched PEI (U) and dex-g-branched PEI (mean ( SD; n ) three independent transfection experiments).
molecular weight of dextran but increased with the increasing number of dextran molecules on PEI except for the complexes prepared at N/P ratio of 2 by using branched PEI grafted with 10 molecules of dex-1500 (Figure 3C). DNA-Polymer Complex Mediated Transfection. Using unmodified branched PEI produced higher transfection efficiency than using unmodified linear PEI did, as shown in Figure 4. When dextran was grafted onto linear PEI, the transfection efficiency was reduced to a minimal level for various grafting degrees in this study (Figure 4A). The transfection efficiency of dex-g-branched PEI depended on the molecular weight of dextran and the grafting degree. Dex10000-g-branched PEI abolished the transfection efficiency for various grafting degrees (Figure 4B). The transfection efficiency, however, was maintained around 15%, a value similar to that of unmodified branched PEI, when branched PEI was grafted with only 10 molecules of dex-1500 and then rapidly declined with the increasing degree of grafting to an undetectable level. Discussion In this study, grafting of dextran onto PEI improved the stability of the complexes in the presence of BSA for branched PEI but not for linear PEI, as indicated by the smaller changes in size, the smaller changes in ζ-potential,
Tseng and Jong
and the lower percentages of DNA release for those complexes prepared with dex-g-branched PEI (Figures 2 and 3). The use of dex-g-linear PEI reduced the changes in size and in ζ-potential under certain conditions (Figure 2), but under most conditions the use of dex-g-branched PEI could further reduce the changes in size by about an order of magnitude and the changes in ζ-potential by about 5-fold, in addition to retaining more DNA within the complexes (Figure 3). The capability of dex-g-branched PEI to retard BSA aggregation depended on the degree of grafting and molecular weight of dextran. Because the molecular size of dex1500 is smaller than that of dex-10000, the improvement of stability required at least 42 molecules of dex-1500 in comparisons with 5 molecules of dex-10000, indicating that the use of dex-10000 was more effective than the use of dex-1500 in retarding the complex aggregation (Figure 3). Less aggregation could prolong the circulation of the complexes because particle size affected the circulation halflife after systemic administration. The stealth liposomes of 100 nm have a half-life of around 48 h in human.26 The complexes prepared with dex-g-PEI were expected to have a shorter circulation time due to the larger average size of 200-300 nm. On the other hand, the particles of size below 300 nm would be able to achieve passive targeting to carcinoma because the passages through the gaps of the leaky endothelium of tumors were found to be around 380-780 nm,27,28 depending on the types and locations of tumors. Grafting dextran onto branched PEI also reduced the amount of DNA released from the complex after exposure to BSA (Figure 3C). The reduction was found most effective at N/P ratio of 2 but not observed at N/P ratio of 9. The increasing number of dextran molecule on PEI caused an increased amount of DNA dissociation from the complexes. The grafted dextran on PEI might enlarge the distance between DNA and the amine groups on PEI and thus reduced the charge interactions. The weakened interactions could not be completely counterbalanced by the shielding effect from the dextran, leading to a slightly loose structure of the complex from which DNA tended to dissociate upon the exposure to BSA. Because most of the secondary and tertiary amine groups were retained on the modified PEI molecule after dextran grafting, the proton-sponge effect was presumably unaffected.29 The reduced transfection efficiency might be caused by the decreased amount of intracellular plasmid. In vitro transfection generally begins with the association of the complexes with the cell membrane, mainly through the interactions of the negatively charged proteoglycan and the positively charged complexes. Because the dextran grafting reduced the surface positive charges of the complexes, as indicated in Table 2, the reduced interactions of the complexes and the cell membrane might decrease the amount of the complexes associated with the membrane and thus allowed less plasmid to be delivered into the cells. When the intracellular plasmid amount fell below a certain threshold, the cells became incapable of carrying out transgene expression.30 Conjugation of a corresponding ligand could enhance the cellular entry of plasmid through receptor-
Stabilization of Polycationic Vector by dex-g-PEI
mediated endocytosis to restore the level of intracellular plasmid for transgene expression. Although both linear and branched PEIs are very effective transfection reagents for in vitro transgene expression and might have the potential to transfect nondividing cells,14,15,31 the complexes prepared with either linear or branched PEI become so unstable that transfection efficiency is greatly reduced after in vivo administration. The formation of aggregates causes most of the complex to be retained within the lung that therefore becomes the tissue of highest transgene expression after intravenous injection of the complex.18 One approach to transfect other organs is to target the complex to a specific organ by conjugation of a corresponding ligand onto the polymer. Based on the molecular size, two major types of ligands are commonly employed: one is a small molecule, such as RGD peptide for targeting to endothelial cells;32 the other is large molecule, such as transferrin for targeting to hepatocytes.21 In this study, dex-1500 and dex-10000 mimicked the small and large ligands, respectively. The incapability of dex-g-linear PEI to enhance the stability of complexes suggested that the use of ligand-conjugated linear PEI might be also incapable of enhancing the in vivo stability of the complexes and hence failed to target the complexes to the corresponding tissue. On the other hand, the potentiality of dex-g-branched PEI to enhance the stability of complex indicated that branched PEI was a better candidate than linear PEI to be conjugated with a corresponding ligand for in vivo targeted delivery. Conclusion Success of targeted nonviral gene delivery depends on how to prolong the half-life of the complexes in the extracellular environment, a major barrier to in vivo administration, and the conjugation of a corresponding ligand for targeting. Both linear and branched PEIs are promising polycationic vectors with high efficiency of in vitro transgene expression, but only branched PEI might serve the purpose for in vivo targeted delivery as indicated by this study. The results of this study also demonstrated that the conjugation of either small or large molecules of dextran onto branched PEI enhanced the stability of complex in the presence of BSA and suggested that successful targeted delivery can be achieved by conjugation of ligands onto the hydroxyl groups of dex-g-branched PEI. Acknowledgment. The authors thank Dr. Tsuei-Yun Fang in the Department of Food Science at the National Taiwan Ocean University for the discussion and suggestion. Our research is supported by the National Health Research Institutes Grant NHRI-EX92-9111SC. References and Notes (1) Miller, A. D. Human gene therapy comes of age. Nature 1992, 357, 455-460. (2) Mulligan, R. C. The basic science of gene therapy. Science 1993, 260, 926-932. (3) Felgner, P. L.; Gadek, T. R.; Holm, M.; Roman, R.; Chan, H. W.; Wenz, M.; Northrop, J. P.; Ringold, G. M.; Danielsen, M. Lipofection: a highly efficient, lipid-mediated DNA-transfection procedure. Proc. Natl. Acad. Sci. U.S.A. 1987, 84, 7413-7417.
Biomacromolecules, Vol. 4, No. 5, 2003 1283 (4) Gao, X.; Huang, L. A novel cationic liposome reagent for efficient transfection of mammalian cells. Biochem. Biophys. Res. Commun. 1991, 179, 280-285. (5) Gao, X.; Huang, L. Cationic liposome-mediated gene transfer. Gene Ther. 1995, 2, 710-722. (6) Gao, X.; Huang, L. Potentiation of cationic liposome-mediated gene delivery by polycations. Biochemistry 1996, 35 (3), 1027-1036. (7) Wolfert, M. A.; Seymour, L. W. Atomic force microscopic analysis of the influence of the molecular weight of poly(L)lysine on the size of polyelectrolyte complexes formed with DNA. Gene Ther. 1996, 3 (3), 269-273. (8) Leong, K. W.; Mao, H. Q.; Truong-Le, V. L.; Roy, K.; Walsh, S. M.; August, J. T. DNA-polycation nanospheres as nonviral gene delivery vehicles. J. Controlled Release 1998, 53 (1-3), 183-193. (9) Boussif, O.; Lezoualc’h, F.; Zanta, M. A.; Mergny, M. D.; Scherman, D.; Demeneix, B.; Behr, J. P. A versatile vector for gene and oligonucleotide transfer into cells in culture and in vivo: polyethylenimine. Proc. Natl. Acad. Sci. U.S.A. 1995, 92, 7297-7301. (10) Ferrari, S.; Moro, E.; Pettenazzo, A.; Behr, J. P.; Zacchello, F.; Scarpa, M. ExGen 500 is an efficient vector for gene delivery to lung epithelial cells in vitro and in vivo. Gene Ther. 1997, 4, 1100-1106. (11) Godbey, W. T.; Wu, K. K.; Mikos, A. G. Poly(ethylenimine) and its role in gene delivery. J. Controlled Release 1999, 60, 149-160. (12) Goula, D.; Remy, J. S.; Erbacher, P.; Wasowicz, M.; Levi, G.; Abdallah, B.; Demeneix, B. A. Size, diffusibility and transfection performance of linear PEI/DNA complexes in the mouse central nervous system. Gene Ther. 1998, 5 (5), 712-717. (13) Tang, M. X.; Szoka, F. C. The influence of polymer structure on the interactions of cationic polymers with DNA and morphology of the resulting complexes. Gene Ther. 1997, 4, 823-832. (14) Pollard, H.; Remy, J. S.; Loussouarn, G.; Demolombe, S.; Behr, J. P.; Escande, D. Polyethylenimine but not cationic lipids promotes transgene delivery to the nucleus in mammalian cells. J. Biol. Chem. 1998, 273, 7507-7511. (15) Brunner, S.; Furtbauer, E.; Sauer, T.; Kursa, M.; Wagner, E. Overcoming the nuclear barrier: cell cycle independent nonviral gene transfer with linear polyethylenimine or electroporation. Mol. Ther. 2002, 5, 80-86. (16) Coll, J. L.; Chollet, P.; Brambilla, E.; Desplanques, D.; Behr, J. P.; Favrot, M. In ViVo delivery to tumors of DNA complexed with linear polyethylenimine. Hum. Gene Ther. 1999, 10, 1659-1666. (17) Tseng, W.-C.; Hunag, L. Liposome-based gene therapy. Pharm. Sci. Technol. Today 1998, 1, 206-213. (18) Li, S.; Tseng, W.-C.; Stolz, D. B.; Wu, S.-P.; Watkins, S. C.; Huang, L. Dynamic changes in the characteristics of cationic lipidic vectors after exposure to mouse serum: implications for intravenous lipofection. Gene Ther. 1999, 6, 585-594. (19) McLean, J. W.; Fox, E. A.; Baluk, P.; Bolton, P. B.; Haskell, A.; Pearlman, R.; Thurston, G.; Umemoto, E. Y.; McDonald, D. M. Organ-specific endothelial cell uptake of cationic liposome-DNA complexes in mice. Am. J. Physiol. 1997, 273, H387-404. (20) Erbacher, P.; Bettinger, T.; Belguise-Valladier, P.; Zou, S.; Coll, J. L.; Behr, J. P.; Remy, J. S. Transfection and physical properties of various saccharide, poly(ethylene glycol), and antibody-derivatized polyethylenimines (PEI). J. Gene Med. 1999, 1 (3), 210-222. (21) Ogris, M.; Brunner, S.; Schuller, S.; Kircheis, R.; Wagner, E. PEGylated DNA/transferrin-PEI complexes: reduced interaction with blood components, extended circulation in blood and potential for systemic gene delivery. Gene Ther. 1999, 6, 595-605. (22) Toncheva, V.; Wolfert, M. A.; Dash, P. R.; Oupicky, D.; Ulbrich, K.; Seymour, L. W.; Schacht, E. H. Novel vectors for gene delivery formed by self-assembly of DNA with poly(L-lysine) grafted with hydrophilic polymers. Biochim. Biophys. Acta 1998, 1380 (3), 354368. (23) Choi, Y. H.; Liu, F.; Park, J. S.; Kim, S. W. Lactose-poly(ethylene glycol)-grafted poly-L-lysine as hepatoma cell-targeted gene carrier. Bioconjug. Chem. 1998, 9 (6), 708-718. (24) Dash, P. R.; Read, M. L.; Barrett, L. B.; Wolfert, M. A.; Seymour, L. W. Factors affecting blood clearance and in vivo distribution of polyelectrolyte complexes for gene delivery. Gene Ther. 1999, 6 (4), 643-650. (25) Li, S.; Rizzo, M. A.; Bhattacharya, S.; Huang, L. Characterization of cationic lipid-protamine-DNA (LPD) complexes for intravenous gene delivery. Gene Ther. 1998, 5 (7), 930-937.
1284
Biomacromolecules, Vol. 4, No. 5, 2003
(26) Gabizon, A.; Isacson, R.; Libson, E.; Kaufman, B.; Uziely, B.; Catane, R.; Ben-Dor, C. G.; Rabello, E.; Cass, Y.; Peretz, T.; Sulkes, A.; Chisin, R.; Barenholz, Y. Clinical studies of liposome-encapsulated doxorubicin. Acta Oncol. 1994, 33, 779-786. (27) Hobbs, S. K.; Monsky, W. L.; Yuan, F.; Roberts, W. G.; Griffith, L.; Torchilin, V. P.; Jain, R. K. Regulation of transport pathways in tumor vessels: role of tumor type and microenvironment. Proc. Natl. Acad. Sci. U.S.A. 1998, 95, 4607-4612. (28) Hashizume, H.; Baluk, P.; Morikawa, S.; McLean, J. W.; Thurston, G.; Roberge, S.; Jain, R. K.; McDonald, D. M. Openings between defective endothelial cells explain tumor vessel leakiness. Am. J. Pathol. 2000, 156, 1363-1380. (29) Thomas, M.; Klibanov, A. M. Enhancing polyethylenimine’s delivery of plasmid DNA into mammalian cells. Proc. Natl. Acad. Sci. U.S.A. 2002, 99, 14640-14645.
Tseng and Jong (30) Tseng, W. C.; Haselton, F. R.; Giorgio, T. D. Transfection by cationic liposomes using simultaneous single cell measurements of plasmid delivery and transgene expression. J. Biol. Chem. 1997, 272, 2564125647. (31) Godbey, W. T.; Wu, K. K.; Mikos, A. G. Tracking the intracellular path of poly(ethylenimine)/DNA complexes for gene delivery. Proc. Natl. Acad. Sci. U.S.A. 1999, 96, 5177-5181. (32) Harbottle, R. P.; Cooper, R. G.; Hart, S. L.; Ladhoff, A.; McKay, T.; Knight, A. M.; Wagner, E.; Miller, A. D.; Coutelle, C. An RGDoligolysine peptide: a prototype construct for integrin-mediated gene delivery. Hum. Gene Ther. 1998, 9 (7), 1037-1047.
BM034083Y
